{
  "best_global_step": 10176,
  "best_metric": 0.48383763432502747,
  "best_model_checkpoint": "saves_multiple/prefix-tuning/llama-3-8b-instruct/train_codealpacapy_42_1760629239/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": 48.71059036254883,
      "learning_rate": 1.179245283018868e-08,
      "loss": 6.6512,
      "num_input_tokens_seen": 2464,
      "step": 5
    },
    {
      "epoch": 0.00589622641509434,
      "grad_norm": 82.15492248535156,
      "learning_rate": 2.6533018867924528e-08,
      "loss": 6.3646,
      "num_input_tokens_seen": 6688,
      "step": 10
    },
    {
      "epoch": 0.00884433962264151,
      "grad_norm": 44.93852233886719,
      "learning_rate": 4.127358490566038e-08,
      "loss": 6.1166,
      "num_input_tokens_seen": 9984,
      "step": 15
    },
    {
      "epoch": 0.01179245283018868,
      "grad_norm": 65.2947998046875,
      "learning_rate": 5.601415094339623e-08,
      "loss": 6.2843,
      "num_input_tokens_seen": 13376,
      "step": 20
    },
    {
      "epoch": 0.01474056603773585,
      "grad_norm": 64.94010162353516,
      "learning_rate": 7.075471698113208e-08,
      "loss": 6.6652,
      "num_input_tokens_seen": 16864,
      "step": 25
    },
    {
      "epoch": 0.01768867924528302,
      "grad_norm": 51.85393524169922,
      "learning_rate": 8.549528301886793e-08,
      "loss": 6.4095,
      "num_input_tokens_seen": 19520,
      "step": 30
    },
    {
      "epoch": 0.020636792452830188,
      "grad_norm": 69.33651733398438,
      "learning_rate": 1.0023584905660378e-07,
      "loss": 6.7411,
      "num_input_tokens_seen": 23360,
      "step": 35
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 54.04358673095703,
      "learning_rate": 1.1497641509433962e-07,
      "loss": 6.5525,
      "num_input_tokens_seen": 27008,
      "step": 40
    },
    {
      "epoch": 0.02653301886792453,
      "grad_norm": 46.656333923339844,
      "learning_rate": 1.297169811320755e-07,
      "loss": 6.4003,
      "num_input_tokens_seen": 30528,
      "step": 45
    },
    {
      "epoch": 0.0294811320754717,
      "grad_norm": 79.08109283447266,
      "learning_rate": 1.4445754716981135e-07,
      "loss": 6.3289,
      "num_input_tokens_seen": 37408,
      "step": 50
    },
    {
      "epoch": 0.03242924528301887,
      "grad_norm": 37.82752990722656,
      "learning_rate": 1.591981132075472e-07,
      "loss": 6.1908,
      "num_input_tokens_seen": 40128,
      "step": 55
    },
    {
      "epoch": 0.03537735849056604,
      "grad_norm": 99.15641021728516,
      "learning_rate": 1.7393867924528304e-07,
      "loss": 6.3195,
      "num_input_tokens_seen": 43392,
      "step": 60
    },
    {
      "epoch": 0.038325471698113206,
      "grad_norm": 50.607093811035156,
      "learning_rate": 1.886792452830189e-07,
      "loss": 6.0541,
      "num_input_tokens_seen": 46336,
      "step": 65
    },
    {
      "epoch": 0.041273584905660375,
      "grad_norm": 46.99235153198242,
      "learning_rate": 2.0341981132075473e-07,
      "loss": 6.2879,
      "num_input_tokens_seen": 49664,
      "step": 70
    },
    {
      "epoch": 0.044221698113207544,
      "grad_norm": 40.56711959838867,
      "learning_rate": 2.1816037735849058e-07,
      "loss": 5.5572,
      "num_input_tokens_seen": 55552,
      "step": 75
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 37.46285629272461,
      "learning_rate": 2.3290094339622643e-07,
      "loss": 6.0275,
      "num_input_tokens_seen": 59072,
      "step": 80
    },
    {
      "epoch": 0.05011792452830189,
      "grad_norm": 87.23681640625,
      "learning_rate": 2.476415094339623e-07,
      "loss": 6.1802,
      "num_input_tokens_seen": 62240,
      "step": 85
    },
    {
      "epoch": 0.05306603773584906,
      "grad_norm": 47.605770111083984,
      "learning_rate": 2.6238207547169815e-07,
      "loss": 6.0592,
      "num_input_tokens_seen": 65600,
      "step": 90
    },
    {
      "epoch": 0.05601415094339623,
      "grad_norm": 39.134613037109375,
      "learning_rate": 2.7712264150943397e-07,
      "loss": 6.2081,
      "num_input_tokens_seen": 68384,
      "step": 95
    },
    {
      "epoch": 0.0589622641509434,
      "grad_norm": 35.796199798583984,
      "learning_rate": 2.9186320754716984e-07,
      "loss": 6.2388,
      "num_input_tokens_seen": 70656,
      "step": 100
    },
    {
      "epoch": 0.061910377358490566,
      "grad_norm": 51.76054382324219,
      "learning_rate": 3.0660377358490567e-07,
      "loss": 5.8466,
      "num_input_tokens_seen": 73632,
      "step": 105
    },
    {
      "epoch": 0.06485849056603774,
      "grad_norm": 56.682796478271484,
      "learning_rate": 3.213443396226416e-07,
      "loss": 5.5644,
      "num_input_tokens_seen": 77152,
      "step": 110
    },
    {
      "epoch": 0.06780660377358491,
      "grad_norm": 40.049007415771484,
      "learning_rate": 3.3608490566037736e-07,
      "loss": 5.4294,
      "num_input_tokens_seen": 80096,
      "step": 115
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 30.174596786499023,
      "learning_rate": 3.508254716981133e-07,
      "loss": 5.765,
      "num_input_tokens_seen": 83488,
      "step": 120
    },
    {
      "epoch": 0.07370283018867925,
      "grad_norm": 33.34933853149414,
      "learning_rate": 3.6556603773584905e-07,
      "loss": 5.5142,
      "num_input_tokens_seen": 86464,
      "step": 125
    },
    {
      "epoch": 0.07665094339622641,
      "grad_norm": 47.40380859375,
      "learning_rate": 3.80306603773585e-07,
      "loss": 5.6643,
      "num_input_tokens_seen": 89984,
      "step": 130
    },
    {
      "epoch": 0.07959905660377359,
      "grad_norm": 56.02983474731445,
      "learning_rate": 3.9504716981132075e-07,
      "loss": 5.4216,
      "num_input_tokens_seen": 93472,
      "step": 135
    },
    {
      "epoch": 0.08254716981132075,
      "grad_norm": 35.633243560791016,
      "learning_rate": 4.097877358490567e-07,
      "loss": 5.3138,
      "num_input_tokens_seen": 96576,
      "step": 140
    },
    {
      "epoch": 0.08549528301886793,
      "grad_norm": 45.071754455566406,
      "learning_rate": 4.2452830188679244e-07,
      "loss": 5.5152,
      "num_input_tokens_seen": 100672,
      "step": 145
    },
    {
      "epoch": 0.08844339622641509,
      "grad_norm": 70.58203887939453,
      "learning_rate": 4.3926886792452837e-07,
      "loss": 5.4596,
      "num_input_tokens_seen": 103584,
      "step": 150
    },
    {
      "epoch": 0.09139150943396226,
      "grad_norm": 38.40453338623047,
      "learning_rate": 4.5400943396226414e-07,
      "loss": 5.5016,
      "num_input_tokens_seen": 107904,
      "step": 155
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 59.74928665161133,
      "learning_rate": 4.6875000000000006e-07,
      "loss": 5.1658,
      "num_input_tokens_seen": 110720,
      "step": 160
    },
    {
      "epoch": 0.0972877358490566,
      "grad_norm": 38.6566276550293,
      "learning_rate": 4.834905660377359e-07,
      "loss": 5.3949,
      "num_input_tokens_seen": 113664,
      "step": 165
    },
    {
      "epoch": 0.10023584905660378,
      "grad_norm": 39.235321044921875,
      "learning_rate": 4.982311320754717e-07,
      "loss": 5.1665,
      "num_input_tokens_seen": 117440,
      "step": 170
    },
    {
      "epoch": 0.10318396226415094,
      "grad_norm": 46.452632904052734,
      "learning_rate": 5.129716981132076e-07,
      "loss": 5.0396,
      "num_input_tokens_seen": 119968,
      "step": 175
    },
    {
      "epoch": 0.10613207547169812,
      "grad_norm": 31.872148513793945,
      "learning_rate": 5.277122641509435e-07,
      "loss": 5.1356,
      "num_input_tokens_seen": 122656,
      "step": 180
    },
    {
      "epoch": 0.10908018867924528,
      "grad_norm": 35.37727355957031,
      "learning_rate": 5.424528301886793e-07,
      "loss": 4.5716,
      "num_input_tokens_seen": 126112,
      "step": 185
    },
    {
      "epoch": 0.11202830188679246,
      "grad_norm": 36.72437286376953,
      "learning_rate": 5.571933962264151e-07,
      "loss": 5.1308,
      "num_input_tokens_seen": 128960,
      "step": 190
    },
    {
      "epoch": 0.11497641509433962,
      "grad_norm": 37.55937576293945,
      "learning_rate": 5.71933962264151e-07,
      "loss": 5.0656,
      "num_input_tokens_seen": 131712,
      "step": 195
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 45.27613067626953,
      "learning_rate": 5.866745283018868e-07,
      "loss": 4.3416,
      "num_input_tokens_seen": 135360,
      "step": 200
    },
    {
      "epoch": 0.12087264150943396,
      "grad_norm": 38.385009765625,
      "learning_rate": 6.014150943396227e-07,
      "loss": 4.1762,
      "num_input_tokens_seen": 138368,
      "step": 205
    },
    {
      "epoch": 0.12382075471698113,
      "grad_norm": 30.53740692138672,
      "learning_rate": 6.161556603773585e-07,
      "loss": 4.3725,
      "num_input_tokens_seen": 141536,
      "step": 210
    },
    {
      "epoch": 0.1267688679245283,
      "grad_norm": 112.80076599121094,
      "learning_rate": 6.308962264150945e-07,
      "loss": 4.0367,
      "num_input_tokens_seen": 144768,
      "step": 215
    },
    {
      "epoch": 0.12971698113207547,
      "grad_norm": 30.331371307373047,
      "learning_rate": 6.456367924528302e-07,
      "loss": 4.4279,
      "num_input_tokens_seen": 148032,
      "step": 220
    },
    {
      "epoch": 0.13266509433962265,
      "grad_norm": 36.1087646484375,
      "learning_rate": 6.603773584905661e-07,
      "loss": 4.2135,
      "num_input_tokens_seen": 151264,
      "step": 225
    },
    {
      "epoch": 0.13561320754716982,
      "grad_norm": 34.845306396484375,
      "learning_rate": 6.75117924528302e-07,
      "loss": 4.2444,
      "num_input_tokens_seen": 154048,
      "step": 230
    },
    {
      "epoch": 0.13856132075471697,
      "grad_norm": 70.9759750366211,
      "learning_rate": 6.898584905660379e-07,
      "loss": 4.2849,
      "num_input_tokens_seen": 157216,
      "step": 235
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 67.03533935546875,
      "learning_rate": 7.045990566037736e-07,
      "loss": 3.7267,
      "num_input_tokens_seen": 160896,
      "step": 240
    },
    {
      "epoch": 0.14445754716981132,
      "grad_norm": 29.55883026123047,
      "learning_rate": 7.193396226415095e-07,
      "loss": 4.1481,
      "num_input_tokens_seen": 165280,
      "step": 245
    },
    {
      "epoch": 0.1474056603773585,
      "grad_norm": 59.074989318847656,
      "learning_rate": 7.340801886792454e-07,
      "loss": 3.8848,
      "num_input_tokens_seen": 170112,
      "step": 250
    },
    {
      "epoch": 0.15035377358490565,
      "grad_norm": 30.45848846435547,
      "learning_rate": 7.488207547169812e-07,
      "loss": 3.49,
      "num_input_tokens_seen": 173408,
      "step": 255
    },
    {
      "epoch": 0.15330188679245282,
      "grad_norm": 46.95903015136719,
      "learning_rate": 7.63561320754717e-07,
      "loss": 3.5115,
      "num_input_tokens_seen": 176896,
      "step": 260
    },
    {
      "epoch": 0.15625,
      "grad_norm": 54.3665771484375,
      "learning_rate": 7.783018867924529e-07,
      "loss": 3.5608,
      "num_input_tokens_seen": 179936,
      "step": 265
    },
    {
      "epoch": 0.15919811320754718,
      "grad_norm": 23.337135314941406,
      "learning_rate": 7.930424528301888e-07,
      "loss": 3.9471,
      "num_input_tokens_seen": 183872,
      "step": 270
    },
    {
      "epoch": 0.16214622641509435,
      "grad_norm": 45.41781997680664,
      "learning_rate": 8.077830188679246e-07,
      "loss": 3.6111,
      "num_input_tokens_seen": 187616,
      "step": 275
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 33.63853073120117,
      "learning_rate": 8.225235849056605e-07,
      "loss": 3.59,
      "num_input_tokens_seen": 190080,
      "step": 280
    },
    {
      "epoch": 0.16804245283018868,
      "grad_norm": 36.62028503417969,
      "learning_rate": 8.372641509433963e-07,
      "loss": 3.2833,
      "num_input_tokens_seen": 193088,
      "step": 285
    },
    {
      "epoch": 0.17099056603773585,
      "grad_norm": 34.628997802734375,
      "learning_rate": 8.520047169811321e-07,
      "loss": 3.7941,
      "num_input_tokens_seen": 196064,
      "step": 290
    },
    {
      "epoch": 0.17393867924528303,
      "grad_norm": 23.633214950561523,
      "learning_rate": 8.66745283018868e-07,
      "loss": 3.9604,
      "num_input_tokens_seen": 198688,
      "step": 295
    },
    {
      "epoch": 0.17688679245283018,
      "grad_norm": 26.187484741210938,
      "learning_rate": 8.814858490566039e-07,
      "loss": 3.6079,
      "num_input_tokens_seen": 201568,
      "step": 300
    },
    {
      "epoch": 0.17983490566037735,
      "grad_norm": 31.048255920410156,
      "learning_rate": 8.962264150943397e-07,
      "loss": 3.4773,
      "num_input_tokens_seen": 205568,
      "step": 305
    },
    {
      "epoch": 0.18278301886792453,
      "grad_norm": 31.47636604309082,
      "learning_rate": 9.109669811320755e-07,
      "loss": 3.1986,
      "num_input_tokens_seen": 208864,
      "step": 310
    },
    {
      "epoch": 0.1857311320754717,
      "grad_norm": 32.68092346191406,
      "learning_rate": 9.257075471698114e-07,
      "loss": 3.4734,
      "num_input_tokens_seen": 211584,
      "step": 315
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 33.412010192871094,
      "learning_rate": 9.404481132075473e-07,
      "loss": 3.3638,
      "num_input_tokens_seen": 214464,
      "step": 320
    },
    {
      "epoch": 0.19162735849056603,
      "grad_norm": 28.543752670288086,
      "learning_rate": 9.551886792452833e-07,
      "loss": 3.0345,
      "num_input_tokens_seen": 217920,
      "step": 325
    },
    {
      "epoch": 0.1945754716981132,
      "grad_norm": 23.40913963317871,
      "learning_rate": 9.699292452830188e-07,
      "loss": 3.2481,
      "num_input_tokens_seen": 220992,
      "step": 330
    },
    {
      "epoch": 0.19752358490566038,
      "grad_norm": 43.49759292602539,
      "learning_rate": 9.846698113207548e-07,
      "loss": 3.1396,
      "num_input_tokens_seen": 224704,
      "step": 335
    },
    {
      "epoch": 0.20047169811320756,
      "grad_norm": 29.57196807861328,
      "learning_rate": 9.994103773584906e-07,
      "loss": 2.7993,
      "num_input_tokens_seen": 227712,
      "step": 340
    },
    {
      "epoch": 0.2034198113207547,
      "grad_norm": 46.249977111816406,
      "learning_rate": 1.0141509433962265e-06,
      "loss": 2.8841,
      "num_input_tokens_seen": 230272,
      "step": 345
    },
    {
      "epoch": 0.20636792452830188,
      "grad_norm": 23.37116050720215,
      "learning_rate": 1.0288915094339623e-06,
      "loss": 2.8042,
      "num_input_tokens_seen": 233600,
      "step": 350
    },
    {
      "epoch": 0.20931603773584906,
      "grad_norm": 50.35514831542969,
      "learning_rate": 1.043632075471698e-06,
      "loss": 2.5032,
      "num_input_tokens_seen": 236576,
      "step": 355
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 29.908647537231445,
      "learning_rate": 1.058372641509434e-06,
      "loss": 2.879,
      "num_input_tokens_seen": 241056,
      "step": 360
    },
    {
      "epoch": 0.21521226415094338,
      "grad_norm": 51.945220947265625,
      "learning_rate": 1.07311320754717e-06,
      "loss": 2.5462,
      "num_input_tokens_seen": 243808,
      "step": 365
    },
    {
      "epoch": 0.21816037735849056,
      "grad_norm": 47.77359390258789,
      "learning_rate": 1.0878537735849056e-06,
      "loss": 2.7807,
      "num_input_tokens_seen": 247520,
      "step": 370
    },
    {
      "epoch": 0.22110849056603774,
      "grad_norm": 22.255592346191406,
      "learning_rate": 1.1025943396226416e-06,
      "loss": 2.4202,
      "num_input_tokens_seen": 251104,
      "step": 375
    },
    {
      "epoch": 0.2240566037735849,
      "grad_norm": 31.116119384765625,
      "learning_rate": 1.1173349056603773e-06,
      "loss": 2.231,
      "num_input_tokens_seen": 254464,
      "step": 380
    },
    {
      "epoch": 0.2270047169811321,
      "grad_norm": 33.76179885864258,
      "learning_rate": 1.1320754716981133e-06,
      "loss": 2.5828,
      "num_input_tokens_seen": 257312,
      "step": 385
    },
    {
      "epoch": 0.22995283018867924,
      "grad_norm": 25.75220489501953,
      "learning_rate": 1.1468160377358493e-06,
      "loss": 2.7209,
      "num_input_tokens_seen": 260480,
      "step": 390
    },
    {
      "epoch": 0.2329009433962264,
      "grad_norm": 34.168331146240234,
      "learning_rate": 1.1615566037735849e-06,
      "loss": 2.8036,
      "num_input_tokens_seen": 264608,
      "step": 395
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 24.65675163269043,
      "learning_rate": 1.1762971698113208e-06,
      "loss": 2.8243,
      "num_input_tokens_seen": 268000,
      "step": 400
    },
    {
      "epoch": 0.23879716981132076,
      "grad_norm": 29.26604652404785,
      "learning_rate": 1.1910377358490568e-06,
      "loss": 2.3111,
      "num_input_tokens_seen": 271936,
      "step": 405
    },
    {
      "epoch": 0.2417452830188679,
      "grad_norm": 31.679262161254883,
      "learning_rate": 1.2057783018867926e-06,
      "loss": 2.5572,
      "num_input_tokens_seen": 274272,
      "step": 410
    },
    {
      "epoch": 0.2446933962264151,
      "grad_norm": 26.455148696899414,
      "learning_rate": 1.2205188679245284e-06,
      "loss": 2.0406,
      "num_input_tokens_seen": 279616,
      "step": 415
    },
    {
      "epoch": 0.24764150943396226,
      "grad_norm": 17.131376266479492,
      "learning_rate": 1.2352594339622641e-06,
      "loss": 2.3823,
      "num_input_tokens_seen": 284480,
      "step": 420
    },
    {
      "epoch": 0.2505896226415094,
      "grad_norm": 33.041114807128906,
      "learning_rate": 1.25e-06,
      "loss": 2.0924,
      "num_input_tokens_seen": 288128,
      "step": 425
    },
    {
      "epoch": 0.2535377358490566,
      "grad_norm": 35.757537841796875,
      "learning_rate": 1.264740566037736e-06,
      "loss": 2.3764,
      "num_input_tokens_seen": 291552,
      "step": 430
    },
    {
      "epoch": 0.25648584905660377,
      "grad_norm": 25.91598129272461,
      "learning_rate": 1.2794811320754718e-06,
      "loss": 2.2371,
      "num_input_tokens_seen": 294688,
      "step": 435
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 44.307861328125,
      "learning_rate": 1.2942216981132078e-06,
      "loss": 1.9278,
      "num_input_tokens_seen": 296992,
      "step": 440
    },
    {
      "epoch": 0.2623820754716981,
      "grad_norm": 24.258953094482422,
      "learning_rate": 1.3089622641509436e-06,
      "loss": 2.312,
      "num_input_tokens_seen": 299712,
      "step": 445
    },
    {
      "epoch": 0.2653301886792453,
      "grad_norm": 37.357696533203125,
      "learning_rate": 1.3237028301886792e-06,
      "loss": 1.7953,
      "num_input_tokens_seen": 302912,
      "step": 450
    },
    {
      "epoch": 0.26827830188679247,
      "grad_norm": 48.12458419799805,
      "learning_rate": 1.3384433962264151e-06,
      "loss": 2.2464,
      "num_input_tokens_seen": 306112,
      "step": 455
    },
    {
      "epoch": 0.27122641509433965,
      "grad_norm": 52.2923469543457,
      "learning_rate": 1.353183962264151e-06,
      "loss": 2.2654,
      "num_input_tokens_seen": 308928,
      "step": 460
    },
    {
      "epoch": 0.27417452830188677,
      "grad_norm": 24.658674240112305,
      "learning_rate": 1.3679245283018869e-06,
      "loss": 1.6971,
      "num_input_tokens_seen": 311712,
      "step": 465
    },
    {
      "epoch": 0.27712264150943394,
      "grad_norm": 32.85096740722656,
      "learning_rate": 1.3826650943396229e-06,
      "loss": 1.5893,
      "num_input_tokens_seen": 315040,
      "step": 470
    },
    {
      "epoch": 0.2800707547169811,
      "grad_norm": 53.20591735839844,
      "learning_rate": 1.3974056603773586e-06,
      "loss": 1.8189,
      "num_input_tokens_seen": 317568,
      "step": 475
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 32.196537017822266,
      "learning_rate": 1.4121462264150946e-06,
      "loss": 1.4128,
      "num_input_tokens_seen": 320576,
      "step": 480
    },
    {
      "epoch": 0.28596698113207547,
      "grad_norm": 30.985126495361328,
      "learning_rate": 1.4268867924528304e-06,
      "loss": 1.1829,
      "num_input_tokens_seen": 323072,
      "step": 485
    },
    {
      "epoch": 0.28891509433962265,
      "grad_norm": 40.506099700927734,
      "learning_rate": 1.4416273584905664e-06,
      "loss": 1.2505,
      "num_input_tokens_seen": 326080,
      "step": 490
    },
    {
      "epoch": 0.2918632075471698,
      "grad_norm": 23.89166259765625,
      "learning_rate": 1.456367924528302e-06,
      "loss": 1.3973,
      "num_input_tokens_seen": 329312,
      "step": 495
    },
    {
      "epoch": 0.294811320754717,
      "grad_norm": 11.195052146911621,
      "learning_rate": 1.4711084905660377e-06,
      "loss": 1.1727,
      "num_input_tokens_seen": 331872,
      "step": 500
    },
    {
      "epoch": 0.2977594339622642,
      "grad_norm": 20.93170166015625,
      "learning_rate": 1.4858490566037737e-06,
      "loss": 1.1246,
      "num_input_tokens_seen": 335168,
      "step": 505
    },
    {
      "epoch": 0.3007075471698113,
      "grad_norm": 21.96256446838379,
      "learning_rate": 1.5005896226415096e-06,
      "loss": 1.3456,
      "num_input_tokens_seen": 338272,
      "step": 510
    },
    {
      "epoch": 0.30365566037735847,
      "grad_norm": 8.379934310913086,
      "learning_rate": 1.5153301886792454e-06,
      "loss": 1.1066,
      "num_input_tokens_seen": 341472,
      "step": 515
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 11.993612289428711,
      "learning_rate": 1.5300707547169814e-06,
      "loss": 0.9691,
      "num_input_tokens_seen": 345152,
      "step": 520
    },
    {
      "epoch": 0.3095518867924528,
      "grad_norm": 18.049999237060547,
      "learning_rate": 1.5448113207547172e-06,
      "loss": 0.9011,
      "num_input_tokens_seen": 349184,
      "step": 525
    },
    {
      "epoch": 0.3125,
      "grad_norm": 12.317525863647461,
      "learning_rate": 1.5595518867924531e-06,
      "loss": 1.075,
      "num_input_tokens_seen": 352160,
      "step": 530
    },
    {
      "epoch": 0.3154481132075472,
      "grad_norm": 10.795198440551758,
      "learning_rate": 1.574292452830189e-06,
      "loss": 0.9761,
      "num_input_tokens_seen": 356000,
      "step": 535
    },
    {
      "epoch": 0.31839622641509435,
      "grad_norm": 10.274962425231934,
      "learning_rate": 1.5890330188679245e-06,
      "loss": 0.9936,
      "num_input_tokens_seen": 359840,
      "step": 540
    },
    {
      "epoch": 0.32134433962264153,
      "grad_norm": 11.443280220031738,
      "learning_rate": 1.6037735849056604e-06,
      "loss": 1.197,
      "num_input_tokens_seen": 364320,
      "step": 545
    },
    {
      "epoch": 0.3242924528301887,
      "grad_norm": 17.257568359375,
      "learning_rate": 1.6185141509433964e-06,
      "loss": 0.8054,
      "num_input_tokens_seen": 367328,
      "step": 550
    },
    {
      "epoch": 0.3272405660377358,
      "grad_norm": 6.4032301902771,
      "learning_rate": 1.6332547169811322e-06,
      "loss": 0.7248,
      "num_input_tokens_seen": 370624,
      "step": 555
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 8.639458656311035,
      "learning_rate": 1.6479952830188682e-06,
      "loss": 0.8802,
      "num_input_tokens_seen": 374368,
      "step": 560
    },
    {
      "epoch": 0.3331367924528302,
      "grad_norm": 10.02137279510498,
      "learning_rate": 1.662735849056604e-06,
      "loss": 0.7962,
      "num_input_tokens_seen": 377504,
      "step": 565
    },
    {
      "epoch": 0.33608490566037735,
      "grad_norm": 11.1533203125,
      "learning_rate": 1.67747641509434e-06,
      "loss": 0.7876,
      "num_input_tokens_seen": 380608,
      "step": 570
    },
    {
      "epoch": 0.33903301886792453,
      "grad_norm": 22.1187801361084,
      "learning_rate": 1.6922169811320757e-06,
      "loss": 0.8277,
      "num_input_tokens_seen": 384512,
      "step": 575
    },
    {
      "epoch": 0.3419811320754717,
      "grad_norm": 12.794469833374023,
      "learning_rate": 1.7069575471698112e-06,
      "loss": 0.913,
      "num_input_tokens_seen": 387584,
      "step": 580
    },
    {
      "epoch": 0.3449292452830189,
      "grad_norm": 15.517716407775879,
      "learning_rate": 1.7216981132075472e-06,
      "loss": 0.9092,
      "num_input_tokens_seen": 391232,
      "step": 585
    },
    {
      "epoch": 0.34787735849056606,
      "grad_norm": 8.473352432250977,
      "learning_rate": 1.736438679245283e-06,
      "loss": 0.9022,
      "num_input_tokens_seen": 394336,
      "step": 590
    },
    {
      "epoch": 0.35082547169811323,
      "grad_norm": 6.469804763793945,
      "learning_rate": 1.751179245283019e-06,
      "loss": 0.831,
      "num_input_tokens_seen": 397568,
      "step": 595
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 19.41082000732422,
      "learning_rate": 1.765919811320755e-06,
      "loss": 0.8677,
      "num_input_tokens_seen": 400608,
      "step": 600
    },
    {
      "epoch": 0.35672169811320753,
      "grad_norm": 13.85968017578125,
      "learning_rate": 1.7806603773584907e-06,
      "loss": 0.8183,
      "num_input_tokens_seen": 404736,
      "step": 605
    },
    {
      "epoch": 0.3596698113207547,
      "grad_norm": 13.261207580566406,
      "learning_rate": 1.7954009433962267e-06,
      "loss": 0.7608,
      "num_input_tokens_seen": 407840,
      "step": 610
    },
    {
      "epoch": 0.3626179245283019,
      "grad_norm": 8.10428524017334,
      "learning_rate": 1.8101415094339625e-06,
      "loss": 0.6245,
      "num_input_tokens_seen": 410752,
      "step": 615
    },
    {
      "epoch": 0.36556603773584906,
      "grad_norm": 4.660794258117676,
      "learning_rate": 1.8248820754716984e-06,
      "loss": 0.7243,
      "num_input_tokens_seen": 413728,
      "step": 620
    },
    {
      "epoch": 0.36851415094339623,
      "grad_norm": 5.193559169769287,
      "learning_rate": 1.839622641509434e-06,
      "loss": 0.9994,
      "num_input_tokens_seen": 416864,
      "step": 625
    },
    {
      "epoch": 0.3714622641509434,
      "grad_norm": 5.949007987976074,
      "learning_rate": 1.8543632075471698e-06,
      "loss": 0.7088,
      "num_input_tokens_seen": 419840,
      "step": 630
    },
    {
      "epoch": 0.3744103773584906,
      "grad_norm": 27.87962532043457,
      "learning_rate": 1.8691037735849057e-06,
      "loss": 0.9749,
      "num_input_tokens_seen": 422272,
      "step": 635
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 5.6117963790893555,
      "learning_rate": 1.8838443396226417e-06,
      "loss": 0.8456,
      "num_input_tokens_seen": 425184,
      "step": 640
    },
    {
      "epoch": 0.3803066037735849,
      "grad_norm": 5.82224702835083,
      "learning_rate": 1.8985849056603775e-06,
      "loss": 0.7637,
      "num_input_tokens_seen": 428224,
      "step": 645
    },
    {
      "epoch": 0.38325471698113206,
      "grad_norm": 4.588514804840088,
      "learning_rate": 1.9133254716981133e-06,
      "loss": 0.748,
      "num_input_tokens_seen": 431040,
      "step": 650
    },
    {
      "epoch": 0.38620283018867924,
      "grad_norm": 5.79656457901001,
      "learning_rate": 1.9280660377358494e-06,
      "loss": 0.8284,
      "num_input_tokens_seen": 433920,
      "step": 655
    },
    {
      "epoch": 0.3891509433962264,
      "grad_norm": 4.868707180023193,
      "learning_rate": 1.9428066037735852e-06,
      "loss": 0.6951,
      "num_input_tokens_seen": 438016,
      "step": 660
    },
    {
      "epoch": 0.3920990566037736,
      "grad_norm": 7.623064994812012,
      "learning_rate": 1.957547169811321e-06,
      "loss": 0.6424,
      "num_input_tokens_seen": 440576,
      "step": 665
    },
    {
      "epoch": 0.39504716981132076,
      "grad_norm": 4.909170150756836,
      "learning_rate": 1.9722877358490568e-06,
      "loss": 0.6532,
      "num_input_tokens_seen": 443872,
      "step": 670
    },
    {
      "epoch": 0.39799528301886794,
      "grad_norm": 7.4576544761657715,
      "learning_rate": 1.9870283018867925e-06,
      "loss": 0.6437,
      "num_input_tokens_seen": 446912,
      "step": 675
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 5.9115986824035645,
      "learning_rate": 2.0017688679245283e-06,
      "loss": 0.7485,
      "num_input_tokens_seen": 451072,
      "step": 680
    },
    {
      "epoch": 0.40389150943396224,
      "grad_norm": 15.431118965148926,
      "learning_rate": 2.0165094339622645e-06,
      "loss": 0.7466,
      "num_input_tokens_seen": 453568,
      "step": 685
    },
    {
      "epoch": 0.4068396226415094,
      "grad_norm": 10.412184715270996,
      "learning_rate": 2.0312500000000002e-06,
      "loss": 0.8747,
      "num_input_tokens_seen": 456416,
      "step": 690
    },
    {
      "epoch": 0.4097877358490566,
      "grad_norm": 7.134666442871094,
      "learning_rate": 2.045990566037736e-06,
      "loss": 0.7353,
      "num_input_tokens_seen": 458816,
      "step": 695
    },
    {
      "epoch": 0.41273584905660377,
      "grad_norm": 8.259252548217773,
      "learning_rate": 2.0607311320754718e-06,
      "loss": 0.7846,
      "num_input_tokens_seen": 461344,
      "step": 700
    },
    {
      "epoch": 0.41568396226415094,
      "grad_norm": 8.173951148986816,
      "learning_rate": 2.075471698113208e-06,
      "loss": 0.5964,
      "num_input_tokens_seen": 464192,
      "step": 705
    },
    {
      "epoch": 0.4186320754716981,
      "grad_norm": 8.26637077331543,
      "learning_rate": 2.0902122641509437e-06,
      "loss": 0.638,
      "num_input_tokens_seen": 467520,
      "step": 710
    },
    {
      "epoch": 0.4215801886792453,
      "grad_norm": 3.905766010284424,
      "learning_rate": 2.1049528301886795e-06,
      "loss": 0.7679,
      "num_input_tokens_seen": 471296,
      "step": 715
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 6.9918293952941895,
      "learning_rate": 2.1196933962264153e-06,
      "loss": 0.6212,
      "num_input_tokens_seen": 474944,
      "step": 720
    },
    {
      "epoch": 0.42747641509433965,
      "grad_norm": 5.137577056884766,
      "learning_rate": 2.134433962264151e-06,
      "loss": 0.7064,
      "num_input_tokens_seen": 477664,
      "step": 725
    },
    {
      "epoch": 0.43042452830188677,
      "grad_norm": 4.444843769073486,
      "learning_rate": 2.149174528301887e-06,
      "loss": 0.5702,
      "num_input_tokens_seen": 481824,
      "step": 730
    },
    {
      "epoch": 0.43337264150943394,
      "grad_norm": 6.798077583312988,
      "learning_rate": 2.163915094339623e-06,
      "loss": 0.7767,
      "num_input_tokens_seen": 485728,
      "step": 735
    },
    {
      "epoch": 0.4363207547169811,
      "grad_norm": 6.888205528259277,
      "learning_rate": 2.1786556603773588e-06,
      "loss": 0.7257,
      "num_input_tokens_seen": 488480,
      "step": 740
    },
    {
      "epoch": 0.4392688679245283,
      "grad_norm": 6.917191028594971,
      "learning_rate": 2.1933962264150945e-06,
      "loss": 0.54,
      "num_input_tokens_seen": 492032,
      "step": 745
    },
    {
      "epoch": 0.44221698113207547,
      "grad_norm": 14.0696439743042,
      "learning_rate": 2.2081367924528303e-06,
      "loss": 0.6328,
      "num_input_tokens_seen": 494496,
      "step": 750
    },
    {
      "epoch": 0.44516509433962265,
      "grad_norm": 4.5550031661987305,
      "learning_rate": 2.2228773584905665e-06,
      "loss": 0.6721,
      "num_input_tokens_seen": 497792,
      "step": 755
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 3.634284257888794,
      "learning_rate": 2.237617924528302e-06,
      "loss": 0.6328,
      "num_input_tokens_seen": 501664,
      "step": 760
    },
    {
      "epoch": 0.451061320754717,
      "grad_norm": 3.3259334564208984,
      "learning_rate": 2.252358490566038e-06,
      "loss": 0.5973,
      "num_input_tokens_seen": 504928,
      "step": 765
    },
    {
      "epoch": 0.4540094339622642,
      "grad_norm": 7.52155876159668,
      "learning_rate": 2.267099056603774e-06,
      "loss": 0.6766,
      "num_input_tokens_seen": 507904,
      "step": 770
    },
    {
      "epoch": 0.4569575471698113,
      "grad_norm": 7.284567832946777,
      "learning_rate": 2.2818396226415096e-06,
      "loss": 0.6994,
      "num_input_tokens_seen": 511648,
      "step": 775
    },
    {
      "epoch": 0.45990566037735847,
      "grad_norm": 5.05999755859375,
      "learning_rate": 2.2965801886792453e-06,
      "loss": 0.6119,
      "num_input_tokens_seen": 515488,
      "step": 780
    },
    {
      "epoch": 0.46285377358490565,
      "grad_norm": 4.1906819343566895,
      "learning_rate": 2.3113207547169815e-06,
      "loss": 0.5788,
      "num_input_tokens_seen": 518624,
      "step": 785
    },
    {
      "epoch": 0.4658018867924528,
      "grad_norm": 9.383630752563477,
      "learning_rate": 2.3260613207547173e-06,
      "loss": 0.8331,
      "num_input_tokens_seen": 521728,
      "step": 790
    },
    {
      "epoch": 0.46875,
      "grad_norm": 7.647820472717285,
      "learning_rate": 2.340801886792453e-06,
      "loss": 0.8239,
      "num_input_tokens_seen": 525248,
      "step": 795
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 5.626327037811279,
      "learning_rate": 2.355542452830189e-06,
      "loss": 0.7549,
      "num_input_tokens_seen": 528640,
      "step": 800
    },
    {
      "epoch": 0.47464622641509435,
      "grad_norm": 10.994980812072754,
      "learning_rate": 2.3702830188679246e-06,
      "loss": 0.546,
      "num_input_tokens_seen": 531776,
      "step": 805
    },
    {
      "epoch": 0.47759433962264153,
      "grad_norm": 6.826515197753906,
      "learning_rate": 2.3850235849056604e-06,
      "loss": 0.6775,
      "num_input_tokens_seen": 535168,
      "step": 810
    },
    {
      "epoch": 0.4805424528301887,
      "grad_norm": 12.07666301727295,
      "learning_rate": 2.3997641509433966e-06,
      "loss": 0.7604,
      "num_input_tokens_seen": 538432,
      "step": 815
    },
    {
      "epoch": 0.4834905660377358,
      "grad_norm": 9.853074073791504,
      "learning_rate": 2.4145047169811323e-06,
      "loss": 0.8166,
      "num_input_tokens_seen": 541184,
      "step": 820
    },
    {
      "epoch": 0.486438679245283,
      "grad_norm": 5.36435604095459,
      "learning_rate": 2.429245283018868e-06,
      "loss": 0.6448,
      "num_input_tokens_seen": 543552,
      "step": 825
    },
    {
      "epoch": 0.4893867924528302,
      "grad_norm": 4.13151216506958,
      "learning_rate": 2.443985849056604e-06,
      "loss": 0.5739,
      "num_input_tokens_seen": 546272,
      "step": 830
    },
    {
      "epoch": 0.49233490566037735,
      "grad_norm": 4.1057538986206055,
      "learning_rate": 2.45872641509434e-06,
      "loss": 0.6901,
      "num_input_tokens_seen": 550336,
      "step": 835
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 2.445622444152832,
      "learning_rate": 2.473466981132076e-06,
      "loss": 0.6293,
      "num_input_tokens_seen": 554048,
      "step": 840
    },
    {
      "epoch": 0.4982311320754717,
      "grad_norm": 4.5358123779296875,
      "learning_rate": 2.4882075471698116e-06,
      "loss": 0.836,
      "num_input_tokens_seen": 556704,
      "step": 845
    },
    {
      "epoch": 0.5011792452830188,
      "grad_norm": 24.46385383605957,
      "learning_rate": 2.5029481132075474e-06,
      "loss": 0.6885,
      "num_input_tokens_seen": 559904,
      "step": 850
    },
    {
      "epoch": 0.504127358490566,
      "grad_norm": 26.36067008972168,
      "learning_rate": 2.517688679245283e-06,
      "loss": 0.8499,
      "num_input_tokens_seen": 562752,
      "step": 855
    },
    {
      "epoch": 0.5070754716981132,
      "grad_norm": 8.24132251739502,
      "learning_rate": 2.532429245283019e-06,
      "loss": 0.9022,
      "num_input_tokens_seen": 565792,
      "step": 860
    },
    {
      "epoch": 0.5100235849056604,
      "grad_norm": 13.84225845336914,
      "learning_rate": 2.547169811320755e-06,
      "loss": 0.7125,
      "num_input_tokens_seen": 569056,
      "step": 865
    },
    {
      "epoch": 0.5129716981132075,
      "grad_norm": 4.6004438400268555,
      "learning_rate": 2.561910377358491e-06,
      "loss": 0.6445,
      "num_input_tokens_seen": 572224,
      "step": 870
    },
    {
      "epoch": 0.5159198113207547,
      "grad_norm": 5.0709147453308105,
      "learning_rate": 2.5766509433962266e-06,
      "loss": 0.6508,
      "num_input_tokens_seen": 576512,
      "step": 875
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 6.475077152252197,
      "learning_rate": 2.5913915094339624e-06,
      "loss": 0.6344,
      "num_input_tokens_seen": 579616,
      "step": 880
    },
    {
      "epoch": 0.5218160377358491,
      "grad_norm": 3.193110704421997,
      "learning_rate": 2.6061320754716986e-06,
      "loss": 0.8628,
      "num_input_tokens_seen": 582784,
      "step": 885
    },
    {
      "epoch": 0.5247641509433962,
      "grad_norm": 6.2870330810546875,
      "learning_rate": 2.6208726415094343e-06,
      "loss": 0.7733,
      "num_input_tokens_seen": 585760,
      "step": 890
    },
    {
      "epoch": 0.5277122641509434,
      "grad_norm": 3.8790974617004395,
      "learning_rate": 2.63561320754717e-06,
      "loss": 0.7089,
      "num_input_tokens_seen": 588672,
      "step": 895
    },
    {
      "epoch": 0.5306603773584906,
      "grad_norm": 4.504887104034424,
      "learning_rate": 2.650353773584906e-06,
      "loss": 0.6357,
      "num_input_tokens_seen": 591520,
      "step": 900
    },
    {
      "epoch": 0.5336084905660378,
      "grad_norm": 6.5464653968811035,
      "learning_rate": 2.665094339622642e-06,
      "loss": 0.5284,
      "num_input_tokens_seen": 594560,
      "step": 905
    },
    {
      "epoch": 0.5365566037735849,
      "grad_norm": 4.315725326538086,
      "learning_rate": 2.679834905660378e-06,
      "loss": 0.6356,
      "num_input_tokens_seen": 601184,
      "step": 910
    },
    {
      "epoch": 0.5395047169811321,
      "grad_norm": 4.365467071533203,
      "learning_rate": 2.694575471698113e-06,
      "loss": 0.6077,
      "num_input_tokens_seen": 604224,
      "step": 915
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 7.08058500289917,
      "learning_rate": 2.709316037735849e-06,
      "loss": 0.5682,
      "num_input_tokens_seen": 607904,
      "step": 920
    },
    {
      "epoch": 0.5454009433962265,
      "grad_norm": 3.742565631866455,
      "learning_rate": 2.724056603773585e-06,
      "loss": 0.6481,
      "num_input_tokens_seen": 610784,
      "step": 925
    },
    {
      "epoch": 0.5483490566037735,
      "grad_norm": 3.8136961460113525,
      "learning_rate": 2.738797169811321e-06,
      "loss": 0.6293,
      "num_input_tokens_seen": 614112,
      "step": 930
    },
    {
      "epoch": 0.5512971698113207,
      "grad_norm": 11.098905563354492,
      "learning_rate": 2.7535377358490567e-06,
      "loss": 0.6597,
      "num_input_tokens_seen": 617792,
      "step": 935
    },
    {
      "epoch": 0.5542452830188679,
      "grad_norm": 4.508008003234863,
      "learning_rate": 2.7682783018867925e-06,
      "loss": 0.7425,
      "num_input_tokens_seen": 620416,
      "step": 940
    },
    {
      "epoch": 0.5571933962264151,
      "grad_norm": 6.828112602233887,
      "learning_rate": 2.7830188679245286e-06,
      "loss": 0.6228,
      "num_input_tokens_seen": 623040,
      "step": 945
    },
    {
      "epoch": 0.5601415094339622,
      "grad_norm": 2.9876842498779297,
      "learning_rate": 2.7977594339622644e-06,
      "loss": 0.5043,
      "num_input_tokens_seen": 626976,
      "step": 950
    },
    {
      "epoch": 0.5630896226415094,
      "grad_norm": 4.164421558380127,
      "learning_rate": 2.8125e-06,
      "loss": 0.6673,
      "num_input_tokens_seen": 630176,
      "step": 955
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 7.771459579467773,
      "learning_rate": 2.827240566037736e-06,
      "loss": 0.708,
      "num_input_tokens_seen": 633408,
      "step": 960
    },
    {
      "epoch": 0.5689858490566038,
      "grad_norm": 4.870665073394775,
      "learning_rate": 2.841981132075472e-06,
      "loss": 0.5939,
      "num_input_tokens_seen": 636000,
      "step": 965
    },
    {
      "epoch": 0.5719339622641509,
      "grad_norm": 5.331478118896484,
      "learning_rate": 2.856721698113208e-06,
      "loss": 0.6315,
      "num_input_tokens_seen": 640000,
      "step": 970
    },
    {
      "epoch": 0.5748820754716981,
      "grad_norm": 4.300553798675537,
      "learning_rate": 2.8714622641509437e-06,
      "loss": 0.6974,
      "num_input_tokens_seen": 643072,
      "step": 975
    },
    {
      "epoch": 0.5778301886792453,
      "grad_norm": 3.346952438354492,
      "learning_rate": 2.8862028301886794e-06,
      "loss": 0.8485,
      "num_input_tokens_seen": 645824,
      "step": 980
    },
    {
      "epoch": 0.5807783018867925,
      "grad_norm": 8.058412551879883,
      "learning_rate": 2.9009433962264156e-06,
      "loss": 0.768,
      "num_input_tokens_seen": 650240,
      "step": 985
    },
    {
      "epoch": 0.5837264150943396,
      "grad_norm": 3.4640181064605713,
      "learning_rate": 2.9156839622641514e-06,
      "loss": 0.5173,
      "num_input_tokens_seen": 652960,
      "step": 990
    },
    {
      "epoch": 0.5866745283018868,
      "grad_norm": 5.996393203735352,
      "learning_rate": 2.930424528301887e-06,
      "loss": 0.6091,
      "num_input_tokens_seen": 655968,
      "step": 995
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 3.2512998580932617,
      "learning_rate": 2.9451650943396225e-06,
      "loss": 0.5206,
      "num_input_tokens_seen": 658752,
      "step": 1000
    },
    {
      "epoch": 0.5925707547169812,
      "grad_norm": 5.460904598236084,
      "learning_rate": 2.9599056603773587e-06,
      "loss": 0.7423,
      "num_input_tokens_seen": 662176,
      "step": 1005
    },
    {
      "epoch": 0.5955188679245284,
      "grad_norm": 5.328415393829346,
      "learning_rate": 2.9746462264150945e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 665760,
      "step": 1010
    },
    {
      "epoch": 0.5984669811320755,
      "grad_norm": 6.513659477233887,
      "learning_rate": 2.9893867924528302e-06,
      "loss": 0.6489,
      "num_input_tokens_seen": 668832,
      "step": 1015
    },
    {
      "epoch": 0.6014150943396226,
      "grad_norm": 3.213778018951416,
      "learning_rate": 3.004127358490566e-06,
      "loss": 0.7294,
      "num_input_tokens_seen": 671936,
      "step": 1020
    },
    {
      "epoch": 0.6043632075471698,
      "grad_norm": 7.343114376068115,
      "learning_rate": 3.018867924528302e-06,
      "loss": 0.6161,
      "num_input_tokens_seen": 675264,
      "step": 1025
    },
    {
      "epoch": 0.6073113207547169,
      "grad_norm": 4.674347400665283,
      "learning_rate": 3.033608490566038e-06,
      "loss": 0.5463,
      "num_input_tokens_seen": 678624,
      "step": 1030
    },
    {
      "epoch": 0.6102594339622641,
      "grad_norm": 2.357262134552002,
      "learning_rate": 3.0483490566037737e-06,
      "loss": 0.6682,
      "num_input_tokens_seen": 683264,
      "step": 1035
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 4.380685329437256,
      "learning_rate": 3.0630896226415095e-06,
      "loss": 0.5731,
      "num_input_tokens_seen": 687552,
      "step": 1040
    },
    {
      "epoch": 0.6161556603773585,
      "grad_norm": 6.210946083068848,
      "learning_rate": 3.0778301886792457e-06,
      "loss": 0.4943,
      "num_input_tokens_seen": 690208,
      "step": 1045
    },
    {
      "epoch": 0.6191037735849056,
      "grad_norm": 2.7070870399475098,
      "learning_rate": 3.0925707547169815e-06,
      "loss": 0.644,
      "num_input_tokens_seen": 693600,
      "step": 1050
    },
    {
      "epoch": 0.6220518867924528,
      "grad_norm": 5.82750940322876,
      "learning_rate": 3.1073113207547172e-06,
      "loss": 0.6085,
      "num_input_tokens_seen": 696160,
      "step": 1055
    },
    {
      "epoch": 0.625,
      "grad_norm": 7.563085556030273,
      "learning_rate": 3.122051886792453e-06,
      "loss": 0.6682,
      "num_input_tokens_seen": 700384,
      "step": 1060
    },
    {
      "epoch": 0.6279481132075472,
      "grad_norm": 2.918795108795166,
      "learning_rate": 3.136792452830189e-06,
      "loss": 0.521,
      "num_input_tokens_seen": 703552,
      "step": 1065
    },
    {
      "epoch": 0.6308962264150944,
      "grad_norm": 6.497398853302002,
      "learning_rate": 3.151533018867925e-06,
      "loss": 0.5971,
      "num_input_tokens_seen": 707360,
      "step": 1070
    },
    {
      "epoch": 0.6338443396226415,
      "grad_norm": 2.6607465744018555,
      "learning_rate": 3.1662735849056607e-06,
      "loss": 0.5919,
      "num_input_tokens_seen": 711232,
      "step": 1075
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 6.521037578582764,
      "learning_rate": 3.181014150943397e-06,
      "loss": 0.6113,
      "num_input_tokens_seen": 713984,
      "step": 1080
    },
    {
      "epoch": 0.6397405660377359,
      "grad_norm": 3.9145472049713135,
      "learning_rate": 3.1957547169811327e-06,
      "loss": 0.6166,
      "num_input_tokens_seen": 717024,
      "step": 1085
    },
    {
      "epoch": 0.6426886792452831,
      "grad_norm": 3.745694637298584,
      "learning_rate": 3.210495283018868e-06,
      "loss": 0.5285,
      "num_input_tokens_seen": 720672,
      "step": 1090
    },
    {
      "epoch": 0.6456367924528302,
      "grad_norm": 2.291426420211792,
      "learning_rate": 3.225235849056604e-06,
      "loss": 0.6873,
      "num_input_tokens_seen": 724960,
      "step": 1095
    },
    {
      "epoch": 0.6485849056603774,
      "grad_norm": 5.659872055053711,
      "learning_rate": 3.2399764150943396e-06,
      "loss": 0.6268,
      "num_input_tokens_seen": 727616,
      "step": 1100
    },
    {
      "epoch": 0.6515330188679245,
      "grad_norm": 9.969420433044434,
      "learning_rate": 3.2547169811320758e-06,
      "loss": 0.6578,
      "num_input_tokens_seen": 730144,
      "step": 1105
    },
    {
      "epoch": 0.6544811320754716,
      "grad_norm": 4.853862285614014,
      "learning_rate": 3.2694575471698115e-06,
      "loss": 0.4531,
      "num_input_tokens_seen": 733696,
      "step": 1110
    },
    {
      "epoch": 0.6574292452830188,
      "grad_norm": 6.840001106262207,
      "learning_rate": 3.2841981132075473e-06,
      "loss": 0.6455,
      "num_input_tokens_seen": 737120,
      "step": 1115
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 4.258190631866455,
      "learning_rate": 3.298938679245283e-06,
      "loss": 0.5823,
      "num_input_tokens_seen": 740512,
      "step": 1120
    },
    {
      "epoch": 0.6633254716981132,
      "grad_norm": 3.404411554336548,
      "learning_rate": 3.3136792452830192e-06,
      "loss": 0.6032,
      "num_input_tokens_seen": 743936,
      "step": 1125
    },
    {
      "epoch": 0.6662735849056604,
      "grad_norm": 6.171648979187012,
      "learning_rate": 3.328419811320755e-06,
      "loss": 0.5163,
      "num_input_tokens_seen": 746176,
      "step": 1130
    },
    {
      "epoch": 0.6692216981132075,
      "grad_norm": 6.43703556060791,
      "learning_rate": 3.3431603773584908e-06,
      "loss": 0.5427,
      "num_input_tokens_seen": 749504,
      "step": 1135
    },
    {
      "epoch": 0.6721698113207547,
      "grad_norm": 5.361392021179199,
      "learning_rate": 3.3579009433962266e-06,
      "loss": 0.5874,
      "num_input_tokens_seen": 752672,
      "step": 1140
    },
    {
      "epoch": 0.6751179245283019,
      "grad_norm": 4.4862470626831055,
      "learning_rate": 3.3726415094339627e-06,
      "loss": 0.6977,
      "num_input_tokens_seen": 755296,
      "step": 1145
    },
    {
      "epoch": 0.6780660377358491,
      "grad_norm": 7.122319221496582,
      "learning_rate": 3.3873820754716985e-06,
      "loss": 0.7821,
      "num_input_tokens_seen": 758048,
      "step": 1150
    },
    {
      "epoch": 0.6810141509433962,
      "grad_norm": 5.9745635986328125,
      "learning_rate": 3.4021226415094343e-06,
      "loss": 0.6409,
      "num_input_tokens_seen": 761024,
      "step": 1155
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 3.079890251159668,
      "learning_rate": 3.4168632075471705e-06,
      "loss": 0.6653,
      "num_input_tokens_seen": 763808,
      "step": 1160
    },
    {
      "epoch": 0.6869103773584906,
      "grad_norm": 2.491684913635254,
      "learning_rate": 3.4316037735849062e-06,
      "loss": 0.5945,
      "num_input_tokens_seen": 766912,
      "step": 1165
    },
    {
      "epoch": 0.6898584905660378,
      "grad_norm": 3.255516529083252,
      "learning_rate": 3.446344339622642e-06,
      "loss": 0.7233,
      "num_input_tokens_seen": 770880,
      "step": 1170
    },
    {
      "epoch": 0.6928066037735849,
      "grad_norm": 13.536747932434082,
      "learning_rate": 3.4610849056603778e-06,
      "loss": 0.5657,
      "num_input_tokens_seen": 774432,
      "step": 1175
    },
    {
      "epoch": 0.6957547169811321,
      "grad_norm": 9.903853416442871,
      "learning_rate": 3.475825471698113e-06,
      "loss": 0.6825,
      "num_input_tokens_seen": 777920,
      "step": 1180
    },
    {
      "epoch": 0.6987028301886793,
      "grad_norm": 3.3338961601257324,
      "learning_rate": 3.4905660377358493e-06,
      "loss": 0.6523,
      "num_input_tokens_seen": 780640,
      "step": 1185
    },
    {
      "epoch": 0.7016509433962265,
      "grad_norm": 4.739099979400635,
      "learning_rate": 3.505306603773585e-06,
      "loss": 0.593,
      "num_input_tokens_seen": 783072,
      "step": 1190
    },
    {
      "epoch": 0.7045990566037735,
      "grad_norm": 5.618747234344482,
      "learning_rate": 3.520047169811321e-06,
      "loss": 0.4876,
      "num_input_tokens_seen": 786944,
      "step": 1195
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 3.2195944786071777,
      "learning_rate": 3.5347877358490566e-06,
      "loss": 0.619,
      "num_input_tokens_seen": 791680,
      "step": 1200
    },
    {
      "epoch": 0.7104952830188679,
      "grad_norm": 2.5906782150268555,
      "learning_rate": 3.549528301886793e-06,
      "loss": 0.6129,
      "num_input_tokens_seen": 794912,
      "step": 1205
    },
    {
      "epoch": 0.7134433962264151,
      "grad_norm": 4.7796478271484375,
      "learning_rate": 3.5642688679245286e-06,
      "loss": 0.5696,
      "num_input_tokens_seen": 797728,
      "step": 1210
    },
    {
      "epoch": 0.7163915094339622,
      "grad_norm": 9.569525718688965,
      "learning_rate": 3.5790094339622643e-06,
      "loss": 0.4925,
      "num_input_tokens_seen": 801632,
      "step": 1215
    },
    {
      "epoch": 0.7193396226415094,
      "grad_norm": 7.355144023895264,
      "learning_rate": 3.59375e-06,
      "loss": 0.6959,
      "num_input_tokens_seen": 804640,
      "step": 1220
    },
    {
      "epoch": 0.7222877358490566,
      "grad_norm": 7.011343955993652,
      "learning_rate": 3.6084905660377363e-06,
      "loss": 0.6877,
      "num_input_tokens_seen": 807008,
      "step": 1225
    },
    {
      "epoch": 0.7252358490566038,
      "grad_norm": 5.464759826660156,
      "learning_rate": 3.623231132075472e-06,
      "loss": 0.6237,
      "num_input_tokens_seen": 810176,
      "step": 1230
    },
    {
      "epoch": 0.7281839622641509,
      "grad_norm": 4.375668048858643,
      "learning_rate": 3.637971698113208e-06,
      "loss": 0.7286,
      "num_input_tokens_seen": 813440,
      "step": 1235
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 2.583573818206787,
      "learning_rate": 3.652712264150944e-06,
      "loss": 0.4679,
      "num_input_tokens_seen": 817728,
      "step": 1240
    },
    {
      "epoch": 0.7340801886792453,
      "grad_norm": 2.480966091156006,
      "learning_rate": 3.66745283018868e-06,
      "loss": 0.5281,
      "num_input_tokens_seen": 820768,
      "step": 1245
    },
    {
      "epoch": 0.7370283018867925,
      "grad_norm": 7.523268699645996,
      "learning_rate": 3.6821933962264156e-06,
      "loss": 0.5894,
      "num_input_tokens_seen": 823392,
      "step": 1250
    },
    {
      "epoch": 0.7399764150943396,
      "grad_norm": 3.030979633331299,
      "learning_rate": 3.6969339622641513e-06,
      "loss": 0.6526,
      "num_input_tokens_seen": 825856,
      "step": 1255
    },
    {
      "epoch": 0.7429245283018868,
      "grad_norm": 4.060668468475342,
      "learning_rate": 3.7116745283018875e-06,
      "loss": 0.6207,
      "num_input_tokens_seen": 828960,
      "step": 1260
    },
    {
      "epoch": 0.745872641509434,
      "grad_norm": 3.184591293334961,
      "learning_rate": 3.726415094339623e-06,
      "loss": 0.6705,
      "num_input_tokens_seen": 832352,
      "step": 1265
    },
    {
      "epoch": 0.7488207547169812,
      "grad_norm": 5.392750263214111,
      "learning_rate": 3.7411556603773586e-06,
      "loss": 0.3802,
      "num_input_tokens_seen": 837056,
      "step": 1270
    },
    {
      "epoch": 0.7517688679245284,
      "grad_norm": 4.639246463775635,
      "learning_rate": 3.7558962264150944e-06,
      "loss": 0.6304,
      "num_input_tokens_seen": 841056,
      "step": 1275
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 8.511612892150879,
      "learning_rate": 3.77063679245283e-06,
      "loss": 0.7641,
      "num_input_tokens_seen": 843232,
      "step": 1280
    },
    {
      "epoch": 0.7576650943396226,
      "grad_norm": 3.3993396759033203,
      "learning_rate": 3.7853773584905664e-06,
      "loss": 0.6184,
      "num_input_tokens_seen": 847264,
      "step": 1285
    },
    {
      "epoch": 0.7606132075471698,
      "grad_norm": 3.0817723274230957,
      "learning_rate": 3.800117924528302e-06,
      "loss": 0.5735,
      "num_input_tokens_seen": 851296,
      "step": 1290
    },
    {
      "epoch": 0.7635613207547169,
      "grad_norm": 5.478623390197754,
      "learning_rate": 3.814858490566038e-06,
      "loss": 0.5515,
      "num_input_tokens_seen": 854336,
      "step": 1295
    },
    {
      "epoch": 0.7665094339622641,
      "grad_norm": 6.094629287719727,
      "learning_rate": 3.829599056603774e-06,
      "loss": 0.5659,
      "num_input_tokens_seen": 856704,
      "step": 1300
    },
    {
      "epoch": 0.7694575471698113,
      "grad_norm": 3.3267412185668945,
      "learning_rate": 3.8443396226415094e-06,
      "loss": 0.7062,
      "num_input_tokens_seen": 860000,
      "step": 1305
    },
    {
      "epoch": 0.7724056603773585,
      "grad_norm": 2.7097973823547363,
      "learning_rate": 3.859080188679246e-06,
      "loss": 0.5802,
      "num_input_tokens_seen": 864160,
      "step": 1310
    },
    {
      "epoch": 0.7753537735849056,
      "grad_norm": 4.457141399383545,
      "learning_rate": 3.873820754716982e-06,
      "loss": 0.8713,
      "num_input_tokens_seen": 866848,
      "step": 1315
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 3.9322586059570312,
      "learning_rate": 3.888561320754717e-06,
      "loss": 0.7524,
      "num_input_tokens_seen": 870176,
      "step": 1320
    },
    {
      "epoch": 0.78125,
      "grad_norm": 3.8864223957061768,
      "learning_rate": 3.903301886792453e-06,
      "loss": 0.6063,
      "num_input_tokens_seen": 874976,
      "step": 1325
    },
    {
      "epoch": 0.7841981132075472,
      "grad_norm": 3.4247708320617676,
      "learning_rate": 3.9180424528301895e-06,
      "loss": 0.6285,
      "num_input_tokens_seen": 878112,
      "step": 1330
    },
    {
      "epoch": 0.7871462264150944,
      "grad_norm": 5.2729363441467285,
      "learning_rate": 3.932783018867925e-06,
      "loss": 0.5765,
      "num_input_tokens_seen": 881248,
      "step": 1335
    },
    {
      "epoch": 0.7900943396226415,
      "grad_norm": 5.952835559844971,
      "learning_rate": 3.947523584905661e-06,
      "loss": 0.5969,
      "num_input_tokens_seen": 884096,
      "step": 1340
    },
    {
      "epoch": 0.7930424528301887,
      "grad_norm": 3.1927132606506348,
      "learning_rate": 3.962264150943396e-06,
      "loss": 0.532,
      "num_input_tokens_seen": 886816,
      "step": 1345
    },
    {
      "epoch": 0.7959905660377359,
      "grad_norm": 4.66749382019043,
      "learning_rate": 3.977004716981133e-06,
      "loss": 0.6226,
      "num_input_tokens_seen": 889952,
      "step": 1350
    },
    {
      "epoch": 0.7989386792452831,
      "grad_norm": 3.9639790058135986,
      "learning_rate": 3.991745283018868e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 892480,
      "step": 1355
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 6.159574031829834,
      "learning_rate": 4.006485849056604e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 895296,
      "step": 1360
    },
    {
      "epoch": 0.8048349056603774,
      "grad_norm": 4.3759541511535645,
      "learning_rate": 4.0212264150943395e-06,
      "loss": 0.5416,
      "num_input_tokens_seen": 898752,
      "step": 1365
    },
    {
      "epoch": 0.8077830188679245,
      "grad_norm": 3.6086649894714355,
      "learning_rate": 4.035966981132076e-06,
      "loss": 0.4249,
      "num_input_tokens_seen": 901600,
      "step": 1370
    },
    {
      "epoch": 0.8107311320754716,
      "grad_norm": 3.9799115657806396,
      "learning_rate": 4.050707547169812e-06,
      "loss": 0.5514,
      "num_input_tokens_seen": 904384,
      "step": 1375
    },
    {
      "epoch": 0.8136792452830188,
      "grad_norm": 3.699873208999634,
      "learning_rate": 4.065448113207547e-06,
      "loss": 0.8923,
      "num_input_tokens_seen": 907264,
      "step": 1380
    },
    {
      "epoch": 0.816627358490566,
      "grad_norm": 11.235692024230957,
      "learning_rate": 4.080188679245283e-06,
      "loss": 0.5869,
      "num_input_tokens_seen": 909760,
      "step": 1385
    },
    {
      "epoch": 0.8195754716981132,
      "grad_norm": 2.4536454677581787,
      "learning_rate": 4.094929245283019e-06,
      "loss": 0.5282,
      "num_input_tokens_seen": 912736,
      "step": 1390
    },
    {
      "epoch": 0.8225235849056604,
      "grad_norm": 8.963294982910156,
      "learning_rate": 4.109669811320755e-06,
      "loss": 0.4604,
      "num_input_tokens_seen": 916384,
      "step": 1395
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 3.29095196723938,
      "learning_rate": 4.124410377358491e-06,
      "loss": 0.7879,
      "num_input_tokens_seen": 919744,
      "step": 1400
    },
    {
      "epoch": 0.8284198113207547,
      "grad_norm": 3.1918609142303467,
      "learning_rate": 4.1391509433962265e-06,
      "loss": 0.5083,
      "num_input_tokens_seen": 922368,
      "step": 1405
    },
    {
      "epoch": 0.8313679245283019,
      "grad_norm": 2.2130563259124756,
      "learning_rate": 4.153891509433963e-06,
      "loss": 0.5873,
      "num_input_tokens_seen": 925152,
      "step": 1410
    },
    {
      "epoch": 0.8343160377358491,
      "grad_norm": 5.1197123527526855,
      "learning_rate": 4.168632075471699e-06,
      "loss": 0.8659,
      "num_input_tokens_seen": 927648,
      "step": 1415
    },
    {
      "epoch": 0.8372641509433962,
      "grad_norm": 2.3207921981811523,
      "learning_rate": 4.183372641509434e-06,
      "loss": 0.5608,
      "num_input_tokens_seen": 930304,
      "step": 1420
    },
    {
      "epoch": 0.8402122641509434,
      "grad_norm": 3.15523099899292,
      "learning_rate": 4.19811320754717e-06,
      "loss": 0.4737,
      "num_input_tokens_seen": 933536,
      "step": 1425
    },
    {
      "epoch": 0.8431603773584906,
      "grad_norm": 2.65962815284729,
      "learning_rate": 4.212853773584907e-06,
      "loss": 0.4669,
      "num_input_tokens_seen": 937504,
      "step": 1430
    },
    {
      "epoch": 0.8461084905660378,
      "grad_norm": 5.112860202789307,
      "learning_rate": 4.227594339622642e-06,
      "loss": 0.8449,
      "num_input_tokens_seen": 940128,
      "step": 1435
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 2.316781759262085,
      "learning_rate": 4.242334905660378e-06,
      "loss": 0.5917,
      "num_input_tokens_seen": 943488,
      "step": 1440
    },
    {
      "epoch": 0.8520047169811321,
      "grad_norm": 4.522274494171143,
      "learning_rate": 4.2570754716981135e-06,
      "loss": 0.6725,
      "num_input_tokens_seen": 946016,
      "step": 1445
    },
    {
      "epoch": 0.8549528301886793,
      "grad_norm": 3.6812198162078857,
      "learning_rate": 4.271816037735849e-06,
      "loss": 0.3664,
      "num_input_tokens_seen": 948672,
      "step": 1450
    },
    {
      "epoch": 0.8579009433962265,
      "grad_norm": 4.211163520812988,
      "learning_rate": 4.286556603773585e-06,
      "loss": 0.5968,
      "num_input_tokens_seen": 951520,
      "step": 1455
    },
    {
      "epoch": 0.8608490566037735,
      "grad_norm": 4.800759792327881,
      "learning_rate": 4.301297169811321e-06,
      "loss": 0.5468,
      "num_input_tokens_seen": 954592,
      "step": 1460
    },
    {
      "epoch": 0.8637971698113207,
      "grad_norm": 4.913768768310547,
      "learning_rate": 4.3160377358490565e-06,
      "loss": 0.4686,
      "num_input_tokens_seen": 957696,
      "step": 1465
    },
    {
      "epoch": 0.8667452830188679,
      "grad_norm": 10.273981094360352,
      "learning_rate": 4.330778301886793e-06,
      "loss": 0.6987,
      "num_input_tokens_seen": 960832,
      "step": 1470
    },
    {
      "epoch": 0.8696933962264151,
      "grad_norm": 12.555068969726562,
      "learning_rate": 4.345518867924529e-06,
      "loss": 0.7999,
      "num_input_tokens_seen": 964608,
      "step": 1475
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 8.242755889892578,
      "learning_rate": 4.360259433962264e-06,
      "loss": 0.5806,
      "num_input_tokens_seen": 967360,
      "step": 1480
    },
    {
      "epoch": 0.8755896226415094,
      "grad_norm": 3.7882580757141113,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.5511,
      "num_input_tokens_seen": 970432,
      "step": 1485
    },
    {
      "epoch": 0.8785377358490566,
      "grad_norm": 9.238855361938477,
      "learning_rate": 4.389740566037737e-06,
      "loss": 0.9184,
      "num_input_tokens_seen": 973920,
      "step": 1490
    },
    {
      "epoch": 0.8814858490566038,
      "grad_norm": 9.596097946166992,
      "learning_rate": 4.404481132075472e-06,
      "loss": 0.5495,
      "num_input_tokens_seen": 977152,
      "step": 1495
    },
    {
      "epoch": 0.8844339622641509,
      "grad_norm": 4.166565418243408,
      "learning_rate": 4.419221698113208e-06,
      "loss": 0.6504,
      "num_input_tokens_seen": 979712,
      "step": 1500
    },
    {
      "epoch": 0.8873820754716981,
      "grad_norm": 6.09000825881958,
      "learning_rate": 4.4339622641509435e-06,
      "loss": 0.6059,
      "num_input_tokens_seen": 982592,
      "step": 1505
    },
    {
      "epoch": 0.8903301886792453,
      "grad_norm": 4.761940002441406,
      "learning_rate": 4.44870283018868e-06,
      "loss": 0.5089,
      "num_input_tokens_seen": 985056,
      "step": 1510
    },
    {
      "epoch": 0.8932783018867925,
      "grad_norm": 4.209322452545166,
      "learning_rate": 4.463443396226416e-06,
      "loss": 0.4899,
      "num_input_tokens_seen": 987904,
      "step": 1515
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 8.132052421569824,
      "learning_rate": 4.478183962264151e-06,
      "loss": 0.4878,
      "num_input_tokens_seen": 990752,
      "step": 1520
    },
    {
      "epoch": 0.8991745283018868,
      "grad_norm": 3.136486053466797,
      "learning_rate": 4.4929245283018875e-06,
      "loss": 0.5886,
      "num_input_tokens_seen": 993344,
      "step": 1525
    },
    {
      "epoch": 0.902122641509434,
      "grad_norm": 2.8332455158233643,
      "learning_rate": 4.507665094339623e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 996896,
      "step": 1530
    },
    {
      "epoch": 0.9050707547169812,
      "grad_norm": 4.6405720710754395,
      "learning_rate": 4.522405660377359e-06,
      "loss": 0.4948,
      "num_input_tokens_seen": 1000672,
      "step": 1535
    },
    {
      "epoch": 0.9080188679245284,
      "grad_norm": 3.8325154781341553,
      "learning_rate": 4.537146226415094e-06,
      "loss": 0.4765,
      "num_input_tokens_seen": 1003808,
      "step": 1540
    },
    {
      "epoch": 0.9109669811320755,
      "grad_norm": 2.709085702896118,
      "learning_rate": 4.5518867924528305e-06,
      "loss": 0.5747,
      "num_input_tokens_seen": 1006720,
      "step": 1545
    },
    {
      "epoch": 0.9139150943396226,
      "grad_norm": 3.058397054672241,
      "learning_rate": 4.566627358490566e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 1009888,
      "step": 1550
    },
    {
      "epoch": 0.9168632075471698,
      "grad_norm": 4.5157389640808105,
      "learning_rate": 4.581367924528302e-06,
      "loss": 0.5392,
      "num_input_tokens_seen": 1012544,
      "step": 1555
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 2.6514673233032227,
      "learning_rate": 4.596108490566038e-06,
      "loss": 0.4825,
      "num_input_tokens_seen": 1015232,
      "step": 1560
    },
    {
      "epoch": 0.9227594339622641,
      "grad_norm": 6.584100723266602,
      "learning_rate": 4.610849056603774e-06,
      "loss": 0.5498,
      "num_input_tokens_seen": 1018496,
      "step": 1565
    },
    {
      "epoch": 0.9257075471698113,
      "grad_norm": 4.574212074279785,
      "learning_rate": 4.62558962264151e-06,
      "loss": 0.6881,
      "num_input_tokens_seen": 1022432,
      "step": 1570
    },
    {
      "epoch": 0.9286556603773585,
      "grad_norm": 4.499818325042725,
      "learning_rate": 4.640330188679246e-06,
      "loss": 0.6306,
      "num_input_tokens_seen": 1025760,
      "step": 1575
    },
    {
      "epoch": 0.9316037735849056,
      "grad_norm": 1.7668083906173706,
      "learning_rate": 4.655070754716981e-06,
      "loss": 0.6061,
      "num_input_tokens_seen": 1030400,
      "step": 1580
    },
    {
      "epoch": 0.9345518867924528,
      "grad_norm": 11.907697677612305,
      "learning_rate": 4.6698113207547175e-06,
      "loss": 0.4505,
      "num_input_tokens_seen": 1033152,
      "step": 1585
    },
    {
      "epoch": 0.9375,
      "grad_norm": 3.887160539627075,
      "learning_rate": 4.684551886792454e-06,
      "loss": 0.5413,
      "num_input_tokens_seen": 1036896,
      "step": 1590
    },
    {
      "epoch": 0.9404481132075472,
      "grad_norm": 2.54729962348938,
      "learning_rate": 4.699292452830189e-06,
      "loss": 0.5893,
      "num_input_tokens_seen": 1040224,
      "step": 1595
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 4.591376304626465,
      "learning_rate": 4.714033018867925e-06,
      "loss": 0.491,
      "num_input_tokens_seen": 1042976,
      "step": 1600
    },
    {
      "epoch": 0.9463443396226415,
      "grad_norm": 2.9473679065704346,
      "learning_rate": 4.728773584905661e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 1046112,
      "step": 1605
    },
    {
      "epoch": 0.9492924528301887,
      "grad_norm": 40.50365447998047,
      "learning_rate": 4.743514150943397e-06,
      "loss": 0.5851,
      "num_input_tokens_seen": 1049152,
      "step": 1610
    },
    {
      "epoch": 0.9522405660377359,
      "grad_norm": 3.8748273849487305,
      "learning_rate": 4.758254716981133e-06,
      "loss": 0.4134,
      "num_input_tokens_seen": 1051968,
      "step": 1615
    },
    {
      "epoch": 0.9551886792452831,
      "grad_norm": 2.571462392807007,
      "learning_rate": 4.772995283018868e-06,
      "loss": 0.5304,
      "num_input_tokens_seen": 1059424,
      "step": 1620
    },
    {
      "epoch": 0.9581367924528302,
      "grad_norm": 3.0791208744049072,
      "learning_rate": 4.787735849056604e-06,
      "loss": 0.6092,
      "num_input_tokens_seen": 1062240,
      "step": 1625
    },
    {
      "epoch": 0.9610849056603774,
      "grad_norm": 3.8257946968078613,
      "learning_rate": 4.80247641509434e-06,
      "loss": 0.5283,
      "num_input_tokens_seen": 1065792,
      "step": 1630
    },
    {
      "epoch": 0.9640330188679245,
      "grad_norm": 2.1022684574127197,
      "learning_rate": 4.817216981132076e-06,
      "loss": 0.6121,
      "num_input_tokens_seen": 1069280,
      "step": 1635
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 4.170267105102539,
      "learning_rate": 4.831957547169811e-06,
      "loss": 0.5424,
      "num_input_tokens_seen": 1071936,
      "step": 1640
    },
    {
      "epoch": 0.9699292452830188,
      "grad_norm": 2.8740952014923096,
      "learning_rate": 4.8466981132075476e-06,
      "loss": 0.6563,
      "num_input_tokens_seen": 1074944,
      "step": 1645
    },
    {
      "epoch": 0.972877358490566,
      "grad_norm": 2.168450117111206,
      "learning_rate": 4.861438679245283e-06,
      "loss": 0.5446,
      "num_input_tokens_seen": 1078368,
      "step": 1650
    },
    {
      "epoch": 0.9758254716981132,
      "grad_norm": 2.4585602283477783,
      "learning_rate": 4.876179245283019e-06,
      "loss": 0.5517,
      "num_input_tokens_seen": 1081312,
      "step": 1655
    },
    {
      "epoch": 0.9787735849056604,
      "grad_norm": 2.7601847648620605,
      "learning_rate": 4.890919811320755e-06,
      "loss": 0.8125,
      "num_input_tokens_seen": 1088704,
      "step": 1660
    },
    {
      "epoch": 0.9817216981132075,
      "grad_norm": 3.5017166137695312,
      "learning_rate": 4.905660377358491e-06,
      "loss": 0.637,
      "num_input_tokens_seen": 1092352,
      "step": 1665
    },
    {
      "epoch": 0.9846698113207547,
      "grad_norm": 4.5479326248168945,
      "learning_rate": 4.920400943396227e-06,
      "loss": 0.6539,
      "num_input_tokens_seen": 1094880,
      "step": 1670
    },
    {
      "epoch": 0.9876179245283019,
      "grad_norm": 3.222949981689453,
      "learning_rate": 4.935141509433963e-06,
      "loss": 0.5469,
      "num_input_tokens_seen": 1098176,
      "step": 1675
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 1.6342679262161255,
      "learning_rate": 4.949882075471698e-06,
      "loss": 0.5497,
      "num_input_tokens_seen": 1101440,
      "step": 1680
    },
    {
      "epoch": 0.9935141509433962,
      "grad_norm": 5.735625267028809,
      "learning_rate": 4.9646226415094346e-06,
      "loss": 0.7427,
      "num_input_tokens_seen": 1104736,
      "step": 1685
    },
    {
      "epoch": 0.9964622641509434,
      "grad_norm": 3.494211435317993,
      "learning_rate": 4.979363207547171e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 1107328,
      "step": 1690
    },
    {
      "epoch": 0.9994103773584906,
      "grad_norm": 3.2509055137634277,
      "learning_rate": 4.994103773584906e-06,
      "loss": 0.4854,
      "num_input_tokens_seen": 1110528,
      "step": 1695
    },
    {
      "epoch": 1.0023584905660377,
      "grad_norm": 2.727750301361084,
      "learning_rate": 5.0088443396226414e-06,
      "loss": 0.566,
      "num_input_tokens_seen": 1113400,
      "step": 1700
    },
    {
      "epoch": 1.005306603773585,
      "grad_norm": 8.881604194641113,
      "learning_rate": 5.023584905660378e-06,
      "loss": 0.5185,
      "num_input_tokens_seen": 1117144,
      "step": 1705
    },
    {
      "epoch": 1.008254716981132,
      "grad_norm": 2.6063976287841797,
      "learning_rate": 5.038325471698113e-06,
      "loss": 0.6226,
      "num_input_tokens_seen": 1120216,
      "step": 1710
    },
    {
      "epoch": 1.0112028301886793,
      "grad_norm": 2.9364309310913086,
      "learning_rate": 5.05306603773585e-06,
      "loss": 0.5095,
      "num_input_tokens_seen": 1122904,
      "step": 1715
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 2.151931047439575,
      "learning_rate": 5.067806603773585e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 1126552,
      "step": 1720
    },
    {
      "epoch": 1.0170990566037736,
      "grad_norm": 2.916461706161499,
      "learning_rate": 5.0825471698113216e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 1130040,
      "step": 1725
    },
    {
      "epoch": 1.0200471698113207,
      "grad_norm": 2.9760499000549316,
      "learning_rate": 5.097287735849057e-06,
      "loss": 0.5636,
      "num_input_tokens_seen": 1133752,
      "step": 1730
    },
    {
      "epoch": 1.022995283018868,
      "grad_norm": 8.922828674316406,
      "learning_rate": 5.112028301886793e-06,
      "loss": 0.6907,
      "num_input_tokens_seen": 1135928,
      "step": 1735
    },
    {
      "epoch": 1.025943396226415,
      "grad_norm": 6.082559108734131,
      "learning_rate": 5.1267688679245284e-06,
      "loss": 0.566,
      "num_input_tokens_seen": 1138712,
      "step": 1740
    },
    {
      "epoch": 1.0288915094339623,
      "grad_norm": 6.1941399574279785,
      "learning_rate": 5.1415094339622655e-06,
      "loss": 0.5239,
      "num_input_tokens_seen": 1141752,
      "step": 1745
    },
    {
      "epoch": 1.0318396226415094,
      "grad_norm": 12.077676773071289,
      "learning_rate": 5.156250000000001e-06,
      "loss": 0.4853,
      "num_input_tokens_seen": 1144952,
      "step": 1750
    },
    {
      "epoch": 1.0347877358490567,
      "grad_norm": 2.5476250648498535,
      "learning_rate": 5.170990566037736e-06,
      "loss": 0.534,
      "num_input_tokens_seen": 1147416,
      "step": 1755
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 2.8608908653259277,
      "learning_rate": 5.185731132075472e-06,
      "loss": 0.6777,
      "num_input_tokens_seen": 1150840,
      "step": 1760
    },
    {
      "epoch": 1.040683962264151,
      "grad_norm": 2.929492950439453,
      "learning_rate": 5.200471698113208e-06,
      "loss": 0.5138,
      "num_input_tokens_seen": 1153592,
      "step": 1765
    },
    {
      "epoch": 1.0436320754716981,
      "grad_norm": 7.823033332824707,
      "learning_rate": 5.215212264150944e-06,
      "loss": 0.4222,
      "num_input_tokens_seen": 1156888,
      "step": 1770
    },
    {
      "epoch": 1.0465801886792452,
      "grad_norm": 14.291205406188965,
      "learning_rate": 5.229952830188679e-06,
      "loss": 0.5446,
      "num_input_tokens_seen": 1159992,
      "step": 1775
    },
    {
      "epoch": 1.0495283018867925,
      "grad_norm": 4.886681079864502,
      "learning_rate": 5.2446933962264154e-06,
      "loss": 0.6806,
      "num_input_tokens_seen": 1163224,
      "step": 1780
    },
    {
      "epoch": 1.0524764150943395,
      "grad_norm": 3.389892816543579,
      "learning_rate": 5.259433962264151e-06,
      "loss": 0.6864,
      "num_input_tokens_seen": 1166744,
      "step": 1785
    },
    {
      "epoch": 1.0554245283018868,
      "grad_norm": 1.4496123790740967,
      "learning_rate": 5.274174528301888e-06,
      "loss": 0.475,
      "num_input_tokens_seen": 1171224,
      "step": 1790
    },
    {
      "epoch": 1.0583726415094339,
      "grad_norm": 2.02174711227417,
      "learning_rate": 5.288915094339623e-06,
      "loss": 0.4934,
      "num_input_tokens_seen": 1175096,
      "step": 1795
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 2.5030105113983154,
      "learning_rate": 5.303655660377359e-06,
      "loss": 0.4973,
      "num_input_tokens_seen": 1178296,
      "step": 1800
    },
    {
      "epoch": 1.0642688679245282,
      "grad_norm": 5.1232404708862305,
      "learning_rate": 5.318396226415095e-06,
      "loss": 0.5081,
      "num_input_tokens_seen": 1181176,
      "step": 1805
    },
    {
      "epoch": 1.0672169811320755,
      "grad_norm": 4.92999267578125,
      "learning_rate": 5.333136792452831e-06,
      "loss": 0.6466,
      "num_input_tokens_seen": 1188376,
      "step": 1810
    },
    {
      "epoch": 1.0701650943396226,
      "grad_norm": 2.0491416454315186,
      "learning_rate": 5.347877358490566e-06,
      "loss": 0.5493,
      "num_input_tokens_seen": 1192344,
      "step": 1815
    },
    {
      "epoch": 1.0731132075471699,
      "grad_norm": 3.5442657470703125,
      "learning_rate": 5.362617924528302e-06,
      "loss": 0.6354,
      "num_input_tokens_seen": 1195416,
      "step": 1820
    },
    {
      "epoch": 1.076061320754717,
      "grad_norm": 3.5353639125823975,
      "learning_rate": 5.377358490566038e-06,
      "loss": 0.6096,
      "num_input_tokens_seen": 1198936,
      "step": 1825
    },
    {
      "epoch": 1.0790094339622642,
      "grad_norm": 6.455561637878418,
      "learning_rate": 5.392099056603775e-06,
      "loss": 0.7194,
      "num_input_tokens_seen": 1201688,
      "step": 1830
    },
    {
      "epoch": 1.0819575471698113,
      "grad_norm": 4.391550540924072,
      "learning_rate": 5.40683962264151e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 1204760,
      "step": 1835
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 3.7178587913513184,
      "learning_rate": 5.4215801886792455e-06,
      "loss": 0.5533,
      "num_input_tokens_seen": 1207416,
      "step": 1840
    },
    {
      "epoch": 1.0878537735849056,
      "grad_norm": 3.764296293258667,
      "learning_rate": 5.436320754716982e-06,
      "loss": 0.8037,
      "num_input_tokens_seen": 1210296,
      "step": 1845
    },
    {
      "epoch": 1.0908018867924527,
      "grad_norm": 1.5686968564987183,
      "learning_rate": 5.451061320754717e-06,
      "loss": 0.6276,
      "num_input_tokens_seen": 1213304,
      "step": 1850
    },
    {
      "epoch": 1.09375,
      "grad_norm": 2.4531240463256836,
      "learning_rate": 5.465801886792453e-06,
      "loss": 0.6061,
      "num_input_tokens_seen": 1216728,
      "step": 1855
    },
    {
      "epoch": 1.0966981132075473,
      "grad_norm": 2.344271183013916,
      "learning_rate": 5.4805424528301886e-06,
      "loss": 0.4811,
      "num_input_tokens_seen": 1220312,
      "step": 1860
    },
    {
      "epoch": 1.0996462264150944,
      "grad_norm": 3.7010281085968018,
      "learning_rate": 5.495283018867925e-06,
      "loss": 0.618,
      "num_input_tokens_seen": 1224152,
      "step": 1865
    },
    {
      "epoch": 1.1025943396226414,
      "grad_norm": 3.5445778369903564,
      "learning_rate": 5.51002358490566e-06,
      "loss": 0.5742,
      "num_input_tokens_seen": 1226488,
      "step": 1870
    },
    {
      "epoch": 1.1055424528301887,
      "grad_norm": 3.5775256156921387,
      "learning_rate": 5.524764150943397e-06,
      "loss": 0.6429,
      "num_input_tokens_seen": 1229912,
      "step": 1875
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 3.600529670715332,
      "learning_rate": 5.5395047169811325e-06,
      "loss": 0.591,
      "num_input_tokens_seen": 1233400,
      "step": 1880
    },
    {
      "epoch": 1.111438679245283,
      "grad_norm": 8.6643705368042,
      "learning_rate": 5.554245283018869e-06,
      "loss": 0.4804,
      "num_input_tokens_seen": 1235832,
      "step": 1885
    },
    {
      "epoch": 1.1143867924528301,
      "grad_norm": 3.163635492324829,
      "learning_rate": 5.568985849056604e-06,
      "loss": 0.5569,
      "num_input_tokens_seen": 1238712,
      "step": 1890
    },
    {
      "epoch": 1.1173349056603774,
      "grad_norm": 8.346482276916504,
      "learning_rate": 5.58372641509434e-06,
      "loss": 0.523,
      "num_input_tokens_seen": 1241912,
      "step": 1895
    },
    {
      "epoch": 1.1202830188679245,
      "grad_norm": 1.8432999849319458,
      "learning_rate": 5.5984669811320755e-06,
      "loss": 0.563,
      "num_input_tokens_seen": 1244984,
      "step": 1900
    },
    {
      "epoch": 1.1232311320754718,
      "grad_norm": 5.057977199554443,
      "learning_rate": 5.613207547169813e-06,
      "loss": 0.552,
      "num_input_tokens_seen": 1249080,
      "step": 1905
    },
    {
      "epoch": 1.1261792452830188,
      "grad_norm": 3.234251022338867,
      "learning_rate": 5.627948113207548e-06,
      "loss": 0.568,
      "num_input_tokens_seen": 1252184,
      "step": 1910
    },
    {
      "epoch": 1.1291273584905661,
      "grad_norm": 3.542262554168701,
      "learning_rate": 5.642688679245284e-06,
      "loss": 0.5306,
      "num_input_tokens_seen": 1255832,
      "step": 1915
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 1.9764870405197144,
      "learning_rate": 5.6574292452830195e-06,
      "loss": 0.4954,
      "num_input_tokens_seen": 1260216,
      "step": 1920
    },
    {
      "epoch": 1.1350235849056605,
      "grad_norm": 1.8039201498031616,
      "learning_rate": 5.672169811320756e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 1263928,
      "step": 1925
    },
    {
      "epoch": 1.1379716981132075,
      "grad_norm": 2.514012098312378,
      "learning_rate": 5.686910377358491e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 1267800,
      "step": 1930
    },
    {
      "epoch": 1.1409198113207548,
      "grad_norm": 2.162351608276367,
      "learning_rate": 5.701650943396226e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 1270712,
      "step": 1935
    },
    {
      "epoch": 1.1438679245283019,
      "grad_norm": 2.157383918762207,
      "learning_rate": 5.7163915094339625e-06,
      "loss": 0.5493,
      "num_input_tokens_seen": 1274008,
      "step": 1940
    },
    {
      "epoch": 1.146816037735849,
      "grad_norm": 2.9947047233581543,
      "learning_rate": 5.731132075471698e-06,
      "loss": 0.548,
      "num_input_tokens_seen": 1276152,
      "step": 1945
    },
    {
      "epoch": 1.1497641509433962,
      "grad_norm": 4.198831081390381,
      "learning_rate": 5.745872641509435e-06,
      "loss": 0.5027,
      "num_input_tokens_seen": 1279064,
      "step": 1950
    },
    {
      "epoch": 1.1527122641509433,
      "grad_norm": 1.9307420253753662,
      "learning_rate": 5.76061320754717e-06,
      "loss": 0.5612,
      "num_input_tokens_seen": 1282264,
      "step": 1955
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 2.0242648124694824,
      "learning_rate": 5.7753537735849065e-06,
      "loss": 0.5651,
      "num_input_tokens_seen": 1285816,
      "step": 1960
    },
    {
      "epoch": 1.1586084905660377,
      "grad_norm": 3.2993099689483643,
      "learning_rate": 5.790094339622642e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 1288280,
      "step": 1965
    },
    {
      "epoch": 1.161556603773585,
      "grad_norm": 2.8574838638305664,
      "learning_rate": 5.804834905660378e-06,
      "loss": 0.6233,
      "num_input_tokens_seen": 1291256,
      "step": 1970
    },
    {
      "epoch": 1.164504716981132,
      "grad_norm": 2.2204315662384033,
      "learning_rate": 5.819575471698113e-06,
      "loss": 0.4851,
      "num_input_tokens_seen": 1294072,
      "step": 1975
    },
    {
      "epoch": 1.1674528301886793,
      "grad_norm": 5.09334135055542,
      "learning_rate": 5.8343160377358495e-06,
      "loss": 0.8181,
      "num_input_tokens_seen": 1298488,
      "step": 1980
    },
    {
      "epoch": 1.1704009433962264,
      "grad_norm": 3.0587069988250732,
      "learning_rate": 5.849056603773585e-06,
      "loss": 0.4212,
      "num_input_tokens_seen": 1301912,
      "step": 1985
    },
    {
      "epoch": 1.1733490566037736,
      "grad_norm": 3.878511905670166,
      "learning_rate": 5.863797169811322e-06,
      "loss": 0.4772,
      "num_input_tokens_seen": 1304312,
      "step": 1990
    },
    {
      "epoch": 1.1762971698113207,
      "grad_norm": 1.8027515411376953,
      "learning_rate": 5.878537735849057e-06,
      "loss": 0.553,
      "num_input_tokens_seen": 1307736,
      "step": 1995
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 1.8634611368179321,
      "learning_rate": 5.8932783018867934e-06,
      "loss": 0.5978,
      "num_input_tokens_seen": 1310904,
      "step": 2000
    },
    {
      "epoch": 1.182193396226415,
      "grad_norm": 3.388397455215454,
      "learning_rate": 5.908018867924529e-06,
      "loss": 0.457,
      "num_input_tokens_seen": 1313208,
      "step": 2005
    },
    {
      "epoch": 1.1851415094339623,
      "grad_norm": 4.648462772369385,
      "learning_rate": 5.922759433962265e-06,
      "loss": 0.5635,
      "num_input_tokens_seen": 1316888,
      "step": 2010
    },
    {
      "epoch": 1.1880896226415094,
      "grad_norm": 3.370392084121704,
      "learning_rate": 5.9375e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 1320120,
      "step": 2015
    },
    {
      "epoch": 1.1910377358490567,
      "grad_norm": 0.5621175169944763,
      "learning_rate": 5.952240566037736e-06,
      "loss": 0.4259,
      "num_input_tokens_seen": 1326168,
      "step": 2020
    },
    {
      "epoch": 1.1939858490566038,
      "grad_norm": 11.912391662597656,
      "learning_rate": 5.966981132075472e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 1329144,
      "step": 2025
    },
    {
      "epoch": 1.196933962264151,
      "grad_norm": 3.366612195968628,
      "learning_rate": 5.981721698113207e-06,
      "loss": 0.4891,
      "num_input_tokens_seen": 1331896,
      "step": 2030
    },
    {
      "epoch": 1.1998820754716981,
      "grad_norm": 2.513620376586914,
      "learning_rate": 5.996462264150944e-06,
      "loss": 0.6328,
      "num_input_tokens_seen": 1335160,
      "step": 2035
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 3.8589208126068115,
      "learning_rate": 6.01120283018868e-06,
      "loss": 0.594,
      "num_input_tokens_seen": 1337848,
      "step": 2040
    },
    {
      "epoch": 1.2057783018867925,
      "grad_norm": 10.620171546936035,
      "learning_rate": 6.025943396226416e-06,
      "loss": 0.5915,
      "num_input_tokens_seen": 1341464,
      "step": 2045
    },
    {
      "epoch": 1.2087264150943395,
      "grad_norm": 2.689403772354126,
      "learning_rate": 6.040683962264151e-06,
      "loss": 0.6467,
      "num_input_tokens_seen": 1344440,
      "step": 2050
    },
    {
      "epoch": 1.2116745283018868,
      "grad_norm": 1.8132679462432861,
      "learning_rate": 6.055424528301887e-06,
      "loss": 0.6749,
      "num_input_tokens_seen": 1347064,
      "step": 2055
    },
    {
      "epoch": 1.2146226415094339,
      "grad_norm": 7.4044718742370605,
      "learning_rate": 6.070165094339623e-06,
      "loss": 0.5188,
      "num_input_tokens_seen": 1349528,
      "step": 2060
    },
    {
      "epoch": 1.2175707547169812,
      "grad_norm": 3.0039288997650146,
      "learning_rate": 6.08490566037736e-06,
      "loss": 0.5857,
      "num_input_tokens_seen": 1352760,
      "step": 2065
    },
    {
      "epoch": 1.2205188679245282,
      "grad_norm": 1.665089726448059,
      "learning_rate": 6.099646226415095e-06,
      "loss": 0.5042,
      "num_input_tokens_seen": 1355512,
      "step": 2070
    },
    {
      "epoch": 1.2234669811320755,
      "grad_norm": 3.6786301136016846,
      "learning_rate": 6.114386792452831e-06,
      "loss": 0.5658,
      "num_input_tokens_seen": 1358328,
      "step": 2075
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 4.06464147567749,
      "learning_rate": 6.129127358490567e-06,
      "loss": 0.5727,
      "num_input_tokens_seen": 1361784,
      "step": 2080
    },
    {
      "epoch": 1.2293632075471699,
      "grad_norm": 2.858006238937378,
      "learning_rate": 6.143867924528303e-06,
      "loss": 0.5784,
      "num_input_tokens_seen": 1364568,
      "step": 2085
    },
    {
      "epoch": 1.232311320754717,
      "grad_norm": 1.9674391746520996,
      "learning_rate": 6.158608490566038e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 1367384,
      "step": 2090
    },
    {
      "epoch": 1.2352594339622642,
      "grad_norm": 2.994668960571289,
      "learning_rate": 6.173349056603774e-06,
      "loss": 0.5664,
      "num_input_tokens_seen": 1370936,
      "step": 2095
    },
    {
      "epoch": 1.2382075471698113,
      "grad_norm": 3.9618070125579834,
      "learning_rate": 6.18808962264151e-06,
      "loss": 0.5463,
      "num_input_tokens_seen": 1374392,
      "step": 2100
    },
    {
      "epoch": 1.2411556603773586,
      "grad_norm": 3.5432379245758057,
      "learning_rate": 6.202830188679245e-06,
      "loss": 0.6779,
      "num_input_tokens_seen": 1377656,
      "step": 2105
    },
    {
      "epoch": 1.2441037735849056,
      "grad_norm": 2.4366848468780518,
      "learning_rate": 6.217570754716982e-06,
      "loss": 0.4946,
      "num_input_tokens_seen": 1380792,
      "step": 2110
    },
    {
      "epoch": 1.2470518867924527,
      "grad_norm": 3.412060499191284,
      "learning_rate": 6.232311320754717e-06,
      "loss": 0.3988,
      "num_input_tokens_seen": 1384184,
      "step": 2115
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5968751907348633,
      "learning_rate": 6.2470518867924536e-06,
      "loss": 0.5362,
      "num_input_tokens_seen": 1388824,
      "step": 2120
    },
    {
      "epoch": 1.2529481132075473,
      "grad_norm": 1.5202233791351318,
      "learning_rate": 6.261792452830189e-06,
      "loss": 0.6328,
      "num_input_tokens_seen": 1392728,
      "step": 2125
    },
    {
      "epoch": 1.2558962264150944,
      "grad_norm": 1.4195961952209473,
      "learning_rate": 6.276533018867925e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 1397016,
      "step": 2130
    },
    {
      "epoch": 1.2588443396226414,
      "grad_norm": 3.664968490600586,
      "learning_rate": 6.2912735849056604e-06,
      "loss": 0.5167,
      "num_input_tokens_seen": 1400824,
      "step": 2135
    },
    {
      "epoch": 1.2617924528301887,
      "grad_norm": 2.1425211429595947,
      "learning_rate": 6.306014150943397e-06,
      "loss": 0.516,
      "num_input_tokens_seen": 1404856,
      "step": 2140
    },
    {
      "epoch": 1.2647405660377358,
      "grad_norm": 7.303549766540527,
      "learning_rate": 6.320754716981132e-06,
      "loss": 0.6553,
      "num_input_tokens_seen": 1407864,
      "step": 2145
    },
    {
      "epoch": 1.267688679245283,
      "grad_norm": 5.857394695281982,
      "learning_rate": 6.335495283018869e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 1411192,
      "step": 2150
    },
    {
      "epoch": 1.2706367924528301,
      "grad_norm": 2.5853240489959717,
      "learning_rate": 6.350235849056604e-06,
      "loss": 0.505,
      "num_input_tokens_seen": 1414136,
      "step": 2155
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 2.362687587738037,
      "learning_rate": 6.3649764150943406e-06,
      "loss": 0.5352,
      "num_input_tokens_seen": 1417112,
      "step": 2160
    },
    {
      "epoch": 1.2765330188679245,
      "grad_norm": 4.412456035614014,
      "learning_rate": 6.379716981132076e-06,
      "loss": 0.5976,
      "num_input_tokens_seen": 1419960,
      "step": 2165
    },
    {
      "epoch": 1.2794811320754718,
      "grad_norm": 3.987879753112793,
      "learning_rate": 6.394457547169812e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 1422968,
      "step": 2170
    },
    {
      "epoch": 1.2824292452830188,
      "grad_norm": 2.9699628353118896,
      "learning_rate": 6.4091981132075474e-06,
      "loss": 0.6564,
      "num_input_tokens_seen": 1426072,
      "step": 2175
    },
    {
      "epoch": 1.2853773584905661,
      "grad_norm": 1.7707409858703613,
      "learning_rate": 6.423938679245284e-06,
      "loss": 0.4747,
      "num_input_tokens_seen": 1428824,
      "step": 2180
    },
    {
      "epoch": 1.2883254716981132,
      "grad_norm": 4.332728385925293,
      "learning_rate": 6.438679245283019e-06,
      "loss": 0.5505,
      "num_input_tokens_seen": 1431192,
      "step": 2185
    },
    {
      "epoch": 1.2912735849056602,
      "grad_norm": 5.0483012199401855,
      "learning_rate": 6.453419811320756e-06,
      "loss": 0.4477,
      "num_input_tokens_seen": 1434456,
      "step": 2190
    },
    {
      "epoch": 1.2942216981132075,
      "grad_norm": 2.6604371070861816,
      "learning_rate": 6.468160377358491e-06,
      "loss": 0.521,
      "num_input_tokens_seen": 1438488,
      "step": 2195
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 6.308269500732422,
      "learning_rate": 6.482900943396227e-06,
      "loss": 0.5384,
      "num_input_tokens_seen": 1441592,
      "step": 2200
    },
    {
      "epoch": 1.3001179245283019,
      "grad_norm": 1.853966236114502,
      "learning_rate": 6.497641509433963e-06,
      "loss": 0.4994,
      "num_input_tokens_seen": 1445496,
      "step": 2205
    },
    {
      "epoch": 1.303066037735849,
      "grad_norm": 1.7769166231155396,
      "learning_rate": 6.512382075471698e-06,
      "loss": 0.6806,
      "num_input_tokens_seen": 1448088,
      "step": 2210
    },
    {
      "epoch": 1.3060141509433962,
      "grad_norm": 1.4387259483337402,
      "learning_rate": 6.5271226415094344e-06,
      "loss": 0.5251,
      "num_input_tokens_seen": 1451352,
      "step": 2215
    },
    {
      "epoch": 1.3089622641509435,
      "grad_norm": 2.946774959564209,
      "learning_rate": 6.54186320754717e-06,
      "loss": 0.4453,
      "num_input_tokens_seen": 1454648,
      "step": 2220
    },
    {
      "epoch": 1.3119103773584906,
      "grad_norm": 12.483369827270508,
      "learning_rate": 6.556603773584907e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 1457400,
      "step": 2225
    },
    {
      "epoch": 1.3148584905660377,
      "grad_norm": 3.435265064239502,
      "learning_rate": 6.571344339622641e-06,
      "loss": 0.5183,
      "num_input_tokens_seen": 1461016,
      "step": 2230
    },
    {
      "epoch": 1.317806603773585,
      "grad_norm": 4.387156963348389,
      "learning_rate": 6.586084905660378e-06,
      "loss": 0.5803,
      "num_input_tokens_seen": 1463736,
      "step": 2235
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 2.8423311710357666,
      "learning_rate": 6.600825471698114e-06,
      "loss": 0.6249,
      "num_input_tokens_seen": 1466552,
      "step": 2240
    },
    {
      "epoch": 1.3237028301886793,
      "grad_norm": 3.1971852779388428,
      "learning_rate": 6.61556603773585e-06,
      "loss": 0.4646,
      "num_input_tokens_seen": 1469944,
      "step": 2245
    },
    {
      "epoch": 1.3266509433962264,
      "grad_norm": 1.8936026096343994,
      "learning_rate": 6.630306603773585e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 1473816,
      "step": 2250
    },
    {
      "epoch": 1.3295990566037736,
      "grad_norm": 2.1839864253997803,
      "learning_rate": 6.645047169811321e-06,
      "loss": 0.3847,
      "num_input_tokens_seen": 1477016,
      "step": 2255
    },
    {
      "epoch": 1.3325471698113207,
      "grad_norm": 2.6927084922790527,
      "learning_rate": 6.659787735849057e-06,
      "loss": 0.5613,
      "num_input_tokens_seen": 1479768,
      "step": 2260
    },
    {
      "epoch": 1.335495283018868,
      "grad_norm": 5.430628299713135,
      "learning_rate": 6.674528301886794e-06,
      "loss": 0.4635,
      "num_input_tokens_seen": 1482904,
      "step": 2265
    },
    {
      "epoch": 1.338443396226415,
      "grad_norm": 2.3409337997436523,
      "learning_rate": 6.689268867924529e-06,
      "loss": 0.4605,
      "num_input_tokens_seen": 1485752,
      "step": 2270
    },
    {
      "epoch": 1.3413915094339623,
      "grad_norm": 2.2950103282928467,
      "learning_rate": 6.704009433962265e-06,
      "loss": 0.4532,
      "num_input_tokens_seen": 1489400,
      "step": 2275
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 4.463829517364502,
      "learning_rate": 6.718750000000001e-06,
      "loss": 0.4467,
      "num_input_tokens_seen": 1492472,
      "step": 2280
    },
    {
      "epoch": 1.3472877358490565,
      "grad_norm": 1.8403592109680176,
      "learning_rate": 6.733490566037736e-06,
      "loss": 0.5131,
      "num_input_tokens_seen": 1494872,
      "step": 2285
    },
    {
      "epoch": 1.3502358490566038,
      "grad_norm": 2.5396196842193604,
      "learning_rate": 6.748231132075472e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 1498712,
      "step": 2290
    },
    {
      "epoch": 1.353183962264151,
      "grad_norm": 5.736464977264404,
      "learning_rate": 6.7629716981132076e-06,
      "loss": 0.4812,
      "num_input_tokens_seen": 1501688,
      "step": 2295
    },
    {
      "epoch": 1.3561320754716981,
      "grad_norm": 3.174868583679199,
      "learning_rate": 6.777712264150944e-06,
      "loss": 0.6799,
      "num_input_tokens_seen": 1505208,
      "step": 2300
    },
    {
      "epoch": 1.3590801886792452,
      "grad_norm": 2.451435089111328,
      "learning_rate": 6.792452830188679e-06,
      "loss": 0.476,
      "num_input_tokens_seen": 1508024,
      "step": 2305
    },
    {
      "epoch": 1.3620283018867925,
      "grad_norm": 85.60828399658203,
      "learning_rate": 6.807193396226416e-06,
      "loss": 0.5381,
      "num_input_tokens_seen": 1510616,
      "step": 2310
    },
    {
      "epoch": 1.3649764150943398,
      "grad_norm": 4.529104232788086,
      "learning_rate": 6.8219339622641515e-06,
      "loss": 0.5126,
      "num_input_tokens_seen": 1514104,
      "step": 2315
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 3.615821599960327,
      "learning_rate": 6.836674528301888e-06,
      "loss": 0.6415,
      "num_input_tokens_seen": 1517240,
      "step": 2320
    },
    {
      "epoch": 1.3708726415094339,
      "grad_norm": 3.6541452407836914,
      "learning_rate": 6.851415094339623e-06,
      "loss": 0.4983,
      "num_input_tokens_seen": 1521464,
      "step": 2325
    },
    {
      "epoch": 1.3738207547169812,
      "grad_norm": 5.30270528793335,
      "learning_rate": 6.866155660377359e-06,
      "loss": 0.6036,
      "num_input_tokens_seen": 1524536,
      "step": 2330
    },
    {
      "epoch": 1.3767688679245282,
      "grad_norm": 4.258907794952393,
      "learning_rate": 6.8808962264150946e-06,
      "loss": 0.5816,
      "num_input_tokens_seen": 1527608,
      "step": 2335
    },
    {
      "epoch": 1.3797169811320755,
      "grad_norm": 2.518812656402588,
      "learning_rate": 6.895636792452831e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 1530616,
      "step": 2340
    },
    {
      "epoch": 1.3826650943396226,
      "grad_norm": 2.3330512046813965,
      "learning_rate": 6.910377358490566e-06,
      "loss": 0.5375,
      "num_input_tokens_seen": 1535512,
      "step": 2345
    },
    {
      "epoch": 1.3856132075471699,
      "grad_norm": 1.5442872047424316,
      "learning_rate": 6.925117924528303e-06,
      "loss": 0.5482,
      "num_input_tokens_seen": 1538456,
      "step": 2350
    },
    {
      "epoch": 1.388561320754717,
      "grad_norm": 2.3406825065612793,
      "learning_rate": 6.9398584905660385e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 1541880,
      "step": 2355
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 4.331757545471191,
      "learning_rate": 6.954599056603775e-06,
      "loss": 0.5307,
      "num_input_tokens_seen": 1544664,
      "step": 2360
    },
    {
      "epoch": 1.3944575471698113,
      "grad_norm": 9.614628791809082,
      "learning_rate": 6.96933962264151e-06,
      "loss": 0.5082,
      "num_input_tokens_seen": 1548120,
      "step": 2365
    },
    {
      "epoch": 1.3974056603773586,
      "grad_norm": 3.876943826675415,
      "learning_rate": 6.984080188679245e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 1551096,
      "step": 2370
    },
    {
      "epoch": 1.4003537735849056,
      "grad_norm": 2.138936996459961,
      "learning_rate": 6.9988207547169815e-06,
      "loss": 0.539,
      "num_input_tokens_seen": 1554360,
      "step": 2375
    },
    {
      "epoch": 1.4033018867924527,
      "grad_norm": 1.9101049900054932,
      "learning_rate": 7.013561320754717e-06,
      "loss": 0.6404,
      "num_input_tokens_seen": 1557240,
      "step": 2380
    },
    {
      "epoch": 1.40625,
      "grad_norm": 3.143528699874878,
      "learning_rate": 7.028301886792454e-06,
      "loss": 0.5188,
      "num_input_tokens_seen": 1560920,
      "step": 2385
    },
    {
      "epoch": 1.4091981132075473,
      "grad_norm": 6.831539154052734,
      "learning_rate": 7.043042452830188e-06,
      "loss": 0.5962,
      "num_input_tokens_seen": 1563288,
      "step": 2390
    },
    {
      "epoch": 1.4121462264150944,
      "grad_norm": 2.2812955379486084,
      "learning_rate": 7.0577830188679255e-06,
      "loss": 0.6344,
      "num_input_tokens_seen": 1566296,
      "step": 2395
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 4.695439338684082,
      "learning_rate": 7.072523584905661e-06,
      "loss": 0.7386,
      "num_input_tokens_seen": 1569752,
      "step": 2400
    },
    {
      "epoch": 1.4180424528301887,
      "grad_norm": 1.6141105890274048,
      "learning_rate": 7.087264150943397e-06,
      "loss": 0.6608,
      "num_input_tokens_seen": 1572536,
      "step": 2405
    },
    {
      "epoch": 1.4209905660377358,
      "grad_norm": 1.252855896949768,
      "learning_rate": 7.102004716981132e-06,
      "loss": 0.4359,
      "num_input_tokens_seen": 1575704,
      "step": 2410
    },
    {
      "epoch": 1.423938679245283,
      "grad_norm": 2.3793206214904785,
      "learning_rate": 7.1167452830188685e-06,
      "loss": 0.4586,
      "num_input_tokens_seen": 1578680,
      "step": 2415
    },
    {
      "epoch": 1.4268867924528301,
      "grad_norm": 3.742922306060791,
      "learning_rate": 7.131485849056604e-06,
      "loss": 0.4817,
      "num_input_tokens_seen": 1581752,
      "step": 2420
    },
    {
      "epoch": 1.4298349056603774,
      "grad_norm": 3.730398654937744,
      "learning_rate": 7.146226415094341e-06,
      "loss": 0.5478,
      "num_input_tokens_seen": 1585016,
      "step": 2425
    },
    {
      "epoch": 1.4327830188679245,
      "grad_norm": 2.5380873680114746,
      "learning_rate": 7.160966981132076e-06,
      "loss": 0.4824,
      "num_input_tokens_seen": 1588440,
      "step": 2430
    },
    {
      "epoch": 1.4357311320754718,
      "grad_norm": 2.7754108905792236,
      "learning_rate": 7.1757075471698125e-06,
      "loss": 0.4705,
      "num_input_tokens_seen": 1591160,
      "step": 2435
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 1.941468358039856,
      "learning_rate": 7.190448113207548e-06,
      "loss": 0.5324,
      "num_input_tokens_seen": 1594712,
      "step": 2440
    },
    {
      "epoch": 1.4416273584905661,
      "grad_norm": 2.8883910179138184,
      "learning_rate": 7.205188679245284e-06,
      "loss": 0.6095,
      "num_input_tokens_seen": 1598168,
      "step": 2445
    },
    {
      "epoch": 1.4445754716981132,
      "grad_norm": 3.1025588512420654,
      "learning_rate": 7.219929245283019e-06,
      "loss": 0.5994,
      "num_input_tokens_seen": 1601752,
      "step": 2450
    },
    {
      "epoch": 1.4475235849056602,
      "grad_norm": 3.1024599075317383,
      "learning_rate": 7.2346698113207555e-06,
      "loss": 0.4669,
      "num_input_tokens_seen": 1604248,
      "step": 2455
    },
    {
      "epoch": 1.4504716981132075,
      "grad_norm": 4.258625030517578,
      "learning_rate": 7.249410377358491e-06,
      "loss": 0.6597,
      "num_input_tokens_seen": 1607352,
      "step": 2460
    },
    {
      "epoch": 1.4534198113207548,
      "grad_norm": 1.4235440492630005,
      "learning_rate": 7.264150943396226e-06,
      "loss": 0.5316,
      "num_input_tokens_seen": 1610264,
      "step": 2465
    },
    {
      "epoch": 1.4563679245283019,
      "grad_norm": 3.493948459625244,
      "learning_rate": 7.278891509433963e-06,
      "loss": 0.663,
      "num_input_tokens_seen": 1613144,
      "step": 2470
    },
    {
      "epoch": 1.459316037735849,
      "grad_norm": 2.1899642944335938,
      "learning_rate": 7.293632075471699e-06,
      "loss": 0.4705,
      "num_input_tokens_seen": 1619448,
      "step": 2475
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 4.526381969451904,
      "learning_rate": 7.308372641509435e-06,
      "loss": 0.509,
      "num_input_tokens_seen": 1622520,
      "step": 2480
    },
    {
      "epoch": 1.4652122641509435,
      "grad_norm": 4.537275314331055,
      "learning_rate": 7.32311320754717e-06,
      "loss": 0.6943,
      "num_input_tokens_seen": 1625592,
      "step": 2485
    },
    {
      "epoch": 1.4681603773584906,
      "grad_norm": 1.7690509557724,
      "learning_rate": 7.337853773584906e-06,
      "loss": 0.5361,
      "num_input_tokens_seen": 1628760,
      "step": 2490
    },
    {
      "epoch": 1.4711084905660377,
      "grad_norm": 2.538558006286621,
      "learning_rate": 7.352594339622642e-06,
      "loss": 0.5753,
      "num_input_tokens_seen": 1631576,
      "step": 2495
    },
    {
      "epoch": 1.474056603773585,
      "grad_norm": 4.208502292633057,
      "learning_rate": 7.367334905660378e-06,
      "loss": 0.5677,
      "num_input_tokens_seen": 1634584,
      "step": 2500
    },
    {
      "epoch": 1.477004716981132,
      "grad_norm": 5.001977920532227,
      "learning_rate": 7.382075471698113e-06,
      "loss": 0.5057,
      "num_input_tokens_seen": 1637304,
      "step": 2505
    },
    {
      "epoch": 1.4799528301886793,
      "grad_norm": 3.8800995349884033,
      "learning_rate": 7.39681603773585e-06,
      "loss": 0.8169,
      "num_input_tokens_seen": 1639640,
      "step": 2510
    },
    {
      "epoch": 1.4829009433962264,
      "grad_norm": 2.8323581218719482,
      "learning_rate": 7.411556603773586e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 1642840,
      "step": 2515
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 2.5109667778015137,
      "learning_rate": 7.426297169811322e-06,
      "loss": 0.7222,
      "num_input_tokens_seen": 1646936,
      "step": 2520
    },
    {
      "epoch": 1.4887971698113207,
      "grad_norm": 1.8484303951263428,
      "learning_rate": 7.441037735849057e-06,
      "loss": 0.7112,
      "num_input_tokens_seen": 1651256,
      "step": 2525
    },
    {
      "epoch": 1.491745283018868,
      "grad_norm": 3.8209269046783447,
      "learning_rate": 7.455778301886793e-06,
      "loss": 0.6471,
      "num_input_tokens_seen": 1654456,
      "step": 2530
    },
    {
      "epoch": 1.494693396226415,
      "grad_norm": 3.0384907722473145,
      "learning_rate": 7.470518867924529e-06,
      "loss": 0.4948,
      "num_input_tokens_seen": 1657176,
      "step": 2535
    },
    {
      "epoch": 1.4976415094339623,
      "grad_norm": 2.4049746990203857,
      "learning_rate": 7.485259433962266e-06,
      "loss": 0.3534,
      "num_input_tokens_seen": 1659416,
      "step": 2540
    },
    {
      "epoch": 1.5005896226415094,
      "grad_norm": 3.772858142852783,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.4647,
      "num_input_tokens_seen": 1662552,
      "step": 2545
    },
    {
      "epoch": 1.5035377358490565,
      "grad_norm": 2.7028937339782715,
      "learning_rate": 7.5147405660377355e-06,
      "loss": 0.4586,
      "num_input_tokens_seen": 1665592,
      "step": 2550
    },
    {
      "epoch": 1.5064858490566038,
      "grad_norm": 6.704731464385986,
      "learning_rate": 7.5294811320754726e-06,
      "loss": 0.372,
      "num_input_tokens_seen": 1667992,
      "step": 2555
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 6.4977312088012695,
      "learning_rate": 7.544221698113208e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 1670840,
      "step": 2560
    },
    {
      "epoch": 1.5123820754716981,
      "grad_norm": 3.440685510635376,
      "learning_rate": 7.558962264150944e-06,
      "loss": 0.5543,
      "num_input_tokens_seen": 1674008,
      "step": 2565
    },
    {
      "epoch": 1.5153301886792452,
      "grad_norm": 1.7390730381011963,
      "learning_rate": 7.5737028301886795e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 1679192,
      "step": 2570
    },
    {
      "epoch": 1.5182783018867925,
      "grad_norm": 1.9058500528335571,
      "learning_rate": 7.588443396226416e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 1682808,
      "step": 2575
    },
    {
      "epoch": 1.5212264150943398,
      "grad_norm": 2.015550374984741,
      "learning_rate": 7.603183962264151e-06,
      "loss": 0.4815,
      "num_input_tokens_seen": 1687352,
      "step": 2580
    },
    {
      "epoch": 1.5241745283018868,
      "grad_norm": 7.000829219818115,
      "learning_rate": 7.617924528301888e-06,
      "loss": 0.6118,
      "num_input_tokens_seen": 1689784,
      "step": 2585
    },
    {
      "epoch": 1.5271226415094339,
      "grad_norm": 3.3565592765808105,
      "learning_rate": 7.632665094339623e-06,
      "loss": 0.6919,
      "num_input_tokens_seen": 1695192,
      "step": 2590
    },
    {
      "epoch": 1.5300707547169812,
      "grad_norm": 9.03105354309082,
      "learning_rate": 7.64740566037736e-06,
      "loss": 0.5491,
      "num_input_tokens_seen": 1698168,
      "step": 2595
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 3.7289204597473145,
      "learning_rate": 7.662146226415095e-06,
      "loss": 0.5908,
      "num_input_tokens_seen": 1701624,
      "step": 2600
    },
    {
      "epoch": 1.5359669811320755,
      "grad_norm": 1.7822237014770508,
      "learning_rate": 7.676886792452832e-06,
      "loss": 0.5148,
      "num_input_tokens_seen": 1705080,
      "step": 2605
    },
    {
      "epoch": 1.5389150943396226,
      "grad_norm": 2.045868158340454,
      "learning_rate": 7.691627358490567e-06,
      "loss": 0.6323,
      "num_input_tokens_seen": 1708280,
      "step": 2610
    },
    {
      "epoch": 1.5418632075471699,
      "grad_norm": 1.643282413482666,
      "learning_rate": 7.706367924528303e-06,
      "loss": 0.5338,
      "num_input_tokens_seen": 1711640,
      "step": 2615
    },
    {
      "epoch": 1.544811320754717,
      "grad_norm": 3.2912795543670654,
      "learning_rate": 7.721108490566038e-06,
      "loss": 0.4385,
      "num_input_tokens_seen": 1714360,
      "step": 2620
    },
    {
      "epoch": 1.547759433962264,
      "grad_norm": 4.204911231994629,
      "learning_rate": 7.735849056603775e-06,
      "loss": 0.4699,
      "num_input_tokens_seen": 1717112,
      "step": 2625
    },
    {
      "epoch": 1.5507075471698113,
      "grad_norm": 2.813568592071533,
      "learning_rate": 7.75058962264151e-06,
      "loss": 0.5202,
      "num_input_tokens_seen": 1720664,
      "step": 2630
    },
    {
      "epoch": 1.5536556603773586,
      "grad_norm": 5.31522274017334,
      "learning_rate": 7.765330188679246e-06,
      "loss": 0.5821,
      "num_input_tokens_seen": 1723800,
      "step": 2635
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 27.995880126953125,
      "learning_rate": 7.780070754716981e-06,
      "loss": 0.821,
      "num_input_tokens_seen": 1726296,
      "step": 2640
    },
    {
      "epoch": 1.5595518867924527,
      "grad_norm": 6.026430606842041,
      "learning_rate": 7.794811320754716e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 1730424,
      "step": 2645
    },
    {
      "epoch": 1.5625,
      "grad_norm": 1.8776435852050781,
      "learning_rate": 7.809551886792453e-06,
      "loss": 0.544,
      "num_input_tokens_seen": 1733176,
      "step": 2650
    },
    {
      "epoch": 1.5654481132075473,
      "grad_norm": 2.2927591800689697,
      "learning_rate": 7.824292452830189e-06,
      "loss": 0.5237,
      "num_input_tokens_seen": 1735992,
      "step": 2655
    },
    {
      "epoch": 1.5683962264150944,
      "grad_norm": 1.5075405836105347,
      "learning_rate": 7.839033018867926e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 1739096,
      "step": 2660
    },
    {
      "epoch": 1.5713443396226414,
      "grad_norm": 3.1318039894104004,
      "learning_rate": 7.853773584905661e-06,
      "loss": 0.5593,
      "num_input_tokens_seen": 1746808,
      "step": 2665
    },
    {
      "epoch": 1.5742924528301887,
      "grad_norm": 2.781938314437866,
      "learning_rate": 7.868514150943397e-06,
      "loss": 0.5755,
      "num_input_tokens_seen": 1750552,
      "step": 2670
    },
    {
      "epoch": 1.577240566037736,
      "grad_norm": 3.840240955352783,
      "learning_rate": 7.883254716981132e-06,
      "loss": 0.5568,
      "num_input_tokens_seen": 1753208,
      "step": 2675
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 1.6603620052337646,
      "learning_rate": 7.897995283018869e-06,
      "loss": 0.4838,
      "num_input_tokens_seen": 1756728,
      "step": 2680
    },
    {
      "epoch": 1.5831367924528301,
      "grad_norm": 1.757144808769226,
      "learning_rate": 7.912735849056604e-06,
      "loss": 0.559,
      "num_input_tokens_seen": 1760472,
      "step": 2685
    },
    {
      "epoch": 1.5860849056603774,
      "grad_norm": 14.298271179199219,
      "learning_rate": 7.927476415094341e-06,
      "loss": 0.3578,
      "num_input_tokens_seen": 1763608,
      "step": 2690
    },
    {
      "epoch": 1.5890330188679245,
      "grad_norm": 3.095339059829712,
      "learning_rate": 7.942216981132077e-06,
      "loss": 0.6648,
      "num_input_tokens_seen": 1766552,
      "step": 2695
    },
    {
      "epoch": 1.5919811320754715,
      "grad_norm": 4.237428188323975,
      "learning_rate": 7.956957547169812e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 1769816,
      "step": 2700
    },
    {
      "epoch": 1.5949292452830188,
      "grad_norm": 2.33149790763855,
      "learning_rate": 7.971698113207547e-06,
      "loss": 0.5028,
      "num_input_tokens_seen": 1773016,
      "step": 2705
    },
    {
      "epoch": 1.5978773584905661,
      "grad_norm": 2.4068198204040527,
      "learning_rate": 7.986438679245284e-06,
      "loss": 0.4626,
      "num_input_tokens_seen": 1775672,
      "step": 2710
    },
    {
      "epoch": 1.6008254716981132,
      "grad_norm": 1.7039551734924316,
      "learning_rate": 8.00117924528302e-06,
      "loss": 0.5417,
      "num_input_tokens_seen": 1778680,
      "step": 2715
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 2.834218978881836,
      "learning_rate": 8.015919811320757e-06,
      "loss": 0.6327,
      "num_input_tokens_seen": 1781528,
      "step": 2720
    },
    {
      "epoch": 1.6067216981132075,
      "grad_norm": 3.051351308822632,
      "learning_rate": 8.030660377358492e-06,
      "loss": 0.6686,
      "num_input_tokens_seen": 1784856,
      "step": 2725
    },
    {
      "epoch": 1.6096698113207548,
      "grad_norm": 3.9004366397857666,
      "learning_rate": 8.045400943396227e-06,
      "loss": 0.46,
      "num_input_tokens_seen": 1787384,
      "step": 2730
    },
    {
      "epoch": 1.6126179245283019,
      "grad_norm": 2.878999710083008,
      "learning_rate": 8.060141509433963e-06,
      "loss": 0.6499,
      "num_input_tokens_seen": 1790712,
      "step": 2735
    },
    {
      "epoch": 1.615566037735849,
      "grad_norm": 2.2776169776916504,
      "learning_rate": 8.074882075471698e-06,
      "loss": 0.5316,
      "num_input_tokens_seen": 1793400,
      "step": 2740
    },
    {
      "epoch": 1.6185141509433962,
      "grad_norm": 2.674333095550537,
      "learning_rate": 8.089622641509435e-06,
      "loss": 0.5906,
      "num_input_tokens_seen": 1796824,
      "step": 2745
    },
    {
      "epoch": 1.6214622641509435,
      "grad_norm": 3.226597785949707,
      "learning_rate": 8.10436320754717e-06,
      "loss": 0.5757,
      "num_input_tokens_seen": 1799608,
      "step": 2750
    },
    {
      "epoch": 1.6244103773584906,
      "grad_norm": 5.390800476074219,
      "learning_rate": 8.119103773584906e-06,
      "loss": 0.5158,
      "num_input_tokens_seen": 1803192,
      "step": 2755
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 1.462003469467163,
      "learning_rate": 8.133844339622641e-06,
      "loss": 0.3916,
      "num_input_tokens_seen": 1806808,
      "step": 2760
    },
    {
      "epoch": 1.630306603773585,
      "grad_norm": 2.1907799243927,
      "learning_rate": 8.148584905660378e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 1809816,
      "step": 2765
    },
    {
      "epoch": 1.6332547169811322,
      "grad_norm": 1.3602794408798218,
      "learning_rate": 8.163325471698114e-06,
      "loss": 0.5403,
      "num_input_tokens_seen": 1812536,
      "step": 2770
    },
    {
      "epoch": 1.6362028301886793,
      "grad_norm": 3.425060272216797,
      "learning_rate": 8.17806603773585e-06,
      "loss": 0.4869,
      "num_input_tokens_seen": 1816824,
      "step": 2775
    },
    {
      "epoch": 1.6391509433962264,
      "grad_norm": 2.7038557529449463,
      "learning_rate": 8.192806603773586e-06,
      "loss": 0.5422,
      "num_input_tokens_seen": 1820312,
      "step": 2780
    },
    {
      "epoch": 1.6420990566037736,
      "grad_norm": 2.6223175525665283,
      "learning_rate": 8.207547169811321e-06,
      "loss": 0.4578,
      "num_input_tokens_seen": 1823320,
      "step": 2785
    },
    {
      "epoch": 1.6450471698113207,
      "grad_norm": 1.4594155550003052,
      "learning_rate": 8.222287735849057e-06,
      "loss": 0.5846,
      "num_input_tokens_seen": 1826552,
      "step": 2790
    },
    {
      "epoch": 1.6479952830188678,
      "grad_norm": 2.1669278144836426,
      "learning_rate": 8.237028301886794e-06,
      "loss": 0.6512,
      "num_input_tokens_seen": 1829432,
      "step": 2795
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 4.189577579498291,
      "learning_rate": 8.251768867924529e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 1832536,
      "step": 2800
    },
    {
      "epoch": 1.6538915094339623,
      "grad_norm": 3.5659003257751465,
      "learning_rate": 8.266509433962266e-06,
      "loss": 0.495,
      "num_input_tokens_seen": 1836600,
      "step": 2805
    },
    {
      "epoch": 1.6568396226415094,
      "grad_norm": 3.5314178466796875,
      "learning_rate": 8.281250000000001e-06,
      "loss": 0.5311,
      "num_input_tokens_seen": 1838808,
      "step": 2810
    },
    {
      "epoch": 1.6597877358490565,
      "grad_norm": 1.6554361581802368,
      "learning_rate": 8.295990566037737e-06,
      "loss": 0.4286,
      "num_input_tokens_seen": 1842232,
      "step": 2815
    },
    {
      "epoch": 1.6627358490566038,
      "grad_norm": 2.18696928024292,
      "learning_rate": 8.310731132075472e-06,
      "loss": 0.5835,
      "num_input_tokens_seen": 1846008,
      "step": 2820
    },
    {
      "epoch": 1.665683962264151,
      "grad_norm": 5.996832847595215,
      "learning_rate": 8.325471698113207e-06,
      "loss": 0.6148,
      "num_input_tokens_seen": 1849272,
      "step": 2825
    },
    {
      "epoch": 1.6686320754716981,
      "grad_norm": 2.098137617111206,
      "learning_rate": 8.340212264150944e-06,
      "loss": 0.5432,
      "num_input_tokens_seen": 1852536,
      "step": 2830
    },
    {
      "epoch": 1.6715801886792452,
      "grad_norm": 2.2328109741210938,
      "learning_rate": 8.35495283018868e-06,
      "loss": 0.4587,
      "num_input_tokens_seen": 1857304,
      "step": 2835
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 4.410581588745117,
      "learning_rate": 8.369693396226415e-06,
      "loss": 0.4895,
      "num_input_tokens_seen": 1861400,
      "step": 2840
    },
    {
      "epoch": 1.6774764150943398,
      "grad_norm": 19.479705810546875,
      "learning_rate": 8.38443396226415e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 1864408,
      "step": 2845
    },
    {
      "epoch": 1.6804245283018868,
      "grad_norm": 2.4968621730804443,
      "learning_rate": 8.399174528301888e-06,
      "loss": 0.4526,
      "num_input_tokens_seen": 1867608,
      "step": 2850
    },
    {
      "epoch": 1.6833726415094339,
      "grad_norm": 3.0936625003814697,
      "learning_rate": 8.413915094339623e-06,
      "loss": 0.446,
      "num_input_tokens_seen": 1870264,
      "step": 2855
    },
    {
      "epoch": 1.6863207547169812,
      "grad_norm": 2.780729293823242,
      "learning_rate": 8.42865566037736e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 1873816,
      "step": 2860
    },
    {
      "epoch": 1.6892688679245285,
      "grad_norm": 1.567718744277954,
      "learning_rate": 8.443396226415095e-06,
      "loss": 0.4416,
      "num_input_tokens_seen": 1877112,
      "step": 2865
    },
    {
      "epoch": 1.6922169811320755,
      "grad_norm": 1.7703508138656616,
      "learning_rate": 8.45813679245283e-06,
      "loss": 0.4891,
      "num_input_tokens_seen": 1880216,
      "step": 2870
    },
    {
      "epoch": 1.6951650943396226,
      "grad_norm": 2.4081647396087646,
      "learning_rate": 8.472877358490566e-06,
      "loss": 0.5221,
      "num_input_tokens_seen": 1883000,
      "step": 2875
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 1.5904195308685303,
      "learning_rate": 8.487617924528303e-06,
      "loss": 0.5907,
      "num_input_tokens_seen": 1885752,
      "step": 2880
    },
    {
      "epoch": 1.701061320754717,
      "grad_norm": 3.138120651245117,
      "learning_rate": 8.502358490566038e-06,
      "loss": 0.5243,
      "num_input_tokens_seen": 1889560,
      "step": 2885
    },
    {
      "epoch": 1.704009433962264,
      "grad_norm": 2.7369465827941895,
      "learning_rate": 8.517099056603775e-06,
      "loss": 0.5486,
      "num_input_tokens_seen": 1893016,
      "step": 2890
    },
    {
      "epoch": 1.7069575471698113,
      "grad_norm": 2.139652729034424,
      "learning_rate": 8.53183962264151e-06,
      "loss": 0.4977,
      "num_input_tokens_seen": 1895960,
      "step": 2895
    },
    {
      "epoch": 1.7099056603773586,
      "grad_norm": 1.5185520648956299,
      "learning_rate": 8.546580188679246e-06,
      "loss": 0.4906,
      "num_input_tokens_seen": 1898968,
      "step": 2900
    },
    {
      "epoch": 1.7128537735849056,
      "grad_norm": 2.112961530685425,
      "learning_rate": 8.561320754716981e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 1902360,
      "step": 2905
    },
    {
      "epoch": 1.7158018867924527,
      "grad_norm": 2.66847825050354,
      "learning_rate": 8.576061320754717e-06,
      "loss": 0.5127,
      "num_input_tokens_seen": 1905336,
      "step": 2910
    },
    {
      "epoch": 1.71875,
      "grad_norm": 1.350322961807251,
      "learning_rate": 8.590801886792454e-06,
      "loss": 0.7356,
      "num_input_tokens_seen": 1908888,
      "step": 2915
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 3.732483148574829,
      "learning_rate": 8.605542452830189e-06,
      "loss": 0.5751,
      "num_input_tokens_seen": 1912952,
      "step": 2920
    },
    {
      "epoch": 1.7246462264150944,
      "grad_norm": 1.9927202463150024,
      "learning_rate": 8.620283018867926e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 1916312,
      "step": 2925
    },
    {
      "epoch": 1.7275943396226414,
      "grad_norm": 2.6339309215545654,
      "learning_rate": 8.635023584905662e-06,
      "loss": 0.4054,
      "num_input_tokens_seen": 1919512,
      "step": 2930
    },
    {
      "epoch": 1.7305424528301887,
      "grad_norm": 5.363759994506836,
      "learning_rate": 8.649764150943397e-06,
      "loss": 0.4923,
      "num_input_tokens_seen": 1922392,
      "step": 2935
    },
    {
      "epoch": 1.733490566037736,
      "grad_norm": 3.5594260692596436,
      "learning_rate": 8.664504716981132e-06,
      "loss": 0.5374,
      "num_input_tokens_seen": 1925720,
      "step": 2940
    },
    {
      "epoch": 1.736438679245283,
      "grad_norm": 6.510471820831299,
      "learning_rate": 8.67924528301887e-06,
      "loss": 0.5583,
      "num_input_tokens_seen": 1928600,
      "step": 2945
    },
    {
      "epoch": 1.7393867924528301,
      "grad_norm": 2.0995020866394043,
      "learning_rate": 8.693985849056605e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 1931288,
      "step": 2950
    },
    {
      "epoch": 1.7423349056603774,
      "grad_norm": 3.5186057090759277,
      "learning_rate": 8.70872641509434e-06,
      "loss": 0.6177,
      "num_input_tokens_seen": 1934392,
      "step": 2955
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 2.7785868644714355,
      "learning_rate": 8.723466981132075e-06,
      "loss": 0.3822,
      "num_input_tokens_seen": 1937304,
      "step": 2960
    },
    {
      "epoch": 1.7482311320754715,
      "grad_norm": 1.6513187885284424,
      "learning_rate": 8.738207547169812e-06,
      "loss": 0.4099,
      "num_input_tokens_seen": 1940536,
      "step": 2965
    },
    {
      "epoch": 1.7511792452830188,
      "grad_norm": 2.6186063289642334,
      "learning_rate": 8.752948113207548e-06,
      "loss": 0.5902,
      "num_input_tokens_seen": 1943128,
      "step": 2970
    },
    {
      "epoch": 1.7541273584905661,
      "grad_norm": 1.8135114908218384,
      "learning_rate": 8.767688679245285e-06,
      "loss": 0.5142,
      "num_input_tokens_seen": 1946136,
      "step": 2975
    },
    {
      "epoch": 1.7570754716981132,
      "grad_norm": 3.118189811706543,
      "learning_rate": 8.78242924528302e-06,
      "loss": 0.4953,
      "num_input_tokens_seen": 1948568,
      "step": 2980
    },
    {
      "epoch": 1.7600235849056602,
      "grad_norm": 2.3354995250701904,
      "learning_rate": 8.797169811320755e-06,
      "loss": 0.7807,
      "num_input_tokens_seen": 1951096,
      "step": 2985
    },
    {
      "epoch": 1.7629716981132075,
      "grad_norm": 2.508113384246826,
      "learning_rate": 8.81191037735849e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 1954264,
      "step": 2990
    },
    {
      "epoch": 1.7659198113207548,
      "grad_norm": 2.108161687850952,
      "learning_rate": 8.826650943396226e-06,
      "loss": 0.4684,
      "num_input_tokens_seen": 1957688,
      "step": 2995
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 2.5507664680480957,
      "learning_rate": 8.841391509433963e-06,
      "loss": 0.637,
      "num_input_tokens_seen": 1960600,
      "step": 3000
    },
    {
      "epoch": 1.771816037735849,
      "grad_norm": 2.8044593334198,
      "learning_rate": 8.856132075471698e-06,
      "loss": 0.537,
      "num_input_tokens_seen": 1964088,
      "step": 3005
    },
    {
      "epoch": 1.7747641509433962,
      "grad_norm": 2.6051383018493652,
      "learning_rate": 8.870872641509435e-06,
      "loss": 0.6031,
      "num_input_tokens_seen": 1967320,
      "step": 3010
    },
    {
      "epoch": 1.7777122641509435,
      "grad_norm": 1.2368626594543457,
      "learning_rate": 8.88561320754717e-06,
      "loss": 0.4097,
      "num_input_tokens_seen": 1972152,
      "step": 3015
    },
    {
      "epoch": 1.7806603773584906,
      "grad_norm": 2.7872304916381836,
      "learning_rate": 8.900353773584906e-06,
      "loss": 0.5527,
      "num_input_tokens_seen": 1975960,
      "step": 3020
    },
    {
      "epoch": 1.7836084905660377,
      "grad_norm": 1.5566701889038086,
      "learning_rate": 8.915094339622642e-06,
      "loss": 0.391,
      "num_input_tokens_seen": 1979256,
      "step": 3025
    },
    {
      "epoch": 1.786556603773585,
      "grad_norm": 1.4559507369995117,
      "learning_rate": 8.929834905660379e-06,
      "loss": 0.6963,
      "num_input_tokens_seen": 1982232,
      "step": 3030
    },
    {
      "epoch": 1.7895047169811322,
      "grad_norm": 0.88956218957901,
      "learning_rate": 8.944575471698114e-06,
      "loss": 0.4936,
      "num_input_tokens_seen": 1986808,
      "step": 3035
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 1.8705368041992188,
      "learning_rate": 8.959316037735851e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 1989240,
      "step": 3040
    },
    {
      "epoch": 1.7954009433962264,
      "grad_norm": 3.7223567962646484,
      "learning_rate": 8.974056603773586e-06,
      "loss": 0.5585,
      "num_input_tokens_seen": 1991960,
      "step": 3045
    },
    {
      "epoch": 1.7983490566037736,
      "grad_norm": 4.903166770935059,
      "learning_rate": 8.988797169811322e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 1994584,
      "step": 3050
    },
    {
      "epoch": 1.8012971698113207,
      "grad_norm": 1.5016978979110718,
      "learning_rate": 9.003537735849057e-06,
      "loss": 0.5391,
      "num_input_tokens_seen": 1998328,
      "step": 3055
    },
    {
      "epoch": 1.8042452830188678,
      "grad_norm": 2.640983819961548,
      "learning_rate": 9.018278301886794e-06,
      "loss": 0.7036,
      "num_input_tokens_seen": 2001528,
      "step": 3060
    },
    {
      "epoch": 1.807193396226415,
      "grad_norm": 1.5730777978897095,
      "learning_rate": 9.03301886792453e-06,
      "loss": 0.5111,
      "num_input_tokens_seen": 2004920,
      "step": 3065
    },
    {
      "epoch": 1.8101415094339623,
      "grad_norm": 1.2213295698165894,
      "learning_rate": 9.047759433962265e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 2008760,
      "step": 3070
    },
    {
      "epoch": 1.8130896226415094,
      "grad_norm": 3.5052170753479004,
      "learning_rate": 9.0625e-06,
      "loss": 0.6961,
      "num_input_tokens_seen": 2011448,
      "step": 3075
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 3.0131208896636963,
      "learning_rate": 9.077240566037735e-06,
      "loss": 0.5515,
      "num_input_tokens_seen": 2014936,
      "step": 3080
    },
    {
      "epoch": 1.8189858490566038,
      "grad_norm": 3.7472498416900635,
      "learning_rate": 9.091981132075472e-06,
      "loss": 0.7468,
      "num_input_tokens_seen": 2017848,
      "step": 3085
    },
    {
      "epoch": 1.821933962264151,
      "grad_norm": 6.211084365844727,
      "learning_rate": 9.106721698113208e-06,
      "loss": 0.5752,
      "num_input_tokens_seen": 2021688,
      "step": 3090
    },
    {
      "epoch": 1.8248820754716981,
      "grad_norm": 4.267122268676758,
      "learning_rate": 9.121462264150945e-06,
      "loss": 0.8491,
      "num_input_tokens_seen": 2024760,
      "step": 3095
    },
    {
      "epoch": 1.8278301886792452,
      "grad_norm": 2.0786874294281006,
      "learning_rate": 9.13620283018868e-06,
      "loss": 0.5528,
      "num_input_tokens_seen": 2028152,
      "step": 3100
    },
    {
      "epoch": 1.8307783018867925,
      "grad_norm": 4.450623989105225,
      "learning_rate": 9.150943396226416e-06,
      "loss": 0.4598,
      "num_input_tokens_seen": 2030904,
      "step": 3105
    },
    {
      "epoch": 1.8337264150943398,
      "grad_norm": 3.1928703784942627,
      "learning_rate": 9.165683962264151e-06,
      "loss": 0.5396,
      "num_input_tokens_seen": 2034328,
      "step": 3110
    },
    {
      "epoch": 1.8366745283018868,
      "grad_norm": 1.864582896232605,
      "learning_rate": 9.180424528301888e-06,
      "loss": 0.7763,
      "num_input_tokens_seen": 2037560,
      "step": 3115
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 2.2890994548797607,
      "learning_rate": 9.195165094339623e-06,
      "loss": 0.5452,
      "num_input_tokens_seen": 2041176,
      "step": 3120
    },
    {
      "epoch": 1.8425707547169812,
      "grad_norm": 3.921854257583618,
      "learning_rate": 9.20990566037736e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 2044472,
      "step": 3125
    },
    {
      "epoch": 1.8455188679245285,
      "grad_norm": 3.6671206951141357,
      "learning_rate": 9.224646226415096e-06,
      "loss": 0.5232,
      "num_input_tokens_seen": 2048088,
      "step": 3130
    },
    {
      "epoch": 1.8484669811320755,
      "grad_norm": 2.812647581100464,
      "learning_rate": 9.239386792452831e-06,
      "loss": 0.4834,
      "num_input_tokens_seen": 2051096,
      "step": 3135
    },
    {
      "epoch": 1.8514150943396226,
      "grad_norm": 1.6622456312179565,
      "learning_rate": 9.254127358490566e-06,
      "loss": 0.4923,
      "num_input_tokens_seen": 2054776,
      "step": 3140
    },
    {
      "epoch": 1.8543632075471699,
      "grad_norm": 3.464712142944336,
      "learning_rate": 9.268867924528303e-06,
      "loss": 0.4875,
      "num_input_tokens_seen": 2057784,
      "step": 3145
    },
    {
      "epoch": 1.857311320754717,
      "grad_norm": 1.3763309717178345,
      "learning_rate": 9.283608490566039e-06,
      "loss": 0.4603,
      "num_input_tokens_seen": 2060504,
      "step": 3150
    },
    {
      "epoch": 1.860259433962264,
      "grad_norm": 2.098742723464966,
      "learning_rate": 9.298349056603774e-06,
      "loss": 0.5476,
      "num_input_tokens_seen": 2063864,
      "step": 3155
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 2.0543158054351807,
      "learning_rate": 9.31308962264151e-06,
      "loss": 0.5082,
      "num_input_tokens_seen": 2067448,
      "step": 3160
    },
    {
      "epoch": 1.8661556603773586,
      "grad_norm": 2.135716438293457,
      "learning_rate": 9.327830188679245e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 2071480,
      "step": 3165
    },
    {
      "epoch": 1.8691037735849056,
      "grad_norm": 3.0030810832977295,
      "learning_rate": 9.342570754716982e-06,
      "loss": 0.5498,
      "num_input_tokens_seen": 2074520,
      "step": 3170
    },
    {
      "epoch": 1.8720518867924527,
      "grad_norm": 3.346625804901123,
      "learning_rate": 9.357311320754717e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 2077848,
      "step": 3175
    },
    {
      "epoch": 1.875,
      "grad_norm": 3.7247564792633057,
      "learning_rate": 9.372051886792454e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 2082136,
      "step": 3180
    },
    {
      "epoch": 1.8779481132075473,
      "grad_norm": 4.084072113037109,
      "learning_rate": 9.38679245283019e-06,
      "loss": 0.5789,
      "num_input_tokens_seen": 2084888,
      "step": 3185
    },
    {
      "epoch": 1.8808962264150944,
      "grad_norm": 2.4903945922851562,
      "learning_rate": 9.401533018867925e-06,
      "loss": 0.6103,
      "num_input_tokens_seen": 2090840,
      "step": 3190
    },
    {
      "epoch": 1.8838443396226414,
      "grad_norm": 1.8782507181167603,
      "learning_rate": 9.41627358490566e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 2094904,
      "step": 3195
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 2.4810802936553955,
      "learning_rate": 9.431014150943397e-06,
      "loss": 0.594,
      "num_input_tokens_seen": 2097688,
      "step": 3200
    },
    {
      "epoch": 1.889740566037736,
      "grad_norm": 3.304565906524658,
      "learning_rate": 9.445754716981133e-06,
      "loss": 0.5259,
      "num_input_tokens_seen": 2101176,
      "step": 3205
    },
    {
      "epoch": 1.892688679245283,
      "grad_norm": 3.5922741889953613,
      "learning_rate": 9.46049528301887e-06,
      "loss": 0.5192,
      "num_input_tokens_seen": 2103992,
      "step": 3210
    },
    {
      "epoch": 1.8956367924528301,
      "grad_norm": 1.0054118633270264,
      "learning_rate": 9.475235849056605e-06,
      "loss": 0.5128,
      "num_input_tokens_seen": 2107896,
      "step": 3215
    },
    {
      "epoch": 1.8985849056603774,
      "grad_norm": 1.2783899307250977,
      "learning_rate": 9.48997641509434e-06,
      "loss": 0.5792,
      "num_input_tokens_seen": 2111608,
      "step": 3220
    },
    {
      "epoch": 1.9015330188679245,
      "grad_norm": 1.2470206022262573,
      "learning_rate": 9.504716981132076e-06,
      "loss": 0.5504,
      "num_input_tokens_seen": 2115064,
      "step": 3225
    },
    {
      "epoch": 1.9044811320754715,
      "grad_norm": 2.263420581817627,
      "learning_rate": 9.519457547169813e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 2118040,
      "step": 3230
    },
    {
      "epoch": 1.9074292452830188,
      "grad_norm": 2.3698678016662598,
      "learning_rate": 9.534198113207548e-06,
      "loss": 0.6842,
      "num_input_tokens_seen": 2120760,
      "step": 3235
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 1.6520953178405762,
      "learning_rate": 9.548938679245285e-06,
      "loss": 0.5488,
      "num_input_tokens_seen": 2124536,
      "step": 3240
    },
    {
      "epoch": 1.9133254716981132,
      "grad_norm": 1.208957314491272,
      "learning_rate": 9.56367924528302e-06,
      "loss": 0.5862,
      "num_input_tokens_seen": 2128184,
      "step": 3245
    },
    {
      "epoch": 1.9162735849056602,
      "grad_norm": 2.1613030433654785,
      "learning_rate": 9.578419811320756e-06,
      "loss": 0.4463,
      "num_input_tokens_seen": 2132184,
      "step": 3250
    },
    {
      "epoch": 1.9192216981132075,
      "grad_norm": 2.2012412548065186,
      "learning_rate": 9.593160377358491e-06,
      "loss": 0.4632,
      "num_input_tokens_seen": 2135288,
      "step": 3255
    },
    {
      "epoch": 1.9221698113207548,
      "grad_norm": 7.033367156982422,
      "learning_rate": 9.607900943396226e-06,
      "loss": 0.5715,
      "num_input_tokens_seen": 2138616,
      "step": 3260
    },
    {
      "epoch": 1.9251179245283019,
      "grad_norm": 2.4463605880737305,
      "learning_rate": 9.622641509433963e-06,
      "loss": 0.5413,
      "num_input_tokens_seen": 2141880,
      "step": 3265
    },
    {
      "epoch": 1.928066037735849,
      "grad_norm": 1.4014438390731812,
      "learning_rate": 9.637382075471699e-06,
      "loss": 0.4785,
      "num_input_tokens_seen": 2144664,
      "step": 3270
    },
    {
      "epoch": 1.9310141509433962,
      "grad_norm": 1.9268380403518677,
      "learning_rate": 9.652122641509434e-06,
      "loss": 0.5875,
      "num_input_tokens_seen": 2148184,
      "step": 3275
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 1.5056350231170654,
      "learning_rate": 9.66686320754717e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 2152568,
      "step": 3280
    },
    {
      "epoch": 1.9369103773584906,
      "grad_norm": 4.365038871765137,
      "learning_rate": 9.681603773584907e-06,
      "loss": 0.5098,
      "num_input_tokens_seen": 2155384,
      "step": 3285
    },
    {
      "epoch": 1.9398584905660377,
      "grad_norm": 2.866955518722534,
      "learning_rate": 9.696344339622642e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 2158744,
      "step": 3290
    },
    {
      "epoch": 1.942806603773585,
      "grad_norm": 2.058398485183716,
      "learning_rate": 9.711084905660379e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 2162200,
      "step": 3295
    },
    {
      "epoch": 1.9457547169811322,
      "grad_norm": 1.8215802907943726,
      "learning_rate": 9.725825471698114e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 2164792,
      "step": 3300
    },
    {
      "epoch": 1.9487028301886793,
      "grad_norm": 1.5940580368041992,
      "learning_rate": 9.74056603773585e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 2167320,
      "step": 3305
    },
    {
      "epoch": 1.9516509433962264,
      "grad_norm": 1.3754923343658447,
      "learning_rate": 9.755306603773585e-06,
      "loss": 0.4264,
      "num_input_tokens_seen": 2169784,
      "step": 3310
    },
    {
      "epoch": 1.9545990566037736,
      "grad_norm": 1.2899601459503174,
      "learning_rate": 9.770047169811322e-06,
      "loss": 0.5689,
      "num_input_tokens_seen": 2173272,
      "step": 3315
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 2.1679368019104004,
      "learning_rate": 9.784787735849057e-06,
      "loss": 0.6058,
      "num_input_tokens_seen": 2176248,
      "step": 3320
    },
    {
      "epoch": 1.9604952830188678,
      "grad_norm": 5.113559246063232,
      "learning_rate": 9.799528301886794e-06,
      "loss": 0.7947,
      "num_input_tokens_seen": 2179704,
      "step": 3325
    },
    {
      "epoch": 1.963443396226415,
      "grad_norm": 3.744192600250244,
      "learning_rate": 9.81426886792453e-06,
      "loss": 0.5105,
      "num_input_tokens_seen": 2183640,
      "step": 3330
    },
    {
      "epoch": 1.9663915094339623,
      "grad_norm": 2.1596715450286865,
      "learning_rate": 9.829009433962265e-06,
      "loss": 0.6292,
      "num_input_tokens_seen": 2187480,
      "step": 3335
    },
    {
      "epoch": 1.9693396226415094,
      "grad_norm": 2.2561075687408447,
      "learning_rate": 9.84375e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 2190360,
      "step": 3340
    },
    {
      "epoch": 1.9722877358490565,
      "grad_norm": 2.32470965385437,
      "learning_rate": 9.858490566037736e-06,
      "loss": 0.6923,
      "num_input_tokens_seen": 2193784,
      "step": 3345
    },
    {
      "epoch": 1.9752358490566038,
      "grad_norm": 1.430677890777588,
      "learning_rate": 9.873231132075473e-06,
      "loss": 0.5737,
      "num_input_tokens_seen": 2196472,
      "step": 3350
    },
    {
      "epoch": 1.978183962264151,
      "grad_norm": 2.0688018798828125,
      "learning_rate": 9.887971698113208e-06,
      "loss": 0.4756,
      "num_input_tokens_seen": 2199032,
      "step": 3355
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 2.4728775024414062,
      "learning_rate": 9.902712264150945e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 2202200,
      "step": 3360
    },
    {
      "epoch": 1.9840801886792452,
      "grad_norm": 5.746274948120117,
      "learning_rate": 9.917452830188679e-06,
      "loss": 0.5698,
      "num_input_tokens_seen": 2205624,
      "step": 3365
    },
    {
      "epoch": 1.9870283018867925,
      "grad_norm": 3.296053886413574,
      "learning_rate": 9.932193396226416e-06,
      "loss": 0.4672,
      "num_input_tokens_seen": 2208344,
      "step": 3370
    },
    {
      "epoch": 1.9899764150943398,
      "grad_norm": 2.7268266677856445,
      "learning_rate": 9.946933962264151e-06,
      "loss": 0.4384,
      "num_input_tokens_seen": 2212280,
      "step": 3375
    },
    {
      "epoch": 1.9929245283018868,
      "grad_norm": 1.3798308372497559,
      "learning_rate": 9.961674528301888e-06,
      "loss": 0.4696,
      "num_input_tokens_seen": 2215160,
      "step": 3380
    },
    {
      "epoch": 1.9958726415094339,
      "grad_norm": 2.6599948406219482,
      "learning_rate": 9.976415094339624e-06,
      "loss": 0.5678,
      "num_input_tokens_seen": 2218520,
      "step": 3385
    },
    {
      "epoch": 1.9988207547169812,
      "grad_norm": 2.516369581222534,
      "learning_rate": 9.991155660377359e-06,
      "loss": 0.5308,
      "num_input_tokens_seen": 2222232,
      "step": 3390
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5022701025009155,
      "eval_runtime": 19.0083,
      "eval_samples_per_second": 89.224,
      "eval_steps_per_second": 22.306,
      "num_input_tokens_seen": 2222792,
      "step": 3392
    },
    {
      "epoch": 2.0017688679245285,
      "grad_norm": 1.620583176612854,
      "learning_rate": 9.999999894098275e-06,
      "loss": 0.4309,
      "num_input_tokens_seen": 2224712,
      "step": 3395
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 2.137629270553589,
      "learning_rate": 9.99999870270391e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 2227720,
      "step": 3400
    },
    {
      "epoch": 2.0076650943396226,
      "grad_norm": 1.7688075304031372,
      "learning_rate": 9.999996187538341e-06,
      "loss": 0.4471,
      "num_input_tokens_seen": 2230824,
      "step": 3405
    },
    {
      "epoch": 2.01061320754717,
      "grad_norm": 2.6755542755126953,
      "learning_rate": 9.999992348602233e-06,
      "loss": 0.5188,
      "num_input_tokens_seen": 2233128,
      "step": 3410
    },
    {
      "epoch": 2.013561320754717,
      "grad_norm": 1.5499404668807983,
      "learning_rate": 9.999987185896598e-06,
      "loss": 0.4745,
      "num_input_tokens_seen": 2236136,
      "step": 3415
    },
    {
      "epoch": 2.016509433962264,
      "grad_norm": 1.7098321914672852,
      "learning_rate": 9.99998069942281e-06,
      "loss": 0.4433,
      "num_input_tokens_seen": 2238792,
      "step": 3420
    },
    {
      "epoch": 2.0194575471698113,
      "grad_norm": 2.5733916759490967,
      "learning_rate": 9.999972889182583e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 2242344,
      "step": 3425
    },
    {
      "epoch": 2.0224056603773586,
      "grad_norm": 2.4553418159484863,
      "learning_rate": 9.999963755177984e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 2245384,
      "step": 3430
    },
    {
      "epoch": 2.025353773584906,
      "grad_norm": 5.669279098510742,
      "learning_rate": 9.999953297411434e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 2248200,
      "step": 3435
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 1.6067826747894287,
      "learning_rate": 9.999941515885699e-06,
      "loss": 0.7137,
      "num_input_tokens_seen": 2250920,
      "step": 3440
    },
    {
      "epoch": 2.03125,
      "grad_norm": 1.7786338329315186,
      "learning_rate": 9.999928410603897e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 2253320,
      "step": 3445
    },
    {
      "epoch": 2.0341981132075473,
      "grad_norm": 1.4499329328536987,
      "learning_rate": 9.999913981569502e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 2257032,
      "step": 3450
    },
    {
      "epoch": 2.037146226415094,
      "grad_norm": 8.010506629943848,
      "learning_rate": 9.999898228786332e-06,
      "loss": 0.5848,
      "num_input_tokens_seen": 2261288,
      "step": 3455
    },
    {
      "epoch": 2.0400943396226414,
      "grad_norm": 4.389882564544678,
      "learning_rate": 9.999881152258557e-06,
      "loss": 0.4886,
      "num_input_tokens_seen": 2264360,
      "step": 3460
    },
    {
      "epoch": 2.0430424528301887,
      "grad_norm": 1.4576832056045532,
      "learning_rate": 9.999862751990697e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 2266920,
      "step": 3465
    },
    {
      "epoch": 2.045990566037736,
      "grad_norm": 1.3836009502410889,
      "learning_rate": 9.999843027987628e-06,
      "loss": 0.5361,
      "num_input_tokens_seen": 2270280,
      "step": 3470
    },
    {
      "epoch": 2.048938679245283,
      "grad_norm": 1.391100525856018,
      "learning_rate": 9.999821980254567e-06,
      "loss": 0.5517,
      "num_input_tokens_seen": 2273256,
      "step": 3475
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 2.0438568592071533,
      "learning_rate": 9.99979960879709e-06,
      "loss": 0.3587,
      "num_input_tokens_seen": 2276040,
      "step": 3480
    },
    {
      "epoch": 2.0548349056603774,
      "grad_norm": 3.128347635269165,
      "learning_rate": 9.99977591362112e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 2278952,
      "step": 3485
    },
    {
      "epoch": 2.0577830188679247,
      "grad_norm": 1.4661860466003418,
      "learning_rate": 9.999750894732927e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 2282760,
      "step": 3490
    },
    {
      "epoch": 2.0607311320754715,
      "grad_norm": 2.1191205978393555,
      "learning_rate": 9.999724552139136e-06,
      "loss": 0.4706,
      "num_input_tokens_seen": 2286664,
      "step": 3495
    },
    {
      "epoch": 2.063679245283019,
      "grad_norm": 0.792425811290741,
      "learning_rate": 9.999696885846724e-06,
      "loss": 0.3701,
      "num_input_tokens_seen": 2289832,
      "step": 3500
    },
    {
      "epoch": 2.066627358490566,
      "grad_norm": 2.5372314453125,
      "learning_rate": 9.999667895863012e-06,
      "loss": 0.4911,
      "num_input_tokens_seen": 2292712,
      "step": 3505
    },
    {
      "epoch": 2.0695754716981134,
      "grad_norm": 3.9704525470733643,
      "learning_rate": 9.99963758219568e-06,
      "loss": 0.4473,
      "num_input_tokens_seen": 2294984,
      "step": 3510
    },
    {
      "epoch": 2.0725235849056602,
      "grad_norm": 3.064893960952759,
      "learning_rate": 9.999605944852749e-06,
      "loss": 0.4913,
      "num_input_tokens_seen": 2298120,
      "step": 3515
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 3.1229116916656494,
      "learning_rate": 9.999572983842599e-06,
      "loss": 0.5291,
      "num_input_tokens_seen": 2300616,
      "step": 3520
    },
    {
      "epoch": 2.078419811320755,
      "grad_norm": 2.5079431533813477,
      "learning_rate": 9.999538699173951e-06,
      "loss": 0.6237,
      "num_input_tokens_seen": 2303816,
      "step": 3525
    },
    {
      "epoch": 2.081367924528302,
      "grad_norm": 1.6968766450881958,
      "learning_rate": 9.99950309085589e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 2306312,
      "step": 3530
    },
    {
      "epoch": 2.084316037735849,
      "grad_norm": 1.7957154512405396,
      "learning_rate": 9.999466158897835e-06,
      "loss": 0.5177,
      "num_input_tokens_seen": 2310280,
      "step": 3535
    },
    {
      "epoch": 2.0872641509433962,
      "grad_norm": 1.5328375101089478,
      "learning_rate": 9.999427903309569e-06,
      "loss": 0.341,
      "num_input_tokens_seen": 2313832,
      "step": 3540
    },
    {
      "epoch": 2.0902122641509435,
      "grad_norm": 1.944772720336914,
      "learning_rate": 9.99938832410122e-06,
      "loss": 0.5864,
      "num_input_tokens_seen": 2316328,
      "step": 3545
    },
    {
      "epoch": 2.0931603773584904,
      "grad_norm": 3.1959779262542725,
      "learning_rate": 9.999347421283267e-06,
      "loss": 0.6311,
      "num_input_tokens_seen": 2320744,
      "step": 3550
    },
    {
      "epoch": 2.0961084905660377,
      "grad_norm": 2.6974313259124756,
      "learning_rate": 9.999305194866538e-06,
      "loss": 0.674,
      "num_input_tokens_seen": 2323720,
      "step": 3555
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 1.9158457517623901,
      "learning_rate": 9.99926164486221e-06,
      "loss": 0.5429,
      "num_input_tokens_seen": 2325736,
      "step": 3560
    },
    {
      "epoch": 2.1020047169811322,
      "grad_norm": 1.4020017385482788,
      "learning_rate": 9.99921677128182e-06,
      "loss": 0.3845,
      "num_input_tokens_seen": 2329000,
      "step": 3565
    },
    {
      "epoch": 2.104952830188679,
      "grad_norm": 2.4075989723205566,
      "learning_rate": 9.99917057413724e-06,
      "loss": 0.6469,
      "num_input_tokens_seen": 2332488,
      "step": 3570
    },
    {
      "epoch": 2.1079009433962264,
      "grad_norm": 1.5627119541168213,
      "learning_rate": 9.99912305344071e-06,
      "loss": 0.6703,
      "num_input_tokens_seen": 2335848,
      "step": 3575
    },
    {
      "epoch": 2.1108490566037736,
      "grad_norm": 1.2948452234268188,
      "learning_rate": 9.999074209204803e-06,
      "loss": 0.4587,
      "num_input_tokens_seen": 2339560,
      "step": 3580
    },
    {
      "epoch": 2.113797169811321,
      "grad_norm": 2.8955752849578857,
      "learning_rate": 9.999024041442455e-06,
      "loss": 0.6341,
      "num_input_tokens_seen": 2343400,
      "step": 3585
    },
    {
      "epoch": 2.1167452830188678,
      "grad_norm": 0.9222820401191711,
      "learning_rate": 9.998972550166948e-06,
      "loss": 0.5312,
      "num_input_tokens_seen": 2346760,
      "step": 3590
    },
    {
      "epoch": 2.119693396226415,
      "grad_norm": 1.886954665184021,
      "learning_rate": 9.998919735391915e-06,
      "loss": 0.4623,
      "num_input_tokens_seen": 2349320,
      "step": 3595
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 1.369951605796814,
      "learning_rate": 9.998865597131336e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 2353480,
      "step": 3600
    },
    {
      "epoch": 2.1255896226415096,
      "grad_norm": 2.2723653316497803,
      "learning_rate": 9.998810135399545e-06,
      "loss": 0.56,
      "num_input_tokens_seen": 2355752,
      "step": 3605
    },
    {
      "epoch": 2.1285377358490565,
      "grad_norm": 20.71672248840332,
      "learning_rate": 9.99875335021123e-06,
      "loss": 0.5512,
      "num_input_tokens_seen": 2359240,
      "step": 3610
    },
    {
      "epoch": 2.1314858490566038,
      "grad_norm": 1.504581332206726,
      "learning_rate": 9.998695241581423e-06,
      "loss": 0.4732,
      "num_input_tokens_seen": 2362600,
      "step": 3615
    },
    {
      "epoch": 2.134433962264151,
      "grad_norm": 2.000737190246582,
      "learning_rate": 9.998635809525504e-06,
      "loss": 0.6404,
      "num_input_tokens_seen": 2366312,
      "step": 3620
    },
    {
      "epoch": 2.137382075471698,
      "grad_norm": 3.522698402404785,
      "learning_rate": 9.998575054059212e-06,
      "loss": 0.4615,
      "num_input_tokens_seen": 2369192,
      "step": 3625
    },
    {
      "epoch": 2.140330188679245,
      "grad_norm": 5.781278610229492,
      "learning_rate": 9.998512975198633e-06,
      "loss": 0.5066,
      "num_input_tokens_seen": 2372328,
      "step": 3630
    },
    {
      "epoch": 2.1432783018867925,
      "grad_norm": 2.1164255142211914,
      "learning_rate": 9.998449572960202e-06,
      "loss": 0.5934,
      "num_input_tokens_seen": 2375944,
      "step": 3635
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 1.5810338258743286,
      "learning_rate": 9.998384847360705e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 2382120,
      "step": 3640
    },
    {
      "epoch": 2.1491745283018866,
      "grad_norm": 1.0848841667175293,
      "learning_rate": 9.998318798417276e-06,
      "loss": 0.5076,
      "num_input_tokens_seen": 2386920,
      "step": 3645
    },
    {
      "epoch": 2.152122641509434,
      "grad_norm": 1.6555845737457275,
      "learning_rate": 9.998251426147403e-06,
      "loss": 0.4953,
      "num_input_tokens_seen": 2389416,
      "step": 3650
    },
    {
      "epoch": 2.155070754716981,
      "grad_norm": 2.0839428901672363,
      "learning_rate": 9.998182730568927e-06,
      "loss": 0.5975,
      "num_input_tokens_seen": 2392808,
      "step": 3655
    },
    {
      "epoch": 2.1580188679245285,
      "grad_norm": 3.612295150756836,
      "learning_rate": 9.998112711700028e-06,
      "loss": 0.5759,
      "num_input_tokens_seen": 2395336,
      "step": 3660
    },
    {
      "epoch": 2.1609669811320753,
      "grad_norm": 1.9112000465393066,
      "learning_rate": 9.99804136955925e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 2398504,
      "step": 3665
    },
    {
      "epoch": 2.1639150943396226,
      "grad_norm": 2.4269864559173584,
      "learning_rate": 9.99796870416548e-06,
      "loss": 0.4951,
      "num_input_tokens_seen": 2402024,
      "step": 3670
    },
    {
      "epoch": 2.16686320754717,
      "grad_norm": 1.1239309310913086,
      "learning_rate": 9.997894715537953e-06,
      "loss": 0.4044,
      "num_input_tokens_seen": 2405288,
      "step": 3675
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 1.3183726072311401,
      "learning_rate": 9.997819403696263e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 2409224,
      "step": 3680
    },
    {
      "epoch": 2.172759433962264,
      "grad_norm": 1.7976112365722656,
      "learning_rate": 9.997742768660345e-06,
      "loss": 0.4844,
      "num_input_tokens_seen": 2412936,
      "step": 3685
    },
    {
      "epoch": 2.1757075471698113,
      "grad_norm": 1.4684852361679077,
      "learning_rate": 9.99766481045049e-06,
      "loss": 0.5275,
      "num_input_tokens_seen": 2416648,
      "step": 3690
    },
    {
      "epoch": 2.1786556603773586,
      "grad_norm": 5.029133319854736,
      "learning_rate": 9.997585529087338e-06,
      "loss": 0.4002,
      "num_input_tokens_seen": 2419784,
      "step": 3695
    },
    {
      "epoch": 2.1816037735849054,
      "grad_norm": 2.139472007751465,
      "learning_rate": 9.997504924591878e-06,
      "loss": 0.4494,
      "num_input_tokens_seen": 2426056,
      "step": 3700
    },
    {
      "epoch": 2.1845518867924527,
      "grad_norm": 1.8799384832382202,
      "learning_rate": 9.99742299698545e-06,
      "loss": 0.6077,
      "num_input_tokens_seen": 2429736,
      "step": 3705
    },
    {
      "epoch": 2.1875,
      "grad_norm": 5.1327900886535645,
      "learning_rate": 9.997339746289749e-06,
      "loss": 0.656,
      "num_input_tokens_seen": 2432648,
      "step": 3710
    },
    {
      "epoch": 2.1904481132075473,
      "grad_norm": 2.1278674602508545,
      "learning_rate": 9.997255172526812e-06,
      "loss": 0.5245,
      "num_input_tokens_seen": 2436072,
      "step": 3715
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 1.8244694471359253,
      "learning_rate": 9.99716927571903e-06,
      "loss": 0.5232,
      "num_input_tokens_seen": 2439080,
      "step": 3720
    },
    {
      "epoch": 2.1963443396226414,
      "grad_norm": 3.742438793182373,
      "learning_rate": 9.997082055889147e-06,
      "loss": 0.4311,
      "num_input_tokens_seen": 2441448,
      "step": 3725
    },
    {
      "epoch": 2.1992924528301887,
      "grad_norm": 1.2605390548706055,
      "learning_rate": 9.996993513060252e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 2444584,
      "step": 3730
    },
    {
      "epoch": 2.202240566037736,
      "grad_norm": 1.7622123956680298,
      "learning_rate": 9.996903647255789e-06,
      "loss": 0.4664,
      "num_input_tokens_seen": 2447848,
      "step": 3735
    },
    {
      "epoch": 2.205188679245283,
      "grad_norm": 1.1874510049819946,
      "learning_rate": 9.99681245849955e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 2450888,
      "step": 3740
    },
    {
      "epoch": 2.20813679245283,
      "grad_norm": 2.222001791000366,
      "learning_rate": 9.996719946815679e-06,
      "loss": 0.7158,
      "num_input_tokens_seen": 2454088,
      "step": 3745
    },
    {
      "epoch": 2.2110849056603774,
      "grad_norm": 3.1075022220611572,
      "learning_rate": 9.996626112228665e-06,
      "loss": 0.5367,
      "num_input_tokens_seen": 2457000,
      "step": 3750
    },
    {
      "epoch": 2.2140330188679247,
      "grad_norm": 1.9559897184371948,
      "learning_rate": 9.996530954763355e-06,
      "loss": 0.5502,
      "num_input_tokens_seen": 2461352,
      "step": 3755
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 1.2056066989898682,
      "learning_rate": 9.99643447444494e-06,
      "loss": 0.4644,
      "num_input_tokens_seen": 2464392,
      "step": 3760
    },
    {
      "epoch": 2.219929245283019,
      "grad_norm": 1.6162651777267456,
      "learning_rate": 9.996336671298965e-06,
      "loss": 0.5216,
      "num_input_tokens_seen": 2466952,
      "step": 3765
    },
    {
      "epoch": 2.222877358490566,
      "grad_norm": 1.7170802354812622,
      "learning_rate": 9.996237545351323e-06,
      "loss": 0.559,
      "num_input_tokens_seen": 2470184,
      "step": 3770
    },
    {
      "epoch": 2.2258254716981134,
      "grad_norm": 1.9885934591293335,
      "learning_rate": 9.996137096628259e-06,
      "loss": 0.3518,
      "num_input_tokens_seen": 2473448,
      "step": 3775
    },
    {
      "epoch": 2.2287735849056602,
      "grad_norm": 3.048759698867798,
      "learning_rate": 9.996035325156366e-06,
      "loss": 0.4367,
      "num_input_tokens_seen": 2476648,
      "step": 3780
    },
    {
      "epoch": 2.2317216981132075,
      "grad_norm": 2.9165189266204834,
      "learning_rate": 9.995932230962589e-06,
      "loss": 0.4627,
      "num_input_tokens_seen": 2479272,
      "step": 3785
    },
    {
      "epoch": 2.234669811320755,
      "grad_norm": 3.609206199645996,
      "learning_rate": 9.995827814074223e-06,
      "loss": 0.5267,
      "num_input_tokens_seen": 2482600,
      "step": 3790
    },
    {
      "epoch": 2.237617924528302,
      "grad_norm": 1.202154517173767,
      "learning_rate": 9.995722074518913e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 2485384,
      "step": 3795
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 1.7875666618347168,
      "learning_rate": 9.99561501232465e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 2488744,
      "step": 3800
    },
    {
      "epoch": 2.2435141509433962,
      "grad_norm": 2.2737772464752197,
      "learning_rate": 9.995506627519786e-06,
      "loss": 0.5034,
      "num_input_tokens_seen": 2491976,
      "step": 3805
    },
    {
      "epoch": 2.2464622641509435,
      "grad_norm": 2.8416855335235596,
      "learning_rate": 9.995396920133012e-06,
      "loss": 0.579,
      "num_input_tokens_seen": 2495080,
      "step": 3810
    },
    {
      "epoch": 2.2494103773584904,
      "grad_norm": 1.6029270887374878,
      "learning_rate": 9.995285890193373e-06,
      "loss": 0.6376,
      "num_input_tokens_seen": 2499144,
      "step": 3815
    },
    {
      "epoch": 2.2523584905660377,
      "grad_norm": 2.7641682624816895,
      "learning_rate": 9.995173537730267e-06,
      "loss": 0.5773,
      "num_input_tokens_seen": 2502088,
      "step": 3820
    },
    {
      "epoch": 2.255306603773585,
      "grad_norm": 2.9080164432525635,
      "learning_rate": 9.99505986277344e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 2504840,
      "step": 3825
    },
    {
      "epoch": 2.2582547169811322,
      "grad_norm": 3.4255166053771973,
      "learning_rate": 9.994944865352986e-06,
      "loss": 0.5211,
      "num_input_tokens_seen": 2508264,
      "step": 3830
    },
    {
      "epoch": 2.261202830188679,
      "grad_norm": 3.4850242137908936,
      "learning_rate": 9.994828545499351e-06,
      "loss": 0.4817,
      "num_input_tokens_seen": 2511592,
      "step": 3835
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 1.5304901599884033,
      "learning_rate": 9.994710903243334e-06,
      "loss": 0.4601,
      "num_input_tokens_seen": 2514504,
      "step": 3840
    },
    {
      "epoch": 2.2670990566037736,
      "grad_norm": 1.016110897064209,
      "learning_rate": 9.994591938616079e-06,
      "loss": 0.4383,
      "num_input_tokens_seen": 2516936,
      "step": 3845
    },
    {
      "epoch": 2.270047169811321,
      "grad_norm": 2.727262020111084,
      "learning_rate": 9.994471651649082e-06,
      "loss": 0.6244,
      "num_input_tokens_seen": 2519752,
      "step": 3850
    },
    {
      "epoch": 2.2729952830188678,
      "grad_norm": 1.8129281997680664,
      "learning_rate": 9.99435004237419e-06,
      "loss": 0.5002,
      "num_input_tokens_seen": 2523240,
      "step": 3855
    },
    {
      "epoch": 2.275943396226415,
      "grad_norm": 2.46368408203125,
      "learning_rate": 9.9942271108236e-06,
      "loss": 0.722,
      "num_input_tokens_seen": 2527528,
      "step": 3860
    },
    {
      "epoch": 2.2788915094339623,
      "grad_norm": 2.801548719406128,
      "learning_rate": 9.994102857029859e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 2531304,
      "step": 3865
    },
    {
      "epoch": 2.2818396226415096,
      "grad_norm": 1.525614857673645,
      "learning_rate": 9.993977281025862e-06,
      "loss": 0.4759,
      "num_input_tokens_seen": 2534760,
      "step": 3870
    },
    {
      "epoch": 2.2847877358490565,
      "grad_norm": 5.436481952667236,
      "learning_rate": 9.993850382844858e-06,
      "loss": 0.4664,
      "num_input_tokens_seen": 2537192,
      "step": 3875
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 2.692549705505371,
      "learning_rate": 9.993722162520443e-06,
      "loss": 0.4889,
      "num_input_tokens_seen": 2540104,
      "step": 3880
    },
    {
      "epoch": 2.290683962264151,
      "grad_norm": 4.584810256958008,
      "learning_rate": 9.993592620086564e-06,
      "loss": 0.5049,
      "num_input_tokens_seen": 2542760,
      "step": 3885
    },
    {
      "epoch": 2.293632075471698,
      "grad_norm": 2.0639450550079346,
      "learning_rate": 9.993461755577518e-06,
      "loss": 0.6388,
      "num_input_tokens_seen": 2545160,
      "step": 3890
    },
    {
      "epoch": 2.296580188679245,
      "grad_norm": 1.564429759979248,
      "learning_rate": 9.99332956902795e-06,
      "loss": 0.5565,
      "num_input_tokens_seen": 2548712,
      "step": 3895
    },
    {
      "epoch": 2.2995283018867925,
      "grad_norm": 2.7065677642822266,
      "learning_rate": 9.993196060472859e-06,
      "loss": 0.395,
      "num_input_tokens_seen": 2551784,
      "step": 3900
    },
    {
      "epoch": 2.3024764150943398,
      "grad_norm": 1.4076188802719116,
      "learning_rate": 9.993061229947591e-06,
      "loss": 0.563,
      "num_input_tokens_seen": 2555496,
      "step": 3905
    },
    {
      "epoch": 2.3054245283018866,
      "grad_norm": 4.543927192687988,
      "learning_rate": 9.992925077487845e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 2558760,
      "step": 3910
    },
    {
      "epoch": 2.308372641509434,
      "grad_norm": 5.009082794189453,
      "learning_rate": 9.992787603129666e-06,
      "loss": 0.5014,
      "num_input_tokens_seen": 2562472,
      "step": 3915
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 2.535106897354126,
      "learning_rate": 9.99264880690945e-06,
      "loss": 0.5016,
      "num_input_tokens_seen": 2564904,
      "step": 3920
    },
    {
      "epoch": 2.3142688679245285,
      "grad_norm": 1.4015475511550903,
      "learning_rate": 9.992508688863947e-06,
      "loss": 0.4367,
      "num_input_tokens_seen": 2568264,
      "step": 3925
    },
    {
      "epoch": 2.3172169811320753,
      "grad_norm": 1.0242820978164673,
      "learning_rate": 9.99236724903025e-06,
      "loss": 0.6327,
      "num_input_tokens_seen": 2571976,
      "step": 3930
    },
    {
      "epoch": 2.3201650943396226,
      "grad_norm": 2.185167074203491,
      "learning_rate": 9.992224487445809e-06,
      "loss": 0.4785,
      "num_input_tokens_seen": 2575560,
      "step": 3935
    },
    {
      "epoch": 2.32311320754717,
      "grad_norm": 1.4828059673309326,
      "learning_rate": 9.99208040414842e-06,
      "loss": 0.5873,
      "num_input_tokens_seen": 2579400,
      "step": 3940
    },
    {
      "epoch": 2.326061320754717,
      "grad_norm": 1.4714442491531372,
      "learning_rate": 9.99193499917623e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 2583464,
      "step": 3945
    },
    {
      "epoch": 2.329009433962264,
      "grad_norm": 2.7186217308044434,
      "learning_rate": 9.991788272567735e-06,
      "loss": 0.5098,
      "num_input_tokens_seen": 2586600,
      "step": 3950
    },
    {
      "epoch": 2.3319575471698113,
      "grad_norm": 1.0144418478012085,
      "learning_rate": 9.991640224361781e-06,
      "loss": 0.4742,
      "num_input_tokens_seen": 2590536,
      "step": 3955
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 2.7273402214050293,
      "learning_rate": 9.991490854597563e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 2593864,
      "step": 3960
    },
    {
      "epoch": 2.3378537735849054,
      "grad_norm": 2.162824869155884,
      "learning_rate": 9.991340163314632e-06,
      "loss": 0.5153,
      "num_input_tokens_seen": 2596584,
      "step": 3965
    },
    {
      "epoch": 2.3408018867924527,
      "grad_norm": 2.5422420501708984,
      "learning_rate": 9.991188150552878e-06,
      "loss": 0.3974,
      "num_input_tokens_seen": 2600904,
      "step": 3970
    },
    {
      "epoch": 2.34375,
      "grad_norm": 6.405313968658447,
      "learning_rate": 9.991034816352553e-06,
      "loss": 0.5755,
      "num_input_tokens_seen": 2603784,
      "step": 3975
    },
    {
      "epoch": 2.3466981132075473,
      "grad_norm": 2.860940456390381,
      "learning_rate": 9.990880160754248e-06,
      "loss": 0.5152,
      "num_input_tokens_seen": 2608264,
      "step": 3980
    },
    {
      "epoch": 2.3496462264150946,
      "grad_norm": 8.067008972167969,
      "learning_rate": 9.990724183798914e-06,
      "loss": 0.5116,
      "num_input_tokens_seen": 2613192,
      "step": 3985
    },
    {
      "epoch": 2.3525943396226414,
      "grad_norm": 1.8643155097961426,
      "learning_rate": 9.990566885527841e-06,
      "loss": 0.3535,
      "num_input_tokens_seen": 2617448,
      "step": 3990
    },
    {
      "epoch": 2.3555424528301887,
      "grad_norm": 2.556971549987793,
      "learning_rate": 9.99040826598268e-06,
      "loss": 0.4352,
      "num_input_tokens_seen": 2620392,
      "step": 3995
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 2.6513607501983643,
      "learning_rate": 9.990248325205418e-06,
      "loss": 0.4443,
      "num_input_tokens_seen": 2623400,
      "step": 4000
    },
    {
      "epoch": 2.361438679245283,
      "grad_norm": 2.1848456859588623,
      "learning_rate": 9.990087063238408e-06,
      "loss": 0.476,
      "num_input_tokens_seen": 2626440,
      "step": 4005
    },
    {
      "epoch": 2.36438679245283,
      "grad_norm": 2.332390785217285,
      "learning_rate": 9.989924480124342e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 2629096,
      "step": 4010
    },
    {
      "epoch": 2.3673349056603774,
      "grad_norm": 2.272407293319702,
      "learning_rate": 9.989760575906264e-06,
      "loss": 0.6129,
      "num_input_tokens_seen": 2631848,
      "step": 4015
    },
    {
      "epoch": 2.3702830188679247,
      "grad_norm": 1.5404633283615112,
      "learning_rate": 9.98959535062757e-06,
      "loss": 0.3529,
      "num_input_tokens_seen": 2635848,
      "step": 4020
    },
    {
      "epoch": 2.3732311320754715,
      "grad_norm": 5.380978107452393,
      "learning_rate": 9.989428804332002e-06,
      "loss": 0.5502,
      "num_input_tokens_seen": 2638952,
      "step": 4025
    },
    {
      "epoch": 2.376179245283019,
      "grad_norm": 2.1804912090301514,
      "learning_rate": 9.989260937063656e-06,
      "loss": 0.4778,
      "num_input_tokens_seen": 2642536,
      "step": 4030
    },
    {
      "epoch": 2.379127358490566,
      "grad_norm": 2.372162342071533,
      "learning_rate": 9.989091748866973e-06,
      "loss": 0.4927,
      "num_input_tokens_seen": 2645800,
      "step": 4035
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 1.259592056274414,
      "learning_rate": 9.988921239786748e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 2649160,
      "step": 4040
    },
    {
      "epoch": 2.3850235849056602,
      "grad_norm": 1.4277245998382568,
      "learning_rate": 9.988749409868124e-06,
      "loss": 0.57,
      "num_input_tokens_seen": 2652232,
      "step": 4045
    },
    {
      "epoch": 2.3879716981132075,
      "grad_norm": 2.675335168838501,
      "learning_rate": 9.988576259156593e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 2655112,
      "step": 4050
    },
    {
      "epoch": 2.390919811320755,
      "grad_norm": 1.9087278842926025,
      "learning_rate": 9.988401787697996e-06,
      "loss": 0.3867,
      "num_input_tokens_seen": 2661640,
      "step": 4055
    },
    {
      "epoch": 2.393867924528302,
      "grad_norm": 1.7638816833496094,
      "learning_rate": 9.98822599553853e-06,
      "loss": 0.5632,
      "num_input_tokens_seen": 2664072,
      "step": 4060
    },
    {
      "epoch": 2.396816037735849,
      "grad_norm": 1.982764720916748,
      "learning_rate": 9.988048882724732e-06,
      "loss": 0.4731,
      "num_input_tokens_seen": 2666984,
      "step": 4065
    },
    {
      "epoch": 2.3997641509433962,
      "grad_norm": 2.363523244857788,
      "learning_rate": 9.987870449303497e-06,
      "loss": 0.6116,
      "num_input_tokens_seen": 2669320,
      "step": 4070
    },
    {
      "epoch": 2.4027122641509435,
      "grad_norm": 1.8666070699691772,
      "learning_rate": 9.98769069532206e-06,
      "loss": 0.6602,
      "num_input_tokens_seen": 2672360,
      "step": 4075
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 1.9655553102493286,
      "learning_rate": 9.98750962082802e-06,
      "loss": 0.5838,
      "num_input_tokens_seen": 2675208,
      "step": 4080
    },
    {
      "epoch": 2.4086084905660377,
      "grad_norm": 1.678615927696228,
      "learning_rate": 9.987327225869312e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 2678600,
      "step": 4085
    },
    {
      "epoch": 2.411556603773585,
      "grad_norm": 3.5674057006835938,
      "learning_rate": 9.987143510494225e-06,
      "loss": 0.5688,
      "num_input_tokens_seen": 2681192,
      "step": 4090
    },
    {
      "epoch": 2.4145047169811322,
      "grad_norm": 1.3060681819915771,
      "learning_rate": 9.9869584747514e-06,
      "loss": 0.4358,
      "num_input_tokens_seen": 2684008,
      "step": 4095
    },
    {
      "epoch": 2.417452830188679,
      "grad_norm": 1.2779580354690552,
      "learning_rate": 9.986772118689828e-06,
      "loss": 0.7018,
      "num_input_tokens_seen": 2687176,
      "step": 4100
    },
    {
      "epoch": 2.4204009433962264,
      "grad_norm": 1.3262113332748413,
      "learning_rate": 9.986584442358845e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 2691752,
      "step": 4105
    },
    {
      "epoch": 2.4233490566037736,
      "grad_norm": 2.0240988731384277,
      "learning_rate": 9.98639544580814e-06,
      "loss": 0.4712,
      "num_input_tokens_seen": 2694888,
      "step": 4110
    },
    {
      "epoch": 2.426297169811321,
      "grad_norm": 2.273996353149414,
      "learning_rate": 9.98620512908775e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 2699240,
      "step": 4115
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 6.603442192077637,
      "learning_rate": 9.986013492248064e-06,
      "loss": 0.6046,
      "num_input_tokens_seen": 2701640,
      "step": 4120
    },
    {
      "epoch": 2.432193396226415,
      "grad_norm": 1.7279762029647827,
      "learning_rate": 9.985820535339817e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 2705640,
      "step": 4125
    },
    {
      "epoch": 2.4351415094339623,
      "grad_norm": 4.428871154785156,
      "learning_rate": 9.985626258414093e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 2708136,
      "step": 4130
    },
    {
      "epoch": 2.4380896226415096,
      "grad_norm": 5.598822593688965,
      "learning_rate": 9.985430661522333e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 2711400,
      "step": 4135
    },
    {
      "epoch": 2.4410377358490565,
      "grad_norm": 1.8064535856246948,
      "learning_rate": 9.98523374471632e-06,
      "loss": 0.4616,
      "num_input_tokens_seen": 2714568,
      "step": 4140
    },
    {
      "epoch": 2.4439858490566038,
      "grad_norm": 1.2782410383224487,
      "learning_rate": 9.985035508048186e-06,
      "loss": 0.3996,
      "num_input_tokens_seen": 2718152,
      "step": 4145
    },
    {
      "epoch": 2.446933962264151,
      "grad_norm": 3.7173972129821777,
      "learning_rate": 9.984835951570417e-06,
      "loss": 0.5505,
      "num_input_tokens_seen": 2720872,
      "step": 4150
    },
    {
      "epoch": 2.449882075471698,
      "grad_norm": 1.6789849996566772,
      "learning_rate": 9.984635075335847e-06,
      "loss": 0.5913,
      "num_input_tokens_seen": 2723912,
      "step": 4155
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 2.536675214767456,
      "learning_rate": 9.984432879397659e-06,
      "loss": 0.5607,
      "num_input_tokens_seen": 2727912,
      "step": 4160
    },
    {
      "epoch": 2.4557783018867925,
      "grad_norm": 2.004117727279663,
      "learning_rate": 9.984229363809383e-06,
      "loss": 0.4696,
      "num_input_tokens_seen": 2731400,
      "step": 4165
    },
    {
      "epoch": 2.4587264150943398,
      "grad_norm": 1.6404134035110474,
      "learning_rate": 9.984024528624904e-06,
      "loss": 0.5481,
      "num_input_tokens_seen": 2734792,
      "step": 4170
    },
    {
      "epoch": 2.4616745283018866,
      "grad_norm": 2.944763660430908,
      "learning_rate": 9.98381837389845e-06,
      "loss": 0.4592,
      "num_input_tokens_seen": 2737704,
      "step": 4175
    },
    {
      "epoch": 2.464622641509434,
      "grad_norm": 1.4043073654174805,
      "learning_rate": 9.983610899684601e-06,
      "loss": 0.5193,
      "num_input_tokens_seen": 2741480,
      "step": 4180
    },
    {
      "epoch": 2.467570754716981,
      "grad_norm": 1.823005199432373,
      "learning_rate": 9.983402106038292e-06,
      "loss": 0.4472,
      "num_input_tokens_seen": 2744008,
      "step": 4185
    },
    {
      "epoch": 2.4705188679245285,
      "grad_norm": 2.1911847591400146,
      "learning_rate": 9.983191993014793e-06,
      "loss": 0.5594,
      "num_input_tokens_seen": 2747368,
      "step": 4190
    },
    {
      "epoch": 2.4734669811320753,
      "grad_norm": 1.4147533178329468,
      "learning_rate": 9.982980560669742e-06,
      "loss": 0.4749,
      "num_input_tokens_seen": 2750856,
      "step": 4195
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 1.380906105041504,
      "learning_rate": 9.98276780905911e-06,
      "loss": 0.446,
      "num_input_tokens_seen": 2753992,
      "step": 4200
    },
    {
      "epoch": 2.47936320754717,
      "grad_norm": 0.9819674491882324,
      "learning_rate": 9.982553738239225e-06,
      "loss": 0.4707,
      "num_input_tokens_seen": 2756968,
      "step": 4205
    },
    {
      "epoch": 2.482311320754717,
      "grad_norm": 1.6193151473999023,
      "learning_rate": 9.982338348266766e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 2760296,
      "step": 4210
    },
    {
      "epoch": 2.485259433962264,
      "grad_norm": 2.3142731189727783,
      "learning_rate": 9.982121639198756e-06,
      "loss": 0.488,
      "num_input_tokens_seen": 2762504,
      "step": 4215
    },
    {
      "epoch": 2.4882075471698113,
      "grad_norm": 3.5965828895568848,
      "learning_rate": 9.98190361109257e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 2765928,
      "step": 4220
    },
    {
      "epoch": 2.4911556603773586,
      "grad_norm": 1.499196171760559,
      "learning_rate": 9.981684264005934e-06,
      "loss": 0.5556,
      "num_input_tokens_seen": 2768776,
      "step": 4225
    },
    {
      "epoch": 2.4941037735849054,
      "grad_norm": 1.18843412399292,
      "learning_rate": 9.981463597996917e-06,
      "loss": 0.4049,
      "num_input_tokens_seen": 2771720,
      "step": 4230
    },
    {
      "epoch": 2.4970518867924527,
      "grad_norm": 1.1523813009262085,
      "learning_rate": 9.981241613123944e-06,
      "loss": 0.4855,
      "num_input_tokens_seen": 2775048,
      "step": 4235
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.7061259746551514,
      "learning_rate": 9.981018309445785e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 2778664,
      "step": 4240
    },
    {
      "epoch": 2.5029481132075473,
      "grad_norm": 1.6477837562561035,
      "learning_rate": 9.980793687021564e-06,
      "loss": 0.5067,
      "num_input_tokens_seen": 2781608,
      "step": 4245
    },
    {
      "epoch": 2.5058962264150946,
      "grad_norm": 4.597771167755127,
      "learning_rate": 9.980567745910746e-06,
      "loss": 0.4568,
      "num_input_tokens_seen": 2785064,
      "step": 4250
    },
    {
      "epoch": 2.5088443396226414,
      "grad_norm": 3.348856210708618,
      "learning_rate": 9.980340486173155e-06,
      "loss": 0.5551,
      "num_input_tokens_seen": 2788296,
      "step": 4255
    },
    {
      "epoch": 2.5117924528301887,
      "grad_norm": 3.647442579269409,
      "learning_rate": 9.980111907868954e-06,
      "loss": 0.5719,
      "num_input_tokens_seen": 2791368,
      "step": 4260
    },
    {
      "epoch": 2.514740566037736,
      "grad_norm": 4.4341864585876465,
      "learning_rate": 9.979882011058662e-06,
      "loss": 0.3975,
      "num_input_tokens_seen": 2794920,
      "step": 4265
    },
    {
      "epoch": 2.517688679245283,
      "grad_norm": 1.3042436838150024,
      "learning_rate": 9.979650795803146e-06,
      "loss": 0.8975,
      "num_input_tokens_seen": 2797672,
      "step": 4270
    },
    {
      "epoch": 2.52063679245283,
      "grad_norm": 1.3254119157791138,
      "learning_rate": 9.979418262163621e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 2800296,
      "step": 4275
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 3.9887142181396484,
      "learning_rate": 9.979184410201652e-06,
      "loss": 0.502,
      "num_input_tokens_seen": 2802856,
      "step": 4280
    },
    {
      "epoch": 2.5265330188679247,
      "grad_norm": 1.935499906539917,
      "learning_rate": 9.97894923997915e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 2805832,
      "step": 4285
    },
    {
      "epoch": 2.5294811320754715,
      "grad_norm": 24.43231773376465,
      "learning_rate": 9.97871275155838e-06,
      "loss": 0.4966,
      "num_input_tokens_seen": 2812648,
      "step": 4290
    },
    {
      "epoch": 2.532429245283019,
      "grad_norm": 1.6646819114685059,
      "learning_rate": 9.978474945001949e-06,
      "loss": 0.5284,
      "num_input_tokens_seen": 2815208,
      "step": 4295
    },
    {
      "epoch": 2.535377358490566,
      "grad_norm": 1.718915581703186,
      "learning_rate": 9.978235820372822e-06,
      "loss": 0.5351,
      "num_input_tokens_seen": 2818664,
      "step": 4300
    },
    {
      "epoch": 2.538325471698113,
      "grad_norm": 2.2351505756378174,
      "learning_rate": 9.977995377734307e-06,
      "loss": 0.4392,
      "num_input_tokens_seen": 2821800,
      "step": 4305
    },
    {
      "epoch": 2.5412735849056602,
      "grad_norm": 4.907851219177246,
      "learning_rate": 9.977753617150061e-06,
      "loss": 0.5042,
      "num_input_tokens_seen": 2824584,
      "step": 4310
    },
    {
      "epoch": 2.5442216981132075,
      "grad_norm": 2.715127944946289,
      "learning_rate": 9.977510538684094e-06,
      "loss": 0.6221,
      "num_input_tokens_seen": 2827496,
      "step": 4315
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 1.516622543334961,
      "learning_rate": 9.977266142400757e-06,
      "loss": 0.5071,
      "num_input_tokens_seen": 2830344,
      "step": 4320
    },
    {
      "epoch": 2.550117924528302,
      "grad_norm": 3.750947952270508,
      "learning_rate": 9.977020428364759e-06,
      "loss": 0.5549,
      "num_input_tokens_seen": 2833352,
      "step": 4325
    },
    {
      "epoch": 2.553066037735849,
      "grad_norm": 3.6755104064941406,
      "learning_rate": 9.976773396641154e-06,
      "loss": 0.4653,
      "num_input_tokens_seen": 2836936,
      "step": 4330
    },
    {
      "epoch": 2.5560141509433962,
      "grad_norm": 1.2179197072982788,
      "learning_rate": 9.976525047295342e-06,
      "loss": 0.5767,
      "num_input_tokens_seen": 2840200,
      "step": 4335
    },
    {
      "epoch": 2.5589622641509435,
      "grad_norm": 1.7859013080596924,
      "learning_rate": 9.976275380393077e-06,
      "loss": 0.5085,
      "num_input_tokens_seen": 2843560,
      "step": 4340
    },
    {
      "epoch": 2.5619103773584904,
      "grad_norm": 1.8894718885421753,
      "learning_rate": 9.976024396000459e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 2846472,
      "step": 4345
    },
    {
      "epoch": 2.5648584905660377,
      "grad_norm": 2.5303242206573486,
      "learning_rate": 9.975772094183935e-06,
      "loss": 0.5003,
      "num_input_tokens_seen": 2849096,
      "step": 4350
    },
    {
      "epoch": 2.567806603773585,
      "grad_norm": 2.590123414993286,
      "learning_rate": 9.975518475010306e-06,
      "loss": 0.525,
      "num_input_tokens_seen": 2852296,
      "step": 4355
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 4.664478778839111,
      "learning_rate": 9.975263538546717e-06,
      "loss": 0.5624,
      "num_input_tokens_seen": 2855912,
      "step": 4360
    },
    {
      "epoch": 2.5737028301886795,
      "grad_norm": 1.6382598876953125,
      "learning_rate": 9.975007284860664e-06,
      "loss": 0.4136,
      "num_input_tokens_seen": 2858952,
      "step": 4365
    },
    {
      "epoch": 2.5766509433962264,
      "grad_norm": 1.8616299629211426,
      "learning_rate": 9.974749714019993e-06,
      "loss": 0.4995,
      "num_input_tokens_seen": 2862056,
      "step": 4370
    },
    {
      "epoch": 2.5795990566037736,
      "grad_norm": 4.01349401473999,
      "learning_rate": 9.974490826092894e-06,
      "loss": 0.5816,
      "num_input_tokens_seen": 2864808,
      "step": 4375
    },
    {
      "epoch": 2.5825471698113205,
      "grad_norm": 1.181229591369629,
      "learning_rate": 9.974230621147907e-06,
      "loss": 0.6295,
      "num_input_tokens_seen": 2867496,
      "step": 4380
    },
    {
      "epoch": 2.5854952830188678,
      "grad_norm": 2.674422025680542,
      "learning_rate": 9.973969099253928e-06,
      "loss": 0.5812,
      "num_input_tokens_seen": 2870440,
      "step": 4385
    },
    {
      "epoch": 2.588443396226415,
      "grad_norm": 1.4279403686523438,
      "learning_rate": 9.973706260480194e-06,
      "loss": 0.4417,
      "num_input_tokens_seen": 2873512,
      "step": 4390
    },
    {
      "epoch": 2.5913915094339623,
      "grad_norm": 1.239289402961731,
      "learning_rate": 9.97344210489629e-06,
      "loss": 0.5264,
      "num_input_tokens_seen": 2877960,
      "step": 4395
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 1.59089994430542,
      "learning_rate": 9.973176632572158e-06,
      "loss": 0.4724,
      "num_input_tokens_seen": 2881096,
      "step": 4400
    },
    {
      "epoch": 2.5972877358490565,
      "grad_norm": 1.0678223371505737,
      "learning_rate": 9.972909843578076e-06,
      "loss": 0.4071,
      "num_input_tokens_seen": 2884136,
      "step": 4405
    },
    {
      "epoch": 2.6002358490566038,
      "grad_norm": 2.9783458709716797,
      "learning_rate": 9.972641737984681e-06,
      "loss": 0.5395,
      "num_input_tokens_seen": 2887464,
      "step": 4410
    },
    {
      "epoch": 2.603183962264151,
      "grad_norm": 1.2662245035171509,
      "learning_rate": 9.972372315862956e-06,
      "loss": 0.4885,
      "num_input_tokens_seen": 2892200,
      "step": 4415
    },
    {
      "epoch": 2.606132075471698,
      "grad_norm": 1.4217298030853271,
      "learning_rate": 9.97210157728423e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 2895272,
      "step": 4420
    },
    {
      "epoch": 2.609080188679245,
      "grad_norm": 1.7170484066009521,
      "learning_rate": 9.971829522320185e-06,
      "loss": 0.6165,
      "num_input_tokens_seen": 2897736,
      "step": 4425
    },
    {
      "epoch": 2.6120283018867925,
      "grad_norm": 0.9830440878868103,
      "learning_rate": 9.971556151042843e-06,
      "loss": 0.546,
      "num_input_tokens_seen": 2901000,
      "step": 4430
    },
    {
      "epoch": 2.6149764150943398,
      "grad_norm": 2.732825756072998,
      "learning_rate": 9.971281463524588e-06,
      "loss": 0.5484,
      "num_input_tokens_seen": 2903176,
      "step": 4435
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 1.5710411071777344,
      "learning_rate": 9.971005459838136e-06,
      "loss": 0.6439,
      "num_input_tokens_seen": 2905896,
      "step": 4440
    },
    {
      "epoch": 2.620872641509434,
      "grad_norm": 1.0999376773834229,
      "learning_rate": 9.970728140056567e-06,
      "loss": 0.5248,
      "num_input_tokens_seen": 2909896,
      "step": 4445
    },
    {
      "epoch": 2.623820754716981,
      "grad_norm": 2.056426525115967,
      "learning_rate": 9.9704495042533e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 2913224,
      "step": 4450
    },
    {
      "epoch": 2.6267688679245285,
      "grad_norm": 2.6993703842163086,
      "learning_rate": 9.970169552502105e-06,
      "loss": 0.5665,
      "num_input_tokens_seen": 2915784,
      "step": 4455
    },
    {
      "epoch": 2.6297169811320753,
      "grad_norm": 1.6140689849853516,
      "learning_rate": 9.969888284877102e-06,
      "loss": 0.4827,
      "num_input_tokens_seen": 2919656,
      "step": 4460
    },
    {
      "epoch": 2.6326650943396226,
      "grad_norm": 1.925849437713623,
      "learning_rate": 9.969605701452757e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 2923112,
      "step": 4465
    },
    {
      "epoch": 2.63561320754717,
      "grad_norm": 7.0577311515808105,
      "learning_rate": 9.969321802303882e-06,
      "loss": 0.57,
      "num_input_tokens_seen": 2925352,
      "step": 4470
    },
    {
      "epoch": 2.638561320754717,
      "grad_norm": 4.190489768981934,
      "learning_rate": 9.969036587505644e-06,
      "loss": 0.6037,
      "num_input_tokens_seen": 2928008,
      "step": 4475
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 2.312328338623047,
      "learning_rate": 9.968750057133555e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 2930504,
      "step": 4480
    },
    {
      "epoch": 2.6444575471698113,
      "grad_norm": 2.36303448677063,
      "learning_rate": 9.968462211263474e-06,
      "loss": 0.4787,
      "num_input_tokens_seen": 2933896,
      "step": 4485
    },
    {
      "epoch": 2.6474056603773586,
      "grad_norm": 1.784894347190857,
      "learning_rate": 9.96817304997161e-06,
      "loss": 0.5688,
      "num_input_tokens_seen": 2937288,
      "step": 4490
    },
    {
      "epoch": 2.6503537735849054,
      "grad_norm": 1.300693392753601,
      "learning_rate": 9.967882573334519e-06,
      "loss": 0.3662,
      "num_input_tokens_seen": 2941480,
      "step": 4495
    },
    {
      "epoch": 2.6533018867924527,
      "grad_norm": 4.16550874710083,
      "learning_rate": 9.967590781429106e-06,
      "loss": 0.3727,
      "num_input_tokens_seen": 2944712,
      "step": 4500
    },
    {
      "epoch": 2.65625,
      "grad_norm": 2.2218925952911377,
      "learning_rate": 9.967297674332625e-06,
      "loss": 0.6108,
      "num_input_tokens_seen": 2947624,
      "step": 4505
    },
    {
      "epoch": 2.6591981132075473,
      "grad_norm": 3.6585025787353516,
      "learning_rate": 9.967003252122675e-06,
      "loss": 0.5262,
      "num_input_tokens_seen": 2950408,
      "step": 4510
    },
    {
      "epoch": 2.6621462264150946,
      "grad_norm": 4.65001916885376,
      "learning_rate": 9.96670751487721e-06,
      "loss": 0.5287,
      "num_input_tokens_seen": 2954952,
      "step": 4515
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 1.918627381324768,
      "learning_rate": 9.966410462674525e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 2958440,
      "step": 4520
    },
    {
      "epoch": 2.6680424528301887,
      "grad_norm": 2.252490282058716,
      "learning_rate": 9.966112095593264e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 2961064,
      "step": 4525
    },
    {
      "epoch": 2.670990566037736,
      "grad_norm": 2.0524322986602783,
      "learning_rate": 9.965812413712425e-06,
      "loss": 0.3999,
      "num_input_tokens_seen": 2964392,
      "step": 4530
    },
    {
      "epoch": 2.673938679245283,
      "grad_norm": 1.2813520431518555,
      "learning_rate": 9.965511417111346e-06,
      "loss": 0.4303,
      "num_input_tokens_seen": 2969032,
      "step": 4535
    },
    {
      "epoch": 2.67688679245283,
      "grad_norm": 1.6433337926864624,
      "learning_rate": 9.96520910586972e-06,
      "loss": 0.48,
      "num_input_tokens_seen": 2973512,
      "step": 4540
    },
    {
      "epoch": 2.6798349056603774,
      "grad_norm": 0.9924342632293701,
      "learning_rate": 9.964905480067585e-06,
      "loss": 0.6018,
      "num_input_tokens_seen": 2977640,
      "step": 4545
    },
    {
      "epoch": 2.6827830188679247,
      "grad_norm": 2.7295196056365967,
      "learning_rate": 9.964600539785328e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 2980968,
      "step": 4550
    },
    {
      "epoch": 2.6857311320754715,
      "grad_norm": 2.2198092937469482,
      "learning_rate": 9.96429428510368e-06,
      "loss": 0.3353,
      "num_input_tokens_seen": 2984072,
      "step": 4555
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 1.506524920463562,
      "learning_rate": 9.963986716103724e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 2988264,
      "step": 4560
    },
    {
      "epoch": 2.691627358490566,
      "grad_norm": 1.4872632026672363,
      "learning_rate": 9.963677832866893e-06,
      "loss": 0.6408,
      "num_input_tokens_seen": 2991880,
      "step": 4565
    },
    {
      "epoch": 2.694575471698113,
      "grad_norm": 2.669271230697632,
      "learning_rate": 9.963367635474962e-06,
      "loss": 0.4877,
      "num_input_tokens_seen": 2994152,
      "step": 4570
    },
    {
      "epoch": 2.6975235849056602,
      "grad_norm": 1.7621372938156128,
      "learning_rate": 9.96305612401006e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 2997448,
      "step": 4575
    },
    {
      "epoch": 2.7004716981132075,
      "grad_norm": 3.288945198059082,
      "learning_rate": 9.96274329855466e-06,
      "loss": 0.3915,
      "num_input_tokens_seen": 3000040,
      "step": 4580
    },
    {
      "epoch": 2.703419811320755,
      "grad_norm": 1.1441066265106201,
      "learning_rate": 9.962429159191583e-06,
      "loss": 0.5408,
      "num_input_tokens_seen": 3002824,
      "step": 4585
    },
    {
      "epoch": 2.706367924528302,
      "grad_norm": 1.7270110845565796,
      "learning_rate": 9.962113706003997e-06,
      "loss": 0.5862,
      "num_input_tokens_seen": 3006504,
      "step": 4590
    },
    {
      "epoch": 2.709316037735849,
      "grad_norm": 1.9242558479309082,
      "learning_rate": 9.961796939075424e-06,
      "loss": 0.5201,
      "num_input_tokens_seen": 3009256,
      "step": 4595
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 1.8738107681274414,
      "learning_rate": 9.961478858489728e-06,
      "loss": 0.4617,
      "num_input_tokens_seen": 3012008,
      "step": 4600
    },
    {
      "epoch": 2.7152122641509435,
      "grad_norm": 1.5541669130325317,
      "learning_rate": 9.961159464331119e-06,
      "loss": 0.4788,
      "num_input_tokens_seen": 3015496,
      "step": 4605
    },
    {
      "epoch": 2.7181603773584904,
      "grad_norm": 2.70796537399292,
      "learning_rate": 9.960838756684161e-06,
      "loss": 0.6475,
      "num_input_tokens_seen": 3018440,
      "step": 4610
    },
    {
      "epoch": 2.7211084905660377,
      "grad_norm": 3.605987787246704,
      "learning_rate": 9.960516735633764e-06,
      "loss": 0.5981,
      "num_input_tokens_seen": 3021288,
      "step": 4615
    },
    {
      "epoch": 2.724056603773585,
      "grad_norm": 2.5952436923980713,
      "learning_rate": 9.960193401265181e-06,
      "loss": 0.4768,
      "num_input_tokens_seen": 3024040,
      "step": 4620
    },
    {
      "epoch": 2.7270047169811322,
      "grad_norm": 1.60612952709198,
      "learning_rate": 9.959868753664018e-06,
      "loss": 0.5306,
      "num_input_tokens_seen": 3027976,
      "step": 4625
    },
    {
      "epoch": 2.7299528301886795,
      "grad_norm": 1.4446420669555664,
      "learning_rate": 9.959542792916227e-06,
      "loss": 0.5723,
      "num_input_tokens_seen": 3030600,
      "step": 4630
    },
    {
      "epoch": 2.7329009433962264,
      "grad_norm": 2.592630624771118,
      "learning_rate": 9.959215519108108e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 3033768,
      "step": 4635
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.8739300966262817,
      "learning_rate": 9.958886932326306e-06,
      "loss": 0.6468,
      "num_input_tokens_seen": 3036904,
      "step": 4640
    },
    {
      "epoch": 2.7387971698113205,
      "grad_norm": 3.1662185192108154,
      "learning_rate": 9.958557032657817e-06,
      "loss": 0.6664,
      "num_input_tokens_seen": 3039944,
      "step": 4645
    },
    {
      "epoch": 2.7417452830188678,
      "grad_norm": 2.2730956077575684,
      "learning_rate": 9.958225820189984e-06,
      "loss": 0.5521,
      "num_input_tokens_seen": 3043208,
      "step": 4650
    },
    {
      "epoch": 2.744693396226415,
      "grad_norm": 1.6620315313339233,
      "learning_rate": 9.957893295010495e-06,
      "loss": 0.4871,
      "num_input_tokens_seen": 3046184,
      "step": 4655
    },
    {
      "epoch": 2.7476415094339623,
      "grad_norm": 1.44398832321167,
      "learning_rate": 9.957559457207391e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 3049256,
      "step": 4660
    },
    {
      "epoch": 2.7505896226415096,
      "grad_norm": 8.704010963439941,
      "learning_rate": 9.957224306869053e-06,
      "loss": 0.5839,
      "num_input_tokens_seen": 3052712,
      "step": 4665
    },
    {
      "epoch": 2.7535377358490565,
      "grad_norm": 1.0286273956298828,
      "learning_rate": 9.956887844084216e-06,
      "loss": 0.5415,
      "num_input_tokens_seen": 3057096,
      "step": 4670
    },
    {
      "epoch": 2.7564858490566038,
      "grad_norm": 4.882441520690918,
      "learning_rate": 9.956550068941958e-06,
      "loss": 0.5298,
      "num_input_tokens_seen": 3060072,
      "step": 4675
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 2.145540475845337,
      "learning_rate": 9.95621098153171e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 3063528,
      "step": 4680
    },
    {
      "epoch": 2.762382075471698,
      "grad_norm": 1.8061028718948364,
      "learning_rate": 9.955870581943243e-06,
      "loss": 0.5337,
      "num_input_tokens_seen": 3066696,
      "step": 4685
    },
    {
      "epoch": 2.765330188679245,
      "grad_norm": 1.9717352390289307,
      "learning_rate": 9.955528870266681e-06,
      "loss": 0.5496,
      "num_input_tokens_seen": 3070056,
      "step": 4690
    },
    {
      "epoch": 2.7682783018867925,
      "grad_norm": 1.2642983198165894,
      "learning_rate": 9.955185846592495e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 3073320,
      "step": 4695
    },
    {
      "epoch": 2.7712264150943398,
      "grad_norm": 1.3834887742996216,
      "learning_rate": 9.9548415110115e-06,
      "loss": 0.5604,
      "num_input_tokens_seen": 3076232,
      "step": 4700
    },
    {
      "epoch": 2.774174528301887,
      "grad_norm": 2.2347700595855713,
      "learning_rate": 9.95449586361486e-06,
      "loss": 0.4881,
      "num_input_tokens_seen": 3079432,
      "step": 4705
    },
    {
      "epoch": 2.777122641509434,
      "grad_norm": 2.302741765975952,
      "learning_rate": 9.954148904494085e-06,
      "loss": 0.4896,
      "num_input_tokens_seen": 3082504,
      "step": 4710
    },
    {
      "epoch": 2.780070754716981,
      "grad_norm": 1.7235546112060547,
      "learning_rate": 9.95380063374104e-06,
      "loss": 0.6258,
      "num_input_tokens_seen": 3085544,
      "step": 4715
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 2.158597469329834,
      "learning_rate": 9.953451051447927e-06,
      "loss": 0.3799,
      "num_input_tokens_seen": 3087848,
      "step": 4720
    },
    {
      "epoch": 2.7859669811320753,
      "grad_norm": 0.9890583157539368,
      "learning_rate": 9.953100157707299e-06,
      "loss": 0.5326,
      "num_input_tokens_seen": 3090888,
      "step": 4725
    },
    {
      "epoch": 2.7889150943396226,
      "grad_norm": 2.284928321838379,
      "learning_rate": 9.952747952612056e-06,
      "loss": 0.3655,
      "num_input_tokens_seen": 3094152,
      "step": 4730
    },
    {
      "epoch": 2.79186320754717,
      "grad_norm": 2.142786741256714,
      "learning_rate": 9.952394436255451e-06,
      "loss": 0.4959,
      "num_input_tokens_seen": 3097256,
      "step": 4735
    },
    {
      "epoch": 2.794811320754717,
      "grad_norm": 1.0626708269119263,
      "learning_rate": 9.952039608731072e-06,
      "loss": 0.5416,
      "num_input_tokens_seen": 3101064,
      "step": 4740
    },
    {
      "epoch": 2.797759433962264,
      "grad_norm": 1.4304851293563843,
      "learning_rate": 9.951683470132868e-06,
      "loss": 0.5712,
      "num_input_tokens_seen": 3104136,
      "step": 4745
    },
    {
      "epoch": 2.8007075471698113,
      "grad_norm": 1.5543434619903564,
      "learning_rate": 9.951326020555122e-06,
      "loss": 0.7464,
      "num_input_tokens_seen": 3107080,
      "step": 4750
    },
    {
      "epoch": 2.8036556603773586,
      "grad_norm": 2.5784034729003906,
      "learning_rate": 9.950967260092473e-06,
      "loss": 0.5466,
      "num_input_tokens_seen": 3110280,
      "step": 4755
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 2.759263753890991,
      "learning_rate": 9.950607188839905e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 3112840,
      "step": 4760
    },
    {
      "epoch": 2.8095518867924527,
      "grad_norm": 1.4480006694793701,
      "learning_rate": 9.950245806892749e-06,
      "loss": 0.4637,
      "num_input_tokens_seen": 3115528,
      "step": 4765
    },
    {
      "epoch": 2.8125,
      "grad_norm": 1.1963011026382446,
      "learning_rate": 9.94988311434668e-06,
      "loss": 0.5564,
      "num_input_tokens_seen": 3118696,
      "step": 4770
    },
    {
      "epoch": 2.8154481132075473,
      "grad_norm": 1.367799997329712,
      "learning_rate": 9.949519111297723e-06,
      "loss": 0.4772,
      "num_input_tokens_seen": 3123336,
      "step": 4775
    },
    {
      "epoch": 2.8183962264150946,
      "grad_norm": 1.6084316968917847,
      "learning_rate": 9.949153797842252e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 3126248,
      "step": 4780
    },
    {
      "epoch": 2.8213443396226414,
      "grad_norm": 1.4331570863723755,
      "learning_rate": 9.948787174076982e-06,
      "loss": 0.5442,
      "num_input_tokens_seen": 3129928,
      "step": 4785
    },
    {
      "epoch": 2.8242924528301887,
      "grad_norm": 1.7055323123931885,
      "learning_rate": 9.948419240098978e-06,
      "loss": 0.5302,
      "num_input_tokens_seen": 3133640,
      "step": 4790
    },
    {
      "epoch": 2.827240566037736,
      "grad_norm": 1.2413452863693237,
      "learning_rate": 9.948049996005657e-06,
      "loss": 0.4855,
      "num_input_tokens_seen": 3136456,
      "step": 4795
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 1.3646422624588013,
      "learning_rate": 9.947679441894773e-06,
      "loss": 0.5353,
      "num_input_tokens_seen": 3139432,
      "step": 4800
    },
    {
      "epoch": 2.83313679245283,
      "grad_norm": 2.125462770462036,
      "learning_rate": 9.947307577864433e-06,
      "loss": 0.5748,
      "num_input_tokens_seen": 3142664,
      "step": 4805
    },
    {
      "epoch": 2.8360849056603774,
      "grad_norm": 1.6998240947723389,
      "learning_rate": 9.94693440401309e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 3146024,
      "step": 4810
    },
    {
      "epoch": 2.8390330188679247,
      "grad_norm": 2.636845111846924,
      "learning_rate": 9.946559920439545e-06,
      "loss": 0.5249,
      "num_input_tokens_seen": 3148424,
      "step": 4815
    },
    {
      "epoch": 2.8419811320754715,
      "grad_norm": 1.6378189325332642,
      "learning_rate": 9.946184127242942e-06,
      "loss": 0.3194,
      "num_input_tokens_seen": 3151112,
      "step": 4820
    },
    {
      "epoch": 2.844929245283019,
      "grad_norm": 2.588013172149658,
      "learning_rate": 9.945807024522774e-06,
      "loss": 0.4474,
      "num_input_tokens_seen": 3154280,
      "step": 4825
    },
    {
      "epoch": 2.847877358490566,
      "grad_norm": 1.1889394521713257,
      "learning_rate": 9.945428612378881e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 3157224,
      "step": 4830
    },
    {
      "epoch": 2.850825471698113,
      "grad_norm": 0.9296377897262573,
      "learning_rate": 9.94504889091145e-06,
      "loss": 0.6202,
      "num_input_tokens_seen": 3160648,
      "step": 4835
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 0.9155482053756714,
      "learning_rate": 9.944667860221013e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 3163816,
      "step": 4840
    },
    {
      "epoch": 2.8567216981132075,
      "grad_norm": 0.9077715277671814,
      "learning_rate": 9.944285520408448e-06,
      "loss": 0.5117,
      "num_input_tokens_seen": 3166952,
      "step": 4845
    },
    {
      "epoch": 2.859669811320755,
      "grad_norm": 2.804185390472412,
      "learning_rate": 9.943901871574984e-06,
      "loss": 0.5421,
      "num_input_tokens_seen": 3169512,
      "step": 4850
    },
    {
      "epoch": 2.862617924528302,
      "grad_norm": 1.7270584106445312,
      "learning_rate": 9.943516913822192e-06,
      "loss": 0.5119,
      "num_input_tokens_seen": 3174280,
      "step": 4855
    },
    {
      "epoch": 2.865566037735849,
      "grad_norm": 1.7630581855773926,
      "learning_rate": 9.943130647251994e-06,
      "loss": 0.4825,
      "num_input_tokens_seen": 3177288,
      "step": 4860
    },
    {
      "epoch": 2.8685141509433962,
      "grad_norm": 1.7554064989089966,
      "learning_rate": 9.94274307196665e-06,
      "loss": 0.4357,
      "num_input_tokens_seen": 3180168,
      "step": 4865
    },
    {
      "epoch": 2.8714622641509435,
      "grad_norm": 2.639463186264038,
      "learning_rate": 9.942354188068778e-06,
      "loss": 0.4306,
      "num_input_tokens_seen": 3182728,
      "step": 4870
    },
    {
      "epoch": 2.8744103773584904,
      "grad_norm": 1.4284615516662598,
      "learning_rate": 9.941963995661333e-06,
      "loss": 0.4662,
      "num_input_tokens_seen": 3185864,
      "step": 4875
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 5.638134956359863,
      "learning_rate": 9.941572494847622e-06,
      "loss": 0.5815,
      "num_input_tokens_seen": 3189288,
      "step": 4880
    },
    {
      "epoch": 2.880306603773585,
      "grad_norm": 7.441652774810791,
      "learning_rate": 9.941179685731297e-06,
      "loss": 0.5668,
      "num_input_tokens_seen": 3191656,
      "step": 4885
    },
    {
      "epoch": 2.8832547169811322,
      "grad_norm": 5.038857936859131,
      "learning_rate": 9.940785568416354e-06,
      "loss": 0.5621,
      "num_input_tokens_seen": 3194760,
      "step": 4890
    },
    {
      "epoch": 2.8862028301886795,
      "grad_norm": 1.3169025182724,
      "learning_rate": 9.940390143007137e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 3197896,
      "step": 4895
    },
    {
      "epoch": 2.8891509433962264,
      "grad_norm": 1.0869367122650146,
      "learning_rate": 9.939993409608339e-06,
      "loss": 0.5072,
      "num_input_tokens_seen": 3201192,
      "step": 4900
    },
    {
      "epoch": 2.8920990566037736,
      "grad_norm": 3.765723943710327,
      "learning_rate": 9.939595368324996e-06,
      "loss": 0.5593,
      "num_input_tokens_seen": 3205064,
      "step": 4905
    },
    {
      "epoch": 2.8950471698113205,
      "grad_norm": 1.3856440782546997,
      "learning_rate": 9.93919601926249e-06,
      "loss": 0.5238,
      "num_input_tokens_seen": 3208328,
      "step": 4910
    },
    {
      "epoch": 2.8979952830188678,
      "grad_norm": 1.671263575553894,
      "learning_rate": 9.938795362526552e-06,
      "loss": 0.5265,
      "num_input_tokens_seen": 3211464,
      "step": 4915
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 1.8862881660461426,
      "learning_rate": 9.938393398223255e-06,
      "loss": 0.4576,
      "num_input_tokens_seen": 3215240,
      "step": 4920
    },
    {
      "epoch": 2.9038915094339623,
      "grad_norm": 1.6615010499954224,
      "learning_rate": 9.937990126459024e-06,
      "loss": 0.4846,
      "num_input_tokens_seen": 3218760,
      "step": 4925
    },
    {
      "epoch": 2.9068396226415096,
      "grad_norm": 3.0793702602386475,
      "learning_rate": 9.937585547340624e-06,
      "loss": 0.4252,
      "num_input_tokens_seen": 3222152,
      "step": 4930
    },
    {
      "epoch": 2.9097877358490565,
      "grad_norm": 1.0689680576324463,
      "learning_rate": 9.937179660975174e-06,
      "loss": 0.4511,
      "num_input_tokens_seen": 3225832,
      "step": 4935
    },
    {
      "epoch": 2.9127358490566038,
      "grad_norm": 1.39501953125,
      "learning_rate": 9.936772467470127e-06,
      "loss": 0.4965,
      "num_input_tokens_seen": 3228776,
      "step": 4940
    },
    {
      "epoch": 2.915683962264151,
      "grad_norm": 1.4777722358703613,
      "learning_rate": 9.936363966933294e-06,
      "loss": 0.3747,
      "num_input_tokens_seen": 3232232,
      "step": 4945
    },
    {
      "epoch": 2.918632075471698,
      "grad_norm": 3.479318857192993,
      "learning_rate": 9.935954159472828e-06,
      "loss": 0.508,
      "num_input_tokens_seen": 3234984,
      "step": 4950
    },
    {
      "epoch": 2.921580188679245,
      "grad_norm": 4.67243766784668,
      "learning_rate": 9.935543045197222e-06,
      "loss": 0.3633,
      "num_input_tokens_seen": 3238088,
      "step": 4955
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 4.435565948486328,
      "learning_rate": 9.935130624215326e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 3241320,
      "step": 4960
    },
    {
      "epoch": 2.9274764150943398,
      "grad_norm": 1.82668137550354,
      "learning_rate": 9.934716896636329e-06,
      "loss": 0.5629,
      "num_input_tokens_seen": 3244872,
      "step": 4965
    },
    {
      "epoch": 2.930424528301887,
      "grad_norm": 1.7953466176986694,
      "learning_rate": 9.934301862569764e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 3247816,
      "step": 4970
    },
    {
      "epoch": 2.933372641509434,
      "grad_norm": 1.3861063718795776,
      "learning_rate": 9.933885522125517e-06,
      "loss": 0.5191,
      "num_input_tokens_seen": 3251688,
      "step": 4975
    },
    {
      "epoch": 2.936320754716981,
      "grad_norm": 1.5534034967422485,
      "learning_rate": 9.933467875413813e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 3254728,
      "step": 4980
    },
    {
      "epoch": 2.9392688679245285,
      "grad_norm": 0.7914652228355408,
      "learning_rate": 9.933048922545227e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 3262728,
      "step": 4985
    },
    {
      "epoch": 2.9422169811320753,
      "grad_norm": 3.399104356765747,
      "learning_rate": 9.932628663630679e-06,
      "loss": 0.4888,
      "num_input_tokens_seen": 3266216,
      "step": 4990
    },
    {
      "epoch": 2.9451650943396226,
      "grad_norm": 1.7984040975570679,
      "learning_rate": 9.932207098781432e-06,
      "loss": 0.6217,
      "num_input_tokens_seen": 3269736,
      "step": 4995
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 1.4967143535614014,
      "learning_rate": 9.931784228109102e-06,
      "loss": 0.5602,
      "num_input_tokens_seen": 3273064,
      "step": 5000
    },
    {
      "epoch": 2.951061320754717,
      "grad_norm": 1.3157926797866821,
      "learning_rate": 9.93136005172564e-06,
      "loss": 0.5518,
      "num_input_tokens_seen": 3276072,
      "step": 5005
    },
    {
      "epoch": 2.954009433962264,
      "grad_norm": 4.00148344039917,
      "learning_rate": 9.930934569743354e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 3278760,
      "step": 5010
    },
    {
      "epoch": 2.9569575471698113,
      "grad_norm": 2.496894598007202,
      "learning_rate": 9.930507782274888e-06,
      "loss": 0.5876,
      "num_input_tokens_seen": 3285352,
      "step": 5015
    },
    {
      "epoch": 2.9599056603773586,
      "grad_norm": 1.4927951097488403,
      "learning_rate": 9.930079689433236e-06,
      "loss": 0.5847,
      "num_input_tokens_seen": 3288520,
      "step": 5020
    },
    {
      "epoch": 2.9628537735849054,
      "grad_norm": 2.139988899230957,
      "learning_rate": 9.92965029133174e-06,
      "loss": 0.4678,
      "num_input_tokens_seen": 3292520,
      "step": 5025
    },
    {
      "epoch": 2.9658018867924527,
      "grad_norm": 1.328861951828003,
      "learning_rate": 9.929219588084084e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 3295880,
      "step": 5030
    },
    {
      "epoch": 2.96875,
      "grad_norm": 2.2210896015167236,
      "learning_rate": 9.9287875798043e-06,
      "loss": 0.6671,
      "num_input_tokens_seen": 3299144,
      "step": 5035
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 2.367943286895752,
      "learning_rate": 9.92835426660676e-06,
      "loss": 0.6987,
      "num_input_tokens_seen": 3302408,
      "step": 5040
    },
    {
      "epoch": 2.9746462264150946,
      "grad_norm": 1.617716908454895,
      "learning_rate": 9.927919648606188e-06,
      "loss": 0.6339,
      "num_input_tokens_seen": 3305128,
      "step": 5045
    },
    {
      "epoch": 2.9775943396226414,
      "grad_norm": 2.1770589351654053,
      "learning_rate": 9.927483725917652e-06,
      "loss": 0.5484,
      "num_input_tokens_seen": 3308104,
      "step": 5050
    },
    {
      "epoch": 2.9805424528301887,
      "grad_norm": 1.454712152481079,
      "learning_rate": 9.927046498656562e-06,
      "loss": 0.5671,
      "num_input_tokens_seen": 3311848,
      "step": 5055
    },
    {
      "epoch": 2.983490566037736,
      "grad_norm": 1.349992275238037,
      "learning_rate": 9.926607966938679e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 3315016,
      "step": 5060
    },
    {
      "epoch": 2.986438679245283,
      "grad_norm": 1.9763844013214111,
      "learning_rate": 9.926168130880103e-06,
      "loss": 0.4367,
      "num_input_tokens_seen": 3317704,
      "step": 5065
    },
    {
      "epoch": 2.98938679245283,
      "grad_norm": 2.551924467086792,
      "learning_rate": 9.925726990597283e-06,
      "loss": 0.6346,
      "num_input_tokens_seen": 3320456,
      "step": 5070
    },
    {
      "epoch": 2.9923349056603774,
      "grad_norm": 1.9989973306655884,
      "learning_rate": 9.925284546207015e-06,
      "loss": 0.5177,
      "num_input_tokens_seen": 3322920,
      "step": 5075
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 1.257860779762268,
      "learning_rate": 9.924840797826436e-06,
      "loss": 0.3891,
      "num_input_tokens_seen": 3326024,
      "step": 5080
    },
    {
      "epoch": 2.9982311320754715,
      "grad_norm": 1.0323781967163086,
      "learning_rate": 9.924395745573029e-06,
      "loss": 0.5474,
      "num_input_tokens_seen": 3329096,
      "step": 5085
    },
    {
      "epoch": 3.001179245283019,
      "grad_norm": 1.705366849899292,
      "learning_rate": 9.923949389564629e-06,
      "loss": 0.4774,
      "num_input_tokens_seen": 3332128,
      "step": 5090
    },
    {
      "epoch": 3.004127358490566,
      "grad_norm": 1.1408157348632812,
      "learning_rate": 9.923501729919404e-06,
      "loss": 0.451,
      "num_input_tokens_seen": 3335488,
      "step": 5095
    },
    {
      "epoch": 3.0070754716981134,
      "grad_norm": 1.3309420347213745,
      "learning_rate": 9.923052766755878e-06,
      "loss": 0.5556,
      "num_input_tokens_seen": 3338240,
      "step": 5100
    },
    {
      "epoch": 3.0100235849056602,
      "grad_norm": 1.6096702814102173,
      "learning_rate": 9.922602500192914e-06,
      "loss": 0.4645,
      "num_input_tokens_seen": 3341568,
      "step": 5105
    },
    {
      "epoch": 3.0129716981132075,
      "grad_norm": 1.6139113903045654,
      "learning_rate": 9.922150930349725e-06,
      "loss": 0.56,
      "num_input_tokens_seen": 3344800,
      "step": 5110
    },
    {
      "epoch": 3.015919811320755,
      "grad_norm": 1.2481412887573242,
      "learning_rate": 9.921698057345863e-06,
      "loss": 0.4766,
      "num_input_tokens_seen": 3347584,
      "step": 5115
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 1.8198604583740234,
      "learning_rate": 9.921243881301229e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 3350720,
      "step": 5120
    },
    {
      "epoch": 3.021816037735849,
      "grad_norm": 0.9110130071640015,
      "learning_rate": 9.920788402336068e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 3353344,
      "step": 5125
    },
    {
      "epoch": 3.0247641509433962,
      "grad_norm": 1.5857529640197754,
      "learning_rate": 9.92033162057097e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 3356672,
      "step": 5130
    },
    {
      "epoch": 3.0277122641509435,
      "grad_norm": 1.6013877391815186,
      "learning_rate": 9.919873536126869e-06,
      "loss": 0.4832,
      "num_input_tokens_seen": 3359168,
      "step": 5135
    },
    {
      "epoch": 3.0306603773584904,
      "grad_norm": 1.7557365894317627,
      "learning_rate": 9.919414149125046e-06,
      "loss": 0.5372,
      "num_input_tokens_seen": 3362208,
      "step": 5140
    },
    {
      "epoch": 3.0336084905660377,
      "grad_norm": 0.9931708574295044,
      "learning_rate": 9.918953459687126e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 3365504,
      "step": 5145
    },
    {
      "epoch": 3.036556603773585,
      "grad_norm": 1.311785340309143,
      "learning_rate": 9.918491467935078e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 3368544,
      "step": 5150
    },
    {
      "epoch": 3.0395047169811322,
      "grad_norm": 2.127629280090332,
      "learning_rate": 9.918028173991218e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 3371488,
      "step": 5155
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 1.1830217838287354,
      "learning_rate": 9.917563577978202e-06,
      "loss": 0.5553,
      "num_input_tokens_seen": 3374464,
      "step": 5160
    },
    {
      "epoch": 3.0454009433962264,
      "grad_norm": 2.083371162414551,
      "learning_rate": 9.917097680019035e-06,
      "loss": 0.5466,
      "num_input_tokens_seen": 3377184,
      "step": 5165
    },
    {
      "epoch": 3.0483490566037736,
      "grad_norm": 1.0516231060028076,
      "learning_rate": 9.916630480237066e-06,
      "loss": 0.5206,
      "num_input_tokens_seen": 3380384,
      "step": 5170
    },
    {
      "epoch": 3.051297169811321,
      "grad_norm": 0.4436565637588501,
      "learning_rate": 9.916161978755988e-06,
      "loss": 0.4604,
      "num_input_tokens_seen": 3385664,
      "step": 5175
    },
    {
      "epoch": 3.0542452830188678,
      "grad_norm": 1.0447680950164795,
      "learning_rate": 9.915692175699838e-06,
      "loss": 0.4847,
      "num_input_tokens_seen": 3388384,
      "step": 5180
    },
    {
      "epoch": 3.057193396226415,
      "grad_norm": 3.093412160873413,
      "learning_rate": 9.915221071193e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 3392096,
      "step": 5185
    },
    {
      "epoch": 3.0601415094339623,
      "grad_norm": 1.1626393795013428,
      "learning_rate": 9.914748665360199e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 3395104,
      "step": 5190
    },
    {
      "epoch": 3.0630896226415096,
      "grad_norm": 0.9592787027359009,
      "learning_rate": 9.914274958326507e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 3399168,
      "step": 5195
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 2.2505605220794678,
      "learning_rate": 9.913799950217341e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 3401824,
      "step": 5200
    },
    {
      "epoch": 3.0689858490566038,
      "grad_norm": 1.9642679691314697,
      "learning_rate": 9.91332364115846e-06,
      "loss": 0.5866,
      "num_input_tokens_seen": 3405568,
      "step": 5205
    },
    {
      "epoch": 3.071933962264151,
      "grad_norm": 1.078163504600525,
      "learning_rate": 9.912846031275972e-06,
      "loss": 0.5217,
      "num_input_tokens_seen": 3409088,
      "step": 5210
    },
    {
      "epoch": 3.074882075471698,
      "grad_norm": 2.7968063354492188,
      "learning_rate": 9.912367120696322e-06,
      "loss": 0.6669,
      "num_input_tokens_seen": 3412160,
      "step": 5215
    },
    {
      "epoch": 3.077830188679245,
      "grad_norm": 2.055645227432251,
      "learning_rate": 9.911886909546307e-06,
      "loss": 0.5173,
      "num_input_tokens_seen": 3415296,
      "step": 5220
    },
    {
      "epoch": 3.0807783018867925,
      "grad_norm": 2.043503999710083,
      "learning_rate": 9.911405397953063e-06,
      "loss": 0.5942,
      "num_input_tokens_seen": 3418848,
      "step": 5225
    },
    {
      "epoch": 3.0837264150943398,
      "grad_norm": 1.1220425367355347,
      "learning_rate": 9.910922586044073e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 3422560,
      "step": 5230
    },
    {
      "epoch": 3.0866745283018866,
      "grad_norm": 0.7859375476837158,
      "learning_rate": 9.910438473947163e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 3425216,
      "step": 5235
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 1.6790266036987305,
      "learning_rate": 9.909953061790506e-06,
      "loss": 0.529,
      "num_input_tokens_seen": 3428128,
      "step": 5240
    },
    {
      "epoch": 3.092570754716981,
      "grad_norm": 1.3805241584777832,
      "learning_rate": 9.909466349702613e-06,
      "loss": 0.46,
      "num_input_tokens_seen": 3431200,
      "step": 5245
    },
    {
      "epoch": 3.0955188679245285,
      "grad_norm": 4.8404765129089355,
      "learning_rate": 9.908978337812348e-06,
      "loss": 0.552,
      "num_input_tokens_seen": 3434240,
      "step": 5250
    },
    {
      "epoch": 3.0984669811320753,
      "grad_norm": 1.8290951251983643,
      "learning_rate": 9.908489026248909e-06,
      "loss": 0.5377,
      "num_input_tokens_seen": 3436992,
      "step": 5255
    },
    {
      "epoch": 3.1014150943396226,
      "grad_norm": 2.0605859756469727,
      "learning_rate": 9.907998415141846e-06,
      "loss": 0.6652,
      "num_input_tokens_seen": 3440288,
      "step": 5260
    },
    {
      "epoch": 3.10436320754717,
      "grad_norm": 2.052490234375,
      "learning_rate": 9.907506504621052e-06,
      "loss": 0.3601,
      "num_input_tokens_seen": 3443008,
      "step": 5265
    },
    {
      "epoch": 3.107311320754717,
      "grad_norm": 1.2243776321411133,
      "learning_rate": 9.907013294816759e-06,
      "loss": 0.5469,
      "num_input_tokens_seen": 3446400,
      "step": 5270
    },
    {
      "epoch": 3.110259433962264,
      "grad_norm": 1.295063853263855,
      "learning_rate": 9.906518785859548e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 3449152,
      "step": 5275
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 1.7371782064437866,
      "learning_rate": 9.906022977880344e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 3452928,
      "step": 5280
    },
    {
      "epoch": 3.1161556603773586,
      "grad_norm": 1.382851004600525,
      "learning_rate": 9.905525871010412e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 3456864,
      "step": 5285
    },
    {
      "epoch": 3.119103773584906,
      "grad_norm": 2.666731834411621,
      "learning_rate": 9.905027465381363e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 3459680,
      "step": 5290
    },
    {
      "epoch": 3.1220518867924527,
      "grad_norm": 2.3106372356414795,
      "learning_rate": 9.904527761125155e-06,
      "loss": 0.5648,
      "num_input_tokens_seen": 3462560,
      "step": 5295
    },
    {
      "epoch": 3.125,
      "grad_norm": 1.2248233556747437,
      "learning_rate": 9.904026758374083e-06,
      "loss": 0.5403,
      "num_input_tokens_seen": 3466816,
      "step": 5300
    },
    {
      "epoch": 3.1279481132075473,
      "grad_norm": 4.964900016784668,
      "learning_rate": 9.903524457260794e-06,
      "loss": 0.5416,
      "num_input_tokens_seen": 3470400,
      "step": 5305
    },
    {
      "epoch": 3.1308962264150946,
      "grad_norm": 1.0680221319198608,
      "learning_rate": 9.90302085791827e-06,
      "loss": 0.6081,
      "num_input_tokens_seen": 3472800,
      "step": 5310
    },
    {
      "epoch": 3.1338443396226414,
      "grad_norm": 1.6968414783477783,
      "learning_rate": 9.902515960479844e-06,
      "loss": 0.4105,
      "num_input_tokens_seen": 3476128,
      "step": 5315
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 2.9375834465026855,
      "learning_rate": 9.902009765079188e-06,
      "loss": 0.5889,
      "num_input_tokens_seen": 3479072,
      "step": 5320
    },
    {
      "epoch": 3.139740566037736,
      "grad_norm": 2.445396900177002,
      "learning_rate": 9.90150227185032e-06,
      "loss": 0.5853,
      "num_input_tokens_seen": 3482496,
      "step": 5325
    },
    {
      "epoch": 3.142688679245283,
      "grad_norm": 2.9998836517333984,
      "learning_rate": 9.900993480927603e-06,
      "loss": 0.4618,
      "num_input_tokens_seen": 3484800,
      "step": 5330
    },
    {
      "epoch": 3.14563679245283,
      "grad_norm": 1.9999650716781616,
      "learning_rate": 9.90048339244574e-06,
      "loss": 0.4694,
      "num_input_tokens_seen": 3488096,
      "step": 5335
    },
    {
      "epoch": 3.1485849056603774,
      "grad_norm": 2.0825369358062744,
      "learning_rate": 9.899972006539776e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 3491552,
      "step": 5340
    },
    {
      "epoch": 3.1515330188679247,
      "grad_norm": 3.4481022357940674,
      "learning_rate": 9.899459323345106e-06,
      "loss": 0.5736,
      "num_input_tokens_seen": 3496064,
      "step": 5345
    },
    {
      "epoch": 3.1544811320754715,
      "grad_norm": 1.7615635395050049,
      "learning_rate": 9.898945342997467e-06,
      "loss": 0.4537,
      "num_input_tokens_seen": 3498752,
      "step": 5350
    },
    {
      "epoch": 3.157429245283019,
      "grad_norm": 2.2619264125823975,
      "learning_rate": 9.898430065632933e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 3501376,
      "step": 5355
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 1.4752082824707031,
      "learning_rate": 9.897913491387929e-06,
      "loss": 0.4711,
      "num_input_tokens_seen": 3505824,
      "step": 5360
    },
    {
      "epoch": 3.1633254716981134,
      "grad_norm": 1.4426079988479614,
      "learning_rate": 9.897395620399219e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 3509632,
      "step": 5365
    },
    {
      "epoch": 3.1662735849056602,
      "grad_norm": 1.3994078636169434,
      "learning_rate": 9.896876452803913e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 3512736,
      "step": 5370
    },
    {
      "epoch": 3.1692216981132075,
      "grad_norm": 3.9959914684295654,
      "learning_rate": 9.896355988739461e-06,
      "loss": 0.3655,
      "num_input_tokens_seen": 3515488,
      "step": 5375
    },
    {
      "epoch": 3.172169811320755,
      "grad_norm": 1.6810894012451172,
      "learning_rate": 9.895834228343658e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 3519168,
      "step": 5380
    },
    {
      "epoch": 3.175117924528302,
      "grad_norm": 1.751395583152771,
      "learning_rate": 9.895311171754644e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 3521760,
      "step": 5385
    },
    {
      "epoch": 3.178066037735849,
      "grad_norm": 2.0702550411224365,
      "learning_rate": 9.8947868191109e-06,
      "loss": 0.5236,
      "num_input_tokens_seen": 3524416,
      "step": 5390
    },
    {
      "epoch": 3.1810141509433962,
      "grad_norm": 2.0399317741394043,
      "learning_rate": 9.894261170551249e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 3529600,
      "step": 5395
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 1.3156852722167969,
      "learning_rate": 9.893734226214861e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 3533120,
      "step": 5400
    },
    {
      "epoch": 3.1869103773584904,
      "grad_norm": 1.1881765127182007,
      "learning_rate": 9.893205986241246e-06,
      "loss": 0.3586,
      "num_input_tokens_seen": 3535552,
      "step": 5405
    },
    {
      "epoch": 3.1898584905660377,
      "grad_norm": 2.5548970699310303,
      "learning_rate": 9.892676450770257e-06,
      "loss": 0.4467,
      "num_input_tokens_seen": 3538432,
      "step": 5410
    },
    {
      "epoch": 3.192806603773585,
      "grad_norm": 1.276084065437317,
      "learning_rate": 9.892145619942092e-06,
      "loss": 0.4814,
      "num_input_tokens_seen": 3542208,
      "step": 5415
    },
    {
      "epoch": 3.1957547169811322,
      "grad_norm": 0.9855227470397949,
      "learning_rate": 9.891613493897289e-06,
      "loss": 0.4137,
      "num_input_tokens_seen": 3546368,
      "step": 5420
    },
    {
      "epoch": 3.198702830188679,
      "grad_norm": 2.312234401702881,
      "learning_rate": 9.891080072776733e-06,
      "loss": 0.6173,
      "num_input_tokens_seen": 3548928,
      "step": 5425
    },
    {
      "epoch": 3.2016509433962264,
      "grad_norm": 1.9669623374938965,
      "learning_rate": 9.890545356721649e-06,
      "loss": 0.4466,
      "num_input_tokens_seen": 3551936,
      "step": 5430
    },
    {
      "epoch": 3.2045990566037736,
      "grad_norm": 1.3534860610961914,
      "learning_rate": 9.890009345873603e-06,
      "loss": 0.4631,
      "num_input_tokens_seen": 3554912,
      "step": 5435
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 4.0622968673706055,
      "learning_rate": 9.889472040374509e-06,
      "loss": 0.4748,
      "num_input_tokens_seen": 3558144,
      "step": 5440
    },
    {
      "epoch": 3.2104952830188678,
      "grad_norm": 2.063538074493408,
      "learning_rate": 9.88893344036662e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 3561376,
      "step": 5445
    },
    {
      "epoch": 3.213443396226415,
      "grad_norm": 2.9676766395568848,
      "learning_rate": 9.888393545992531e-06,
      "loss": 0.4844,
      "num_input_tokens_seen": 3563936,
      "step": 5450
    },
    {
      "epoch": 3.2163915094339623,
      "grad_norm": 1.3016362190246582,
      "learning_rate": 9.887852357395184e-06,
      "loss": 0.5768,
      "num_input_tokens_seen": 3567424,
      "step": 5455
    },
    {
      "epoch": 3.2193396226415096,
      "grad_norm": 7.063687324523926,
      "learning_rate": 9.88730987471786e-06,
      "loss": 0.6118,
      "num_input_tokens_seen": 3570432,
      "step": 5460
    },
    {
      "epoch": 3.2222877358490565,
      "grad_norm": 2.668668031692505,
      "learning_rate": 9.886766098104183e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 3573312,
      "step": 5465
    },
    {
      "epoch": 3.2252358490566038,
      "grad_norm": 1.0627461671829224,
      "learning_rate": 9.886221027698122e-06,
      "loss": 0.4988,
      "num_input_tokens_seen": 3577440,
      "step": 5470
    },
    {
      "epoch": 3.228183962264151,
      "grad_norm": 1.0813132524490356,
      "learning_rate": 9.885674663643983e-06,
      "loss": 0.4694,
      "num_input_tokens_seen": 3581504,
      "step": 5475
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 1.2965104579925537,
      "learning_rate": 9.885127006086423e-06,
      "loss": 0.4134,
      "num_input_tokens_seen": 3585248,
      "step": 5480
    },
    {
      "epoch": 3.234080188679245,
      "grad_norm": 2.155576229095459,
      "learning_rate": 9.884578055170434e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 3588000,
      "step": 5485
    },
    {
      "epoch": 3.2370283018867925,
      "grad_norm": 2.4048752784729004,
      "learning_rate": 9.884027811041353e-06,
      "loss": 0.5547,
      "num_input_tokens_seen": 3590400,
      "step": 5490
    },
    {
      "epoch": 3.2399764150943398,
      "grad_norm": 2.9630963802337646,
      "learning_rate": 9.883476273844861e-06,
      "loss": 0.658,
      "num_input_tokens_seen": 3598112,
      "step": 5495
    },
    {
      "epoch": 3.2429245283018866,
      "grad_norm": 2.1405656337738037,
      "learning_rate": 9.882923443726977e-06,
      "loss": 0.5553,
      "num_input_tokens_seen": 3600544,
      "step": 5500
    },
    {
      "epoch": 3.245872641509434,
      "grad_norm": 2.820202589035034,
      "learning_rate": 9.882369320834068e-06,
      "loss": 0.4888,
      "num_input_tokens_seen": 3604544,
      "step": 5505
    },
    {
      "epoch": 3.248820754716981,
      "grad_norm": 1.0060725212097168,
      "learning_rate": 9.88181390531284e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 3608096,
      "step": 5510
    },
    {
      "epoch": 3.2517688679245285,
      "grad_norm": 2.2366461753845215,
      "learning_rate": 9.88125719731034e-06,
      "loss": 0.4567,
      "num_input_tokens_seen": 3611072,
      "step": 5515
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 1.4462004899978638,
      "learning_rate": 9.880699196973962e-06,
      "loss": 0.4658,
      "num_input_tokens_seen": 3613696,
      "step": 5520
    },
    {
      "epoch": 3.2576650943396226,
      "grad_norm": 1.980528712272644,
      "learning_rate": 9.880139904451436e-06,
      "loss": 0.3692,
      "num_input_tokens_seen": 3615840,
      "step": 5525
    },
    {
      "epoch": 3.26061320754717,
      "grad_norm": 2.885554313659668,
      "learning_rate": 9.879579319890838e-06,
      "loss": 0.5636,
      "num_input_tokens_seen": 3618656,
      "step": 5530
    },
    {
      "epoch": 3.263561320754717,
      "grad_norm": 1.3495756387710571,
      "learning_rate": 9.879017443440584e-06,
      "loss": 0.5318,
      "num_input_tokens_seen": 3622560,
      "step": 5535
    },
    {
      "epoch": 3.266509433962264,
      "grad_norm": 3.12746262550354,
      "learning_rate": 9.878454275249436e-06,
      "loss": 0.604,
      "num_input_tokens_seen": 3625056,
      "step": 5540
    },
    {
      "epoch": 3.2694575471698113,
      "grad_norm": 0.8229339718818665,
      "learning_rate": 9.877889815466493e-06,
      "loss": 0.4353,
      "num_input_tokens_seen": 3628032,
      "step": 5545
    },
    {
      "epoch": 3.2724056603773586,
      "grad_norm": 1.1381443738937378,
      "learning_rate": 9.877324064241198e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 3632384,
      "step": 5550
    },
    {
      "epoch": 3.2753537735849054,
      "grad_norm": 1.3924790620803833,
      "learning_rate": 9.876757021723338e-06,
      "loss": 0.4494,
      "num_input_tokens_seen": 3635776,
      "step": 5555
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 2.650437355041504,
      "learning_rate": 9.876188688063038e-06,
      "loss": 0.6159,
      "num_input_tokens_seen": 3638720,
      "step": 5560
    },
    {
      "epoch": 3.28125,
      "grad_norm": 1.9412283897399902,
      "learning_rate": 9.875619063410768e-06,
      "loss": 0.4679,
      "num_input_tokens_seen": 3642688,
      "step": 5565
    },
    {
      "epoch": 3.2841981132075473,
      "grad_norm": 1.7628861665725708,
      "learning_rate": 9.875048147917339e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 3645600,
      "step": 5570
    },
    {
      "epoch": 3.2871462264150946,
      "grad_norm": 2.3113176822662354,
      "learning_rate": 9.874475941733902e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 3648320,
      "step": 5575
    },
    {
      "epoch": 3.2900943396226414,
      "grad_norm": 1.1149119138717651,
      "learning_rate": 9.873902445011952e-06,
      "loss": 0.5065,
      "num_input_tokens_seen": 3651264,
      "step": 5580
    },
    {
      "epoch": 3.2930424528301887,
      "grad_norm": 1.8580138683319092,
      "learning_rate": 9.873327657903324e-06,
      "loss": 0.4161,
      "num_input_tokens_seen": 3654528,
      "step": 5585
    },
    {
      "epoch": 3.295990566037736,
      "grad_norm": 3.66279673576355,
      "learning_rate": 9.872751580560194e-06,
      "loss": 0.4402,
      "num_input_tokens_seen": 3657376,
      "step": 5590
    },
    {
      "epoch": 3.298938679245283,
      "grad_norm": 2.7742903232574463,
      "learning_rate": 9.872174213135084e-06,
      "loss": 0.602,
      "num_input_tokens_seen": 3660832,
      "step": 5595
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 1.1665987968444824,
      "learning_rate": 9.871595555780855e-06,
      "loss": 0.514,
      "num_input_tokens_seen": 3664256,
      "step": 5600
    },
    {
      "epoch": 3.3048349056603774,
      "grad_norm": 1.7539814710617065,
      "learning_rate": 9.871015608650705e-06,
      "loss": 0.5073,
      "num_input_tokens_seen": 3667008,
      "step": 5605
    },
    {
      "epoch": 3.3077830188679247,
      "grad_norm": 1.6499611139297485,
      "learning_rate": 9.870434371898182e-06,
      "loss": 0.5488,
      "num_input_tokens_seen": 3670112,
      "step": 5610
    },
    {
      "epoch": 3.3107311320754715,
      "grad_norm": 1.9590286016464233,
      "learning_rate": 9.869851845677165e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 3672896,
      "step": 5615
    },
    {
      "epoch": 3.313679245283019,
      "grad_norm": 1.447089433670044,
      "learning_rate": 9.869268030141886e-06,
      "loss": 0.4934,
      "num_input_tokens_seen": 3676032,
      "step": 5620
    },
    {
      "epoch": 3.316627358490566,
      "grad_norm": 1.4614351987838745,
      "learning_rate": 9.86868292544691e-06,
      "loss": 0.5459,
      "num_input_tokens_seen": 3679840,
      "step": 5625
    },
    {
      "epoch": 3.3195754716981134,
      "grad_norm": 1.412472128868103,
      "learning_rate": 9.868096531747149e-06,
      "loss": 0.4931,
      "num_input_tokens_seen": 3682816,
      "step": 5630
    },
    {
      "epoch": 3.3225235849056602,
      "grad_norm": 1.4861077070236206,
      "learning_rate": 9.867508849197848e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 3687136,
      "step": 5635
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 1.7424558401107788,
      "learning_rate": 9.866919877954602e-06,
      "loss": 0.6988,
      "num_input_tokens_seen": 3691616,
      "step": 5640
    },
    {
      "epoch": 3.328419811320755,
      "grad_norm": 1.4001785516738892,
      "learning_rate": 9.866329618173344e-06,
      "loss": 0.4636,
      "num_input_tokens_seen": 3695712,
      "step": 5645
    },
    {
      "epoch": 3.331367924528302,
      "grad_norm": 1.2370343208312988,
      "learning_rate": 9.865738070010346e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 3699616,
      "step": 5650
    },
    {
      "epoch": 3.334316037735849,
      "grad_norm": 3.707430124282837,
      "learning_rate": 9.865145233622223e-06,
      "loss": 0.5399,
      "num_input_tokens_seen": 3703360,
      "step": 5655
    },
    {
      "epoch": 3.3372641509433962,
      "grad_norm": 5.823843002319336,
      "learning_rate": 9.864551109165935e-06,
      "loss": 0.5035,
      "num_input_tokens_seen": 3706432,
      "step": 5660
    },
    {
      "epoch": 3.3402122641509435,
      "grad_norm": 1.0574711561203003,
      "learning_rate": 9.863955696798773e-06,
      "loss": 0.5072,
      "num_input_tokens_seen": 3710880,
      "step": 5665
    },
    {
      "epoch": 3.3431603773584904,
      "grad_norm": 1.7641799449920654,
      "learning_rate": 9.863358996678378e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 3713536,
      "step": 5670
    },
    {
      "epoch": 3.3461084905660377,
      "grad_norm": 0.9829106330871582,
      "learning_rate": 9.86276100896273e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 3716896,
      "step": 5675
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 1.8848925828933716,
      "learning_rate": 9.862161733810147e-06,
      "loss": 0.5691,
      "num_input_tokens_seen": 3719808,
      "step": 5680
    },
    {
      "epoch": 3.3520047169811322,
      "grad_norm": 2.108386754989624,
      "learning_rate": 9.86156117137929e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 3722784,
      "step": 5685
    },
    {
      "epoch": 3.354952830188679,
      "grad_norm": 2.638390064239502,
      "learning_rate": 9.860959321829159e-06,
      "loss": 0.4444,
      "num_input_tokens_seen": 3725632,
      "step": 5690
    },
    {
      "epoch": 3.3579009433962264,
      "grad_norm": 1.0299745798110962,
      "learning_rate": 9.860356185319102e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 3728608,
      "step": 5695
    },
    {
      "epoch": 3.3608490566037736,
      "grad_norm": 2.68991756439209,
      "learning_rate": 9.859751762008796e-06,
      "loss": 0.5376,
      "num_input_tokens_seen": 3731488,
      "step": 5700
    },
    {
      "epoch": 3.363797169811321,
      "grad_norm": 2.626784086227417,
      "learning_rate": 9.859146052058266e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 3733952,
      "step": 5705
    },
    {
      "epoch": 3.3667452830188678,
      "grad_norm": 1.6710686683654785,
      "learning_rate": 9.858539055627876e-06,
      "loss": 0.4966,
      "num_input_tokens_seen": 3737088,
      "step": 5710
    },
    {
      "epoch": 3.369693396226415,
      "grad_norm": 2.2778468132019043,
      "learning_rate": 9.857930772878333e-06,
      "loss": 0.4752,
      "num_input_tokens_seen": 3740000,
      "step": 5715
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 3.082333564758301,
      "learning_rate": 9.857321203970682e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 3742784,
      "step": 5720
    },
    {
      "epoch": 3.3755896226415096,
      "grad_norm": 1.407512903213501,
      "learning_rate": 9.856710349066307e-06,
      "loss": 0.502,
      "num_input_tokens_seen": 3746400,
      "step": 5725
    },
    {
      "epoch": 3.3785377358490565,
      "grad_norm": 1.6729859113693237,
      "learning_rate": 9.856098208326937e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 3749440,
      "step": 5730
    },
    {
      "epoch": 3.3814858490566038,
      "grad_norm": 3.554318904876709,
      "learning_rate": 9.855484781914639e-06,
      "loss": 0.5554,
      "num_input_tokens_seen": 3752352,
      "step": 5735
    },
    {
      "epoch": 3.384433962264151,
      "grad_norm": 1.325243353843689,
      "learning_rate": 9.854870069991817e-06,
      "loss": 0.4817,
      "num_input_tokens_seen": 3756192,
      "step": 5740
    },
    {
      "epoch": 3.387382075471698,
      "grad_norm": 3.772488832473755,
      "learning_rate": 9.854254072721222e-06,
      "loss": 0.5249,
      "num_input_tokens_seen": 3759392,
      "step": 5745
    },
    {
      "epoch": 3.390330188679245,
      "grad_norm": 2.0203943252563477,
      "learning_rate": 9.853636790265938e-06,
      "loss": 0.6764,
      "num_input_tokens_seen": 3761824,
      "step": 5750
    },
    {
      "epoch": 3.3932783018867925,
      "grad_norm": 1.4606796503067017,
      "learning_rate": 9.853018222789397e-06,
      "loss": 0.3494,
      "num_input_tokens_seen": 3765504,
      "step": 5755
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 1.5235087871551514,
      "learning_rate": 9.852398370455367e-06,
      "loss": 0.5427,
      "num_input_tokens_seen": 3768288,
      "step": 5760
    },
    {
      "epoch": 3.3991745283018866,
      "grad_norm": 3.182023763656616,
      "learning_rate": 9.851777233427955e-06,
      "loss": 0.6825,
      "num_input_tokens_seen": 3771264,
      "step": 5765
    },
    {
      "epoch": 3.402122641509434,
      "grad_norm": 0.8177620768547058,
      "learning_rate": 9.85115481187161e-06,
      "loss": 0.5174,
      "num_input_tokens_seen": 3773952,
      "step": 5770
    },
    {
      "epoch": 3.405070754716981,
      "grad_norm": 1.7739533185958862,
      "learning_rate": 9.850531105951123e-06,
      "loss": 0.629,
      "num_input_tokens_seen": 3776896,
      "step": 5775
    },
    {
      "epoch": 3.4080188679245285,
      "grad_norm": 0.965461254119873,
      "learning_rate": 9.84990611583162e-06,
      "loss": 0.4403,
      "num_input_tokens_seen": 3780000,
      "step": 5780
    },
    {
      "epoch": 3.4109669811320753,
      "grad_norm": 1.8973751068115234,
      "learning_rate": 9.849279841678572e-06,
      "loss": 0.5215,
      "num_input_tokens_seen": 3782720,
      "step": 5785
    },
    {
      "epoch": 3.4139150943396226,
      "grad_norm": 1.9839820861816406,
      "learning_rate": 9.848652283657785e-06,
      "loss": 0.4353,
      "num_input_tokens_seen": 3785984,
      "step": 5790
    },
    {
      "epoch": 3.41686320754717,
      "grad_norm": 3.5479331016540527,
      "learning_rate": 9.848023441935411e-06,
      "loss": 0.4412,
      "num_input_tokens_seen": 3788448,
      "step": 5795
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 1.5580400228500366,
      "learning_rate": 9.847393316677935e-06,
      "loss": 0.5021,
      "num_input_tokens_seen": 3791456,
      "step": 5800
    },
    {
      "epoch": 3.422759433962264,
      "grad_norm": 2.377774715423584,
      "learning_rate": 9.846761908052188e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 3794432,
      "step": 5805
    },
    {
      "epoch": 3.4257075471698113,
      "grad_norm": 2.0717790126800537,
      "learning_rate": 9.846129216225338e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 3798048,
      "step": 5810
    },
    {
      "epoch": 3.4286556603773586,
      "grad_norm": 1.220272183418274,
      "learning_rate": 9.845495241364892e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 3801120,
      "step": 5815
    },
    {
      "epoch": 3.4316037735849054,
      "grad_norm": 0.8439756631851196,
      "learning_rate": 9.844859983638696e-06,
      "loss": 0.5324,
      "num_input_tokens_seen": 3804320,
      "step": 5820
    },
    {
      "epoch": 3.4345518867924527,
      "grad_norm": 1.1022052764892578,
      "learning_rate": 9.844223443214942e-06,
      "loss": 0.5378,
      "num_input_tokens_seen": 3808864,
      "step": 5825
    },
    {
      "epoch": 3.4375,
      "grad_norm": 1.9363981485366821,
      "learning_rate": 9.843585620262153e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 3812064,
      "step": 5830
    },
    {
      "epoch": 3.4404481132075473,
      "grad_norm": 0.8281024694442749,
      "learning_rate": 9.842946514949197e-06,
      "loss": 0.401,
      "num_input_tokens_seen": 3815776,
      "step": 5835
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 3.4893798828125,
      "learning_rate": 9.842306127445279e-06,
      "loss": 0.4925,
      "num_input_tokens_seen": 3818624,
      "step": 5840
    },
    {
      "epoch": 3.4463443396226414,
      "grad_norm": 1.8219276666641235,
      "learning_rate": 9.841664457919944e-06,
      "loss": 0.5639,
      "num_input_tokens_seen": 3821024,
      "step": 5845
    },
    {
      "epoch": 3.4492924528301887,
      "grad_norm": 1.2562804222106934,
      "learning_rate": 9.841021506543079e-06,
      "loss": 0.3847,
      "num_input_tokens_seen": 3823488,
      "step": 5850
    },
    {
      "epoch": 3.452240566037736,
      "grad_norm": 1.0920352935791016,
      "learning_rate": 9.840377273484904e-06,
      "loss": 0.4157,
      "num_input_tokens_seen": 3826592,
      "step": 5855
    },
    {
      "epoch": 3.455188679245283,
      "grad_norm": 1.8910034894943237,
      "learning_rate": 9.839731758915986e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 3828928,
      "step": 5860
    },
    {
      "epoch": 3.45813679245283,
      "grad_norm": 5.32118558883667,
      "learning_rate": 9.839084963007226e-06,
      "loss": 0.6921,
      "num_input_tokens_seen": 3832160,
      "step": 5865
    },
    {
      "epoch": 3.4610849056603774,
      "grad_norm": 1.7797826528549194,
      "learning_rate": 9.838436885929868e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 3834688,
      "step": 5870
    },
    {
      "epoch": 3.4640330188679247,
      "grad_norm": 2.6195404529571533,
      "learning_rate": 9.837787527855492e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 3837984,
      "step": 5875
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 3.876128911972046,
      "learning_rate": 9.837136888956017e-06,
      "loss": 0.5588,
      "num_input_tokens_seen": 3844704,
      "step": 5880
    },
    {
      "epoch": 3.469929245283019,
      "grad_norm": 1.2786006927490234,
      "learning_rate": 9.836484969403705e-06,
      "loss": 0.4646,
      "num_input_tokens_seen": 3847776,
      "step": 5885
    },
    {
      "epoch": 3.472877358490566,
      "grad_norm": 1.4999171495437622,
      "learning_rate": 9.835831769371152e-06,
      "loss": 0.4844,
      "num_input_tokens_seen": 3850816,
      "step": 5890
    },
    {
      "epoch": 3.4758254716981134,
      "grad_norm": 2.2422657012939453,
      "learning_rate": 9.835177289031298e-06,
      "loss": 0.5663,
      "num_input_tokens_seen": 3853920,
      "step": 5895
    },
    {
      "epoch": 3.4787735849056602,
      "grad_norm": 3.2343759536743164,
      "learning_rate": 9.834521528557419e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 3857120,
      "step": 5900
    },
    {
      "epoch": 3.4817216981132075,
      "grad_norm": 1.5392374992370605,
      "learning_rate": 9.833864488123128e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 3860192,
      "step": 5905
    },
    {
      "epoch": 3.484669811320755,
      "grad_norm": 5.111044406890869,
      "learning_rate": 9.83320616790238e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 3862976,
      "step": 5910
    },
    {
      "epoch": 3.487617924528302,
      "grad_norm": 2.379945993423462,
      "learning_rate": 9.832546568069472e-06,
      "loss": 0.5132,
      "num_input_tokens_seen": 3866752,
      "step": 5915
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 1.5224056243896484,
      "learning_rate": 9.831885688799031e-06,
      "loss": 0.5099,
      "num_input_tokens_seen": 3870528,
      "step": 5920
    },
    {
      "epoch": 3.4935141509433962,
      "grad_norm": 1.731109857559204,
      "learning_rate": 9.83122353026603e-06,
      "loss": 0.4681,
      "num_input_tokens_seen": 3873920,
      "step": 5925
    },
    {
      "epoch": 3.4964622641509435,
      "grad_norm": 1.6047178506851196,
      "learning_rate": 9.830560092645778e-06,
      "loss": 0.4892,
      "num_input_tokens_seen": 3876832,
      "step": 5930
    },
    {
      "epoch": 3.4994103773584904,
      "grad_norm": 1.1299241781234741,
      "learning_rate": 9.829895376113923e-06,
      "loss": 0.4166,
      "num_input_tokens_seen": 3879712,
      "step": 5935
    },
    {
      "epoch": 3.5023584905660377,
      "grad_norm": 1.6172032356262207,
      "learning_rate": 9.829229380846452e-06,
      "loss": 0.5034,
      "num_input_tokens_seen": 3883072,
      "step": 5940
    },
    {
      "epoch": 3.505306603773585,
      "grad_norm": 1.7562639713287354,
      "learning_rate": 9.82856210701969e-06,
      "loss": 0.4938,
      "num_input_tokens_seen": 3886336,
      "step": 5945
    },
    {
      "epoch": 3.5082547169811322,
      "grad_norm": 1.5856565237045288,
      "learning_rate": 9.827893554810298e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 3889280,
      "step": 5950
    },
    {
      "epoch": 3.5112028301886795,
      "grad_norm": 0.8660147190093994,
      "learning_rate": 9.827223724395281e-06,
      "loss": 0.5074,
      "num_input_tokens_seen": 3893856,
      "step": 5955
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 1.7920117378234863,
      "learning_rate": 9.82655261595198e-06,
      "loss": 0.5925,
      "num_input_tokens_seen": 3897056,
      "step": 5960
    },
    {
      "epoch": 3.5170990566037736,
      "grad_norm": 1.8953392505645752,
      "learning_rate": 9.825880229658073e-06,
      "loss": 0.4429,
      "num_input_tokens_seen": 3900256,
      "step": 5965
    },
    {
      "epoch": 3.5200471698113205,
      "grad_norm": 1.2079137563705444,
      "learning_rate": 9.825206565691576e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 3903360,
      "step": 5970
    },
    {
      "epoch": 3.5229952830188678,
      "grad_norm": 2.659231424331665,
      "learning_rate": 9.824531624230844e-06,
      "loss": 0.602,
      "num_input_tokens_seen": 3907104,
      "step": 5975
    },
    {
      "epoch": 3.525943396226415,
      "grad_norm": 3.112238645553589,
      "learning_rate": 9.823855405454573e-06,
      "loss": 0.4271,
      "num_input_tokens_seen": 3910592,
      "step": 5980
    },
    {
      "epoch": 3.5288915094339623,
      "grad_norm": 2.490907669067383,
      "learning_rate": 9.823177909541795e-06,
      "loss": 0.4253,
      "num_input_tokens_seen": 3913952,
      "step": 5985
    },
    {
      "epoch": 3.5318396226415096,
      "grad_norm": 2.247433662414551,
      "learning_rate": 9.822499136671877e-06,
      "loss": 0.645,
      "num_input_tokens_seen": 3916896,
      "step": 5990
    },
    {
      "epoch": 3.5347877358490565,
      "grad_norm": 1.3860961198806763,
      "learning_rate": 9.82181908702453e-06,
      "loss": 0.6131,
      "num_input_tokens_seen": 3920224,
      "step": 5995
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 1.1410797834396362,
      "learning_rate": 9.821137760779797e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 3923840,
      "step": 6000
    },
    {
      "epoch": 3.540683962264151,
      "grad_norm": 1.946249008178711,
      "learning_rate": 9.820455158118065e-06,
      "loss": 0.5482,
      "num_input_tokens_seen": 3927040,
      "step": 6005
    },
    {
      "epoch": 3.543632075471698,
      "grad_norm": 1.451001763343811,
      "learning_rate": 9.819771279220053e-06,
      "loss": 0.5104,
      "num_input_tokens_seen": 3931264,
      "step": 6010
    },
    {
      "epoch": 3.546580188679245,
      "grad_norm": 1.3526102304458618,
      "learning_rate": 9.819086124266825e-06,
      "loss": 0.422,
      "num_input_tokens_seen": 3935392,
      "step": 6015
    },
    {
      "epoch": 3.5495283018867925,
      "grad_norm": 1.2305052280426025,
      "learning_rate": 9.818399693439778e-06,
      "loss": 0.518,
      "num_input_tokens_seen": 3938240,
      "step": 6020
    },
    {
      "epoch": 3.5524764150943398,
      "grad_norm": 3.292396068572998,
      "learning_rate": 9.817711986920644e-06,
      "loss": 0.5828,
      "num_input_tokens_seen": 3941280,
      "step": 6025
    },
    {
      "epoch": 3.555424528301887,
      "grad_norm": 4.04533052444458,
      "learning_rate": 9.817023004891497e-06,
      "loss": 0.5739,
      "num_input_tokens_seen": 3943936,
      "step": 6030
    },
    {
      "epoch": 3.558372641509434,
      "grad_norm": 1.6901971101760864,
      "learning_rate": 9.816332747534752e-06,
      "loss": 0.627,
      "num_input_tokens_seen": 3947360,
      "step": 6035
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 1.4621602296829224,
      "learning_rate": 9.815641215033153e-06,
      "loss": 0.4623,
      "num_input_tokens_seen": 3951200,
      "step": 6040
    },
    {
      "epoch": 3.5642688679245285,
      "grad_norm": 2.0119452476501465,
      "learning_rate": 9.814948407569789e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 3954752,
      "step": 6045
    },
    {
      "epoch": 3.5672169811320753,
      "grad_norm": 2.5273189544677734,
      "learning_rate": 9.814254325328082e-06,
      "loss": 0.4474,
      "num_input_tokens_seen": 3957984,
      "step": 6050
    },
    {
      "epoch": 3.5701650943396226,
      "grad_norm": 2.740997076034546,
      "learning_rate": 9.813558968491794e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 3961600,
      "step": 6055
    },
    {
      "epoch": 3.57311320754717,
      "grad_norm": 6.524359226226807,
      "learning_rate": 9.812862337245024e-06,
      "loss": 0.5149,
      "num_input_tokens_seen": 3965408,
      "step": 6060
    },
    {
      "epoch": 3.576061320754717,
      "grad_norm": 2.79864501953125,
      "learning_rate": 9.812164431772208e-06,
      "loss": 0.5762,
      "num_input_tokens_seen": 3968928,
      "step": 6065
    },
    {
      "epoch": 3.579009433962264,
      "grad_norm": 2.6403040885925293,
      "learning_rate": 9.81146525225812e-06,
      "loss": 0.4692,
      "num_input_tokens_seen": 3971872,
      "step": 6070
    },
    {
      "epoch": 3.5819575471698113,
      "grad_norm": 1.7062780857086182,
      "learning_rate": 9.810764798887868e-06,
      "loss": 0.4345,
      "num_input_tokens_seen": 3975136,
      "step": 6075
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 2.0908071994781494,
      "learning_rate": 9.810063071846905e-06,
      "loss": 0.5521,
      "num_input_tokens_seen": 3977888,
      "step": 6080
    },
    {
      "epoch": 3.5878537735849054,
      "grad_norm": 1.5211453437805176,
      "learning_rate": 9.809360071321013e-06,
      "loss": 0.4402,
      "num_input_tokens_seen": 3981344,
      "step": 6085
    },
    {
      "epoch": 3.5908018867924527,
      "grad_norm": 2.3381199836730957,
      "learning_rate": 9.808655797496314e-06,
      "loss": 0.5731,
      "num_input_tokens_seen": 3984000,
      "step": 6090
    },
    {
      "epoch": 3.59375,
      "grad_norm": 1.350576400756836,
      "learning_rate": 9.807950250559268e-06,
      "loss": 0.6154,
      "num_input_tokens_seen": 3987584,
      "step": 6095
    },
    {
      "epoch": 3.5966981132075473,
      "grad_norm": 2.221839666366577,
      "learning_rate": 9.807243430696673e-06,
      "loss": 0.5387,
      "num_input_tokens_seen": 3990560,
      "step": 6100
    },
    {
      "epoch": 3.5996462264150946,
      "grad_norm": 2.050088405609131,
      "learning_rate": 9.806535338095661e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 3993728,
      "step": 6105
    },
    {
      "epoch": 3.6025943396226414,
      "grad_norm": 2.1210544109344482,
      "learning_rate": 9.805825972943706e-06,
      "loss": 0.748,
      "num_input_tokens_seen": 3996256,
      "step": 6110
    },
    {
      "epoch": 3.6055424528301887,
      "grad_norm": 0.8280351161956787,
      "learning_rate": 9.80511533542861e-06,
      "loss": 0.5014,
      "num_input_tokens_seen": 3999040,
      "step": 6115
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 1.7262794971466064,
      "learning_rate": 9.80440342573852e-06,
      "loss": 0.4947,
      "num_input_tokens_seen": 4001568,
      "step": 6120
    },
    {
      "epoch": 3.611438679245283,
      "grad_norm": 1.927152156829834,
      "learning_rate": 9.803690244061919e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 4004160,
      "step": 6125
    },
    {
      "epoch": 3.61438679245283,
      "grad_norm": 2.1269006729125977,
      "learning_rate": 9.802975790587621e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 4006784,
      "step": 6130
    },
    {
      "epoch": 3.6173349056603774,
      "grad_norm": 6.769991397857666,
      "learning_rate": 9.802260065504783e-06,
      "loss": 0.4276,
      "num_input_tokens_seen": 4009312,
      "step": 6135
    },
    {
      "epoch": 3.6202830188679247,
      "grad_norm": 1.0379269123077393,
      "learning_rate": 9.801543069002897e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 4014016,
      "step": 6140
    },
    {
      "epoch": 3.6232311320754715,
      "grad_norm": 1.5005934238433838,
      "learning_rate": 9.80082480127179e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 4017760,
      "step": 6145
    },
    {
      "epoch": 3.626179245283019,
      "grad_norm": 1.603636622428894,
      "learning_rate": 9.800105262501628e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 4020128,
      "step": 6150
    },
    {
      "epoch": 3.629127358490566,
      "grad_norm": 2.256763219833374,
      "learning_rate": 9.799384452882907e-06,
      "loss": 0.4383,
      "num_input_tokens_seen": 4023712,
      "step": 6155
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 1.5814650058746338,
      "learning_rate": 9.798662372606469e-06,
      "loss": 0.5614,
      "num_input_tokens_seen": 4026432,
      "step": 6160
    },
    {
      "epoch": 3.6350235849056602,
      "grad_norm": 1.5546823740005493,
      "learning_rate": 9.797939021863487e-06,
      "loss": 0.4624,
      "num_input_tokens_seen": 4029472,
      "step": 6165
    },
    {
      "epoch": 3.6379716981132075,
      "grad_norm": 1.0595825910568237,
      "learning_rate": 9.797214400845472e-06,
      "loss": 0.4869,
      "num_input_tokens_seen": 4032896,
      "step": 6170
    },
    {
      "epoch": 3.640919811320755,
      "grad_norm": 1.8683580160140991,
      "learning_rate": 9.796488509744269e-06,
      "loss": 0.4231,
      "num_input_tokens_seen": 4036608,
      "step": 6175
    },
    {
      "epoch": 3.643867924528302,
      "grad_norm": 1.4995150566101074,
      "learning_rate": 9.79576134875206e-06,
      "loss": 0.5165,
      "num_input_tokens_seen": 4039872,
      "step": 6180
    },
    {
      "epoch": 3.646816037735849,
      "grad_norm": 2.6108646392822266,
      "learning_rate": 9.795032918061367e-06,
      "loss": 0.3904,
      "num_input_tokens_seen": 4043424,
      "step": 6185
    },
    {
      "epoch": 3.6497641509433962,
      "grad_norm": 1.7615370750427246,
      "learning_rate": 9.794303217865041e-06,
      "loss": 0.4765,
      "num_input_tokens_seen": 4046240,
      "step": 6190
    },
    {
      "epoch": 3.6527122641509435,
      "grad_norm": 1.4838383197784424,
      "learning_rate": 9.79357224835628e-06,
      "loss": 0.4902,
      "num_input_tokens_seen": 4049248,
      "step": 6195
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 2.101064682006836,
      "learning_rate": 9.792840009728605e-06,
      "loss": 0.4794,
      "num_input_tokens_seen": 4052192,
      "step": 6200
    },
    {
      "epoch": 3.6586084905660377,
      "grad_norm": 1.4491682052612305,
      "learning_rate": 9.79210650217588e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 4056256,
      "step": 6205
    },
    {
      "epoch": 3.661556603773585,
      "grad_norm": 1.7391060590744019,
      "learning_rate": 9.791371725892307e-06,
      "loss": 0.3802,
      "num_input_tokens_seen": 4059808,
      "step": 6210
    },
    {
      "epoch": 3.6645047169811322,
      "grad_norm": 1.849182367324829,
      "learning_rate": 9.79063568107242e-06,
      "loss": 0.369,
      "num_input_tokens_seen": 4061920,
      "step": 6215
    },
    {
      "epoch": 3.6674528301886795,
      "grad_norm": 2.1498515605926514,
      "learning_rate": 9.78989836791109e-06,
      "loss": 0.4229,
      "num_input_tokens_seen": 4065632,
      "step": 6220
    },
    {
      "epoch": 3.6704009433962264,
      "grad_norm": 1.441758632659912,
      "learning_rate": 9.789159786603524e-06,
      "loss": 0.4507,
      "num_input_tokens_seen": 4068768,
      "step": 6225
    },
    {
      "epoch": 3.6733490566037736,
      "grad_norm": 3.593398332595825,
      "learning_rate": 9.788419937345263e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 4071616,
      "step": 6230
    },
    {
      "epoch": 3.6762971698113205,
      "grad_norm": 1.5819224119186401,
      "learning_rate": 9.787678820332188e-06,
      "loss": 0.5158,
      "num_input_tokens_seen": 4074368,
      "step": 6235
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 1.3901891708374023,
      "learning_rate": 9.78693643576051e-06,
      "loss": 0.4926,
      "num_input_tokens_seen": 4078912,
      "step": 6240
    },
    {
      "epoch": 3.682193396226415,
      "grad_norm": 4.482086658477783,
      "learning_rate": 9.786192783826782e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 4082048,
      "step": 6245
    },
    {
      "epoch": 3.6851415094339623,
      "grad_norm": 1.5613305568695068,
      "learning_rate": 9.785447864727887e-06,
      "loss": 0.5067,
      "num_input_tokens_seen": 4084960,
      "step": 6250
    },
    {
      "epoch": 3.6880896226415096,
      "grad_norm": 1.3843508958816528,
      "learning_rate": 9.784701678661045e-06,
      "loss": 0.419,
      "num_input_tokens_seen": 4087360,
      "step": 6255
    },
    {
      "epoch": 3.6910377358490565,
      "grad_norm": 1.0762572288513184,
      "learning_rate": 9.783954225823813e-06,
      "loss": 0.414,
      "num_input_tokens_seen": 4090752,
      "step": 6260
    },
    {
      "epoch": 3.6939858490566038,
      "grad_norm": 1.0805782079696655,
      "learning_rate": 9.783205506414082e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 4093824,
      "step": 6265
    },
    {
      "epoch": 3.696933962264151,
      "grad_norm": 3.538618326187134,
      "learning_rate": 9.782455520630079e-06,
      "loss": 0.7462,
      "num_input_tokens_seen": 4096416,
      "step": 6270
    },
    {
      "epoch": 3.699882075471698,
      "grad_norm": 1.0099101066589355,
      "learning_rate": 9.781704268670364e-06,
      "loss": 0.4778,
      "num_input_tokens_seen": 4100064,
      "step": 6275
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 1.929429292678833,
      "learning_rate": 9.780951750733837e-06,
      "loss": 0.4025,
      "num_input_tokens_seen": 4103168,
      "step": 6280
    },
    {
      "epoch": 3.7057783018867925,
      "grad_norm": 1.1872811317443848,
      "learning_rate": 9.780197967019728e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 4106272,
      "step": 6285
    },
    {
      "epoch": 3.7087264150943398,
      "grad_norm": 3.2629570960998535,
      "learning_rate": 9.779442917727608e-06,
      "loss": 0.6309,
      "num_input_tokens_seen": 4108960,
      "step": 6290
    },
    {
      "epoch": 3.711674528301887,
      "grad_norm": 1.8601330518722534,
      "learning_rate": 9.778686603057377e-06,
      "loss": 0.4165,
      "num_input_tokens_seen": 4111680,
      "step": 6295
    },
    {
      "epoch": 3.714622641509434,
      "grad_norm": 1.799086332321167,
      "learning_rate": 9.777929023209271e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 4114976,
      "step": 6300
    },
    {
      "epoch": 3.717570754716981,
      "grad_norm": 1.6830500364303589,
      "learning_rate": 9.777170178383866e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 4118656,
      "step": 6305
    },
    {
      "epoch": 3.7205188679245285,
      "grad_norm": 2.1540603637695312,
      "learning_rate": 9.776410068782068e-06,
      "loss": 0.4427,
      "num_input_tokens_seen": 4121184,
      "step": 6310
    },
    {
      "epoch": 3.7234669811320753,
      "grad_norm": 1.8403667211532593,
      "learning_rate": 9.775648694605118e-06,
      "loss": 0.4124,
      "num_input_tokens_seen": 4124384,
      "step": 6315
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 0.9677417278289795,
      "learning_rate": 9.774886056054593e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 4127904,
      "step": 6320
    },
    {
      "epoch": 3.72936320754717,
      "grad_norm": 1.3664913177490234,
      "learning_rate": 9.774122153332408e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 4130912,
      "step": 6325
    },
    {
      "epoch": 3.732311320754717,
      "grad_norm": 1.0303430557250977,
      "learning_rate": 9.773356986640807e-06,
      "loss": 0.5729,
      "num_input_tokens_seen": 4133536,
      "step": 6330
    },
    {
      "epoch": 3.735259433962264,
      "grad_norm": 2.2690205574035645,
      "learning_rate": 9.772590556182373e-06,
      "loss": 0.5655,
      "num_input_tokens_seen": 4136768,
      "step": 6335
    },
    {
      "epoch": 3.7382075471698113,
      "grad_norm": 1.1081606149673462,
      "learning_rate": 9.77182286216002e-06,
      "loss": 0.5542,
      "num_input_tokens_seen": 4139392,
      "step": 6340
    },
    {
      "epoch": 3.7411556603773586,
      "grad_norm": 1.9214736223220825,
      "learning_rate": 9.771053904776998e-06,
      "loss": 0.495,
      "num_input_tokens_seen": 4142784,
      "step": 6345
    },
    {
      "epoch": 3.7441037735849054,
      "grad_norm": 2.0866923332214355,
      "learning_rate": 9.770283684236891e-06,
      "loss": 0.4776,
      "num_input_tokens_seen": 4146464,
      "step": 6350
    },
    {
      "epoch": 3.7470518867924527,
      "grad_norm": 2.737359046936035,
      "learning_rate": 9.769512200743623e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 4149664,
      "step": 6355
    },
    {
      "epoch": 3.75,
      "grad_norm": 2.0262043476104736,
      "learning_rate": 9.768739454501444e-06,
      "loss": 0.381,
      "num_input_tokens_seen": 4152640,
      "step": 6360
    },
    {
      "epoch": 3.7529481132075473,
      "grad_norm": 1.3444408178329468,
      "learning_rate": 9.76796544571494e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 4156352,
      "step": 6365
    },
    {
      "epoch": 3.7558962264150946,
      "grad_norm": 3.4963226318359375,
      "learning_rate": 9.767190174589036e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 4159200,
      "step": 6370
    },
    {
      "epoch": 3.7588443396226414,
      "grad_norm": 1.485098123550415,
      "learning_rate": 9.76641364132899e-06,
      "loss": 0.4416,
      "num_input_tokens_seen": 4162688,
      "step": 6375
    },
    {
      "epoch": 3.7617924528301887,
      "grad_norm": 5.593234062194824,
      "learning_rate": 9.765635846140389e-06,
      "loss": 0.3587,
      "num_input_tokens_seen": 4165696,
      "step": 6380
    },
    {
      "epoch": 3.764740566037736,
      "grad_norm": 1.6745392084121704,
      "learning_rate": 9.764856789229157e-06,
      "loss": 0.5526,
      "num_input_tokens_seen": 4168736,
      "step": 6385
    },
    {
      "epoch": 3.767688679245283,
      "grad_norm": 1.6972877979278564,
      "learning_rate": 9.764076470801557e-06,
      "loss": 0.4749,
      "num_input_tokens_seen": 4176160,
      "step": 6390
    },
    {
      "epoch": 3.77063679245283,
      "grad_norm": 1.031673789024353,
      "learning_rate": 9.763294891064182e-06,
      "loss": 0.5179,
      "num_input_tokens_seen": 4179552,
      "step": 6395
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 1.4192754030227661,
      "learning_rate": 9.762512050223951e-06,
      "loss": 0.4013,
      "num_input_tokens_seen": 4182528,
      "step": 6400
    },
    {
      "epoch": 3.7765330188679247,
      "grad_norm": 1.7241017818450928,
      "learning_rate": 9.761727948488132e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 4185344,
      "step": 6405
    },
    {
      "epoch": 3.7794811320754715,
      "grad_norm": 1.225419521331787,
      "learning_rate": 9.760942586064315e-06,
      "loss": 0.5265,
      "num_input_tokens_seen": 4189024,
      "step": 6410
    },
    {
      "epoch": 3.782429245283019,
      "grad_norm": 2.088930130004883,
      "learning_rate": 9.760155963160431e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 4192000,
      "step": 6415
    },
    {
      "epoch": 3.785377358490566,
      "grad_norm": 1.4585833549499512,
      "learning_rate": 9.759368079984741e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 4194560,
      "step": 6420
    },
    {
      "epoch": 3.788325471698113,
      "grad_norm": 1.2453200817108154,
      "learning_rate": 9.758578936745839e-06,
      "loss": 0.4016,
      "num_input_tokens_seen": 4199072,
      "step": 6425
    },
    {
      "epoch": 3.7912735849056602,
      "grad_norm": 1.317825198173523,
      "learning_rate": 9.757788533652656e-06,
      "loss": 0.6209,
      "num_input_tokens_seen": 4201696,
      "step": 6430
    },
    {
      "epoch": 3.7942216981132075,
      "grad_norm": 2.302766799926758,
      "learning_rate": 9.756996870914454e-06,
      "loss": 0.5418,
      "num_input_tokens_seen": 4206080,
      "step": 6435
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 1.1977827548980713,
      "learning_rate": 9.756203948740828e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 4208672,
      "step": 6440
    },
    {
      "epoch": 3.800117924528302,
      "grad_norm": 0.8242406845092773,
      "learning_rate": 9.755409767341709e-06,
      "loss": 0.447,
      "num_input_tokens_seen": 4211360,
      "step": 6445
    },
    {
      "epoch": 3.803066037735849,
      "grad_norm": 1.4522291421890259,
      "learning_rate": 9.75461432692736e-06,
      "loss": 0.5037,
      "num_input_tokens_seen": 4215456,
      "step": 6450
    },
    {
      "epoch": 3.8060141509433962,
      "grad_norm": 1.6402428150177002,
      "learning_rate": 9.753817627708375e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 4218912,
      "step": 6455
    },
    {
      "epoch": 3.8089622641509435,
      "grad_norm": 0.7990967035293579,
      "learning_rate": 9.753019669895686e-06,
      "loss": 0.4834,
      "num_input_tokens_seen": 4223264,
      "step": 6460
    },
    {
      "epoch": 3.8119103773584904,
      "grad_norm": 2.4298441410064697,
      "learning_rate": 9.752220453700556e-06,
      "loss": 0.5319,
      "num_input_tokens_seen": 4225600,
      "step": 6465
    },
    {
      "epoch": 3.8148584905660377,
      "grad_norm": 2.1529860496520996,
      "learning_rate": 9.75141997933458e-06,
      "loss": 0.4209,
      "num_input_tokens_seen": 4228320,
      "step": 6470
    },
    {
      "epoch": 3.817806603773585,
      "grad_norm": 2.374169111251831,
      "learning_rate": 9.750618247009685e-06,
      "loss": 0.4836,
      "num_input_tokens_seen": 4231040,
      "step": 6475
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 1.8573365211486816,
      "learning_rate": 9.749815256938138e-06,
      "loss": 0.5099,
      "num_input_tokens_seen": 4234944,
      "step": 6480
    },
    {
      "epoch": 3.8237028301886795,
      "grad_norm": 3.1929380893707275,
      "learning_rate": 9.749011009332529e-06,
      "loss": 0.6765,
      "num_input_tokens_seen": 4237984,
      "step": 6485
    },
    {
      "epoch": 3.8266509433962264,
      "grad_norm": 3.2139370441436768,
      "learning_rate": 9.748205504405787e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 4240832,
      "step": 6490
    },
    {
      "epoch": 3.8295990566037736,
      "grad_norm": 3.171604633331299,
      "learning_rate": 9.747398742371177e-06,
      "loss": 0.5004,
      "num_input_tokens_seen": 4244608,
      "step": 6495
    },
    {
      "epoch": 3.8325471698113205,
      "grad_norm": 1.4513798952102661,
      "learning_rate": 9.746590723442289e-06,
      "loss": 0.5012,
      "num_input_tokens_seen": 4248416,
      "step": 6500
    },
    {
      "epoch": 3.8354952830188678,
      "grad_norm": 1.7050694227218628,
      "learning_rate": 9.745781447833049e-06,
      "loss": 0.4254,
      "num_input_tokens_seen": 4250848,
      "step": 6505
    },
    {
      "epoch": 3.838443396226415,
      "grad_norm": 1.0940502882003784,
      "learning_rate": 9.74497091575772e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 4254112,
      "step": 6510
    },
    {
      "epoch": 3.8413915094339623,
      "grad_norm": 1.9161567687988281,
      "learning_rate": 9.744159127430888e-06,
      "loss": 0.4751,
      "num_input_tokens_seen": 4258080,
      "step": 6515
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 1.1551562547683716,
      "learning_rate": 9.743346083067482e-06,
      "loss": 0.4724,
      "num_input_tokens_seen": 4261504,
      "step": 6520
    },
    {
      "epoch": 3.8472877358490565,
      "grad_norm": 1.4184552431106567,
      "learning_rate": 9.742531782882758e-06,
      "loss": 0.3629,
      "num_input_tokens_seen": 4267904,
      "step": 6525
    },
    {
      "epoch": 3.8502358490566038,
      "grad_norm": 2.0865414142608643,
      "learning_rate": 9.741716227092305e-06,
      "loss": 0.5104,
      "num_input_tokens_seen": 4270592,
      "step": 6530
    },
    {
      "epoch": 3.853183962264151,
      "grad_norm": 1.1559494733810425,
      "learning_rate": 9.740899415912048e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 4273216,
      "step": 6535
    },
    {
      "epoch": 3.856132075471698,
      "grad_norm": 1.060171365737915,
      "learning_rate": 9.740081349558236e-06,
      "loss": 0.5404,
      "num_input_tokens_seen": 4276960,
      "step": 6540
    },
    {
      "epoch": 3.859080188679245,
      "grad_norm": 2.117844343185425,
      "learning_rate": 9.739262028247459e-06,
      "loss": 0.4446,
      "num_input_tokens_seen": 4279424,
      "step": 6545
    },
    {
      "epoch": 3.8620283018867925,
      "grad_norm": 0.959248960018158,
      "learning_rate": 9.738441452196633e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 4283488,
      "step": 6550
    },
    {
      "epoch": 3.8649764150943398,
      "grad_norm": 1.136558175086975,
      "learning_rate": 9.737619621623013e-06,
      "loss": 0.5603,
      "num_input_tokens_seen": 4286944,
      "step": 6555
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 0.9683400392532349,
      "learning_rate": 9.73679653674418e-06,
      "loss": 0.5362,
      "num_input_tokens_seen": 4289600,
      "step": 6560
    },
    {
      "epoch": 3.870872641509434,
      "grad_norm": 1.1359095573425293,
      "learning_rate": 9.735972197778047e-06,
      "loss": 0.4715,
      "num_input_tokens_seen": 4292928,
      "step": 6565
    },
    {
      "epoch": 3.873820754716981,
      "grad_norm": 2.6346774101257324,
      "learning_rate": 9.735146604942867e-06,
      "loss": 0.5285,
      "num_input_tokens_seen": 4295584,
      "step": 6570
    },
    {
      "epoch": 3.8767688679245285,
      "grad_norm": 1.9369008541107178,
      "learning_rate": 9.734319758457214e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 4298304,
      "step": 6575
    },
    {
      "epoch": 3.8797169811320753,
      "grad_norm": 1.3324795961380005,
      "learning_rate": 9.733491658540001e-06,
      "loss": 0.4762,
      "num_input_tokens_seen": 4301600,
      "step": 6580
    },
    {
      "epoch": 3.8826650943396226,
      "grad_norm": 1.4051011800765991,
      "learning_rate": 9.732662305410474e-06,
      "loss": 0.4444,
      "num_input_tokens_seen": 4305792,
      "step": 6585
    },
    {
      "epoch": 3.88561320754717,
      "grad_norm": 1.6451137065887451,
      "learning_rate": 9.731831699288203e-06,
      "loss": 0.5579,
      "num_input_tokens_seen": 4308448,
      "step": 6590
    },
    {
      "epoch": 3.888561320754717,
      "grad_norm": 2.025876760482788,
      "learning_rate": 9.730999840393096e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 4312128,
      "step": 6595
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 1.3969205617904663,
      "learning_rate": 9.730166728945391e-06,
      "loss": 0.5054,
      "num_input_tokens_seen": 4316032,
      "step": 6600
    },
    {
      "epoch": 3.8944575471698113,
      "grad_norm": 1.3159922361373901,
      "learning_rate": 9.72933236516566e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 4319136,
      "step": 6605
    },
    {
      "epoch": 3.8974056603773586,
      "grad_norm": 1.6401150226593018,
      "learning_rate": 9.728496749274806e-06,
      "loss": 0.5363,
      "num_input_tokens_seen": 4321696,
      "step": 6610
    },
    {
      "epoch": 3.9003537735849054,
      "grad_norm": 1.5129942893981934,
      "learning_rate": 9.727659881494054e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 4324576,
      "step": 6615
    },
    {
      "epoch": 3.9033018867924527,
      "grad_norm": 2.395235776901245,
      "learning_rate": 9.726821762044975e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 4327584,
      "step": 6620
    },
    {
      "epoch": 3.90625,
      "grad_norm": 1.5721921920776367,
      "learning_rate": 9.725982391149465e-06,
      "loss": 0.439,
      "num_input_tokens_seen": 4330784,
      "step": 6625
    },
    {
      "epoch": 3.9091981132075473,
      "grad_norm": 0.9298762679100037,
      "learning_rate": 9.725141769029747e-06,
      "loss": 0.5596,
      "num_input_tokens_seen": 4333984,
      "step": 6630
    },
    {
      "epoch": 3.9121462264150946,
      "grad_norm": 1.7009413242340088,
      "learning_rate": 9.72429989590838e-06,
      "loss": 0.4699,
      "num_input_tokens_seen": 4337632,
      "step": 6635
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 1.2141597270965576,
      "learning_rate": 9.723456772008257e-06,
      "loss": 0.3391,
      "num_input_tokens_seen": 4340544,
      "step": 6640
    },
    {
      "epoch": 3.9180424528301887,
      "grad_norm": 1.0265843868255615,
      "learning_rate": 9.722612397552598e-06,
      "loss": 0.4008,
      "num_input_tokens_seen": 4343136,
      "step": 6645
    },
    {
      "epoch": 3.920990566037736,
      "grad_norm": 1.1616767644882202,
      "learning_rate": 9.72176677276495e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 4346336,
      "step": 6650
    },
    {
      "epoch": 3.923938679245283,
      "grad_norm": 1.5487358570098877,
      "learning_rate": 9.7209198978692e-06,
      "loss": 0.6319,
      "num_input_tokens_seen": 4350560,
      "step": 6655
    },
    {
      "epoch": 3.92688679245283,
      "grad_norm": 1.2958273887634277,
      "learning_rate": 9.720071773089564e-06,
      "loss": 0.5134,
      "num_input_tokens_seen": 4354528,
      "step": 6660
    },
    {
      "epoch": 3.9298349056603774,
      "grad_norm": 1.6039414405822754,
      "learning_rate": 9.71922239865058e-06,
      "loss": 0.4497,
      "num_input_tokens_seen": 4358112,
      "step": 6665
    },
    {
      "epoch": 3.9327830188679247,
      "grad_norm": 0.9181948900222778,
      "learning_rate": 9.718371774777131e-06,
      "loss": 0.387,
      "num_input_tokens_seen": 4361888,
      "step": 6670
    },
    {
      "epoch": 3.9357311320754715,
      "grad_norm": 1.8333975076675415,
      "learning_rate": 9.717519901694416e-06,
      "loss": 0.6064,
      "num_input_tokens_seen": 4365088,
      "step": 6675
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 3.0031418800354004,
      "learning_rate": 9.716666779627978e-06,
      "loss": 0.4618,
      "num_input_tokens_seen": 4368000,
      "step": 6680
    },
    {
      "epoch": 3.941627358490566,
      "grad_norm": 1.4607669115066528,
      "learning_rate": 9.715812408803681e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 4370816,
      "step": 6685
    },
    {
      "epoch": 3.944575471698113,
      "grad_norm": 2.631556987762451,
      "learning_rate": 9.714956789447726e-06,
      "loss": 0.5228,
      "num_input_tokens_seen": 4374048,
      "step": 6690
    },
    {
      "epoch": 3.9475235849056602,
      "grad_norm": 2.4600791931152344,
      "learning_rate": 9.71409992178664e-06,
      "loss": 0.5366,
      "num_input_tokens_seen": 4377888,
      "step": 6695
    },
    {
      "epoch": 3.9504716981132075,
      "grad_norm": 2.461810350418091,
      "learning_rate": 9.713241806047282e-06,
      "loss": 0.5543,
      "num_input_tokens_seen": 4380640,
      "step": 6700
    },
    {
      "epoch": 3.953419811320755,
      "grad_norm": 1.9499585628509521,
      "learning_rate": 9.712382442456845e-06,
      "loss": 0.7052,
      "num_input_tokens_seen": 4383296,
      "step": 6705
    },
    {
      "epoch": 3.956367924528302,
      "grad_norm": 1.3980119228363037,
      "learning_rate": 9.711521831242846e-06,
      "loss": 0.4783,
      "num_input_tokens_seen": 4386432,
      "step": 6710
    },
    {
      "epoch": 3.959316037735849,
      "grad_norm": 2.140974283218384,
      "learning_rate": 9.710659972633137e-06,
      "loss": 0.5973,
      "num_input_tokens_seen": 4388704,
      "step": 6715
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 1.50264573097229,
      "learning_rate": 9.709796866855899e-06,
      "loss": 0.3276,
      "num_input_tokens_seen": 4391744,
      "step": 6720
    },
    {
      "epoch": 3.9652122641509435,
      "grad_norm": 1.5842195749282837,
      "learning_rate": 9.70893251413964e-06,
      "loss": 0.4903,
      "num_input_tokens_seen": 4395552,
      "step": 6725
    },
    {
      "epoch": 3.9681603773584904,
      "grad_norm": 1.1841042041778564,
      "learning_rate": 9.708066914713205e-06,
      "loss": 0.6713,
      "num_input_tokens_seen": 4398848,
      "step": 6730
    },
    {
      "epoch": 3.9711084905660377,
      "grad_norm": 1.655017375946045,
      "learning_rate": 9.707200068805764e-06,
      "loss": 0.4412,
      "num_input_tokens_seen": 4404064,
      "step": 6735
    },
    {
      "epoch": 3.974056603773585,
      "grad_norm": 1.5305254459381104,
      "learning_rate": 9.706331976646817e-06,
      "loss": 0.3734,
      "num_input_tokens_seen": 4406560,
      "step": 6740
    },
    {
      "epoch": 3.9770047169811322,
      "grad_norm": 2.3772377967834473,
      "learning_rate": 9.705462638466197e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 4409440,
      "step": 6745
    },
    {
      "epoch": 3.9799528301886795,
      "grad_norm": 1.6252679824829102,
      "learning_rate": 9.704592054494065e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 4413376,
      "step": 6750
    },
    {
      "epoch": 3.9829009433962264,
      "grad_norm": 1.9509201049804688,
      "learning_rate": 9.703720224960909e-06,
      "loss": 0.4809,
      "num_input_tokens_seen": 4417248,
      "step": 6755
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 1.4275671243667603,
      "learning_rate": 9.702847150097552e-06,
      "loss": 0.3747,
      "num_input_tokens_seen": 4419904,
      "step": 6760
    },
    {
      "epoch": 3.9887971698113205,
      "grad_norm": 1.2485082149505615,
      "learning_rate": 9.701972830135143e-06,
      "loss": 0.3791,
      "num_input_tokens_seen": 4423360,
      "step": 6765
    },
    {
      "epoch": 3.9917452830188678,
      "grad_norm": 2.6905627250671387,
      "learning_rate": 9.701097265305164e-06,
      "loss": 0.4864,
      "num_input_tokens_seen": 4426656,
      "step": 6770
    },
    {
      "epoch": 3.994693396226415,
      "grad_norm": 5.6421613693237305,
      "learning_rate": 9.700220455839422e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 4429440,
      "step": 6775
    },
    {
      "epoch": 3.9976415094339623,
      "grad_norm": 1.8210076093673706,
      "learning_rate": 9.69934240197006e-06,
      "loss": 0.6077,
      "num_input_tokens_seen": 4433344,
      "step": 6780
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.48438429832458496,
      "eval_runtime": 19.0962,
      "eval_samples_per_second": 88.814,
      "eval_steps_per_second": 22.203,
      "num_input_tokens_seen": 4435584,
      "step": 6784
    },
    {
      "epoch": 4.00058962264151,
      "grad_norm": 1.072381854057312,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.372,
      "num_input_tokens_seen": 4436224,
      "step": 6785
    },
    {
      "epoch": 4.003537735849057,
      "grad_norm": 1.4454569816589355,
      "learning_rate": 9.697582561950669e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 4439552,
      "step": 6790
    },
    {
      "epoch": 4.006485849056604,
      "grad_norm": 1.3843566179275513,
      "learning_rate": 9.696700776266568e-06,
      "loss": 0.3422,
      "num_input_tokens_seen": 4442496,
      "step": 6795
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 1.5517284870147705,
      "learning_rate": 9.695817747110694e-06,
      "loss": 0.5398,
      "num_input_tokens_seen": 4445440,
      "step": 6800
    },
    {
      "epoch": 4.012382075471698,
      "grad_norm": 1.336027979850769,
      "learning_rate": 9.694933474716831e-06,
      "loss": 0.4497,
      "num_input_tokens_seen": 4448640,
      "step": 6805
    },
    {
      "epoch": 4.015330188679245,
      "grad_norm": 2.1746997833251953,
      "learning_rate": 9.6940479593191e-06,
      "loss": 0.3685,
      "num_input_tokens_seen": 4452736,
      "step": 6810
    },
    {
      "epoch": 4.0182783018867925,
      "grad_norm": 1.889971137046814,
      "learning_rate": 9.693161201151942e-06,
      "loss": 0.559,
      "num_input_tokens_seen": 4455808,
      "step": 6815
    },
    {
      "epoch": 4.02122641509434,
      "grad_norm": 1.5759683847427368,
      "learning_rate": 9.692273200450128e-06,
      "loss": 0.3939,
      "num_input_tokens_seen": 4458272,
      "step": 6820
    },
    {
      "epoch": 4.024174528301887,
      "grad_norm": 1.9908459186553955,
      "learning_rate": 9.69138395744876e-06,
      "loss": 0.5289,
      "num_input_tokens_seen": 4461024,
      "step": 6825
    },
    {
      "epoch": 4.027122641509434,
      "grad_norm": 2.7365169525146484,
      "learning_rate": 9.690493472383274e-06,
      "loss": 0.5613,
      "num_input_tokens_seen": 4463744,
      "step": 6830
    },
    {
      "epoch": 4.030070754716981,
      "grad_norm": 1.0328547954559326,
      "learning_rate": 9.689601745489423e-06,
      "loss": 0.4967,
      "num_input_tokens_seen": 4466560,
      "step": 6835
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 1.5940300226211548,
      "learning_rate": 9.6887087770033e-06,
      "loss": 0.5322,
      "num_input_tokens_seen": 4469408,
      "step": 6840
    },
    {
      "epoch": 4.035966981132075,
      "grad_norm": 1.4302631616592407,
      "learning_rate": 9.687814567161322e-06,
      "loss": 0.5631,
      "num_input_tokens_seen": 4472672,
      "step": 6845
    },
    {
      "epoch": 4.038915094339623,
      "grad_norm": 1.9695943593978882,
      "learning_rate": 9.686919116200232e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 4475520,
      "step": 6850
    },
    {
      "epoch": 4.04186320754717,
      "grad_norm": 2.710556745529175,
      "learning_rate": 9.686022424357108e-06,
      "loss": 0.3558,
      "num_input_tokens_seen": 4479328,
      "step": 6855
    },
    {
      "epoch": 4.044811320754717,
      "grad_norm": 1.2896437644958496,
      "learning_rate": 9.685124491869353e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 4483328,
      "step": 6860
    },
    {
      "epoch": 4.0477594339622645,
      "grad_norm": 1.8668121099472046,
      "learning_rate": 9.684225318974696e-06,
      "loss": 0.5363,
      "num_input_tokens_seen": 4486752,
      "step": 6865
    },
    {
      "epoch": 4.050707547169812,
      "grad_norm": 1.5209589004516602,
      "learning_rate": 9.683324905911197e-06,
      "loss": 0.5311,
      "num_input_tokens_seen": 4489792,
      "step": 6870
    },
    {
      "epoch": 4.053655660377358,
      "grad_norm": 1.5439870357513428,
      "learning_rate": 9.682423252917245e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 4492672,
      "step": 6875
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 1.071082353591919,
      "learning_rate": 9.681520360231557e-06,
      "loss": 0.3487,
      "num_input_tokens_seen": 4496096,
      "step": 6880
    },
    {
      "epoch": 4.059551886792453,
      "grad_norm": 1.3899497985839844,
      "learning_rate": 9.680616228093178e-06,
      "loss": 0.5202,
      "num_input_tokens_seen": 4499040,
      "step": 6885
    },
    {
      "epoch": 4.0625,
      "grad_norm": 1.5542653799057007,
      "learning_rate": 9.67971085674148e-06,
      "loss": 0.4509,
      "num_input_tokens_seen": 4502016,
      "step": 6890
    },
    {
      "epoch": 4.065448113207547,
      "grad_norm": 2.4277191162109375,
      "learning_rate": 9.678804246416164e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 4504960,
      "step": 6895
    },
    {
      "epoch": 4.068396226415095,
      "grad_norm": 0.9027440547943115,
      "learning_rate": 9.677896397357259e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 4508704,
      "step": 6900
    },
    {
      "epoch": 4.071344339622642,
      "grad_norm": 1.1059106588363647,
      "learning_rate": 9.676987309805121e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 4512064,
      "step": 6905
    },
    {
      "epoch": 4.074292452830188,
      "grad_norm": 2.0204622745513916,
      "learning_rate": 9.67607698400044e-06,
      "loss": 0.4803,
      "num_input_tokens_seen": 4516256,
      "step": 6910
    },
    {
      "epoch": 4.0772405660377355,
      "grad_norm": 1.3957586288452148,
      "learning_rate": 9.67516542018422e-06,
      "loss": 0.49,
      "num_input_tokens_seen": 4519552,
      "step": 6915
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 1.3008168935775757,
      "learning_rate": 9.67425261859781e-06,
      "loss": 0.5735,
      "num_input_tokens_seen": 4522912,
      "step": 6920
    },
    {
      "epoch": 4.08313679245283,
      "grad_norm": 1.3227643966674805,
      "learning_rate": 9.673338579482871e-06,
      "loss": 0.3723,
      "num_input_tokens_seen": 4526208,
      "step": 6925
    },
    {
      "epoch": 4.086084905660377,
      "grad_norm": 2.177189588546753,
      "learning_rate": 9.672423303081404e-06,
      "loss": 0.4642,
      "num_input_tokens_seen": 4528736,
      "step": 6930
    },
    {
      "epoch": 4.089033018867925,
      "grad_norm": 1.254641056060791,
      "learning_rate": 9.67150678963573e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 4531712,
      "step": 6935
    },
    {
      "epoch": 4.091981132075472,
      "grad_norm": 3.1974799633026123,
      "learning_rate": 9.670589039388501e-06,
      "loss": 0.6676,
      "num_input_tokens_seen": 4535296,
      "step": 6940
    },
    {
      "epoch": 4.094929245283019,
      "grad_norm": 1.7800627946853638,
      "learning_rate": 9.669670052582695e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 4537792,
      "step": 6945
    },
    {
      "epoch": 4.097877358490566,
      "grad_norm": 1.6227556467056274,
      "learning_rate": 9.668749829461617e-06,
      "loss": 0.4433,
      "num_input_tokens_seen": 4542144,
      "step": 6950
    },
    {
      "epoch": 4.100825471698113,
      "grad_norm": 2.2978737354278564,
      "learning_rate": 9.667828370268898e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 4544672,
      "step": 6955
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 1.5856572389602661,
      "learning_rate": 9.666905675248505e-06,
      "loss": 0.7013,
      "num_input_tokens_seen": 4548128,
      "step": 6960
    },
    {
      "epoch": 4.1067216981132075,
      "grad_norm": 1.4127230644226074,
      "learning_rate": 9.66598174464472e-06,
      "loss": 0.5045,
      "num_input_tokens_seen": 4553696,
      "step": 6965
    },
    {
      "epoch": 4.109669811320755,
      "grad_norm": 1.507616400718689,
      "learning_rate": 9.665056578702157e-06,
      "loss": 0.4734,
      "num_input_tokens_seen": 4556512,
      "step": 6970
    },
    {
      "epoch": 4.112617924528302,
      "grad_norm": 1.6608260869979858,
      "learning_rate": 9.66413017766576e-06,
      "loss": 0.4957,
      "num_input_tokens_seen": 4559680,
      "step": 6975
    },
    {
      "epoch": 4.115566037735849,
      "grad_norm": 2.2198104858398438,
      "learning_rate": 9.663202541780799e-06,
      "loss": 0.547,
      "num_input_tokens_seen": 4562560,
      "step": 6980
    },
    {
      "epoch": 4.118514150943396,
      "grad_norm": 1.7912169694900513,
      "learning_rate": 9.662273671292866e-06,
      "loss": 0.4066,
      "num_input_tokens_seen": 4566624,
      "step": 6985
    },
    {
      "epoch": 4.121462264150943,
      "grad_norm": 1.8897771835327148,
      "learning_rate": 9.661343566447886e-06,
      "loss": 0.549,
      "num_input_tokens_seen": 4570176,
      "step": 6990
    },
    {
      "epoch": 4.12441037735849,
      "grad_norm": 0.9435240626335144,
      "learning_rate": 9.660412227492107e-06,
      "loss": 0.4069,
      "num_input_tokens_seen": 4573632,
      "step": 6995
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 2.4019994735717773,
      "learning_rate": 9.659479654672106e-06,
      "loss": 0.6018,
      "num_input_tokens_seen": 4576960,
      "step": 7000
    },
    {
      "epoch": 4.130306603773585,
      "grad_norm": 0.4784364104270935,
      "learning_rate": 9.658545848234784e-06,
      "loss": 0.5107,
      "num_input_tokens_seen": 4585184,
      "step": 7005
    },
    {
      "epoch": 4.133254716981132,
      "grad_norm": 3.1462862491607666,
      "learning_rate": 9.657610808427372e-06,
      "loss": 0.4108,
      "num_input_tokens_seen": 4587808,
      "step": 7010
    },
    {
      "epoch": 4.1362028301886795,
      "grad_norm": 1.3664734363555908,
      "learning_rate": 9.656674535497425e-06,
      "loss": 0.5263,
      "num_input_tokens_seen": 4590464,
      "step": 7015
    },
    {
      "epoch": 4.139150943396227,
      "grad_norm": 2.0244321823120117,
      "learning_rate": 9.655737029692827e-06,
      "loss": 0.4374,
      "num_input_tokens_seen": 4593472,
      "step": 7020
    },
    {
      "epoch": 4.142099056603773,
      "grad_norm": 4.052973747253418,
      "learning_rate": 9.654798291261785e-06,
      "loss": 0.452,
      "num_input_tokens_seen": 4595968,
      "step": 7025
    },
    {
      "epoch": 4.1450471698113205,
      "grad_norm": 1.386129379272461,
      "learning_rate": 9.653858320452833e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 4598880,
      "step": 7030
    },
    {
      "epoch": 4.147995283018868,
      "grad_norm": 4.375176906585693,
      "learning_rate": 9.652917117514836e-06,
      "loss": 0.3795,
      "num_input_tokens_seen": 4602336,
      "step": 7035
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 3.6257481575012207,
      "learning_rate": 9.651974682696975e-06,
      "loss": 0.5091,
      "num_input_tokens_seen": 4605632,
      "step": 7040
    },
    {
      "epoch": 4.153891509433962,
      "grad_norm": 1.6394401788711548,
      "learning_rate": 9.651031016248773e-06,
      "loss": 0.428,
      "num_input_tokens_seen": 4609440,
      "step": 7045
    },
    {
      "epoch": 4.15683962264151,
      "grad_norm": 1.7606271505355835,
      "learning_rate": 9.65008611842006e-06,
      "loss": 0.5875,
      "num_input_tokens_seen": 4612256,
      "step": 7050
    },
    {
      "epoch": 4.159787735849057,
      "grad_norm": 1.872793436050415,
      "learning_rate": 9.64913998946101e-06,
      "loss": 0.4694,
      "num_input_tokens_seen": 4615552,
      "step": 7055
    },
    {
      "epoch": 4.162735849056604,
      "grad_norm": 1.174114465713501,
      "learning_rate": 9.648192629622109e-06,
      "loss": 0.5508,
      "num_input_tokens_seen": 4618688,
      "step": 7060
    },
    {
      "epoch": 4.165683962264151,
      "grad_norm": 1.1568256616592407,
      "learning_rate": 9.647244039154178e-06,
      "loss": 0.4389,
      "num_input_tokens_seen": 4621632,
      "step": 7065
    },
    {
      "epoch": 4.168632075471698,
      "grad_norm": 1.6358710527420044,
      "learning_rate": 9.64629421830836e-06,
      "loss": 0.4974,
      "num_input_tokens_seen": 4624704,
      "step": 7070
    },
    {
      "epoch": 4.171580188679245,
      "grad_norm": 1.855725884437561,
      "learning_rate": 9.64534316733612e-06,
      "loss": 0.3256,
      "num_input_tokens_seen": 4628768,
      "step": 7075
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 1.8478025197982788,
      "learning_rate": 9.644390886489258e-06,
      "loss": 0.5891,
      "num_input_tokens_seen": 4632000,
      "step": 7080
    },
    {
      "epoch": 4.17747641509434,
      "grad_norm": 1.4135563373565674,
      "learning_rate": 9.643437376019893e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 4634592,
      "step": 7085
    },
    {
      "epoch": 4.180424528301887,
      "grad_norm": 0.7546047568321228,
      "learning_rate": 9.64248263618047e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 4640064,
      "step": 7090
    },
    {
      "epoch": 4.183372641509434,
      "grad_norm": 1.5586471557617188,
      "learning_rate": 9.64152666722376e-06,
      "loss": 0.4739,
      "num_input_tokens_seen": 4644000,
      "step": 7095
    },
    {
      "epoch": 4.186320754716981,
      "grad_norm": 5.412915229797363,
      "learning_rate": 9.640569469402863e-06,
      "loss": 0.4231,
      "num_input_tokens_seen": 4647328,
      "step": 7100
    },
    {
      "epoch": 4.189268867924528,
      "grad_norm": 1.656447172164917,
      "learning_rate": 9.639611042971198e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 4650432,
      "step": 7105
    },
    {
      "epoch": 4.192216981132075,
      "grad_norm": 1.6821658611297607,
      "learning_rate": 9.638651388182514e-06,
      "loss": 0.5613,
      "num_input_tokens_seen": 4652960,
      "step": 7110
    },
    {
      "epoch": 4.195165094339623,
      "grad_norm": 1.298686146736145,
      "learning_rate": 9.637690505290884e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 4655872,
      "step": 7115
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 2.294571876525879,
      "learning_rate": 9.636728394550705e-06,
      "loss": 0.5486,
      "num_input_tokens_seen": 4660288,
      "step": 7120
    },
    {
      "epoch": 4.201061320754717,
      "grad_norm": 1.4789689779281616,
      "learning_rate": 9.6357650562167e-06,
      "loss": 0.5309,
      "num_input_tokens_seen": 4663360,
      "step": 7125
    },
    {
      "epoch": 4.2040094339622645,
      "grad_norm": 1.0539246797561646,
      "learning_rate": 9.634800490543918e-06,
      "loss": 0.4112,
      "num_input_tokens_seen": 4667200,
      "step": 7130
    },
    {
      "epoch": 4.206957547169812,
      "grad_norm": 1.1202205419540405,
      "learning_rate": 9.633834697787731e-06,
      "loss": 0.4568,
      "num_input_tokens_seen": 4670912,
      "step": 7135
    },
    {
      "epoch": 4.209905660377358,
      "grad_norm": 1.0558562278747559,
      "learning_rate": 9.632867678203836e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 4673568,
      "step": 7140
    },
    {
      "epoch": 4.212853773584905,
      "grad_norm": 3.6966497898101807,
      "learning_rate": 9.631899432048258e-06,
      "loss": 0.5507,
      "num_input_tokens_seen": 4676032,
      "step": 7145
    },
    {
      "epoch": 4.215801886792453,
      "grad_norm": 4.106484889984131,
      "learning_rate": 9.630929959577343e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 4678784,
      "step": 7150
    },
    {
      "epoch": 4.21875,
      "grad_norm": 1.469151258468628,
      "learning_rate": 9.629959261047764e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 4681696,
      "step": 7155
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 1.7285774946212769,
      "learning_rate": 9.628987336716513e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 4686816,
      "step": 7160
    },
    {
      "epoch": 4.224646226415095,
      "grad_norm": 1.7863441705703735,
      "learning_rate": 9.628014186840918e-06,
      "loss": 0.4143,
      "num_input_tokens_seen": 4689472,
      "step": 7165
    },
    {
      "epoch": 4.227594339622642,
      "grad_norm": 0.9573928713798523,
      "learning_rate": 9.62703981167862e-06,
      "loss": 0.4141,
      "num_input_tokens_seen": 4692288,
      "step": 7170
    },
    {
      "epoch": 4.230542452830188,
      "grad_norm": 1.2845207452774048,
      "learning_rate": 9.626064211487592e-06,
      "loss": 0.4743,
      "num_input_tokens_seen": 4696320,
      "step": 7175
    },
    {
      "epoch": 4.2334905660377355,
      "grad_norm": 1.5434110164642334,
      "learning_rate": 9.625087386526125e-06,
      "loss": 0.4948,
      "num_input_tokens_seen": 4699424,
      "step": 7180
    },
    {
      "epoch": 4.236438679245283,
      "grad_norm": 1.187532663345337,
      "learning_rate": 9.624109337052839e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 4702720,
      "step": 7185
    },
    {
      "epoch": 4.23938679245283,
      "grad_norm": 1.23360013961792,
      "learning_rate": 9.623130063326678e-06,
      "loss": 0.373,
      "num_input_tokens_seen": 4706080,
      "step": 7190
    },
    {
      "epoch": 4.242334905660377,
      "grad_norm": 2.0539867877960205,
      "learning_rate": 9.622149565606909e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 4709632,
      "step": 7195
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 1.868472695350647,
      "learning_rate": 9.621167844153122e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 4712640,
      "step": 7200
    },
    {
      "epoch": 4.248231132075472,
      "grad_norm": 2.8843917846679688,
      "learning_rate": 9.620184899225231e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 4715200,
      "step": 7205
    },
    {
      "epoch": 4.251179245283019,
      "grad_norm": 1.8268054723739624,
      "learning_rate": 9.619200731083477e-06,
      "loss": 0.409,
      "num_input_tokens_seen": 4718912,
      "step": 7210
    },
    {
      "epoch": 4.254127358490566,
      "grad_norm": 1.6688611507415771,
      "learning_rate": 9.618215339988422e-06,
      "loss": 0.4729,
      "num_input_tokens_seen": 4722400,
      "step": 7215
    },
    {
      "epoch": 4.257075471698113,
      "grad_norm": 1.3193798065185547,
      "learning_rate": 9.617228726200951e-06,
      "loss": 0.5856,
      "num_input_tokens_seen": 4726048,
      "step": 7220
    },
    {
      "epoch": 4.26002358490566,
      "grad_norm": 1.9256333112716675,
      "learning_rate": 9.616240889982277e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 4729568,
      "step": 7225
    },
    {
      "epoch": 4.2629716981132075,
      "grad_norm": 1.1934884786605835,
      "learning_rate": 9.61525183159393e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 4733024,
      "step": 7230
    },
    {
      "epoch": 4.265919811320755,
      "grad_norm": 1.4925769567489624,
      "learning_rate": 9.614261551297774e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 4735776,
      "step": 7235
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 1.766947865486145,
      "learning_rate": 9.613270049355983e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 4738720,
      "step": 7240
    },
    {
      "epoch": 4.271816037735849,
      "grad_norm": 3.1467254161834717,
      "learning_rate": 9.612277326031065e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 4741472,
      "step": 7245
    },
    {
      "epoch": 4.274764150943396,
      "grad_norm": 2.0735387802124023,
      "learning_rate": 9.611283381585848e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 4744896,
      "step": 7250
    },
    {
      "epoch": 4.277712264150943,
      "grad_norm": 1.1981290578842163,
      "learning_rate": 9.61028821628348e-06,
      "loss": 0.4379,
      "num_input_tokens_seen": 4748576,
      "step": 7255
    },
    {
      "epoch": 4.28066037735849,
      "grad_norm": 1.3056575059890747,
      "learning_rate": 9.609291830387439e-06,
      "loss": 0.4932,
      "num_input_tokens_seen": 4752000,
      "step": 7260
    },
    {
      "epoch": 4.283608490566038,
      "grad_norm": 1.3185089826583862,
      "learning_rate": 9.608294224161523e-06,
      "loss": 0.2798,
      "num_input_tokens_seen": 4755840,
      "step": 7265
    },
    {
      "epoch": 4.286556603773585,
      "grad_norm": 1.1648133993148804,
      "learning_rate": 9.607295397869847e-06,
      "loss": 0.3109,
      "num_input_tokens_seen": 4759136,
      "step": 7270
    },
    {
      "epoch": 4.289504716981132,
      "grad_norm": 2.056828498840332,
      "learning_rate": 9.60629535177686e-06,
      "loss": 0.3127,
      "num_input_tokens_seen": 4763264,
      "step": 7275
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 1.9875962734222412,
      "learning_rate": 9.605294086147325e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 4766528,
      "step": 7280
    },
    {
      "epoch": 4.295400943396227,
      "grad_norm": 1.6677964925765991,
      "learning_rate": 9.604291601246333e-06,
      "loss": 0.4663,
      "num_input_tokens_seen": 4769216,
      "step": 7285
    },
    {
      "epoch": 4.298349056603773,
      "grad_norm": 2.173290491104126,
      "learning_rate": 9.603287897339299e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 4772416,
      "step": 7290
    },
    {
      "epoch": 4.3012971698113205,
      "grad_norm": 2.1238534450531006,
      "learning_rate": 9.602282974691953e-06,
      "loss": 0.4954,
      "num_input_tokens_seen": 4775840,
      "step": 7295
    },
    {
      "epoch": 4.304245283018868,
      "grad_norm": 1.0428217649459839,
      "learning_rate": 9.601276833570355e-06,
      "loss": 0.4932,
      "num_input_tokens_seen": 4779296,
      "step": 7300
    },
    {
      "epoch": 4.307193396226415,
      "grad_norm": 1.297163963317871,
      "learning_rate": 9.600269474240885e-06,
      "loss": 0.3881,
      "num_input_tokens_seen": 4782400,
      "step": 7305
    },
    {
      "epoch": 4.310141509433962,
      "grad_norm": 0.9883877038955688,
      "learning_rate": 9.599260896970246e-06,
      "loss": 0.4987,
      "num_input_tokens_seen": 4785792,
      "step": 7310
    },
    {
      "epoch": 4.31308962264151,
      "grad_norm": 1.7175856828689575,
      "learning_rate": 9.598251102025463e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 4788736,
      "step": 7315
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 2.0105319023132324,
      "learning_rate": 9.597240089673882e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 4792032,
      "step": 7320
    },
    {
      "epoch": 4.318985849056604,
      "grad_norm": 2.4700934886932373,
      "learning_rate": 9.596227860183175e-06,
      "loss": 0.4691,
      "num_input_tokens_seen": 4795424,
      "step": 7325
    },
    {
      "epoch": 4.321933962264151,
      "grad_norm": 1.4187638759613037,
      "learning_rate": 9.595214413821334e-06,
      "loss": 0.443,
      "num_input_tokens_seen": 4798144,
      "step": 7330
    },
    {
      "epoch": 4.324882075471698,
      "grad_norm": 1.401833176612854,
      "learning_rate": 9.59419975085667e-06,
      "loss": 0.5425,
      "num_input_tokens_seen": 4800736,
      "step": 7335
    },
    {
      "epoch": 4.327830188679245,
      "grad_norm": 0.9488842487335205,
      "learning_rate": 9.593183871557826e-06,
      "loss": 0.506,
      "num_input_tokens_seen": 4804832,
      "step": 7340
    },
    {
      "epoch": 4.3307783018867925,
      "grad_norm": 1.0184153318405151,
      "learning_rate": 9.592166776193754e-06,
      "loss": 0.4857,
      "num_input_tokens_seen": 4807488,
      "step": 7345
    },
    {
      "epoch": 4.33372641509434,
      "grad_norm": 1.3929153680801392,
      "learning_rate": 9.591148465033738e-06,
      "loss": 0.658,
      "num_input_tokens_seen": 4812192,
      "step": 7350
    },
    {
      "epoch": 4.336674528301887,
      "grad_norm": 3.6355948448181152,
      "learning_rate": 9.590128938347378e-06,
      "loss": 0.5659,
      "num_input_tokens_seen": 4814592,
      "step": 7355
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 2.960806131362915,
      "learning_rate": 9.589108196404599e-06,
      "loss": 0.3076,
      "num_input_tokens_seen": 4817344,
      "step": 7360
    },
    {
      "epoch": 4.342570754716981,
      "grad_norm": 1.8832556009292603,
      "learning_rate": 9.588086239475649e-06,
      "loss": 0.4505,
      "num_input_tokens_seen": 4819616,
      "step": 7365
    },
    {
      "epoch": 4.345518867924528,
      "grad_norm": 2.365912437438965,
      "learning_rate": 9.587063067831092e-06,
      "loss": 0.5254,
      "num_input_tokens_seen": 4822272,
      "step": 7370
    },
    {
      "epoch": 4.348466981132075,
      "grad_norm": 1.4596527814865112,
      "learning_rate": 9.586038681741818e-06,
      "loss": 0.3274,
      "num_input_tokens_seen": 4824960,
      "step": 7375
    },
    {
      "epoch": 4.351415094339623,
      "grad_norm": 2.328187942504883,
      "learning_rate": 9.58501308147904e-06,
      "loss": 0.5573,
      "num_input_tokens_seen": 4827968,
      "step": 7380
    },
    {
      "epoch": 4.35436320754717,
      "grad_norm": 2.731288194656372,
      "learning_rate": 9.583986267314288e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 4831776,
      "step": 7385
    },
    {
      "epoch": 4.357311320754717,
      "grad_norm": 2.644049882888794,
      "learning_rate": 9.582958239519416e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 4834272,
      "step": 7390
    },
    {
      "epoch": 4.3602594339622645,
      "grad_norm": 2.0315990447998047,
      "learning_rate": 9.581928998366597e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 4836928,
      "step": 7395
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 4.9690728187561035,
      "learning_rate": 9.58089854412833e-06,
      "loss": 0.618,
      "num_input_tokens_seen": 4840224,
      "step": 7400
    },
    {
      "epoch": 4.366155660377358,
      "grad_norm": 1.2889043092727661,
      "learning_rate": 9.579866877077431e-06,
      "loss": 0.395,
      "num_input_tokens_seen": 4843680,
      "step": 7405
    },
    {
      "epoch": 4.369103773584905,
      "grad_norm": 0.8798941969871521,
      "learning_rate": 9.578833997487038e-06,
      "loss": 0.5152,
      "num_input_tokens_seen": 4847424,
      "step": 7410
    },
    {
      "epoch": 4.372051886792453,
      "grad_norm": 2.0116095542907715,
      "learning_rate": 9.57779990563061e-06,
      "loss": 0.4004,
      "num_input_tokens_seen": 4850752,
      "step": 7415
    },
    {
      "epoch": 4.375,
      "grad_norm": 2.6594152450561523,
      "learning_rate": 9.576764601781928e-06,
      "loss": 0.4124,
      "num_input_tokens_seen": 4853408,
      "step": 7420
    },
    {
      "epoch": 4.377948113207547,
      "grad_norm": 1.1749051809310913,
      "learning_rate": 9.575728086215093e-06,
      "loss": 0.4824,
      "num_input_tokens_seen": 4856640,
      "step": 7425
    },
    {
      "epoch": 4.380896226415095,
      "grad_norm": 1.948190689086914,
      "learning_rate": 9.574690359204527e-06,
      "loss": 0.5818,
      "num_input_tokens_seen": 4859872,
      "step": 7430
    },
    {
      "epoch": 4.383844339622642,
      "grad_norm": 1.0961192846298218,
      "learning_rate": 9.573651421024972e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 4863360,
      "step": 7435
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 1.5679632425308228,
      "learning_rate": 9.572611271951494e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 4865792,
      "step": 7440
    },
    {
      "epoch": 4.3897405660377355,
      "grad_norm": 2.2737958431243896,
      "learning_rate": 9.571569912259473e-06,
      "loss": 0.5899,
      "num_input_tokens_seen": 4868960,
      "step": 7445
    },
    {
      "epoch": 4.392688679245283,
      "grad_norm": 1.4222983121871948,
      "learning_rate": 9.570527342224614e-06,
      "loss": 0.5029,
      "num_input_tokens_seen": 4872608,
      "step": 7450
    },
    {
      "epoch": 4.39563679245283,
      "grad_norm": 1.5546033382415771,
      "learning_rate": 9.569483562122945e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 4878912,
      "step": 7455
    },
    {
      "epoch": 4.398584905660377,
      "grad_norm": 2.138932943344116,
      "learning_rate": 9.568438572230811e-06,
      "loss": 0.4954,
      "num_input_tokens_seen": 4882368,
      "step": 7460
    },
    {
      "epoch": 4.401533018867925,
      "grad_norm": 2.7993390560150146,
      "learning_rate": 9.567392372824873e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 4885152,
      "step": 7465
    },
    {
      "epoch": 4.404481132075472,
      "grad_norm": 1.3708701133728027,
      "learning_rate": 9.566344964182123e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 4888160,
      "step": 7470
    },
    {
      "epoch": 4.407429245283019,
      "grad_norm": 1.7168391942977905,
      "learning_rate": 9.565296346579862e-06,
      "loss": 0.3637,
      "num_input_tokens_seen": 4891520,
      "step": 7475
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 1.879428744316101,
      "learning_rate": 9.564246520295719e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 4894144,
      "step": 7480
    },
    {
      "epoch": 4.413325471698113,
      "grad_norm": 1.2213250398635864,
      "learning_rate": 9.563195485607638e-06,
      "loss": 0.4617,
      "num_input_tokens_seen": 4897856,
      "step": 7485
    },
    {
      "epoch": 4.41627358490566,
      "grad_norm": 3.351194143295288,
      "learning_rate": 9.562143242793885e-06,
      "loss": 0.6249,
      "num_input_tokens_seen": 4900320,
      "step": 7490
    },
    {
      "epoch": 4.4192216981132075,
      "grad_norm": 1.273376703262329,
      "learning_rate": 9.561089792133048e-06,
      "loss": 0.4638,
      "num_input_tokens_seen": 4903168,
      "step": 7495
    },
    {
      "epoch": 4.422169811320755,
      "grad_norm": 2.005100727081299,
      "learning_rate": 9.560035133904031e-06,
      "loss": 0.6846,
      "num_input_tokens_seen": 4906368,
      "step": 7500
    },
    {
      "epoch": 4.425117924528302,
      "grad_norm": 0.8993720412254333,
      "learning_rate": 9.55897926838606e-06,
      "loss": 0.5293,
      "num_input_tokens_seen": 4909760,
      "step": 7505
    },
    {
      "epoch": 4.428066037735849,
      "grad_norm": 0.8910791873931885,
      "learning_rate": 9.55792219585868e-06,
      "loss": 0.5721,
      "num_input_tokens_seen": 4912896,
      "step": 7510
    },
    {
      "epoch": 4.431014150943396,
      "grad_norm": 1.4665204286575317,
      "learning_rate": 9.556863916601754e-06,
      "loss": 0.4971,
      "num_input_tokens_seen": 4915168,
      "step": 7515
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 1.1714715957641602,
      "learning_rate": 9.555804430895467e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 4918592,
      "step": 7520
    },
    {
      "epoch": 4.43691037735849,
      "grad_norm": 1.4948314428329468,
      "learning_rate": 9.554743739020325e-06,
      "loss": 0.545,
      "num_input_tokens_seen": 4921600,
      "step": 7525
    },
    {
      "epoch": 4.439858490566038,
      "grad_norm": 1.4543814659118652,
      "learning_rate": 9.553681841257146e-06,
      "loss": 0.4525,
      "num_input_tokens_seen": 4924448,
      "step": 7530
    },
    {
      "epoch": 4.442806603773585,
      "grad_norm": 1.6533616781234741,
      "learning_rate": 9.552618737887073e-06,
      "loss": 0.4904,
      "num_input_tokens_seen": 4927296,
      "step": 7535
    },
    {
      "epoch": 4.445754716981132,
      "grad_norm": 2.668494462966919,
      "learning_rate": 9.55155442919157e-06,
      "loss": 0.5529,
      "num_input_tokens_seen": 4930080,
      "step": 7540
    },
    {
      "epoch": 4.4487028301886795,
      "grad_norm": 1.4162874221801758,
      "learning_rate": 9.550488915452416e-06,
      "loss": 0.5001,
      "num_input_tokens_seen": 4932800,
      "step": 7545
    },
    {
      "epoch": 4.451650943396227,
      "grad_norm": 1.8471534252166748,
      "learning_rate": 9.54942219695171e-06,
      "loss": 0.5797,
      "num_input_tokens_seen": 4935840,
      "step": 7550
    },
    {
      "epoch": 4.454599056603773,
      "grad_norm": 1.9524219036102295,
      "learning_rate": 9.54835427397187e-06,
      "loss": 0.4935,
      "num_input_tokens_seen": 4938464,
      "step": 7555
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 1.3183122873306274,
      "learning_rate": 9.547285146795634e-06,
      "loss": 0.3968,
      "num_input_tokens_seen": 4941856,
      "step": 7560
    },
    {
      "epoch": 4.460495283018868,
      "grad_norm": 1.1411452293395996,
      "learning_rate": 9.546214815706059e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 4945120,
      "step": 7565
    },
    {
      "epoch": 4.463443396226415,
      "grad_norm": 0.869899570941925,
      "learning_rate": 9.545143280986518e-06,
      "loss": 0.5127,
      "num_input_tokens_seen": 4948800,
      "step": 7570
    },
    {
      "epoch": 4.466391509433962,
      "grad_norm": 1.4152759313583374,
      "learning_rate": 9.544070542920703e-06,
      "loss": 0.7132,
      "num_input_tokens_seen": 4951424,
      "step": 7575
    },
    {
      "epoch": 4.46933962264151,
      "grad_norm": 1.3487770557403564,
      "learning_rate": 9.542996601792629e-06,
      "loss": 0.6072,
      "num_input_tokens_seen": 4954784,
      "step": 7580
    },
    {
      "epoch": 4.472287735849057,
      "grad_norm": 1.6704012155532837,
      "learning_rate": 9.541921457886624e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 4958944,
      "step": 7585
    },
    {
      "epoch": 4.475235849056604,
      "grad_norm": 2.5303385257720947,
      "learning_rate": 9.54084511148734e-06,
      "loss": 0.4257,
      "num_input_tokens_seen": 4961888,
      "step": 7590
    },
    {
      "epoch": 4.478183962264151,
      "grad_norm": 1.6728098392486572,
      "learning_rate": 9.539767562879742e-06,
      "loss": 0.4518,
      "num_input_tokens_seen": 4964704,
      "step": 7595
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 1.370964765548706,
      "learning_rate": 9.538688812349118e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 4968192,
      "step": 7600
    },
    {
      "epoch": 4.484080188679245,
      "grad_norm": 1.8719149827957153,
      "learning_rate": 9.537608860181069e-06,
      "loss": 0.5836,
      "num_input_tokens_seen": 4971712,
      "step": 7605
    },
    {
      "epoch": 4.4870283018867925,
      "grad_norm": 1.514885425567627,
      "learning_rate": 9.536527706661519e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 4975520,
      "step": 7610
    },
    {
      "epoch": 4.48997641509434,
      "grad_norm": 3.534165143966675,
      "learning_rate": 9.535445352076707e-06,
      "loss": 0.4631,
      "num_input_tokens_seen": 4978720,
      "step": 7615
    },
    {
      "epoch": 4.492924528301887,
      "grad_norm": 2.0152809619903564,
      "learning_rate": 9.534361796713191e-06,
      "loss": 0.4985,
      "num_input_tokens_seen": 4981600,
      "step": 7620
    },
    {
      "epoch": 4.495872641509434,
      "grad_norm": 1.7527351379394531,
      "learning_rate": 9.533277040857847e-06,
      "loss": 0.5136,
      "num_input_tokens_seen": 4985888,
      "step": 7625
    },
    {
      "epoch": 4.498820754716981,
      "grad_norm": 1.4030812978744507,
      "learning_rate": 9.53219108479787e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 4989056,
      "step": 7630
    },
    {
      "epoch": 4.501768867924528,
      "grad_norm": 4.67748498916626,
      "learning_rate": 9.53110392882077e-06,
      "loss": 0.5194,
      "num_input_tokens_seen": 4991840,
      "step": 7635
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 2.1798205375671387,
      "learning_rate": 9.530015573214378e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 4994560,
      "step": 7640
    },
    {
      "epoch": 4.507665094339623,
      "grad_norm": 1.038201928138733,
      "learning_rate": 9.528926018266837e-06,
      "loss": 0.376,
      "num_input_tokens_seen": 4999104,
      "step": 7645
    },
    {
      "epoch": 4.51061320754717,
      "grad_norm": 0.9823158383369446,
      "learning_rate": 9.527835264266617e-06,
      "loss": 0.4926,
      "num_input_tokens_seen": 5002304,
      "step": 7650
    },
    {
      "epoch": 4.513561320754717,
      "grad_norm": 1.060989499092102,
      "learning_rate": 9.526743311502496e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 5006432,
      "step": 7655
    },
    {
      "epoch": 4.5165094339622645,
      "grad_norm": 2.811574935913086,
      "learning_rate": 9.525650160263573e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 5009632,
      "step": 7660
    },
    {
      "epoch": 4.519457547169811,
      "grad_norm": 3.55657696723938,
      "learning_rate": 9.524555810839267e-06,
      "loss": 0.4701,
      "num_input_tokens_seen": 5013376,
      "step": 7665
    },
    {
      "epoch": 4.522405660377358,
      "grad_norm": 1.4186691045761108,
      "learning_rate": 9.523460263519309e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 5015936,
      "step": 7670
    },
    {
      "epoch": 4.525353773584905,
      "grad_norm": 1.0253254175186157,
      "learning_rate": 9.522363518593753e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 5019872,
      "step": 7675
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 1.3352330923080444,
      "learning_rate": 9.521265576352963e-06,
      "loss": 0.3809,
      "num_input_tokens_seen": 5022976,
      "step": 7680
    },
    {
      "epoch": 4.53125,
      "grad_norm": 1.780050277709961,
      "learning_rate": 9.520166437087628e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 5025504,
      "step": 7685
    },
    {
      "epoch": 4.534198113207547,
      "grad_norm": 1.486417531967163,
      "learning_rate": 9.519066101088748e-06,
      "loss": 0.4822,
      "num_input_tokens_seen": 5028992,
      "step": 7690
    },
    {
      "epoch": 4.537146226415095,
      "grad_norm": 1.9587833881378174,
      "learning_rate": 9.51796456864764e-06,
      "loss": 0.5305,
      "num_input_tokens_seen": 5032448,
      "step": 7695
    },
    {
      "epoch": 4.540094339622642,
      "grad_norm": 1.958592176437378,
      "learning_rate": 9.516861840055942e-06,
      "loss": 0.3578,
      "num_input_tokens_seen": 5035392,
      "step": 7700
    },
    {
      "epoch": 4.543042452830189,
      "grad_norm": 1.202128291130066,
      "learning_rate": 9.515757915605604e-06,
      "loss": 0.4019,
      "num_input_tokens_seen": 5038624,
      "step": 7705
    },
    {
      "epoch": 4.5459905660377355,
      "grad_norm": 0.9313827753067017,
      "learning_rate": 9.514652795588899e-06,
      "loss": 0.3518,
      "num_input_tokens_seen": 5042144,
      "step": 7710
    },
    {
      "epoch": 4.548938679245283,
      "grad_norm": 2.1812996864318848,
      "learning_rate": 9.513546480298405e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 5044960,
      "step": 7715
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 1.4487873315811157,
      "learning_rate": 9.512438970027032e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 5048160,
      "step": 7720
    },
    {
      "epoch": 4.554834905660377,
      "grad_norm": 3.8466107845306396,
      "learning_rate": 9.511330265067992e-06,
      "loss": 0.4777,
      "num_input_tokens_seen": 5050560,
      "step": 7725
    },
    {
      "epoch": 4.557783018867925,
      "grad_norm": 1.5872455835342407,
      "learning_rate": 9.510220365714822e-06,
      "loss": 0.4729,
      "num_input_tokens_seen": 5053152,
      "step": 7730
    },
    {
      "epoch": 4.560731132075472,
      "grad_norm": 1.235642910003662,
      "learning_rate": 9.509109272261373e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 5055776,
      "step": 7735
    },
    {
      "epoch": 4.563679245283019,
      "grad_norm": 2.5548903942108154,
      "learning_rate": 9.50799698500181e-06,
      "loss": 0.5961,
      "num_input_tokens_seen": 5059040,
      "step": 7740
    },
    {
      "epoch": 4.566627358490566,
      "grad_norm": 1.9291280508041382,
      "learning_rate": 9.506883504230618e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 5062496,
      "step": 7745
    },
    {
      "epoch": 4.569575471698113,
      "grad_norm": 4.368766784667969,
      "learning_rate": 9.505768830242593e-06,
      "loss": 0.4099,
      "num_input_tokens_seen": 5066112,
      "step": 7750
    },
    {
      "epoch": 4.57252358490566,
      "grad_norm": 3.015547752380371,
      "learning_rate": 9.504652963332852e-06,
      "loss": 0.5469,
      "num_input_tokens_seen": 5068960,
      "step": 7755
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 1.163577914237976,
      "learning_rate": 9.503535903796825e-06,
      "loss": 0.476,
      "num_input_tokens_seen": 5072672,
      "step": 7760
    },
    {
      "epoch": 4.578419811320755,
      "grad_norm": 1.774612545967102,
      "learning_rate": 9.50241765193026e-06,
      "loss": 0.4627,
      "num_input_tokens_seen": 5075264,
      "step": 7765
    },
    {
      "epoch": 4.581367924528302,
      "grad_norm": 1.298682689666748,
      "learning_rate": 9.501298208029214e-06,
      "loss": 0.3792,
      "num_input_tokens_seen": 5078752,
      "step": 7770
    },
    {
      "epoch": 4.584316037735849,
      "grad_norm": 1.6226075887680054,
      "learning_rate": 9.500177572390071e-06,
      "loss": 0.4834,
      "num_input_tokens_seen": 5082176,
      "step": 7775
    },
    {
      "epoch": 4.587264150943396,
      "grad_norm": 2.454296827316284,
      "learning_rate": 9.49905574530952e-06,
      "loss": 0.5557,
      "num_input_tokens_seen": 5085280,
      "step": 7780
    },
    {
      "epoch": 4.590212264150943,
      "grad_norm": 1.6259760856628418,
      "learning_rate": 9.497932727084571e-06,
      "loss": 0.4412,
      "num_input_tokens_seen": 5087872,
      "step": 7785
    },
    {
      "epoch": 4.59316037735849,
      "grad_norm": 2.048954725265503,
      "learning_rate": 9.496808518012545e-06,
      "loss": 0.5154,
      "num_input_tokens_seen": 5091168,
      "step": 7790
    },
    {
      "epoch": 4.596108490566038,
      "grad_norm": 1.2450222969055176,
      "learning_rate": 9.495683118391087e-06,
      "loss": 0.634,
      "num_input_tokens_seen": 5094144,
      "step": 7795
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 1.6107733249664307,
      "learning_rate": 9.494556528518146e-06,
      "loss": 0.5734,
      "num_input_tokens_seen": 5097280,
      "step": 7800
    },
    {
      "epoch": 4.602004716981132,
      "grad_norm": 2.358375072479248,
      "learning_rate": 9.493428748691995e-06,
      "loss": 0.5568,
      "num_input_tokens_seen": 5099840,
      "step": 7805
    },
    {
      "epoch": 4.6049528301886795,
      "grad_norm": 1.3987880945205688,
      "learning_rate": 9.492299779211215e-06,
      "loss": 0.5105,
      "num_input_tokens_seen": 5103104,
      "step": 7810
    },
    {
      "epoch": 4.607900943396227,
      "grad_norm": 1.827114224433899,
      "learning_rate": 9.49116962037471e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 5105408,
      "step": 7815
    },
    {
      "epoch": 4.610849056603773,
      "grad_norm": 2.394301652908325,
      "learning_rate": 9.490038272481691e-06,
      "loss": 0.4825,
      "num_input_tokens_seen": 5108064,
      "step": 7820
    },
    {
      "epoch": 4.6137971698113205,
      "grad_norm": 0.9063048362731934,
      "learning_rate": 9.488905735831689e-06,
      "loss": 0.3351,
      "num_input_tokens_seen": 5111264,
      "step": 7825
    },
    {
      "epoch": 4.616745283018868,
      "grad_norm": 1.5732303857803345,
      "learning_rate": 9.487772010724548e-06,
      "loss": 0.4402,
      "num_input_tokens_seen": 5113984,
      "step": 7830
    },
    {
      "epoch": 4.619693396226415,
      "grad_norm": 1.0210626125335693,
      "learning_rate": 9.486637097460425e-06,
      "loss": 0.4481,
      "num_input_tokens_seen": 5117088,
      "step": 7835
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 1.8335503339767456,
      "learning_rate": 9.485500996339793e-06,
      "loss": 0.3325,
      "num_input_tokens_seen": 5120256,
      "step": 7840
    },
    {
      "epoch": 4.62558962264151,
      "grad_norm": 1.8555923700332642,
      "learning_rate": 9.484363707663443e-06,
      "loss": 0.4518,
      "num_input_tokens_seen": 5123136,
      "step": 7845
    },
    {
      "epoch": 4.628537735849057,
      "grad_norm": 7.985414981842041,
      "learning_rate": 9.483225231732474e-06,
      "loss": 0.5435,
      "num_input_tokens_seen": 5125952,
      "step": 7850
    },
    {
      "epoch": 4.631485849056604,
      "grad_norm": 1.590451717376709,
      "learning_rate": 9.482085568848302e-06,
      "loss": 0.4554,
      "num_input_tokens_seen": 5128896,
      "step": 7855
    },
    {
      "epoch": 4.634433962264151,
      "grad_norm": 1.7939743995666504,
      "learning_rate": 9.480944719312659e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 5132000,
      "step": 7860
    },
    {
      "epoch": 4.637382075471698,
      "grad_norm": 1.45896577835083,
      "learning_rate": 9.47980268342759e-06,
      "loss": 0.576,
      "num_input_tokens_seen": 5134816,
      "step": 7865
    },
    {
      "epoch": 4.640330188679245,
      "grad_norm": 1.287460207939148,
      "learning_rate": 9.478659461495456e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 5137216,
      "step": 7870
    },
    {
      "epoch": 4.6432783018867925,
      "grad_norm": 1.529208779335022,
      "learning_rate": 9.477515053818926e-06,
      "loss": 0.466,
      "num_input_tokens_seen": 5141248,
      "step": 7875
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 1.6803432703018188,
      "learning_rate": 9.476369460700988e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 5143616,
      "step": 7880
    },
    {
      "epoch": 4.649174528301887,
      "grad_norm": 2.576446294784546,
      "learning_rate": 9.475222682444944e-06,
      "loss": 0.3567,
      "num_input_tokens_seen": 5146976,
      "step": 7885
    },
    {
      "epoch": 4.652122641509434,
      "grad_norm": 1.2117456197738647,
      "learning_rate": 9.474074719354406e-06,
      "loss": 0.4678,
      "num_input_tokens_seen": 5149952,
      "step": 7890
    },
    {
      "epoch": 4.655070754716981,
      "grad_norm": 0.9689227342605591,
      "learning_rate": 9.472925571733306e-06,
      "loss": 0.5356,
      "num_input_tokens_seen": 5152864,
      "step": 7895
    },
    {
      "epoch": 4.658018867924528,
      "grad_norm": 2.755690336227417,
      "learning_rate": 9.471775239885883e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 5156192,
      "step": 7900
    },
    {
      "epoch": 4.660966981132075,
      "grad_norm": 3.22027325630188,
      "learning_rate": 9.470623724116693e-06,
      "loss": 0.4296,
      "num_input_tokens_seen": 5159584,
      "step": 7905
    },
    {
      "epoch": 4.663915094339623,
      "grad_norm": 1.3719979524612427,
      "learning_rate": 9.469471024730606e-06,
      "loss": 0.328,
      "num_input_tokens_seen": 5162144,
      "step": 7910
    },
    {
      "epoch": 4.66686320754717,
      "grad_norm": 0.8866376280784607,
      "learning_rate": 9.4683171420328e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 5165824,
      "step": 7915
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 1.3963021039962769,
      "learning_rate": 9.467162076328776e-06,
      "loss": 0.4792,
      "num_input_tokens_seen": 5169248,
      "step": 7920
    },
    {
      "epoch": 4.6727594339622645,
      "grad_norm": 1.6422536373138428,
      "learning_rate": 9.466005827924337e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 5172384,
      "step": 7925
    },
    {
      "epoch": 4.675707547169811,
      "grad_norm": 2.300344705581665,
      "learning_rate": 9.46484839712561e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 5175424,
      "step": 7930
    },
    {
      "epoch": 4.678655660377358,
      "grad_norm": 2.0406808853149414,
      "learning_rate": 9.463689784239026e-06,
      "loss": 0.4847,
      "num_input_tokens_seen": 5182368,
      "step": 7935
    },
    {
      "epoch": 4.681603773584905,
      "grad_norm": 1.6918795108795166,
      "learning_rate": 9.462529989571334e-06,
      "loss": 0.4806,
      "num_input_tokens_seen": 5186112,
      "step": 7940
    },
    {
      "epoch": 4.684551886792453,
      "grad_norm": 1.409661054611206,
      "learning_rate": 9.461369013429595e-06,
      "loss": 0.4372,
      "num_input_tokens_seen": 5189664,
      "step": 7945
    },
    {
      "epoch": 4.6875,
      "grad_norm": 0.9480183720588684,
      "learning_rate": 9.460206856121183e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 5193440,
      "step": 7950
    },
    {
      "epoch": 4.690448113207547,
      "grad_norm": 1.435253381729126,
      "learning_rate": 9.459043517953786e-06,
      "loss": 0.3982,
      "num_input_tokens_seen": 5196768,
      "step": 7955
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 2.3356757164001465,
      "learning_rate": 9.457878999235396e-06,
      "loss": 0.5648,
      "num_input_tokens_seen": 5201056,
      "step": 7960
    },
    {
      "epoch": 4.696344339622642,
      "grad_norm": 1.4380927085876465,
      "learning_rate": 9.45671330027433e-06,
      "loss": 0.3439,
      "num_input_tokens_seen": 5204192,
      "step": 7965
    },
    {
      "epoch": 4.699292452830189,
      "grad_norm": 1.8361990451812744,
      "learning_rate": 9.45554642137921e-06,
      "loss": 0.4079,
      "num_input_tokens_seen": 5207840,
      "step": 7970
    },
    {
      "epoch": 4.7022405660377355,
      "grad_norm": 1.429263710975647,
      "learning_rate": 9.454378362858974e-06,
      "loss": 0.4532,
      "num_input_tokens_seen": 5211136,
      "step": 7975
    },
    {
      "epoch": 4.705188679245283,
      "grad_norm": 2.4700374603271484,
      "learning_rate": 9.453209125022867e-06,
      "loss": 0.5271,
      "num_input_tokens_seen": 5213408,
      "step": 7980
    },
    {
      "epoch": 4.70813679245283,
      "grad_norm": 1.676206111907959,
      "learning_rate": 9.452038708180453e-06,
      "loss": 0.5278,
      "num_input_tokens_seen": 5217664,
      "step": 7985
    },
    {
      "epoch": 4.711084905660377,
      "grad_norm": 1.2887953519821167,
      "learning_rate": 9.450867112641603e-06,
      "loss": 0.3929,
      "num_input_tokens_seen": 5221600,
      "step": 7990
    },
    {
      "epoch": 4.714033018867925,
      "grad_norm": 2.0029635429382324,
      "learning_rate": 9.449694338716506e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 5225472,
      "step": 7995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 2.219892740249634,
      "learning_rate": 9.448520386715653e-06,
      "loss": 0.4014,
      "num_input_tokens_seen": 5228256,
      "step": 8000
    },
    {
      "epoch": 4.719929245283019,
      "grad_norm": 1.4878082275390625,
      "learning_rate": 9.447345256949855e-06,
      "loss": 0.5785,
      "num_input_tokens_seen": 5231904,
      "step": 8005
    },
    {
      "epoch": 4.722877358490566,
      "grad_norm": 2.475067377090454,
      "learning_rate": 9.446168949730234e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 5235232,
      "step": 8010
    },
    {
      "epoch": 4.725825471698113,
      "grad_norm": 1.7630001306533813,
      "learning_rate": 9.444991465368223e-06,
      "loss": 0.553,
      "num_input_tokens_seen": 5237856,
      "step": 8015
    },
    {
      "epoch": 4.72877358490566,
      "grad_norm": 2.227893114089966,
      "learning_rate": 9.443812804175562e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 5240736,
      "step": 8020
    },
    {
      "epoch": 4.7317216981132075,
      "grad_norm": 1.518802285194397,
      "learning_rate": 9.44263296646431e-06,
      "loss": 0.5533,
      "num_input_tokens_seen": 5243712,
      "step": 8025
    },
    {
      "epoch": 4.734669811320755,
      "grad_norm": 1.8206170797348022,
      "learning_rate": 9.441451952546835e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 5247520,
      "step": 8030
    },
    {
      "epoch": 4.737617924528302,
      "grad_norm": 2.6229798793792725,
      "learning_rate": 9.440269762735814e-06,
      "loss": 0.3283,
      "num_input_tokens_seen": 5252224,
      "step": 8035
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 1.9982359409332275,
      "learning_rate": 9.439086397344236e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 5255168,
      "step": 8040
    },
    {
      "epoch": 4.743514150943396,
      "grad_norm": 2.1413938999176025,
      "learning_rate": 9.437901856685404e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 5258080,
      "step": 8045
    },
    {
      "epoch": 4.746462264150943,
      "grad_norm": 1.2891086339950562,
      "learning_rate": 9.436716141072925e-06,
      "loss": 0.5021,
      "num_input_tokens_seen": 5261216,
      "step": 8050
    },
    {
      "epoch": 4.74941037735849,
      "grad_norm": 1.1813912391662598,
      "learning_rate": 9.435529250820732e-06,
      "loss": 0.538,
      "num_input_tokens_seen": 5264704,
      "step": 8055
    },
    {
      "epoch": 4.752358490566038,
      "grad_norm": 1.3717230558395386,
      "learning_rate": 9.43434118624305e-06,
      "loss": 0.5762,
      "num_input_tokens_seen": 5267808,
      "step": 8060
    },
    {
      "epoch": 4.755306603773585,
      "grad_norm": 1.4407525062561035,
      "learning_rate": 9.433151947654428e-06,
      "loss": 0.4013,
      "num_input_tokens_seen": 5271488,
      "step": 8065
    },
    {
      "epoch": 4.758254716981132,
      "grad_norm": 0.8528293967247009,
      "learning_rate": 9.431961535369724e-06,
      "loss": 0.4244,
      "num_input_tokens_seen": 5274784,
      "step": 8070
    },
    {
      "epoch": 4.7612028301886795,
      "grad_norm": 2.7706401348114014,
      "learning_rate": 9.430769949704103e-06,
      "loss": 0.5836,
      "num_input_tokens_seen": 5277504,
      "step": 8075
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 1.5826739072799683,
      "learning_rate": 9.42957719097304e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 5280896,
      "step": 8080
    },
    {
      "epoch": 4.767099056603773,
      "grad_norm": 1.643378734588623,
      "learning_rate": 9.42838325949233e-06,
      "loss": 0.3492,
      "num_input_tokens_seen": 5284256,
      "step": 8085
    },
    {
      "epoch": 4.7700471698113205,
      "grad_norm": 1.7598609924316406,
      "learning_rate": 9.427188155578062e-06,
      "loss": 0.551,
      "num_input_tokens_seen": 5286528,
      "step": 8090
    },
    {
      "epoch": 4.772995283018868,
      "grad_norm": 1.2895909547805786,
      "learning_rate": 9.42599187954665e-06,
      "loss": 0.5062,
      "num_input_tokens_seen": 5289376,
      "step": 8095
    },
    {
      "epoch": 4.775943396226415,
      "grad_norm": 0.9510688781738281,
      "learning_rate": 9.424794431714814e-06,
      "loss": 0.3663,
      "num_input_tokens_seen": 5292768,
      "step": 8100
    },
    {
      "epoch": 4.778891509433962,
      "grad_norm": 1.0903135538101196,
      "learning_rate": 9.423595812399581e-06,
      "loss": 0.3279,
      "num_input_tokens_seen": 5299808,
      "step": 8105
    },
    {
      "epoch": 4.78183962264151,
      "grad_norm": 2.1231489181518555,
      "learning_rate": 9.422396021918296e-06,
      "loss": 0.494,
      "num_input_tokens_seen": 5302624,
      "step": 8110
    },
    {
      "epoch": 4.784787735849057,
      "grad_norm": 1.4040777683258057,
      "learning_rate": 9.421195060588602e-06,
      "loss": 0.4723,
      "num_input_tokens_seen": 5305888,
      "step": 8115
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 1.5806814432144165,
      "learning_rate": 9.419992928728461e-06,
      "loss": 0.5457,
      "num_input_tokens_seen": 5310208,
      "step": 8120
    },
    {
      "epoch": 4.790683962264151,
      "grad_norm": 1.0235157012939453,
      "learning_rate": 9.418789626656144e-06,
      "loss": 0.3334,
      "num_input_tokens_seen": 5313760,
      "step": 8125
    },
    {
      "epoch": 4.793632075471698,
      "grad_norm": 1.8617626428604126,
      "learning_rate": 9.417585154690229e-06,
      "loss": 0.3699,
      "num_input_tokens_seen": 5316736,
      "step": 8130
    },
    {
      "epoch": 4.796580188679245,
      "grad_norm": 1.3455277681350708,
      "learning_rate": 9.416379513149605e-06,
      "loss": 0.4507,
      "num_input_tokens_seen": 5320512,
      "step": 8135
    },
    {
      "epoch": 4.7995283018867925,
      "grad_norm": 2.5582973957061768,
      "learning_rate": 9.415172702353471e-06,
      "loss": 0.4512,
      "num_input_tokens_seen": 5323168,
      "step": 8140
    },
    {
      "epoch": 4.80247641509434,
      "grad_norm": 2.1141610145568848,
      "learning_rate": 9.413964722621339e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 5326400,
      "step": 8145
    },
    {
      "epoch": 4.805424528301887,
      "grad_norm": 1.5132012367248535,
      "learning_rate": 9.41275557427302e-06,
      "loss": 0.5359,
      "num_input_tokens_seen": 5333216,
      "step": 8150
    },
    {
      "epoch": 4.808372641509434,
      "grad_norm": 1.1381033658981323,
      "learning_rate": 9.411545257628646e-06,
      "loss": 0.4912,
      "num_input_tokens_seen": 5336992,
      "step": 8155
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 2.0048437118530273,
      "learning_rate": 9.410333773008653e-06,
      "loss": 0.4564,
      "num_input_tokens_seen": 5339904,
      "step": 8160
    },
    {
      "epoch": 4.814268867924528,
      "grad_norm": 1.3509811162948608,
      "learning_rate": 9.409121120733784e-06,
      "loss": 0.5169,
      "num_input_tokens_seen": 5344192,
      "step": 8165
    },
    {
      "epoch": 4.817216981132075,
      "grad_norm": 1.932436227798462,
      "learning_rate": 9.4079073011251e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 5347232,
      "step": 8170
    },
    {
      "epoch": 4.820165094339623,
      "grad_norm": 2.1171276569366455,
      "learning_rate": 9.406692314503956e-06,
      "loss": 0.4468,
      "num_input_tokens_seen": 5349728,
      "step": 8175
    },
    {
      "epoch": 4.82311320754717,
      "grad_norm": 2.8441503047943115,
      "learning_rate": 9.405476161192033e-06,
      "loss": 0.5212,
      "num_input_tokens_seen": 5352480,
      "step": 8180
    },
    {
      "epoch": 4.826061320754717,
      "grad_norm": 2.5249924659729004,
      "learning_rate": 9.40425884151131e-06,
      "loss": 0.3964,
      "num_input_tokens_seen": 5354880,
      "step": 8185
    },
    {
      "epoch": 4.8290094339622645,
      "grad_norm": 1.3938353061676025,
      "learning_rate": 9.403040355784076e-06,
      "loss": 0.4651,
      "num_input_tokens_seen": 5357568,
      "step": 8190
    },
    {
      "epoch": 4.831957547169811,
      "grad_norm": 8.202880859375,
      "learning_rate": 9.401820704332932e-06,
      "loss": 0.5333,
      "num_input_tokens_seen": 5360000,
      "step": 8195
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 1.3870265483856201,
      "learning_rate": 9.400599887480786e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 5363808,
      "step": 8200
    },
    {
      "epoch": 4.837853773584905,
      "grad_norm": 2.4023022651672363,
      "learning_rate": 9.399377905550854e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 5366560,
      "step": 8205
    },
    {
      "epoch": 4.840801886792453,
      "grad_norm": 1.3850816488265991,
      "learning_rate": 9.398154758866662e-06,
      "loss": 0.573,
      "num_input_tokens_seen": 5369856,
      "step": 8210
    },
    {
      "epoch": 4.84375,
      "grad_norm": 1.409345269203186,
      "learning_rate": 9.396930447752041e-06,
      "loss": 0.4617,
      "num_input_tokens_seen": 5372736,
      "step": 8215
    },
    {
      "epoch": 4.846698113207547,
      "grad_norm": 1.5833823680877686,
      "learning_rate": 9.395704972531137e-06,
      "loss": 0.5104,
      "num_input_tokens_seen": 5375872,
      "step": 8220
    },
    {
      "epoch": 4.849646226415095,
      "grad_norm": 3.234210729598999,
      "learning_rate": 9.394478333528396e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 5378208,
      "step": 8225
    },
    {
      "epoch": 4.852594339622642,
      "grad_norm": 1.0982325077056885,
      "learning_rate": 9.393250531068576e-06,
      "loss": 0.5065,
      "num_input_tokens_seen": 5381600,
      "step": 8230
    },
    {
      "epoch": 4.855542452830189,
      "grad_norm": 1.4776220321655273,
      "learning_rate": 9.392021565476744e-06,
      "loss": 0.5241,
      "num_input_tokens_seen": 5384512,
      "step": 8235
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 1.1546037197113037,
      "learning_rate": 9.390791437078274e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 5387456,
      "step": 8240
    },
    {
      "epoch": 4.861438679245283,
      "grad_norm": 1.8213926553726196,
      "learning_rate": 9.38956014619885e-06,
      "loss": 0.6648,
      "num_input_tokens_seen": 5391360,
      "step": 8245
    },
    {
      "epoch": 4.86438679245283,
      "grad_norm": 2.8283865451812744,
      "learning_rate": 9.388327693164456e-06,
      "loss": 0.525,
      "num_input_tokens_seen": 5394048,
      "step": 8250
    },
    {
      "epoch": 4.867334905660377,
      "grad_norm": 1.211863398551941,
      "learning_rate": 9.387094078301395e-06,
      "loss": 0.5494,
      "num_input_tokens_seen": 5397440,
      "step": 8255
    },
    {
      "epoch": 4.870283018867925,
      "grad_norm": 5.16395378112793,
      "learning_rate": 9.385859301936269e-06,
      "loss": 0.4895,
      "num_input_tokens_seen": 5401056,
      "step": 8260
    },
    {
      "epoch": 4.873231132075472,
      "grad_norm": 2.0897209644317627,
      "learning_rate": 9.38462336439599e-06,
      "loss": 0.6269,
      "num_input_tokens_seen": 5403872,
      "step": 8265
    },
    {
      "epoch": 4.876179245283019,
      "grad_norm": 2.5733871459960938,
      "learning_rate": 9.383386266007779e-06,
      "loss": 0.525,
      "num_input_tokens_seen": 5406592,
      "step": 8270
    },
    {
      "epoch": 4.879127358490566,
      "grad_norm": 1.6193221807479858,
      "learning_rate": 9.382148007099164e-06,
      "loss": 0.4537,
      "num_input_tokens_seen": 5409536,
      "step": 8275
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 1.8796823024749756,
      "learning_rate": 9.380908587997977e-06,
      "loss": 0.4511,
      "num_input_tokens_seen": 5412192,
      "step": 8280
    },
    {
      "epoch": 4.88502358490566,
      "grad_norm": 2.599104166030884,
      "learning_rate": 9.37966800903236e-06,
      "loss": 0.3612,
      "num_input_tokens_seen": 5415104,
      "step": 8285
    },
    {
      "epoch": 4.8879716981132075,
      "grad_norm": 1.1825108528137207,
      "learning_rate": 9.378426270530762e-06,
      "loss": 0.5087,
      "num_input_tokens_seen": 5417440,
      "step": 8290
    },
    {
      "epoch": 4.890919811320755,
      "grad_norm": 1.8377580642700195,
      "learning_rate": 9.37718337282194e-06,
      "loss": 0.5512,
      "num_input_tokens_seen": 5419680,
      "step": 8295
    },
    {
      "epoch": 4.893867924528302,
      "grad_norm": 1.470245361328125,
      "learning_rate": 9.375939316234956e-06,
      "loss": 0.4443,
      "num_input_tokens_seen": 5422720,
      "step": 8300
    },
    {
      "epoch": 4.896816037735849,
      "grad_norm": 1.0184236764907837,
      "learning_rate": 9.374694101099178e-06,
      "loss": 0.4829,
      "num_input_tokens_seen": 5426496,
      "step": 8305
    },
    {
      "epoch": 4.899764150943396,
      "grad_norm": 1.8521127700805664,
      "learning_rate": 9.373447727744282e-06,
      "loss": 0.5001,
      "num_input_tokens_seen": 5430752,
      "step": 8310
    },
    {
      "epoch": 4.902712264150943,
      "grad_norm": 1.7610279321670532,
      "learning_rate": 9.372200196500253e-06,
      "loss": 0.6339,
      "num_input_tokens_seen": 5434176,
      "step": 8315
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 2.1384332180023193,
      "learning_rate": 9.37095150769738e-06,
      "loss": 0.4356,
      "num_input_tokens_seen": 5437120,
      "step": 8320
    },
    {
      "epoch": 4.908608490566038,
      "grad_norm": 1.8401201963424683,
      "learning_rate": 9.369701661666255e-06,
      "loss": 0.5899,
      "num_input_tokens_seen": 5440512,
      "step": 8325
    },
    {
      "epoch": 4.911556603773585,
      "grad_norm": 1.1352607011795044,
      "learning_rate": 9.368450658737782e-06,
      "loss": 0.5315,
      "num_input_tokens_seen": 5444928,
      "step": 8330
    },
    {
      "epoch": 4.914504716981132,
      "grad_norm": 1.3849295377731323,
      "learning_rate": 9.367198499243173e-06,
      "loss": 0.5428,
      "num_input_tokens_seen": 5450080,
      "step": 8335
    },
    {
      "epoch": 4.9174528301886795,
      "grad_norm": 1.1126539707183838,
      "learning_rate": 9.365945183513938e-06,
      "loss": 0.5089,
      "num_input_tokens_seen": 5452512,
      "step": 8340
    },
    {
      "epoch": 4.920400943396227,
      "grad_norm": 1.7832050323486328,
      "learning_rate": 9.3646907118819e-06,
      "loss": 0.5167,
      "num_input_tokens_seen": 5455616,
      "step": 8345
    },
    {
      "epoch": 4.923349056603773,
      "grad_norm": 1.153149127960205,
      "learning_rate": 9.363435084679185e-06,
      "loss": 0.5304,
      "num_input_tokens_seen": 5458560,
      "step": 8350
    },
    {
      "epoch": 4.9262971698113205,
      "grad_norm": 2.018462657928467,
      "learning_rate": 9.362178302238227e-06,
      "loss": 0.4524,
      "num_input_tokens_seen": 5461504,
      "step": 8355
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 1.0325111150741577,
      "learning_rate": 9.360920364891762e-06,
      "loss": 0.3779,
      "num_input_tokens_seen": 5465664,
      "step": 8360
    },
    {
      "epoch": 4.932193396226415,
      "grad_norm": 1.6108322143554688,
      "learning_rate": 9.359661272972836e-06,
      "loss": 0.3596,
      "num_input_tokens_seen": 5468832,
      "step": 8365
    },
    {
      "epoch": 4.935141509433962,
      "grad_norm": 1.1578460931777954,
      "learning_rate": 9.3584010268148e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 5472736,
      "step": 8370
    },
    {
      "epoch": 4.93808962264151,
      "grad_norm": 1.2657461166381836,
      "learning_rate": 9.357139626751308e-06,
      "loss": 0.4697,
      "num_input_tokens_seen": 5475360,
      "step": 8375
    },
    {
      "epoch": 4.941037735849057,
      "grad_norm": 2.4177167415618896,
      "learning_rate": 9.355877073116321e-06,
      "loss": 0.4566,
      "num_input_tokens_seen": 5478656,
      "step": 8380
    },
    {
      "epoch": 4.943985849056604,
      "grad_norm": 2.4435830116271973,
      "learning_rate": 9.354613366244108e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 5480992,
      "step": 8385
    },
    {
      "epoch": 4.946933962264151,
      "grad_norm": 1.1529597043991089,
      "learning_rate": 9.353348506469236e-06,
      "loss": 0.7178,
      "num_input_tokens_seen": 5484384,
      "step": 8390
    },
    {
      "epoch": 4.949882075471698,
      "grad_norm": 2.0200259685516357,
      "learning_rate": 9.352082494126586e-06,
      "loss": 0.6049,
      "num_input_tokens_seen": 5487328,
      "step": 8395
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 1.5889414548873901,
      "learning_rate": 9.350815329551341e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 5490464,
      "step": 8400
    },
    {
      "epoch": 4.9557783018867925,
      "grad_norm": 1.808859944343567,
      "learning_rate": 9.349547013078986e-06,
      "loss": 0.5245,
      "num_input_tokens_seen": 5493792,
      "step": 8405
    },
    {
      "epoch": 4.95872641509434,
      "grad_norm": 1.674601674079895,
      "learning_rate": 9.348277545045312e-06,
      "loss": 0.5096,
      "num_input_tokens_seen": 5496928,
      "step": 8410
    },
    {
      "epoch": 4.961674528301887,
      "grad_norm": 3.477091073989868,
      "learning_rate": 9.347006925786418e-06,
      "loss": 0.5909,
      "num_input_tokens_seen": 5499616,
      "step": 8415
    },
    {
      "epoch": 4.964622641509434,
      "grad_norm": 3.228248357772827,
      "learning_rate": 9.34573515563871e-06,
      "loss": 0.4355,
      "num_input_tokens_seen": 5502944,
      "step": 8420
    },
    {
      "epoch": 4.967570754716981,
      "grad_norm": 2.6045312881469727,
      "learning_rate": 9.344462234938885e-06,
      "loss": 0.4535,
      "num_input_tokens_seen": 5505312,
      "step": 8425
    },
    {
      "epoch": 4.970518867924528,
      "grad_norm": 1.9931674003601074,
      "learning_rate": 9.343188164023962e-06,
      "loss": 0.4991,
      "num_input_tokens_seen": 5507968,
      "step": 8430
    },
    {
      "epoch": 4.973466981132075,
      "grad_norm": 2.0450499057769775,
      "learning_rate": 9.341912943231256e-06,
      "loss": 0.5883,
      "num_input_tokens_seen": 5510848,
      "step": 8435
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 1.1032357215881348,
      "learning_rate": 9.340636572898383e-06,
      "loss": 0.4083,
      "num_input_tokens_seen": 5514944,
      "step": 8440
    },
    {
      "epoch": 4.97936320754717,
      "grad_norm": 1.2248214483261108,
      "learning_rate": 9.339359053363272e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 5518080,
      "step": 8445
    },
    {
      "epoch": 4.982311320754717,
      "grad_norm": 1.1182721853256226,
      "learning_rate": 9.338080384964148e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 5520384,
      "step": 8450
    },
    {
      "epoch": 4.9852594339622645,
      "grad_norm": 1.1931827068328857,
      "learning_rate": 9.336800568039548e-06,
      "loss": 0.4816,
      "num_input_tokens_seen": 5523680,
      "step": 8455
    },
    {
      "epoch": 4.988207547169811,
      "grad_norm": 1.6295688152313232,
      "learning_rate": 9.335519602928307e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 5525888,
      "step": 8460
    },
    {
      "epoch": 4.991155660377358,
      "grad_norm": 2.0451438426971436,
      "learning_rate": 9.334237489969565e-06,
      "loss": 0.5685,
      "num_input_tokens_seen": 5529344,
      "step": 8465
    },
    {
      "epoch": 4.994103773584905,
      "grad_norm": 0.9389410614967346,
      "learning_rate": 9.332954229502768e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 5532416,
      "step": 8470
    },
    {
      "epoch": 4.997051886792453,
      "grad_norm": 1.2936044931411743,
      "learning_rate": 9.331669821867665e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 5536416,
      "step": 8475
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.652873992919922,
      "learning_rate": 9.33038426740431e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 5538872,
      "step": 8480
    },
    {
      "epoch": 5.002948113207547,
      "grad_norm": 2.7420942783355713,
      "learning_rate": 9.329097566453055e-06,
      "loss": 0.4943,
      "num_input_tokens_seen": 5541240,
      "step": 8485
    },
    {
      "epoch": 5.005896226415095,
      "grad_norm": 1.4358949661254883,
      "learning_rate": 9.327809719354564e-06,
      "loss": 0.3597,
      "num_input_tokens_seen": 5544120,
      "step": 8490
    },
    {
      "epoch": 5.008844339622642,
      "grad_norm": 1.2289319038391113,
      "learning_rate": 9.326520726449795e-06,
      "loss": 0.3915,
      "num_input_tokens_seen": 5547384,
      "step": 8495
    },
    {
      "epoch": 5.011792452830188,
      "grad_norm": 1.4814389944076538,
      "learning_rate": 9.32523058808002e-06,
      "loss": 0.317,
      "num_input_tokens_seen": 5553304,
      "step": 8500
    },
    {
      "epoch": 5.0147405660377355,
      "grad_norm": 1.505171537399292,
      "learning_rate": 9.323939304586806e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 5557176,
      "step": 8505
    },
    {
      "epoch": 5.017688679245283,
      "grad_norm": 3.0605287551879883,
      "learning_rate": 9.322646876312025e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 5560120,
      "step": 8510
    },
    {
      "epoch": 5.02063679245283,
      "grad_norm": 2.013603687286377,
      "learning_rate": 9.321353303597854e-06,
      "loss": 0.4791,
      "num_input_tokens_seen": 5563576,
      "step": 8515
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 1.6632239818572998,
      "learning_rate": 9.320058586786771e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 5566488,
      "step": 8520
    },
    {
      "epoch": 5.026533018867925,
      "grad_norm": 1.6491384506225586,
      "learning_rate": 9.31876272622156e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 5569176,
      "step": 8525
    },
    {
      "epoch": 5.029481132075472,
      "grad_norm": 2.8502814769744873,
      "learning_rate": 9.317465722245305e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 5571704,
      "step": 8530
    },
    {
      "epoch": 5.032429245283019,
      "grad_norm": 1.1415776014328003,
      "learning_rate": 9.316167575201391e-06,
      "loss": 0.4824,
      "num_input_tokens_seen": 5575384,
      "step": 8535
    },
    {
      "epoch": 5.035377358490566,
      "grad_norm": 1.2614654302597046,
      "learning_rate": 9.31486828543351e-06,
      "loss": 0.497,
      "num_input_tokens_seen": 5579704,
      "step": 8540
    },
    {
      "epoch": 5.038325471698113,
      "grad_norm": 3.065140724182129,
      "learning_rate": 9.313567853285656e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 5583000,
      "step": 8545
    },
    {
      "epoch": 5.04127358490566,
      "grad_norm": 4.103222846984863,
      "learning_rate": 9.31226627910212e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 5586584,
      "step": 8550
    },
    {
      "epoch": 5.0442216981132075,
      "grad_norm": 1.4378769397735596,
      "learning_rate": 9.310963563227504e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 5589592,
      "step": 8555
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 1.3345752954483032,
      "learning_rate": 9.309659706006704e-06,
      "loss": 0.5483,
      "num_input_tokens_seen": 5592440,
      "step": 8560
    },
    {
      "epoch": 5.050117924528302,
      "grad_norm": 1.4868515729904175,
      "learning_rate": 9.308354707784925e-06,
      "loss": 0.505,
      "num_input_tokens_seen": 5595256,
      "step": 8565
    },
    {
      "epoch": 5.053066037735849,
      "grad_norm": 1.6190009117126465,
      "learning_rate": 9.307048568907669e-06,
      "loss": 0.4534,
      "num_input_tokens_seen": 5598648,
      "step": 8570
    },
    {
      "epoch": 5.056014150943396,
      "grad_norm": 0.2477325052022934,
      "learning_rate": 9.30574128972074e-06,
      "loss": 0.2385,
      "num_input_tokens_seen": 5604472,
      "step": 8575
    },
    {
      "epoch": 5.058962264150943,
      "grad_norm": 1.7263487577438354,
      "learning_rate": 9.304432870570247e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 5607192,
      "step": 8580
    },
    {
      "epoch": 5.06191037735849,
      "grad_norm": 3.8622288703918457,
      "learning_rate": 9.303123311802605e-06,
      "loss": 0.5469,
      "num_input_tokens_seen": 5609944,
      "step": 8585
    },
    {
      "epoch": 5.064858490566038,
      "grad_norm": 1.4418171644210815,
      "learning_rate": 9.301812613764516e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 5612568,
      "step": 8590
    },
    {
      "epoch": 5.067806603773585,
      "grad_norm": 1.7910279035568237,
      "learning_rate": 9.300500776803001e-06,
      "loss": 0.4942,
      "num_input_tokens_seen": 5615384,
      "step": 8595
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 2.4552786350250244,
      "learning_rate": 9.29918780126537e-06,
      "loss": 0.4092,
      "num_input_tokens_seen": 5618872,
      "step": 8600
    },
    {
      "epoch": 5.0737028301886795,
      "grad_norm": 1.772151231765747,
      "learning_rate": 9.297873687499239e-06,
      "loss": 0.2762,
      "num_input_tokens_seen": 5622264,
      "step": 8605
    },
    {
      "epoch": 5.076650943396227,
      "grad_norm": 1.9424874782562256,
      "learning_rate": 9.296558435852528e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 5624408,
      "step": 8610
    },
    {
      "epoch": 5.079599056603773,
      "grad_norm": 1.103755235671997,
      "learning_rate": 9.295242046673454e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 5627160,
      "step": 8615
    },
    {
      "epoch": 5.0825471698113205,
      "grad_norm": 2.3753349781036377,
      "learning_rate": 9.293924520310535e-06,
      "loss": 0.3993,
      "num_input_tokens_seen": 5630296,
      "step": 8620
    },
    {
      "epoch": 5.085495283018868,
      "grad_norm": 2.034350633621216,
      "learning_rate": 9.292605857112595e-06,
      "loss": 0.4376,
      "num_input_tokens_seen": 5633592,
      "step": 8625
    },
    {
      "epoch": 5.088443396226415,
      "grad_norm": 0.9719741344451904,
      "learning_rate": 9.291286057428755e-06,
      "loss": 0.3015,
      "num_input_tokens_seen": 5636280,
      "step": 8630
    },
    {
      "epoch": 5.091391509433962,
      "grad_norm": 1.348941445350647,
      "learning_rate": 9.289965121608436e-06,
      "loss": 0.4866,
      "num_input_tokens_seen": 5639416,
      "step": 8635
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 1.5383740663528442,
      "learning_rate": 9.288643050001362e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 5643672,
      "step": 8640
    },
    {
      "epoch": 5.097287735849057,
      "grad_norm": 2.6494710445404053,
      "learning_rate": 9.287319842957557e-06,
      "loss": 0.5276,
      "num_input_tokens_seen": 5646680,
      "step": 8645
    },
    {
      "epoch": 5.100235849056604,
      "grad_norm": 2.1352341175079346,
      "learning_rate": 9.285995500827348e-06,
      "loss": 0.3884,
      "num_input_tokens_seen": 5649368,
      "step": 8650
    },
    {
      "epoch": 5.103183962264151,
      "grad_norm": 1.113258719444275,
      "learning_rate": 9.284670023961355e-06,
      "loss": 0.4642,
      "num_input_tokens_seen": 5652728,
      "step": 8655
    },
    {
      "epoch": 5.106132075471698,
      "grad_norm": 3.3507814407348633,
      "learning_rate": 9.28334341271051e-06,
      "loss": 0.4174,
      "num_input_tokens_seen": 5655928,
      "step": 8660
    },
    {
      "epoch": 5.109080188679245,
      "grad_norm": 1.5923197269439697,
      "learning_rate": 9.282015667426036e-06,
      "loss": 0.5015,
      "num_input_tokens_seen": 5659352,
      "step": 8665
    },
    {
      "epoch": 5.1120283018867925,
      "grad_norm": 1.702967882156372,
      "learning_rate": 9.280686788459461e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 5662776,
      "step": 8670
    },
    {
      "epoch": 5.11497641509434,
      "grad_norm": 1.205973744392395,
      "learning_rate": 9.279356776162606e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 5665912,
      "step": 8675
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 1.4448063373565674,
      "learning_rate": 9.278025630887607e-06,
      "loss": 0.5039,
      "num_input_tokens_seen": 5669208,
      "step": 8680
    },
    {
      "epoch": 5.120872641509434,
      "grad_norm": 1.4047986268997192,
      "learning_rate": 9.27669335298688e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 5672632,
      "step": 8685
    },
    {
      "epoch": 5.123820754716981,
      "grad_norm": 0.9138341546058655,
      "learning_rate": 9.275359942813158e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 5675768,
      "step": 8690
    },
    {
      "epoch": 5.126768867924528,
      "grad_norm": 2.8905889987945557,
      "learning_rate": 9.274025400719466e-06,
      "loss": 0.5892,
      "num_input_tokens_seen": 5678648,
      "step": 8695
    },
    {
      "epoch": 5.129716981132075,
      "grad_norm": 1.1102213859558105,
      "learning_rate": 9.27268972705913e-06,
      "loss": 0.403,
      "num_input_tokens_seen": 5682584,
      "step": 8700
    },
    {
      "epoch": 5.132665094339623,
      "grad_norm": 2.4295380115509033,
      "learning_rate": 9.271352922185772e-06,
      "loss": 0.399,
      "num_input_tokens_seen": 5685048,
      "step": 8705
    },
    {
      "epoch": 5.13561320754717,
      "grad_norm": 1.2704472541809082,
      "learning_rate": 9.270014986453321e-06,
      "loss": 0.3896,
      "num_input_tokens_seen": 5688536,
      "step": 8710
    },
    {
      "epoch": 5.138561320754717,
      "grad_norm": 2.0620484352111816,
      "learning_rate": 9.268675920215999e-06,
      "loss": 0.4512,
      "num_input_tokens_seen": 5691352,
      "step": 8715
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 2.3789825439453125,
      "learning_rate": 9.26733572382833e-06,
      "loss": 0.501,
      "num_input_tokens_seen": 5694104,
      "step": 8720
    },
    {
      "epoch": 5.144457547169812,
      "grad_norm": 1.7674195766448975,
      "learning_rate": 9.265994397645137e-06,
      "loss": 0.4212,
      "num_input_tokens_seen": 5697624,
      "step": 8725
    },
    {
      "epoch": 5.147405660377358,
      "grad_norm": 1.447295069694519,
      "learning_rate": 9.264651942021543e-06,
      "loss": 0.5757,
      "num_input_tokens_seen": 5701048,
      "step": 8730
    },
    {
      "epoch": 5.150353773584905,
      "grad_norm": 3.5669946670532227,
      "learning_rate": 9.263308357312966e-06,
      "loss": 0.4364,
      "num_input_tokens_seen": 5705656,
      "step": 8735
    },
    {
      "epoch": 5.153301886792453,
      "grad_norm": 3.009891986846924,
      "learning_rate": 9.26196364387513e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 5708216,
      "step": 8740
    },
    {
      "epoch": 5.15625,
      "grad_norm": 2.5396504402160645,
      "learning_rate": 9.26061780206405e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 5711480,
      "step": 8745
    },
    {
      "epoch": 5.159198113207547,
      "grad_norm": 2.1485722064971924,
      "learning_rate": 9.259270832236043e-06,
      "loss": 0.4677,
      "num_input_tokens_seen": 5714520,
      "step": 8750
    },
    {
      "epoch": 5.162146226415095,
      "grad_norm": 1.5069303512573242,
      "learning_rate": 9.257922734747729e-06,
      "loss": 0.6097,
      "num_input_tokens_seen": 5717720,
      "step": 8755
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 1.5910019874572754,
      "learning_rate": 9.256573509956018e-06,
      "loss": 0.6674,
      "num_input_tokens_seen": 5721080,
      "step": 8760
    },
    {
      "epoch": 5.168042452830188,
      "grad_norm": 1.956933617591858,
      "learning_rate": 9.255223158218127e-06,
      "loss": 0.4903,
      "num_input_tokens_seen": 5723480,
      "step": 8765
    },
    {
      "epoch": 5.1709905660377355,
      "grad_norm": 2.054112672805786,
      "learning_rate": 9.253871679891566e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 5726424,
      "step": 8770
    },
    {
      "epoch": 5.173938679245283,
      "grad_norm": 2.397785186767578,
      "learning_rate": 9.252519075334143e-06,
      "loss": 0.5639,
      "num_input_tokens_seen": 5729464,
      "step": 8775
    },
    {
      "epoch": 5.17688679245283,
      "grad_norm": 2.0727624893188477,
      "learning_rate": 9.251165344903969e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 5732696,
      "step": 8780
    },
    {
      "epoch": 5.179834905660377,
      "grad_norm": 2.5727663040161133,
      "learning_rate": 9.249810488959448e-06,
      "loss": 0.592,
      "num_input_tokens_seen": 5735704,
      "step": 8785
    },
    {
      "epoch": 5.182783018867925,
      "grad_norm": 1.6739391088485718,
      "learning_rate": 9.248454507859285e-06,
      "loss": 0.3844,
      "num_input_tokens_seen": 5738264,
      "step": 8790
    },
    {
      "epoch": 5.185731132075472,
      "grad_norm": 1.4829350709915161,
      "learning_rate": 9.247097401962482e-06,
      "loss": 0.4369,
      "num_input_tokens_seen": 5742264,
      "step": 8795
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 2.080010414123535,
      "learning_rate": 9.245739171628335e-06,
      "loss": 0.5004,
      "num_input_tokens_seen": 5745400,
      "step": 8800
    },
    {
      "epoch": 5.191627358490566,
      "grad_norm": 1.9254777431488037,
      "learning_rate": 9.244379817216447e-06,
      "loss": 0.4156,
      "num_input_tokens_seen": 5749432,
      "step": 8805
    },
    {
      "epoch": 5.194575471698113,
      "grad_norm": 2.354182004928589,
      "learning_rate": 9.243019339086708e-06,
      "loss": 0.5674,
      "num_input_tokens_seen": 5753432,
      "step": 8810
    },
    {
      "epoch": 5.19752358490566,
      "grad_norm": 1.6701116561889648,
      "learning_rate": 9.241657737599313e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 5757464,
      "step": 8815
    },
    {
      "epoch": 5.2004716981132075,
      "grad_norm": 2.3173279762268066,
      "learning_rate": 9.240295013114752e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 5760088,
      "step": 8820
    },
    {
      "epoch": 5.203419811320755,
      "grad_norm": 1.2014950513839722,
      "learning_rate": 9.238931165993811e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 5763320,
      "step": 8825
    },
    {
      "epoch": 5.206367924528302,
      "grad_norm": 1.155423641204834,
      "learning_rate": 9.237566196597577e-06,
      "loss": 0.4584,
      "num_input_tokens_seen": 5766360,
      "step": 8830
    },
    {
      "epoch": 5.209316037735849,
      "grad_norm": 1.2854483127593994,
      "learning_rate": 9.236200105287427e-06,
      "loss": 0.3325,
      "num_input_tokens_seen": 5769592,
      "step": 8835
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 1.6690623760223389,
      "learning_rate": 9.234832892425042e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 5773368,
      "step": 8840
    },
    {
      "epoch": 5.215212264150943,
      "grad_norm": 2.8791377544403076,
      "learning_rate": 9.2334645583724e-06,
      "loss": 0.5025,
      "num_input_tokens_seen": 5776184,
      "step": 8845
    },
    {
      "epoch": 5.21816037735849,
      "grad_norm": 1.6586191654205322,
      "learning_rate": 9.23209510349177e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 5778488,
      "step": 8850
    },
    {
      "epoch": 5.221108490566038,
      "grad_norm": 2.5717110633850098,
      "learning_rate": 9.230724528145722e-06,
      "loss": 0.3653,
      "num_input_tokens_seen": 5781912,
      "step": 8855
    },
    {
      "epoch": 5.224056603773585,
      "grad_norm": 1.4521546363830566,
      "learning_rate": 9.229352832697122e-06,
      "loss": 0.4758,
      "num_input_tokens_seen": 5784984,
      "step": 8860
    },
    {
      "epoch": 5.227004716981132,
      "grad_norm": 1.9902127981185913,
      "learning_rate": 9.22798001750913e-06,
      "loss": 0.4223,
      "num_input_tokens_seen": 5788280,
      "step": 8865
    },
    {
      "epoch": 5.2299528301886795,
      "grad_norm": 2.837940216064453,
      "learning_rate": 9.226606082945209e-06,
      "loss": 0.3876,
      "num_input_tokens_seen": 5790744,
      "step": 8870
    },
    {
      "epoch": 5.232900943396227,
      "grad_norm": 1.9621179103851318,
      "learning_rate": 9.225231029369112e-06,
      "loss": 0.5413,
      "num_input_tokens_seen": 5793368,
      "step": 8875
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 2.779245376586914,
      "learning_rate": 9.22385485714489e-06,
      "loss": 0.5528,
      "num_input_tokens_seen": 5796376,
      "step": 8880
    },
    {
      "epoch": 5.2387971698113205,
      "grad_norm": 1.7538864612579346,
      "learning_rate": 9.222477566636889e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 5799256,
      "step": 8885
    },
    {
      "epoch": 5.241745283018868,
      "grad_norm": 1.9255648851394653,
      "learning_rate": 9.221099158209757e-06,
      "loss": 0.4267,
      "num_input_tokens_seen": 5803384,
      "step": 8890
    },
    {
      "epoch": 5.244693396226415,
      "grad_norm": 1.1398993730545044,
      "learning_rate": 9.219719632228429e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 5806712,
      "step": 8895
    },
    {
      "epoch": 5.247641509433962,
      "grad_norm": 1.4057310819625854,
      "learning_rate": 9.218338989058141e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 5809336,
      "step": 8900
    },
    {
      "epoch": 5.25058962264151,
      "grad_norm": 1.684432864189148,
      "learning_rate": 9.21695722906443e-06,
      "loss": 0.5003,
      "num_input_tokens_seen": 5812312,
      "step": 8905
    },
    {
      "epoch": 5.253537735849057,
      "grad_norm": 3.1074512004852295,
      "learning_rate": 9.215574352613115e-06,
      "loss": 0.5285,
      "num_input_tokens_seen": 5815832,
      "step": 8910
    },
    {
      "epoch": 5.256485849056604,
      "grad_norm": 2.716393232345581,
      "learning_rate": 9.214190360070323e-06,
      "loss": 0.3981,
      "num_input_tokens_seen": 5820120,
      "step": 8915
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 1.4069578647613525,
      "learning_rate": 9.212805251802471e-06,
      "loss": 0.3829,
      "num_input_tokens_seen": 5823160,
      "step": 8920
    },
    {
      "epoch": 5.262382075471698,
      "grad_norm": 0.9861044883728027,
      "learning_rate": 9.211419028176273e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 5825592,
      "step": 8925
    },
    {
      "epoch": 5.265330188679245,
      "grad_norm": 1.8644124269485474,
      "learning_rate": 9.210031689558738e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 5828280,
      "step": 8930
    },
    {
      "epoch": 5.2682783018867925,
      "grad_norm": 1.5836007595062256,
      "learning_rate": 9.208643236317166e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 5831544,
      "step": 8935
    },
    {
      "epoch": 5.27122641509434,
      "grad_norm": 1.3701446056365967,
      "learning_rate": 9.20725366881916e-06,
      "loss": 0.4397,
      "num_input_tokens_seen": 5835224,
      "step": 8940
    },
    {
      "epoch": 5.274174528301887,
      "grad_norm": 1.9015079736709595,
      "learning_rate": 9.205862987432614e-06,
      "loss": 0.4388,
      "num_input_tokens_seen": 5838264,
      "step": 8945
    },
    {
      "epoch": 5.277122641509434,
      "grad_norm": 1.67697274684906,
      "learning_rate": 9.204471192525715e-06,
      "loss": 0.5433,
      "num_input_tokens_seen": 5841368,
      "step": 8950
    },
    {
      "epoch": 5.280070754716981,
      "grad_norm": 2.042046308517456,
      "learning_rate": 9.203078284466949e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 5845048,
      "step": 8955
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 1.7114808559417725,
      "learning_rate": 9.201684263625091e-06,
      "loss": 0.5398,
      "num_input_tokens_seen": 5847928,
      "step": 8960
    },
    {
      "epoch": 5.285966981132075,
      "grad_norm": 1.5776622295379639,
      "learning_rate": 9.200289130369218e-06,
      "loss": 0.381,
      "num_input_tokens_seen": 5850776,
      "step": 8965
    },
    {
      "epoch": 5.288915094339623,
      "grad_norm": 3.4823622703552246,
      "learning_rate": 9.198892885068693e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 5853912,
      "step": 8970
    },
    {
      "epoch": 5.29186320754717,
      "grad_norm": 1.1605238914489746,
      "learning_rate": 9.197495528093182e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 5856952,
      "step": 8975
    },
    {
      "epoch": 5.294811320754717,
      "grad_norm": 1.2419577836990356,
      "learning_rate": 9.196097059812639e-06,
      "loss": 0.5837,
      "num_input_tokens_seen": 5860920,
      "step": 8980
    },
    {
      "epoch": 5.2977594339622645,
      "grad_norm": 1.7890515327453613,
      "learning_rate": 9.194697480597316e-06,
      "loss": 0.4083,
      "num_input_tokens_seen": 5864248,
      "step": 8985
    },
    {
      "epoch": 5.300707547169811,
      "grad_norm": 2.0319530963897705,
      "learning_rate": 9.193296790817755e-06,
      "loss": 0.4178,
      "num_input_tokens_seen": 5866840,
      "step": 8990
    },
    {
      "epoch": 5.303655660377358,
      "grad_norm": 2.3416240215301514,
      "learning_rate": 9.1918949908448e-06,
      "loss": 0.4744,
      "num_input_tokens_seen": 5870008,
      "step": 8995
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 3.2975070476531982,
      "learning_rate": 9.190492081049578e-06,
      "loss": 0.4673,
      "num_input_tokens_seen": 5872888,
      "step": 9000
    },
    {
      "epoch": 5.309551886792453,
      "grad_norm": 1.0939768552780151,
      "learning_rate": 9.189088061803517e-06,
      "loss": 0.4812,
      "num_input_tokens_seen": 5876568,
      "step": 9005
    },
    {
      "epoch": 5.3125,
      "grad_norm": 2.096231460571289,
      "learning_rate": 9.187682933478337e-06,
      "loss": 0.4125,
      "num_input_tokens_seen": 5879768,
      "step": 9010
    },
    {
      "epoch": 5.315448113207547,
      "grad_norm": 1.2858150005340576,
      "learning_rate": 9.186276696446054e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 5882648,
      "step": 9015
    },
    {
      "epoch": 5.318396226415095,
      "grad_norm": 1.413573145866394,
      "learning_rate": 9.184869351078974e-06,
      "loss": 0.305,
      "num_input_tokens_seen": 5885176,
      "step": 9020
    },
    {
      "epoch": 5.321344339622642,
      "grad_norm": 2.517014265060425,
      "learning_rate": 9.183460897749697e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 5888280,
      "step": 9025
    },
    {
      "epoch": 5.324292452830189,
      "grad_norm": 1.326893925666809,
      "learning_rate": 9.182051336831117e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 5894456,
      "step": 9030
    },
    {
      "epoch": 5.3272405660377355,
      "grad_norm": 1.6250325441360474,
      "learning_rate": 9.180640668696424e-06,
      "loss": 0.3617,
      "num_input_tokens_seen": 5897272,
      "step": 9035
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 1.876648187637329,
      "learning_rate": 9.179228893719094e-06,
      "loss": 0.5168,
      "num_input_tokens_seen": 5900344,
      "step": 9040
    },
    {
      "epoch": 5.33313679245283,
      "grad_norm": 1.16081964969635,
      "learning_rate": 9.177816012272904e-06,
      "loss": 0.3139,
      "num_input_tokens_seen": 5904440,
      "step": 9045
    },
    {
      "epoch": 5.336084905660377,
      "grad_norm": 1.3900142908096313,
      "learning_rate": 9.17640202473192e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 5907672,
      "step": 9050
    },
    {
      "epoch": 5.339033018867925,
      "grad_norm": 2.376675605773926,
      "learning_rate": 9.1749869314705e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 5909880,
      "step": 9055
    },
    {
      "epoch": 5.341981132075472,
      "grad_norm": 1.5803892612457275,
      "learning_rate": 9.173570732863295e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 5912504,
      "step": 9060
    },
    {
      "epoch": 5.344929245283019,
      "grad_norm": 1.5988380908966064,
      "learning_rate": 9.172153429285254e-06,
      "loss": 0.4768,
      "num_input_tokens_seen": 5915416,
      "step": 9065
    },
    {
      "epoch": 5.347877358490566,
      "grad_norm": 4.629351615905762,
      "learning_rate": 9.17073502111161e-06,
      "loss": 0.6215,
      "num_input_tokens_seen": 5919384,
      "step": 9070
    },
    {
      "epoch": 5.350825471698113,
      "grad_norm": 1.60885488986969,
      "learning_rate": 9.169315508717895e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 5923160,
      "step": 9075
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 1.5132442712783813,
      "learning_rate": 9.167894892479932e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 5926200,
      "step": 9080
    },
    {
      "epoch": 5.3567216981132075,
      "grad_norm": 1.8697545528411865,
      "learning_rate": 9.16647317277383e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 5929144,
      "step": 9085
    },
    {
      "epoch": 5.359669811320755,
      "grad_norm": 1.505454421043396,
      "learning_rate": 9.165050349976002e-06,
      "loss": 0.4116,
      "num_input_tokens_seen": 5931800,
      "step": 9090
    },
    {
      "epoch": 5.362617924528302,
      "grad_norm": 1.9919071197509766,
      "learning_rate": 9.16362642446314e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 5935992,
      "step": 9095
    },
    {
      "epoch": 5.365566037735849,
      "grad_norm": 1.1345899105072021,
      "learning_rate": 9.162201396612242e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 5939736,
      "step": 9100
    },
    {
      "epoch": 5.368514150943396,
      "grad_norm": 3.4386613368988037,
      "learning_rate": 9.160775266800583e-06,
      "loss": 0.4122,
      "num_input_tokens_seen": 5942200,
      "step": 9105
    },
    {
      "epoch": 5.371462264150943,
      "grad_norm": 1.8309788703918457,
      "learning_rate": 9.159348035405742e-06,
      "loss": 0.515,
      "num_input_tokens_seen": 5946040,
      "step": 9110
    },
    {
      "epoch": 5.37441037735849,
      "grad_norm": 1.300981044769287,
      "learning_rate": 9.157919702805582e-06,
      "loss": 0.4849,
      "num_input_tokens_seen": 5950744,
      "step": 9115
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 2.2341055870056152,
      "learning_rate": 9.156490269378262e-06,
      "loss": 0.6686,
      "num_input_tokens_seen": 5954392,
      "step": 9120
    },
    {
      "epoch": 5.380306603773585,
      "grad_norm": 1.7851070165634155,
      "learning_rate": 9.15505973550223e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 5957368,
      "step": 9125
    },
    {
      "epoch": 5.383254716981132,
      "grad_norm": 1.8383705615997314,
      "learning_rate": 9.153628101556223e-06,
      "loss": 0.4066,
      "num_input_tokens_seen": 5959992,
      "step": 9130
    },
    {
      "epoch": 5.3862028301886795,
      "grad_norm": 2.2004008293151855,
      "learning_rate": 9.152195367919277e-06,
      "loss": 0.5295,
      "num_input_tokens_seen": 5962968,
      "step": 9135
    },
    {
      "epoch": 5.389150943396227,
      "grad_norm": 1.5030440092086792,
      "learning_rate": 9.150761534970713e-06,
      "loss": 0.4429,
      "num_input_tokens_seen": 5967000,
      "step": 9140
    },
    {
      "epoch": 5.392099056603773,
      "grad_norm": 1.53957200050354,
      "learning_rate": 9.149326603090144e-06,
      "loss": 0.6063,
      "num_input_tokens_seen": 5969624,
      "step": 9145
    },
    {
      "epoch": 5.3950471698113205,
      "grad_norm": 1.4748774766921997,
      "learning_rate": 9.147890572657471e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 5972376,
      "step": 9150
    },
    {
      "epoch": 5.397995283018868,
      "grad_norm": 1.9818493127822876,
      "learning_rate": 9.146453444052895e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 5974648,
      "step": 9155
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 1.5779138803482056,
      "learning_rate": 9.145015217656899e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 5977784,
      "step": 9160
    },
    {
      "epoch": 5.403891509433962,
      "grad_norm": 2.8589470386505127,
      "learning_rate": 9.14357589385026e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 5980696,
      "step": 9165
    },
    {
      "epoch": 5.40683962264151,
      "grad_norm": 1.0811729431152344,
      "learning_rate": 9.142135473014046e-06,
      "loss": 0.5024,
      "num_input_tokens_seen": 5984248,
      "step": 9170
    },
    {
      "epoch": 5.409787735849057,
      "grad_norm": 2.356567621231079,
      "learning_rate": 9.140693955529614e-06,
      "loss": 0.5288,
      "num_input_tokens_seen": 5987032,
      "step": 9175
    },
    {
      "epoch": 5.412735849056604,
      "grad_norm": 2.3432393074035645,
      "learning_rate": 9.13925134177861e-06,
      "loss": 0.4381,
      "num_input_tokens_seen": 5990616,
      "step": 9180
    },
    {
      "epoch": 5.415683962264151,
      "grad_norm": 2.484057664871216,
      "learning_rate": 9.137807632142977e-06,
      "loss": 0.593,
      "num_input_tokens_seen": 5993176,
      "step": 9185
    },
    {
      "epoch": 5.418632075471698,
      "grad_norm": 2.206685781478882,
      "learning_rate": 9.136362827004937e-06,
      "loss": 0.6016,
      "num_input_tokens_seen": 5996472,
      "step": 9190
    },
    {
      "epoch": 5.421580188679245,
      "grad_norm": 2.1147615909576416,
      "learning_rate": 9.134916926747015e-06,
      "loss": 0.4756,
      "num_input_tokens_seen": 5999960,
      "step": 9195
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 0.9989801049232483,
      "learning_rate": 9.133469931752016e-06,
      "loss": 0.5293,
      "num_input_tokens_seen": 6003160,
      "step": 9200
    },
    {
      "epoch": 5.42747641509434,
      "grad_norm": 1.825844645500183,
      "learning_rate": 9.132021842403035e-06,
      "loss": 0.4726,
      "num_input_tokens_seen": 6006168,
      "step": 9205
    },
    {
      "epoch": 5.430424528301887,
      "grad_norm": 1.243045449256897,
      "learning_rate": 9.130572659083465e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 6009336,
      "step": 9210
    },
    {
      "epoch": 5.433372641509434,
      "grad_norm": 2.8012053966522217,
      "learning_rate": 9.129122382176982e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 6011992,
      "step": 9215
    },
    {
      "epoch": 5.436320754716981,
      "grad_norm": 2.7644503116607666,
      "learning_rate": 9.127671012067554e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 6015288,
      "step": 9220
    },
    {
      "epoch": 5.439268867924528,
      "grad_norm": 1.675557017326355,
      "learning_rate": 9.126218549139434e-06,
      "loss": 0.4856,
      "num_input_tokens_seen": 6018680,
      "step": 9225
    },
    {
      "epoch": 5.442216981132075,
      "grad_norm": 2.1388590335845947,
      "learning_rate": 9.124764993777171e-06,
      "loss": 0.5528,
      "num_input_tokens_seen": 6021304,
      "step": 9230
    },
    {
      "epoch": 5.445165094339623,
      "grad_norm": 1.6079766750335693,
      "learning_rate": 9.1233103463656e-06,
      "loss": 0.6229,
      "num_input_tokens_seen": 6025336,
      "step": 9235
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 1.450141429901123,
      "learning_rate": 9.121854607289842e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 6028920,
      "step": 9240
    },
    {
      "epoch": 5.451061320754717,
      "grad_norm": 1.9899173974990845,
      "learning_rate": 9.120397776935314e-06,
      "loss": 0.4463,
      "num_input_tokens_seen": 6031576,
      "step": 9245
    },
    {
      "epoch": 5.4540094339622645,
      "grad_norm": 3.0140907764434814,
      "learning_rate": 9.118939855687717e-06,
      "loss": 0.5476,
      "num_input_tokens_seen": 6034296,
      "step": 9250
    },
    {
      "epoch": 5.456957547169811,
      "grad_norm": 1.1145292520523071,
      "learning_rate": 9.117480843933043e-06,
      "loss": 0.5568,
      "num_input_tokens_seen": 6037976,
      "step": 9255
    },
    {
      "epoch": 5.459905660377358,
      "grad_norm": 1.4633828401565552,
      "learning_rate": 9.116020742057567e-06,
      "loss": 0.6079,
      "num_input_tokens_seen": 6041048,
      "step": 9260
    },
    {
      "epoch": 5.462853773584905,
      "grad_norm": 0.9874786138534546,
      "learning_rate": 9.114559550447863e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 6044632,
      "step": 9265
    },
    {
      "epoch": 5.465801886792453,
      "grad_norm": 1.4796861410140991,
      "learning_rate": 9.113097269490784e-06,
      "loss": 0.4941,
      "num_input_tokens_seen": 6047960,
      "step": 9270
    },
    {
      "epoch": 5.46875,
      "grad_norm": 2.6763792037963867,
      "learning_rate": 9.111633899573476e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 6051544,
      "step": 9275
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 1.6188429594039917,
      "learning_rate": 9.110169441083374e-06,
      "loss": 0.4269,
      "num_input_tokens_seen": 6054552,
      "step": 9280
    },
    {
      "epoch": 5.474646226415095,
      "grad_norm": 1.9624098539352417,
      "learning_rate": 9.108703894408198e-06,
      "loss": 0.3915,
      "num_input_tokens_seen": 6057752,
      "step": 9285
    },
    {
      "epoch": 5.477594339622642,
      "grad_norm": 1.8661527633666992,
      "learning_rate": 9.107237259935959e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 6060984,
      "step": 9290
    },
    {
      "epoch": 5.480542452830189,
      "grad_norm": 1.7536612749099731,
      "learning_rate": 9.105769538054954e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 6063800,
      "step": 9295
    },
    {
      "epoch": 5.4834905660377355,
      "grad_norm": 2.0628578662872314,
      "learning_rate": 9.104300729153768e-06,
      "loss": 0.4073,
      "num_input_tokens_seen": 6066872,
      "step": 9300
    },
    {
      "epoch": 5.486438679245283,
      "grad_norm": 1.2360578775405884,
      "learning_rate": 9.102830833621277e-06,
      "loss": 0.6932,
      "num_input_tokens_seen": 6069688,
      "step": 9305
    },
    {
      "epoch": 5.48938679245283,
      "grad_norm": 1.2822805643081665,
      "learning_rate": 9.101359851846639e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 6073112,
      "step": 9310
    },
    {
      "epoch": 5.492334905660377,
      "grad_norm": 0.98284512758255,
      "learning_rate": 9.099887784219305e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 6076408,
      "step": 9315
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 2.500035285949707,
      "learning_rate": 9.098414631129012e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 6079896,
      "step": 9320
    },
    {
      "epoch": 5.498231132075472,
      "grad_norm": 1.5253368616104126,
      "learning_rate": 9.09694039296578e-06,
      "loss": 0.3718,
      "num_input_tokens_seen": 6084088,
      "step": 9325
    },
    {
      "epoch": 5.501179245283019,
      "grad_norm": 3.082324266433716,
      "learning_rate": 9.095465070119924e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 6088568,
      "step": 9330
    },
    {
      "epoch": 5.504127358490566,
      "grad_norm": 2.325551748275757,
      "learning_rate": 9.09398866298204e-06,
      "loss": 0.4616,
      "num_input_tokens_seen": 6091000,
      "step": 9335
    },
    {
      "epoch": 5.507075471698113,
      "grad_norm": 1.2468445301055908,
      "learning_rate": 9.09251117194301e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 6094840,
      "step": 9340
    },
    {
      "epoch": 5.51002358490566,
      "grad_norm": 3.4928975105285645,
      "learning_rate": 9.091032597394012e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 6096920,
      "step": 9345
    },
    {
      "epoch": 5.5129716981132075,
      "grad_norm": 2.8167800903320312,
      "learning_rate": 9.089552939726503e-06,
      "loss": 0.7163,
      "num_input_tokens_seen": 6101272,
      "step": 9350
    },
    {
      "epoch": 5.515919811320755,
      "grad_norm": 1.4385277032852173,
      "learning_rate": 9.088072199332227e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 6104536,
      "step": 9355
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 1.5333603620529175,
      "learning_rate": 9.08659037660322e-06,
      "loss": 0.4503,
      "num_input_tokens_seen": 6108120,
      "step": 9360
    },
    {
      "epoch": 5.521816037735849,
      "grad_norm": 1.4738523960113525,
      "learning_rate": 9.085107471931797e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 6112312,
      "step": 9365
    },
    {
      "epoch": 5.524764150943396,
      "grad_norm": 1.6129785776138306,
      "learning_rate": 9.083623485710564e-06,
      "loss": 0.3002,
      "num_input_tokens_seen": 6115768,
      "step": 9370
    },
    {
      "epoch": 5.527712264150943,
      "grad_norm": 1.0850787162780762,
      "learning_rate": 9.082138418332416e-06,
      "loss": 0.4415,
      "num_input_tokens_seen": 6119544,
      "step": 9375
    },
    {
      "epoch": 5.53066037735849,
      "grad_norm": 1.5544534921646118,
      "learning_rate": 9.080652270190527e-06,
      "loss": 0.4981,
      "num_input_tokens_seen": 6122072,
      "step": 9380
    },
    {
      "epoch": 5.533608490566038,
      "grad_norm": 1.1695162057876587,
      "learning_rate": 9.079165041678363e-06,
      "loss": 0.4336,
      "num_input_tokens_seen": 6125208,
      "step": 9385
    },
    {
      "epoch": 5.536556603773585,
      "grad_norm": 3.282097816467285,
      "learning_rate": 9.077676733189675e-06,
      "loss": 0.4468,
      "num_input_tokens_seen": 6127928,
      "step": 9390
    },
    {
      "epoch": 5.539504716981132,
      "grad_norm": 1.2967989444732666,
      "learning_rate": 9.076187345118496e-06,
      "loss": 0.4118,
      "num_input_tokens_seen": 6131224,
      "step": 9395
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 3.71042537689209,
      "learning_rate": 9.074696877859152e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 6133848,
      "step": 9400
    },
    {
      "epoch": 5.545400943396227,
      "grad_norm": 0.7070104479789734,
      "learning_rate": 9.073205331806248e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 6138008,
      "step": 9405
    },
    {
      "epoch": 5.548349056603773,
      "grad_norm": 2.4920735359191895,
      "learning_rate": 9.071712707354676e-06,
      "loss": 0.5318,
      "num_input_tokens_seen": 6141560,
      "step": 9410
    },
    {
      "epoch": 5.5512971698113205,
      "grad_norm": 2.4607324600219727,
      "learning_rate": 9.070219004899618e-06,
      "loss": 0.531,
      "num_input_tokens_seen": 6145016,
      "step": 9415
    },
    {
      "epoch": 5.554245283018868,
      "grad_norm": 5.740159511566162,
      "learning_rate": 9.068724224836538e-06,
      "loss": 0.509,
      "num_input_tokens_seen": 6148888,
      "step": 9420
    },
    {
      "epoch": 5.557193396226415,
      "grad_norm": 1.633988857269287,
      "learning_rate": 9.067228367561182e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 6152280,
      "step": 9425
    },
    {
      "epoch": 5.560141509433962,
      "grad_norm": 1.5052180290222168,
      "learning_rate": 9.06573143346959e-06,
      "loss": 0.4815,
      "num_input_tokens_seen": 6155160,
      "step": 9430
    },
    {
      "epoch": 5.56308962264151,
      "grad_norm": 2.1565279960632324,
      "learning_rate": 9.064233422958078e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 6159768,
      "step": 9435
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 2.828359603881836,
      "learning_rate": 9.062734336423248e-06,
      "loss": 0.5432,
      "num_input_tokens_seen": 6162776,
      "step": 9440
    },
    {
      "epoch": 5.568985849056604,
      "grad_norm": 3.066452980041504,
      "learning_rate": 9.061234174261998e-06,
      "loss": 0.5734,
      "num_input_tokens_seen": 6165528,
      "step": 9445
    },
    {
      "epoch": 5.571933962264151,
      "grad_norm": 1.8290003538131714,
      "learning_rate": 9.059732936871493e-06,
      "loss": 0.4458,
      "num_input_tokens_seen": 6168760,
      "step": 9450
    },
    {
      "epoch": 5.574882075471698,
      "grad_norm": 1.2739454507827759,
      "learning_rate": 9.058230624649198e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 6172440,
      "step": 9455
    },
    {
      "epoch": 5.577830188679245,
      "grad_norm": 1.5867599248886108,
      "learning_rate": 9.056727237992856e-06,
      "loss": 0.4685,
      "num_input_tokens_seen": 6174840,
      "step": 9460
    },
    {
      "epoch": 5.5807783018867925,
      "grad_norm": 1.065354824066162,
      "learning_rate": 9.055222777300493e-06,
      "loss": 0.3732,
      "num_input_tokens_seen": 6179032,
      "step": 9465
    },
    {
      "epoch": 5.58372641509434,
      "grad_norm": 2.5281238555908203,
      "learning_rate": 9.053717242970423e-06,
      "loss": 0.4928,
      "num_input_tokens_seen": 6182104,
      "step": 9470
    },
    {
      "epoch": 5.586674528301887,
      "grad_norm": 1.4441630840301514,
      "learning_rate": 9.052210635401244e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 6185240,
      "step": 9475
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 2.1855344772338867,
      "learning_rate": 9.050702954991833e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 6188504,
      "step": 9480
    },
    {
      "epoch": 5.592570754716981,
      "grad_norm": 1.5344091653823853,
      "learning_rate": 9.049194202141358e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 6192760,
      "step": 9485
    },
    {
      "epoch": 5.595518867924528,
      "grad_norm": 2.2318220138549805,
      "learning_rate": 9.047684377249267e-06,
      "loss": 0.5078,
      "num_input_tokens_seen": 6195800,
      "step": 9490
    },
    {
      "epoch": 5.598466981132075,
      "grad_norm": 1.5282148122787476,
      "learning_rate": 9.046173480715292e-06,
      "loss": 0.6103,
      "num_input_tokens_seen": 6198680,
      "step": 9495
    },
    {
      "epoch": 5.601415094339623,
      "grad_norm": 1.4158135652542114,
      "learning_rate": 9.044661512939451e-06,
      "loss": 0.524,
      "num_input_tokens_seen": 6203000,
      "step": 9500
    },
    {
      "epoch": 5.60436320754717,
      "grad_norm": 1.366776943206787,
      "learning_rate": 9.043148474322043e-06,
      "loss": 0.46,
      "num_input_tokens_seen": 6206776,
      "step": 9505
    },
    {
      "epoch": 5.607311320754717,
      "grad_norm": 1.4267454147338867,
      "learning_rate": 9.04163436526365e-06,
      "loss": 0.3999,
      "num_input_tokens_seen": 6209816,
      "step": 9510
    },
    {
      "epoch": 5.6102594339622645,
      "grad_norm": 1.3689570426940918,
      "learning_rate": 9.040119186165142e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 6212792,
      "step": 9515
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 2.5191845893859863,
      "learning_rate": 9.038602937427665e-06,
      "loss": 0.5092,
      "num_input_tokens_seen": 6216696,
      "step": 9520
    },
    {
      "epoch": 5.616155660377358,
      "grad_norm": 1.2067756652832031,
      "learning_rate": 9.037085619452658e-06,
      "loss": 0.4173,
      "num_input_tokens_seen": 6219832,
      "step": 9525
    },
    {
      "epoch": 5.619103773584905,
      "grad_norm": 0.9765531420707703,
      "learning_rate": 9.035567232641833e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 6223032,
      "step": 9530
    },
    {
      "epoch": 5.622051886792453,
      "grad_norm": 1.5662314891815186,
      "learning_rate": 9.03404777739719e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 6225688,
      "step": 9535
    },
    {
      "epoch": 5.625,
      "grad_norm": 1.366266131401062,
      "learning_rate": 9.032527254121013e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 6228600,
      "step": 9540
    },
    {
      "epoch": 5.627948113207547,
      "grad_norm": 1.2653448581695557,
      "learning_rate": 9.031005663215867e-06,
      "loss": 0.3978,
      "num_input_tokens_seen": 6232632,
      "step": 9545
    },
    {
      "epoch": 5.630896226415095,
      "grad_norm": 1.9471909999847412,
      "learning_rate": 9.029483005084595e-06,
      "loss": 0.4635,
      "num_input_tokens_seen": 6235992,
      "step": 9550
    },
    {
      "epoch": 5.633844339622642,
      "grad_norm": 1.7848467826843262,
      "learning_rate": 9.027959280130337e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 6238712,
      "step": 9555
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 1.9503623247146606,
      "learning_rate": 9.026434488756496e-06,
      "loss": 0.4392,
      "num_input_tokens_seen": 6241912,
      "step": 9560
    },
    {
      "epoch": 5.6397405660377355,
      "grad_norm": 1.8083369731903076,
      "learning_rate": 9.024908631366774e-06,
      "loss": 0.536,
      "num_input_tokens_seen": 6245112,
      "step": 9565
    },
    {
      "epoch": 5.642688679245283,
      "grad_norm": 1.4834867715835571,
      "learning_rate": 9.023381708365143e-06,
      "loss": 0.4291,
      "num_input_tokens_seen": 6248472,
      "step": 9570
    },
    {
      "epoch": 5.64563679245283,
      "grad_norm": 1.3561058044433594,
      "learning_rate": 9.021853720155866e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 6250584,
      "step": 9575
    },
    {
      "epoch": 5.648584905660377,
      "grad_norm": 0.80008465051651,
      "learning_rate": 9.020324667143483e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 6258072,
      "step": 9580
    },
    {
      "epoch": 5.651533018867925,
      "grad_norm": 2.6839540004730225,
      "learning_rate": 9.018794549732819e-06,
      "loss": 0.4591,
      "num_input_tokens_seen": 6260280,
      "step": 9585
    },
    {
      "epoch": 5.654481132075472,
      "grad_norm": 1.7686944007873535,
      "learning_rate": 9.017263368328977e-06,
      "loss": 0.4792,
      "num_input_tokens_seen": 6263672,
      "step": 9590
    },
    {
      "epoch": 5.657429245283019,
      "grad_norm": 1.3399150371551514,
      "learning_rate": 9.015731123337344e-06,
      "loss": 0.5509,
      "num_input_tokens_seen": 6266840,
      "step": 9595
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 1.5934486389160156,
      "learning_rate": 9.01419781516359e-06,
      "loss": 0.5395,
      "num_input_tokens_seen": 6269784,
      "step": 9600
    },
    {
      "epoch": 5.663325471698113,
      "grad_norm": 3.4706313610076904,
      "learning_rate": 9.012663444213664e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 6272440,
      "step": 9605
    },
    {
      "epoch": 5.66627358490566,
      "grad_norm": 2.095283269882202,
      "learning_rate": 9.011128010893797e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 6275640,
      "step": 9610
    },
    {
      "epoch": 5.6692216981132075,
      "grad_norm": 0.9666507840156555,
      "learning_rate": 9.009591515610503e-06,
      "loss": 0.5474,
      "num_input_tokens_seen": 6279480,
      "step": 9615
    },
    {
      "epoch": 5.672169811320755,
      "grad_norm": 1.4456658363342285,
      "learning_rate": 9.008053958770575e-06,
      "loss": 0.5755,
      "num_input_tokens_seen": 6282168,
      "step": 9620
    },
    {
      "epoch": 5.675117924528302,
      "grad_norm": 5.761960983276367,
      "learning_rate": 9.006515340781087e-06,
      "loss": 0.5576,
      "num_input_tokens_seen": 6285048,
      "step": 9625
    },
    {
      "epoch": 5.678066037735849,
      "grad_norm": 1.6227318048477173,
      "learning_rate": 9.004975662049396e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 6287832,
      "step": 9630
    },
    {
      "epoch": 5.681014150943396,
      "grad_norm": 5.075666427612305,
      "learning_rate": 9.003434922983138e-06,
      "loss": 0.4294,
      "num_input_tokens_seen": 6290648,
      "step": 9635
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 1.8767173290252686,
      "learning_rate": 9.00189312399023e-06,
      "loss": 0.3455,
      "num_input_tokens_seen": 6292952,
      "step": 9640
    },
    {
      "epoch": 5.68691037735849,
      "grad_norm": 3.2433676719665527,
      "learning_rate": 9.00035026547887e-06,
      "loss": 0.5575,
      "num_input_tokens_seen": 6295768,
      "step": 9645
    },
    {
      "epoch": 5.689858490566038,
      "grad_norm": 1.4572858810424805,
      "learning_rate": 8.998806347857537e-06,
      "loss": 0.5664,
      "num_input_tokens_seen": 6299000,
      "step": 9650
    },
    {
      "epoch": 5.692806603773585,
      "grad_norm": 1.2767596244812012,
      "learning_rate": 8.99726137153499e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 6302968,
      "step": 9655
    },
    {
      "epoch": 5.695754716981132,
      "grad_norm": 1.311949610710144,
      "learning_rate": 8.995715336920266e-06,
      "loss": 0.4803,
      "num_input_tokens_seen": 6306488,
      "step": 9660
    },
    {
      "epoch": 5.6987028301886795,
      "grad_norm": 2.539301872253418,
      "learning_rate": 8.994168244422687e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 6309080,
      "step": 9665
    },
    {
      "epoch": 5.701650943396227,
      "grad_norm": 1.175352692604065,
      "learning_rate": 8.992620094451852e-06,
      "loss": 0.501,
      "num_input_tokens_seen": 6312792,
      "step": 9670
    },
    {
      "epoch": 5.704599056603773,
      "grad_norm": 1.7095028162002563,
      "learning_rate": 8.991070887417639e-06,
      "loss": 0.4826,
      "num_input_tokens_seen": 6315512,
      "step": 9675
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 2.1093029975891113,
      "learning_rate": 8.989520623730208e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 6318424,
      "step": 9680
    },
    {
      "epoch": 5.710495283018868,
      "grad_norm": 1.559562087059021,
      "learning_rate": 8.987969303799998e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 6321016,
      "step": 9685
    },
    {
      "epoch": 5.713443396226415,
      "grad_norm": 1.6024647951126099,
      "learning_rate": 8.986416928037728e-06,
      "loss": 0.5049,
      "num_input_tokens_seen": 6324344,
      "step": 9690
    },
    {
      "epoch": 5.716391509433962,
      "grad_norm": 1.4956096410751343,
      "learning_rate": 8.984863496854395e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 6328664,
      "step": 9695
    },
    {
      "epoch": 5.71933962264151,
      "grad_norm": 1.0884099006652832,
      "learning_rate": 8.983309010661279e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 6331928,
      "step": 9700
    },
    {
      "epoch": 5.722287735849057,
      "grad_norm": 1.1198208332061768,
      "learning_rate": 8.981753469869934e-06,
      "loss": 0.4381,
      "num_input_tokens_seen": 6334904,
      "step": 9705
    },
    {
      "epoch": 5.725235849056604,
      "grad_norm": 1.5786765813827515,
      "learning_rate": 8.980196874892198e-06,
      "loss": 0.4505,
      "num_input_tokens_seen": 6338776,
      "step": 9710
    },
    {
      "epoch": 5.728183962264151,
      "grad_norm": 1.7544306516647339,
      "learning_rate": 8.978639226140184e-06,
      "loss": 0.4382,
      "num_input_tokens_seen": 6341624,
      "step": 9715
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 2.553950786590576,
      "learning_rate": 8.977080524026289e-06,
      "loss": 0.5207,
      "num_input_tokens_seen": 6344440,
      "step": 9720
    },
    {
      "epoch": 5.734080188679245,
      "grad_norm": 1.5306864976882935,
      "learning_rate": 8.975520768963186e-06,
      "loss": 0.4178,
      "num_input_tokens_seen": 6348472,
      "step": 9725
    },
    {
      "epoch": 5.7370283018867925,
      "grad_norm": 1.4117215871810913,
      "learning_rate": 8.973959961363825e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 6351640,
      "step": 9730
    },
    {
      "epoch": 5.73997641509434,
      "grad_norm": 1.8969188928604126,
      "learning_rate": 8.972398101641438e-06,
      "loss": 0.4278,
      "num_input_tokens_seen": 6355160,
      "step": 9735
    },
    {
      "epoch": 5.742924528301887,
      "grad_norm": 1.7587002515792847,
      "learning_rate": 8.970835190209532e-06,
      "loss": 0.5854,
      "num_input_tokens_seen": 6357592,
      "step": 9740
    },
    {
      "epoch": 5.745872641509434,
      "grad_norm": 1.4103196859359741,
      "learning_rate": 8.969271227481899e-06,
      "loss": 0.4823,
      "num_input_tokens_seen": 6359992,
      "step": 9745
    },
    {
      "epoch": 5.748820754716981,
      "grad_norm": 1.641676664352417,
      "learning_rate": 8.967706213872599e-06,
      "loss": 0.4262,
      "num_input_tokens_seen": 6362616,
      "step": 9750
    },
    {
      "epoch": 5.751768867924528,
      "grad_norm": 1.3094441890716553,
      "learning_rate": 8.966140149795981e-06,
      "loss": 0.3532,
      "num_input_tokens_seen": 6366040,
      "step": 9755
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 1.0604348182678223,
      "learning_rate": 8.964573035666663e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 6369112,
      "step": 9760
    },
    {
      "epoch": 5.757665094339623,
      "grad_norm": 1.1641610860824585,
      "learning_rate": 8.96300487189955e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 6373240,
      "step": 9765
    },
    {
      "epoch": 5.76061320754717,
      "grad_norm": 1.1274014711380005,
      "learning_rate": 8.961435658909816e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 6375960,
      "step": 9770
    },
    {
      "epoch": 5.763561320754717,
      "grad_norm": 3.0470855236053467,
      "learning_rate": 8.959865397112918e-06,
      "loss": 0.523,
      "num_input_tokens_seen": 6379224,
      "step": 9775
    },
    {
      "epoch": 5.7665094339622645,
      "grad_norm": 1.3585205078125,
      "learning_rate": 8.95829408692459e-06,
      "loss": 0.5124,
      "num_input_tokens_seen": 6382232,
      "step": 9780
    },
    {
      "epoch": 5.769457547169811,
      "grad_norm": 1.9035520553588867,
      "learning_rate": 8.956721728760845e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 6385656,
      "step": 9785
    },
    {
      "epoch": 5.772405660377358,
      "grad_norm": 1.7590277194976807,
      "learning_rate": 8.95514832303797e-06,
      "loss": 0.4891,
      "num_input_tokens_seen": 6388216,
      "step": 9790
    },
    {
      "epoch": 5.775353773584905,
      "grad_norm": 1.1858161687850952,
      "learning_rate": 8.953573870172528e-06,
      "loss": 0.3018,
      "num_input_tokens_seen": 6392664,
      "step": 9795
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 1.7944549322128296,
      "learning_rate": 8.951998370581368e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 6395416,
      "step": 9800
    },
    {
      "epoch": 5.78125,
      "grad_norm": 2.6107659339904785,
      "learning_rate": 8.950421824681605e-06,
      "loss": 0.3013,
      "num_input_tokens_seen": 6397816,
      "step": 9805
    },
    {
      "epoch": 5.784198113207547,
      "grad_norm": 1.4293004274368286,
      "learning_rate": 8.948844232890638e-06,
      "loss": 0.5675,
      "num_input_tokens_seen": 6401112,
      "step": 9810
    },
    {
      "epoch": 5.787146226415095,
      "grad_norm": 1.4626777172088623,
      "learning_rate": 8.947265595626144e-06,
      "loss": 0.5237,
      "num_input_tokens_seen": 6405304,
      "step": 9815
    },
    {
      "epoch": 5.790094339622642,
      "grad_norm": 1.465264916419983,
      "learning_rate": 8.945685913306071e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 6408120,
      "step": 9820
    },
    {
      "epoch": 5.793042452830189,
      "grad_norm": 1.7241607904434204,
      "learning_rate": 8.944105186348646e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 6411320,
      "step": 9825
    },
    {
      "epoch": 5.7959905660377355,
      "grad_norm": 2.253093957901001,
      "learning_rate": 8.942523415172377e-06,
      "loss": 0.5376,
      "num_input_tokens_seen": 6414712,
      "step": 9830
    },
    {
      "epoch": 5.798938679245283,
      "grad_norm": 5.096747875213623,
      "learning_rate": 8.94094060019604e-06,
      "loss": 0.4366,
      "num_input_tokens_seen": 6417912,
      "step": 9835
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 1.9329639673233032,
      "learning_rate": 8.939356741838696e-06,
      "loss": 0.5726,
      "num_input_tokens_seen": 6420824,
      "step": 9840
    },
    {
      "epoch": 5.804834905660377,
      "grad_norm": 2.8602206707000732,
      "learning_rate": 8.937771840519677e-06,
      "loss": 0.4937,
      "num_input_tokens_seen": 6423480,
      "step": 9845
    },
    {
      "epoch": 5.807783018867925,
      "grad_norm": 3.5938045978546143,
      "learning_rate": 8.936185896658593e-06,
      "loss": 0.548,
      "num_input_tokens_seen": 6426232,
      "step": 9850
    },
    {
      "epoch": 5.810731132075472,
      "grad_norm": 2.013221263885498,
      "learning_rate": 8.934598910675329e-06,
      "loss": 0.5019,
      "num_input_tokens_seen": 6429560,
      "step": 9855
    },
    {
      "epoch": 5.813679245283019,
      "grad_norm": 1.2644932270050049,
      "learning_rate": 8.933010882990044e-06,
      "loss": 0.4467,
      "num_input_tokens_seen": 6433368,
      "step": 9860
    },
    {
      "epoch": 5.816627358490566,
      "grad_norm": 1.586609125137329,
      "learning_rate": 8.93142181402318e-06,
      "loss": 0.4783,
      "num_input_tokens_seen": 6437048,
      "step": 9865
    },
    {
      "epoch": 5.819575471698113,
      "grad_norm": 1.7761478424072266,
      "learning_rate": 8.929831704195445e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 6439992,
      "step": 9870
    },
    {
      "epoch": 5.82252358490566,
      "grad_norm": 2.919567346572876,
      "learning_rate": 8.928240553927831e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 6445048,
      "step": 9875
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 2.6349339485168457,
      "learning_rate": 8.926648363641602e-06,
      "loss": 0.5456,
      "num_input_tokens_seen": 6447608,
      "step": 9880
    },
    {
      "epoch": 5.828419811320755,
      "grad_norm": 1.6749918460845947,
      "learning_rate": 8.925055133758294e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 6451992,
      "step": 9885
    },
    {
      "epoch": 5.831367924528302,
      "grad_norm": 2.646953821182251,
      "learning_rate": 8.923460864699723e-06,
      "loss": 0.4222,
      "num_input_tokens_seen": 6454904,
      "step": 9890
    },
    {
      "epoch": 5.834316037735849,
      "grad_norm": 2.3030388355255127,
      "learning_rate": 8.921865556887979e-06,
      "loss": 0.493,
      "num_input_tokens_seen": 6458136,
      "step": 9895
    },
    {
      "epoch": 5.837264150943396,
      "grad_norm": 1.3023337125778198,
      "learning_rate": 8.920269210745426e-06,
      "loss": 0.5337,
      "num_input_tokens_seen": 6461592,
      "step": 9900
    },
    {
      "epoch": 5.840212264150943,
      "grad_norm": 1.3240606784820557,
      "learning_rate": 8.918671826694704e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 6464984,
      "step": 9905
    },
    {
      "epoch": 5.84316037735849,
      "grad_norm": 3.571124315261841,
      "learning_rate": 8.91707340515873e-06,
      "loss": 0.4903,
      "num_input_tokens_seen": 6467544,
      "step": 9910
    },
    {
      "epoch": 5.846108490566038,
      "grad_norm": 1.3651524782180786,
      "learning_rate": 8.915473946560688e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 6470168,
      "step": 9915
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 1.433719515800476,
      "learning_rate": 8.913873451324044e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 6473880,
      "step": 9920
    },
    {
      "epoch": 5.852004716981132,
      "grad_norm": 1.5205668210983276,
      "learning_rate": 8.912271919872538e-06,
      "loss": 0.3487,
      "num_input_tokens_seen": 6476952,
      "step": 9925
    },
    {
      "epoch": 5.8549528301886795,
      "grad_norm": 2.1003284454345703,
      "learning_rate": 8.910669352630176e-06,
      "loss": 0.6355,
      "num_input_tokens_seen": 6479736,
      "step": 9930
    },
    {
      "epoch": 5.857900943396227,
      "grad_norm": 2.1194207668304443,
      "learning_rate": 8.909065750021253e-06,
      "loss": 0.6229,
      "num_input_tokens_seen": 6482936,
      "step": 9935
    },
    {
      "epoch": 5.860849056603773,
      "grad_norm": 1.418034315109253,
      "learning_rate": 8.907461112470323e-06,
      "loss": 0.609,
      "num_input_tokens_seen": 6485848,
      "step": 9940
    },
    {
      "epoch": 5.8637971698113205,
      "grad_norm": 2.073967933654785,
      "learning_rate": 8.905855440402225e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 6489496,
      "step": 9945
    },
    {
      "epoch": 5.866745283018868,
      "grad_norm": 2.0763418674468994,
      "learning_rate": 8.904248734242065e-06,
      "loss": 0.4118,
      "num_input_tokens_seen": 6493080,
      "step": 9950
    },
    {
      "epoch": 5.869693396226415,
      "grad_norm": 0.9696897864341736,
      "learning_rate": 8.902640994415226e-06,
      "loss": 0.4012,
      "num_input_tokens_seen": 6496408,
      "step": 9955
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 2.6535019874572754,
      "learning_rate": 8.901032221347364e-06,
      "loss": 0.374,
      "num_input_tokens_seen": 6499576,
      "step": 9960
    },
    {
      "epoch": 5.87558962264151,
      "grad_norm": 1.2540382146835327,
      "learning_rate": 8.899422415464409e-06,
      "loss": 0.4787,
      "num_input_tokens_seen": 6502808,
      "step": 9965
    },
    {
      "epoch": 5.878537735849057,
      "grad_norm": 1.477772831916809,
      "learning_rate": 8.897811577192565e-06,
      "loss": 0.3635,
      "num_input_tokens_seen": 6506296,
      "step": 9970
    },
    {
      "epoch": 5.881485849056604,
      "grad_norm": 1.4254497289657593,
      "learning_rate": 8.896199706958306e-06,
      "loss": 0.4696,
      "num_input_tokens_seen": 6509176,
      "step": 9975
    },
    {
      "epoch": 5.884433962264151,
      "grad_norm": 4.718723773956299,
      "learning_rate": 8.894586805188384e-06,
      "loss": 0.3674,
      "num_input_tokens_seen": 6511832,
      "step": 9980
    },
    {
      "epoch": 5.887382075471698,
      "grad_norm": 1.0747472047805786,
      "learning_rate": 8.892972872309821e-06,
      "loss": 0.4465,
      "num_input_tokens_seen": 6514712,
      "step": 9985
    },
    {
      "epoch": 5.890330188679245,
      "grad_norm": 1.2221148014068604,
      "learning_rate": 8.89135790874991e-06,
      "loss": 0.4855,
      "num_input_tokens_seen": 6518456,
      "step": 9990
    },
    {
      "epoch": 5.8932783018867925,
      "grad_norm": 3.9369680881500244,
      "learning_rate": 8.889741914936224e-06,
      "loss": 0.5032,
      "num_input_tokens_seen": 6521528,
      "step": 9995
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 7.876033782958984,
      "learning_rate": 8.888124891296602e-06,
      "loss": 0.3246,
      "num_input_tokens_seen": 6524696,
      "step": 10000
    },
    {
      "epoch": 5.899174528301887,
      "grad_norm": 1.2290730476379395,
      "learning_rate": 8.886506838259156e-06,
      "loss": 0.3844,
      "num_input_tokens_seen": 6527672,
      "step": 10005
    },
    {
      "epoch": 5.902122641509434,
      "grad_norm": 1.317514419555664,
      "learning_rate": 8.884887756252279e-06,
      "loss": 0.4958,
      "num_input_tokens_seen": 6534968,
      "step": 10010
    },
    {
      "epoch": 5.905070754716981,
      "grad_norm": 1.6785964965820312,
      "learning_rate": 8.88326764570462e-06,
      "loss": 0.4356,
      "num_input_tokens_seen": 6539032,
      "step": 10015
    },
    {
      "epoch": 5.908018867924528,
      "grad_norm": 1.6879326105117798,
      "learning_rate": 8.88164650704512e-06,
      "loss": 0.5145,
      "num_input_tokens_seen": 6542904,
      "step": 10020
    },
    {
      "epoch": 5.910966981132075,
      "grad_norm": 1.6137535572052002,
      "learning_rate": 8.880024340702978e-06,
      "loss": 0.4273,
      "num_input_tokens_seen": 6545720,
      "step": 10025
    },
    {
      "epoch": 5.913915094339623,
      "grad_norm": 1.2747035026550293,
      "learning_rate": 8.878401147107667e-06,
      "loss": 0.4787,
      "num_input_tokens_seen": 6549080,
      "step": 10030
    },
    {
      "epoch": 5.91686320754717,
      "grad_norm": 2.492171287536621,
      "learning_rate": 8.87677692668894e-06,
      "loss": 0.4707,
      "num_input_tokens_seen": 6552248,
      "step": 10035
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 2.699409008026123,
      "learning_rate": 8.875151679876813e-06,
      "loss": 0.5262,
      "num_input_tokens_seen": 6554360,
      "step": 10040
    },
    {
      "epoch": 5.9227594339622645,
      "grad_norm": 1.7275655269622803,
      "learning_rate": 8.873525407101577e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 6556792,
      "step": 10045
    },
    {
      "epoch": 5.925707547169811,
      "grad_norm": 1.0764330625534058,
      "learning_rate": 8.871898108793796e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 6559704,
      "step": 10050
    },
    {
      "epoch": 5.928655660377358,
      "grad_norm": 1.58961820602417,
      "learning_rate": 8.870269785384304e-06,
      "loss": 0.4433,
      "num_input_tokens_seen": 6562872,
      "step": 10055
    },
    {
      "epoch": 5.931603773584905,
      "grad_norm": 1.218278408050537,
      "learning_rate": 8.868640437304206e-06,
      "loss": 0.432,
      "num_input_tokens_seen": 6566552,
      "step": 10060
    },
    {
      "epoch": 5.934551886792453,
      "grad_norm": 1.1819274425506592,
      "learning_rate": 8.86701006498488e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 6570072,
      "step": 10065
    },
    {
      "epoch": 5.9375,
      "grad_norm": 1.6576875448226929,
      "learning_rate": 8.865378668857972e-06,
      "loss": 0.5256,
      "num_input_tokens_seen": 6572952,
      "step": 10070
    },
    {
      "epoch": 5.940448113207547,
      "grad_norm": 3.084968090057373,
      "learning_rate": 8.863746249355404e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 6576056,
      "step": 10075
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 2.5933618545532227,
      "learning_rate": 8.862112806909365e-06,
      "loss": 0.4917,
      "num_input_tokens_seen": 6579544,
      "step": 10080
    },
    {
      "epoch": 5.946344339622642,
      "grad_norm": 2.375326633453369,
      "learning_rate": 8.860478341952314e-06,
      "loss": 0.4647,
      "num_input_tokens_seen": 6583224,
      "step": 10085
    },
    {
      "epoch": 5.949292452830189,
      "grad_norm": 2.8612711429595947,
      "learning_rate": 8.858842854916985e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 6586360,
      "step": 10090
    },
    {
      "epoch": 5.9522405660377355,
      "grad_norm": 1.7050399780273438,
      "learning_rate": 8.85720634623638e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 6589336,
      "step": 10095
    },
    {
      "epoch": 5.955188679245283,
      "grad_norm": 2.692549705505371,
      "learning_rate": 8.855568816343769e-06,
      "loss": 0.5095,
      "num_input_tokens_seen": 6592056,
      "step": 10100
    },
    {
      "epoch": 5.95813679245283,
      "grad_norm": 1.3385995626449585,
      "learning_rate": 8.8539302656727e-06,
      "loss": 0.4419,
      "num_input_tokens_seen": 6595448,
      "step": 10105
    },
    {
      "epoch": 5.961084905660377,
      "grad_norm": 1.5005338191986084,
      "learning_rate": 8.852290694656983e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 6599352,
      "step": 10110
    },
    {
      "epoch": 5.964033018867925,
      "grad_norm": 2.403186082839966,
      "learning_rate": 8.8506501037307e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 6606392,
      "step": 10115
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 1.3847852945327759,
      "learning_rate": 8.849008493328209e-06,
      "loss": 0.5353,
      "num_input_tokens_seen": 6609624,
      "step": 10120
    },
    {
      "epoch": 5.969929245283019,
      "grad_norm": 2.0584378242492676,
      "learning_rate": 8.847365863884131e-06,
      "loss": 0.436,
      "num_input_tokens_seen": 6612760,
      "step": 10125
    },
    {
      "epoch": 5.972877358490566,
      "grad_norm": 2.2479701042175293,
      "learning_rate": 8.845722215833359e-06,
      "loss": 0.4758,
      "num_input_tokens_seen": 6616408,
      "step": 10130
    },
    {
      "epoch": 5.975825471698113,
      "grad_norm": 1.2362128496170044,
      "learning_rate": 8.844077549611056e-06,
      "loss": 0.3583,
      "num_input_tokens_seen": 6619480,
      "step": 10135
    },
    {
      "epoch": 5.97877358490566,
      "grad_norm": 4.061662197113037,
      "learning_rate": 8.842431865652654e-06,
      "loss": 0.694,
      "num_input_tokens_seen": 6622040,
      "step": 10140
    },
    {
      "epoch": 5.9817216981132075,
      "grad_norm": 1.1201255321502686,
      "learning_rate": 8.840785164393858e-06,
      "loss": 0.3268,
      "num_input_tokens_seen": 6625432,
      "step": 10145
    },
    {
      "epoch": 5.984669811320755,
      "grad_norm": 4.726824760437012,
      "learning_rate": 8.839137446270634e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 6629080,
      "step": 10150
    },
    {
      "epoch": 5.987617924528302,
      "grad_norm": 2.7790212631225586,
      "learning_rate": 8.837488711719226e-06,
      "loss": 0.6432,
      "num_input_tokens_seen": 6631896,
      "step": 10155
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 2.8213279247283936,
      "learning_rate": 8.835838961176143e-06,
      "loss": 0.5769,
      "num_input_tokens_seen": 6634584,
      "step": 10160
    },
    {
      "epoch": 5.993514150943396,
      "grad_norm": 2.135136842727661,
      "learning_rate": 8.834188195078164e-06,
      "loss": 0.4742,
      "num_input_tokens_seen": 6637464,
      "step": 10165
    },
    {
      "epoch": 5.996462264150943,
      "grad_norm": 0.9019614458084106,
      "learning_rate": 8.832536413862337e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 6640280,
      "step": 10170
    },
    {
      "epoch": 5.99941037735849,
      "grad_norm": 1.9355944395065308,
      "learning_rate": 8.830883617965976e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 6643704,
      "step": 10175
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.48383763432502747,
      "eval_runtime": 19.0536,
      "eval_samples_per_second": 89.012,
      "eval_steps_per_second": 22.253,
      "num_input_tokens_seen": 6643816,
      "step": 10176
    },
    {
      "epoch": 6.002358490566038,
      "grad_norm": 1.374747395515442,
      "learning_rate": 8.829229807826665e-06,
      "loss": 0.4788,
      "num_input_tokens_seen": 6646472,
      "step": 10180
    },
    {
      "epoch": 6.005306603773585,
      "grad_norm": 1.8040485382080078,
      "learning_rate": 8.82757498388226e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 6649256,
      "step": 10185
    },
    {
      "epoch": 6.008254716981132,
      "grad_norm": 3.2622287273406982,
      "learning_rate": 8.825919146570884e-06,
      "loss": 0.445,
      "num_input_tokens_seen": 6653160,
      "step": 10190
    },
    {
      "epoch": 6.0112028301886795,
      "grad_norm": 3.609764814376831,
      "learning_rate": 8.824262296330925e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 6656808,
      "step": 10195
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 1.1366406679153442,
      "learning_rate": 8.822604433601041e-06,
      "loss": 0.326,
      "num_input_tokens_seen": 6659528,
      "step": 10200
    },
    {
      "epoch": 6.017099056603773,
      "grad_norm": 1.7176588773727417,
      "learning_rate": 8.820945558820158e-06,
      "loss": 0.5696,
      "num_input_tokens_seen": 6663400,
      "step": 10205
    },
    {
      "epoch": 6.0200471698113205,
      "grad_norm": 2.0193300247192383,
      "learning_rate": 8.81928567242747e-06,
      "loss": 0.3481,
      "num_input_tokens_seen": 6666856,
      "step": 10210
    },
    {
      "epoch": 6.022995283018868,
      "grad_norm": 2.3732550144195557,
      "learning_rate": 8.817624774862443e-06,
      "loss": 0.269,
      "num_input_tokens_seen": 6669864,
      "step": 10215
    },
    {
      "epoch": 6.025943396226415,
      "grad_norm": 1.108096957206726,
      "learning_rate": 8.815962866564803e-06,
      "loss": 0.4209,
      "num_input_tokens_seen": 6674664,
      "step": 10220
    },
    {
      "epoch": 6.028891509433962,
      "grad_norm": 3.712996006011963,
      "learning_rate": 8.814299947974547e-06,
      "loss": 0.406,
      "num_input_tokens_seen": 6677320,
      "step": 10225
    },
    {
      "epoch": 6.03183962264151,
      "grad_norm": 1.2370606660842896,
      "learning_rate": 8.812636019531942e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 6679976,
      "step": 10230
    },
    {
      "epoch": 6.034787735849057,
      "grad_norm": 6.240902900695801,
      "learning_rate": 8.810971081677517e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 6683816,
      "step": 10235
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 2.2323808670043945,
      "learning_rate": 8.809305134852076e-06,
      "loss": 0.5079,
      "num_input_tokens_seen": 6686984,
      "step": 10240
    },
    {
      "epoch": 6.040683962264151,
      "grad_norm": 1.2000513076782227,
      "learning_rate": 8.807638179496684e-06,
      "loss": 0.4176,
      "num_input_tokens_seen": 6689800,
      "step": 10245
    },
    {
      "epoch": 6.043632075471698,
      "grad_norm": 6.4935688972473145,
      "learning_rate": 8.805970216052673e-06,
      "loss": 0.4327,
      "num_input_tokens_seen": 6692744,
      "step": 10250
    },
    {
      "epoch": 6.046580188679245,
      "grad_norm": 1.6586312055587769,
      "learning_rate": 8.804301244961645e-06,
      "loss": 0.4296,
      "num_input_tokens_seen": 6695912,
      "step": 10255
    },
    {
      "epoch": 6.0495283018867925,
      "grad_norm": 3.49442982673645,
      "learning_rate": 8.802631266665465e-06,
      "loss": 0.4165,
      "num_input_tokens_seen": 6699432,
      "step": 10260
    },
    {
      "epoch": 6.05247641509434,
      "grad_norm": 2.2777843475341797,
      "learning_rate": 8.80096028160627e-06,
      "loss": 0.3891,
      "num_input_tokens_seen": 6702056,
      "step": 10265
    },
    {
      "epoch": 6.055424528301887,
      "grad_norm": 2.6969666481018066,
      "learning_rate": 8.799288290226457e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 6706024,
      "step": 10270
    },
    {
      "epoch": 6.058372641509434,
      "grad_norm": 1.5798810720443726,
      "learning_rate": 8.797615292968698e-06,
      "loss": 0.4024,
      "num_input_tokens_seen": 6709288,
      "step": 10275
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 0.9074104428291321,
      "learning_rate": 8.795941290275923e-06,
      "loss": 0.3171,
      "num_input_tokens_seen": 6713096,
      "step": 10280
    },
    {
      "epoch": 6.064268867924528,
      "grad_norm": 1.533441424369812,
      "learning_rate": 8.79426628259133e-06,
      "loss": 0.4656,
      "num_input_tokens_seen": 6715624,
      "step": 10285
    },
    {
      "epoch": 6.067216981132075,
      "grad_norm": 1.947579264640808,
      "learning_rate": 8.792590270358389e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 6718824,
      "step": 10290
    },
    {
      "epoch": 6.070165094339623,
      "grad_norm": 1.7584617137908936,
      "learning_rate": 8.790913254020827e-06,
      "loss": 0.4546,
      "num_input_tokens_seen": 6722472,
      "step": 10295
    },
    {
      "epoch": 6.07311320754717,
      "grad_norm": 1.5465511083602905,
      "learning_rate": 8.789235234022643e-06,
      "loss": 0.3033,
      "num_input_tokens_seen": 6728712,
      "step": 10300
    },
    {
      "epoch": 6.076061320754717,
      "grad_norm": 1.394791603088379,
      "learning_rate": 8.787556210808101e-06,
      "loss": 0.3218,
      "num_input_tokens_seen": 6731784,
      "step": 10305
    },
    {
      "epoch": 6.0790094339622645,
      "grad_norm": 1.5550930500030518,
      "learning_rate": 8.78587618482173e-06,
      "loss": 0.4114,
      "num_input_tokens_seen": 6734664,
      "step": 10310
    },
    {
      "epoch": 6.081957547169812,
      "grad_norm": 2.7411856651306152,
      "learning_rate": 8.78419515650832e-06,
      "loss": 0.4554,
      "num_input_tokens_seen": 6741384,
      "step": 10315
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 2.870938777923584,
      "learning_rate": 8.782513126312934e-06,
      "loss": 0.4963,
      "num_input_tokens_seen": 6745352,
      "step": 10320
    },
    {
      "epoch": 6.087853773584905,
      "grad_norm": 1.3814818859100342,
      "learning_rate": 8.780830094680897e-06,
      "loss": 0.4598,
      "num_input_tokens_seen": 6749160,
      "step": 10325
    },
    {
      "epoch": 6.090801886792453,
      "grad_norm": 2.393259048461914,
      "learning_rate": 8.779146062057797e-06,
      "loss": 0.4889,
      "num_input_tokens_seen": 6753512,
      "step": 10330
    },
    {
      "epoch": 6.09375,
      "grad_norm": 1.9891831874847412,
      "learning_rate": 8.777461028889492e-06,
      "loss": 0.4521,
      "num_input_tokens_seen": 6757096,
      "step": 10335
    },
    {
      "epoch": 6.096698113207547,
      "grad_norm": 1.1162190437316895,
      "learning_rate": 8.775774995622097e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 6760840,
      "step": 10340
    },
    {
      "epoch": 6.099646226415095,
      "grad_norm": 2.406433343887329,
      "learning_rate": 8.774087962702e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 6763944,
      "step": 10345
    },
    {
      "epoch": 6.102594339622642,
      "grad_norm": 2.682136058807373,
      "learning_rate": 8.772399930575849e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 6767176,
      "step": 10350
    },
    {
      "epoch": 6.105542452830188,
      "grad_norm": 2.0697097778320312,
      "learning_rate": 8.77071089969056e-06,
      "loss": 0.4902,
      "num_input_tokens_seen": 6769768,
      "step": 10355
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 2.0213942527770996,
      "learning_rate": 8.769020870493309e-06,
      "loss": 0.3833,
      "num_input_tokens_seen": 6773736,
      "step": 10360
    },
    {
      "epoch": 6.111438679245283,
      "grad_norm": 1.6370025873184204,
      "learning_rate": 8.767329843431537e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 6776680,
      "step": 10365
    },
    {
      "epoch": 6.11438679245283,
      "grad_norm": 1.9768359661102295,
      "learning_rate": 8.765637818952954e-06,
      "loss": 0.452,
      "num_input_tokens_seen": 6780200,
      "step": 10370
    },
    {
      "epoch": 6.117334905660377,
      "grad_norm": 1.574218988418579,
      "learning_rate": 8.76394479750553e-06,
      "loss": 0.3225,
      "num_input_tokens_seen": 6783112,
      "step": 10375
    },
    {
      "epoch": 6.120283018867925,
      "grad_norm": 2.0369973182678223,
      "learning_rate": 8.762250779537499e-06,
      "loss": 0.5251,
      "num_input_tokens_seen": 6785832,
      "step": 10380
    },
    {
      "epoch": 6.123231132075472,
      "grad_norm": 2.184579372406006,
      "learning_rate": 8.760555765497358e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 6788232,
      "step": 10385
    },
    {
      "epoch": 6.126179245283019,
      "grad_norm": 2.9179022312164307,
      "learning_rate": 8.758859755833873e-06,
      "loss": 0.4959,
      "num_input_tokens_seen": 6790952,
      "step": 10390
    },
    {
      "epoch": 6.129127358490566,
      "grad_norm": 1.4735593795776367,
      "learning_rate": 8.757162750996066e-06,
      "loss": 0.5326,
      "num_input_tokens_seen": 6794920,
      "step": 10395
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 2.0168797969818115,
      "learning_rate": 8.755464751433229e-06,
      "loss": 0.5761,
      "num_input_tokens_seen": 6798088,
      "step": 10400
    },
    {
      "epoch": 6.13502358490566,
      "grad_norm": 3.3761091232299805,
      "learning_rate": 8.753765757594915e-06,
      "loss": 0.5356,
      "num_input_tokens_seen": 6800744,
      "step": 10405
    },
    {
      "epoch": 6.1379716981132075,
      "grad_norm": 1.492551565170288,
      "learning_rate": 8.752065769930938e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 6804008,
      "step": 10410
    },
    {
      "epoch": 6.140919811320755,
      "grad_norm": 2.037916898727417,
      "learning_rate": 8.75036478889138e-06,
      "loss": 0.4065,
      "num_input_tokens_seen": 6807144,
      "step": 10415
    },
    {
      "epoch": 6.143867924528302,
      "grad_norm": 2.7841808795928955,
      "learning_rate": 8.748662814926576e-06,
      "loss": 0.4893,
      "num_input_tokens_seen": 6814408,
      "step": 10420
    },
    {
      "epoch": 6.146816037735849,
      "grad_norm": 1.5850087404251099,
      "learning_rate": 8.746959848487139e-06,
      "loss": 0.578,
      "num_input_tokens_seen": 6817992,
      "step": 10425
    },
    {
      "epoch": 6.149764150943396,
      "grad_norm": 1.3914048671722412,
      "learning_rate": 8.745255890023934e-06,
      "loss": 0.4797,
      "num_input_tokens_seen": 6820712,
      "step": 10430
    },
    {
      "epoch": 6.152712264150943,
      "grad_norm": 1.09807288646698,
      "learning_rate": 8.74355093998809e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 6824840,
      "step": 10435
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 2.9785921573638916,
      "learning_rate": 8.741844998831001e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 6827976,
      "step": 10440
    },
    {
      "epoch": 6.158608490566038,
      "grad_norm": 1.360189437866211,
      "learning_rate": 8.740138067004323e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 6830824,
      "step": 10445
    },
    {
      "epoch": 6.161556603773585,
      "grad_norm": 1.6515066623687744,
      "learning_rate": 8.738430144959973e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 6834472,
      "step": 10450
    },
    {
      "epoch": 6.164504716981132,
      "grad_norm": 2.239821195602417,
      "learning_rate": 8.73672123315013e-06,
      "loss": 0.3605,
      "num_input_tokens_seen": 6838216,
      "step": 10455
    },
    {
      "epoch": 6.1674528301886795,
      "grad_norm": 1.5015143156051636,
      "learning_rate": 8.735011332027234e-06,
      "loss": 0.3801,
      "num_input_tokens_seen": 6841096,
      "step": 10460
    },
    {
      "epoch": 6.170400943396227,
      "grad_norm": 2.324664831161499,
      "learning_rate": 8.733300442043993e-06,
      "loss": 0.448,
      "num_input_tokens_seen": 6844808,
      "step": 10465
    },
    {
      "epoch": 6.173349056603773,
      "grad_norm": 2.6038501262664795,
      "learning_rate": 8.73158856365337e-06,
      "loss": 0.4253,
      "num_input_tokens_seen": 6847464,
      "step": 10470
    },
    {
      "epoch": 6.1762971698113205,
      "grad_norm": 2.6956946849823,
      "learning_rate": 8.729875697308592e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 6850088,
      "step": 10475
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 1.9941338300704956,
      "learning_rate": 8.728161843463148e-06,
      "loss": 0.526,
      "num_input_tokens_seen": 6853000,
      "step": 10480
    },
    {
      "epoch": 6.182193396226415,
      "grad_norm": 1.3956106901168823,
      "learning_rate": 8.726447002570791e-06,
      "loss": 0.4867,
      "num_input_tokens_seen": 6856616,
      "step": 10485
    },
    {
      "epoch": 6.185141509433962,
      "grad_norm": 3.318361282348633,
      "learning_rate": 8.724731175085526e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 6859816,
      "step": 10490
    },
    {
      "epoch": 6.18808962264151,
      "grad_norm": 2.103188991546631,
      "learning_rate": 8.723014361461633e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 6862600,
      "step": 10495
    },
    {
      "epoch": 6.191037735849057,
      "grad_norm": 1.5342081785202026,
      "learning_rate": 8.72129656215364e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 6866632,
      "step": 10500
    },
    {
      "epoch": 6.193985849056604,
      "grad_norm": 2.9051411151885986,
      "learning_rate": 8.719577777616347e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 6868968,
      "step": 10505
    },
    {
      "epoch": 6.196933962264151,
      "grad_norm": 1.1959784030914307,
      "learning_rate": 8.717858008304804e-06,
      "loss": 0.3746,
      "num_input_tokens_seen": 6872680,
      "step": 10510
    },
    {
      "epoch": 6.199882075471698,
      "grad_norm": 3.4626851081848145,
      "learning_rate": 8.71613725467433e-06,
      "loss": 0.6696,
      "num_input_tokens_seen": 6875528,
      "step": 10515
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 1.4649503231048584,
      "learning_rate": 8.714415517180506e-06,
      "loss": 0.6105,
      "num_input_tokens_seen": 6878472,
      "step": 10520
    },
    {
      "epoch": 6.2057783018867925,
      "grad_norm": 2.2525105476379395,
      "learning_rate": 8.712692796279164e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 6881032,
      "step": 10525
    },
    {
      "epoch": 6.20872641509434,
      "grad_norm": 2.3236258029937744,
      "learning_rate": 8.710969092426401e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 6884456,
      "step": 10530
    },
    {
      "epoch": 6.211674528301887,
      "grad_norm": 1.425258994102478,
      "learning_rate": 8.70924440607858e-06,
      "loss": 0.5364,
      "num_input_tokens_seen": 6887880,
      "step": 10535
    },
    {
      "epoch": 6.214622641509434,
      "grad_norm": 1.0380324125289917,
      "learning_rate": 8.707518737692315e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 6890440,
      "step": 10540
    },
    {
      "epoch": 6.217570754716981,
      "grad_norm": 2.0063095092773438,
      "learning_rate": 8.705792087724485e-06,
      "loss": 0.5842,
      "num_input_tokens_seen": 6893672,
      "step": 10545
    },
    {
      "epoch": 6.220518867924528,
      "grad_norm": 2.2466189861297607,
      "learning_rate": 8.704064456632231e-06,
      "loss": 0.466,
      "num_input_tokens_seen": 6896712,
      "step": 10550
    },
    {
      "epoch": 6.223466981132075,
      "grad_norm": 2.520476818084717,
      "learning_rate": 8.702335844872946e-06,
      "loss": 0.5148,
      "num_input_tokens_seen": 6900296,
      "step": 10555
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 1.4400898218154907,
      "learning_rate": 8.700606252904293e-06,
      "loss": 0.4221,
      "num_input_tokens_seen": 6904296,
      "step": 10560
    },
    {
      "epoch": 6.22936320754717,
      "grad_norm": 1.5614081621170044,
      "learning_rate": 8.698875681184183e-06,
      "loss": 0.398,
      "num_input_tokens_seen": 6907752,
      "step": 10565
    },
    {
      "epoch": 6.232311320754717,
      "grad_norm": 2.1250522136688232,
      "learning_rate": 8.697144130170797e-06,
      "loss": 0.3325,
      "num_input_tokens_seen": 6910248,
      "step": 10570
    },
    {
      "epoch": 6.2352594339622645,
      "grad_norm": 1.0618118047714233,
      "learning_rate": 8.695411600322568e-06,
      "loss": 0.4518,
      "num_input_tokens_seen": 6916360,
      "step": 10575
    },
    {
      "epoch": 6.238207547169812,
      "grad_norm": 1.5392694473266602,
      "learning_rate": 8.693678092098191e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 6919720,
      "step": 10580
    },
    {
      "epoch": 6.241155660377358,
      "grad_norm": 1.8455537557601929,
      "learning_rate": 8.691943605956621e-06,
      "loss": 0.5344,
      "num_input_tokens_seen": 6923304,
      "step": 10585
    },
    {
      "epoch": 6.244103773584905,
      "grad_norm": 1.991140604019165,
      "learning_rate": 8.690208142357069e-06,
      "loss": 0.4883,
      "num_input_tokens_seen": 6926184,
      "step": 10590
    },
    {
      "epoch": 6.247051886792453,
      "grad_norm": 1.527840495109558,
      "learning_rate": 8.68847170175901e-06,
      "loss": 0.4598,
      "num_input_tokens_seen": 6929512,
      "step": 10595
    },
    {
      "epoch": 6.25,
      "grad_norm": 4.950246334075928,
      "learning_rate": 8.686734284622168e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 6933032,
      "step": 10600
    },
    {
      "epoch": 6.252948113207547,
      "grad_norm": 1.8747286796569824,
      "learning_rate": 8.684995891406537e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 6935560,
      "step": 10605
    },
    {
      "epoch": 6.255896226415095,
      "grad_norm": 2.4182589054107666,
      "learning_rate": 8.683256522572362e-06,
      "loss": 0.5317,
      "num_input_tokens_seen": 6938728,
      "step": 10610
    },
    {
      "epoch": 6.258844339622642,
      "grad_norm": 3.788961887359619,
      "learning_rate": 8.68151617858015e-06,
      "loss": 0.5258,
      "num_input_tokens_seen": 6940936,
      "step": 10615
    },
    {
      "epoch": 6.261792452830189,
      "grad_norm": 1.3752750158309937,
      "learning_rate": 8.67977485989066e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 6945064,
      "step": 10620
    },
    {
      "epoch": 6.2647405660377355,
      "grad_norm": 1.4527148008346558,
      "learning_rate": 8.67803256696492e-06,
      "loss": 0.3833,
      "num_input_tokens_seen": 6947976,
      "step": 10625
    },
    {
      "epoch": 6.267688679245283,
      "grad_norm": 1.587907314300537,
      "learning_rate": 8.676289300264205e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 6950760,
      "step": 10630
    },
    {
      "epoch": 6.27063679245283,
      "grad_norm": 1.4983513355255127,
      "learning_rate": 8.674545060250054e-06,
      "loss": 0.4375,
      "num_input_tokens_seen": 6954152,
      "step": 10635
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 1.6739470958709717,
      "learning_rate": 8.672799847384263e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 6957256,
      "step": 10640
    },
    {
      "epoch": 6.276533018867925,
      "grad_norm": 2.9461820125579834,
      "learning_rate": 8.671053662128883e-06,
      "loss": 0.3315,
      "num_input_tokens_seen": 6960808,
      "step": 10645
    },
    {
      "epoch": 6.279481132075472,
      "grad_norm": 1.720680832862854,
      "learning_rate": 8.669306504946223e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 6963848,
      "step": 10650
    },
    {
      "epoch": 6.282429245283019,
      "grad_norm": 1.2292002439498901,
      "learning_rate": 8.667558376298854e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 6966824,
      "step": 10655
    },
    {
      "epoch": 6.285377358490566,
      "grad_norm": 1.258919358253479,
      "learning_rate": 8.665809276649597e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 6969832,
      "step": 10660
    },
    {
      "epoch": 6.288325471698113,
      "grad_norm": 1.657394289970398,
      "learning_rate": 8.664059206461537e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 6973256,
      "step": 10665
    },
    {
      "epoch": 6.29127358490566,
      "grad_norm": 1.2802711725234985,
      "learning_rate": 8.662308166198009e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 6976648,
      "step": 10670
    },
    {
      "epoch": 6.2942216981132075,
      "grad_norm": 1.6997356414794922,
      "learning_rate": 8.660556156322611e-06,
      "loss": 0.4959,
      "num_input_tokens_seen": 6980552,
      "step": 10675
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 1.895542025566101,
      "learning_rate": 8.658803177299196e-06,
      "loss": 0.6417,
      "num_input_tokens_seen": 6983176,
      "step": 10680
    },
    {
      "epoch": 6.300117924528302,
      "grad_norm": 1.4546973705291748,
      "learning_rate": 8.65704922959187e-06,
      "loss": 0.4539,
      "num_input_tokens_seen": 6986728,
      "step": 10685
    },
    {
      "epoch": 6.303066037735849,
      "grad_norm": 1.8358277082443237,
      "learning_rate": 8.655294313664998e-06,
      "loss": 0.3323,
      "num_input_tokens_seen": 6989480,
      "step": 10690
    },
    {
      "epoch": 6.306014150943396,
      "grad_norm": 2.0670907497406006,
      "learning_rate": 8.653538429983204e-06,
      "loss": 0.4083,
      "num_input_tokens_seen": 6992264,
      "step": 10695
    },
    {
      "epoch": 6.308962264150943,
      "grad_norm": 1.8590391874313354,
      "learning_rate": 8.651781579011366e-06,
      "loss": 0.5907,
      "num_input_tokens_seen": 6995144,
      "step": 10700
    },
    {
      "epoch": 6.31191037735849,
      "grad_norm": 2.4423580169677734,
      "learning_rate": 8.650023761214615e-06,
      "loss": 0.4247,
      "num_input_tokens_seen": 6997512,
      "step": 10705
    },
    {
      "epoch": 6.314858490566038,
      "grad_norm": 1.1746540069580078,
      "learning_rate": 8.648264977058344e-06,
      "loss": 0.3825,
      "num_input_tokens_seen": 7001544,
      "step": 10710
    },
    {
      "epoch": 6.317806603773585,
      "grad_norm": 1.7019646167755127,
      "learning_rate": 8.646505227008197e-06,
      "loss": 0.4579,
      "num_input_tokens_seen": 7005288,
      "step": 10715
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 2.3581931591033936,
      "learning_rate": 8.644744511530074e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 7008232,
      "step": 10720
    },
    {
      "epoch": 6.3237028301886795,
      "grad_norm": 2.824728488922119,
      "learning_rate": 8.642982831090135e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 7011080,
      "step": 10725
    },
    {
      "epoch": 6.326650943396227,
      "grad_norm": 1.6422390937805176,
      "learning_rate": 8.64122018615479e-06,
      "loss": 0.3994,
      "num_input_tokens_seen": 7014888,
      "step": 10730
    },
    {
      "epoch": 6.329599056603773,
      "grad_norm": 2.2509095668792725,
      "learning_rate": 8.63945657719071e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 7018376,
      "step": 10735
    },
    {
      "epoch": 6.3325471698113205,
      "grad_norm": 1.9439877271652222,
      "learning_rate": 8.637692004664816e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 7021448,
      "step": 10740
    },
    {
      "epoch": 6.335495283018868,
      "grad_norm": 1.754699945449829,
      "learning_rate": 8.635926469044284e-06,
      "loss": 0.4071,
      "num_input_tokens_seen": 7024488,
      "step": 10745
    },
    {
      "epoch": 6.338443396226415,
      "grad_norm": 1.814724087715149,
      "learning_rate": 8.63415997079655e-06,
      "loss": 0.3854,
      "num_input_tokens_seen": 7028840,
      "step": 10750
    },
    {
      "epoch": 6.341391509433962,
      "grad_norm": 2.249955177307129,
      "learning_rate": 8.6323925103893e-06,
      "loss": 0.4785,
      "num_input_tokens_seen": 7033704,
      "step": 10755
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 1.227712631225586,
      "learning_rate": 8.63062408829048e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 7037384,
      "step": 10760
    },
    {
      "epoch": 6.347287735849057,
      "grad_norm": 2.8285059928894043,
      "learning_rate": 8.628854704968285e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 7039880,
      "step": 10765
    },
    {
      "epoch": 6.350235849056604,
      "grad_norm": 1.8627907037734985,
      "learning_rate": 8.62708436089117e-06,
      "loss": 0.3938,
      "num_input_tokens_seen": 7042856,
      "step": 10770
    },
    {
      "epoch": 6.353183962264151,
      "grad_norm": 2.45833158493042,
      "learning_rate": 8.625313056527836e-06,
      "loss": 0.5834,
      "num_input_tokens_seen": 7045448,
      "step": 10775
    },
    {
      "epoch": 6.356132075471698,
      "grad_norm": 3.6617274284362793,
      "learning_rate": 8.623540792347244e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 7048776,
      "step": 10780
    },
    {
      "epoch": 6.359080188679245,
      "grad_norm": 6.544510841369629,
      "learning_rate": 8.621767568818614e-06,
      "loss": 0.5189,
      "num_input_tokens_seen": 7051656,
      "step": 10785
    },
    {
      "epoch": 6.3620283018867925,
      "grad_norm": 1.32256281375885,
      "learning_rate": 8.619993386411409e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 7053960,
      "step": 10790
    },
    {
      "epoch": 6.36497641509434,
      "grad_norm": 2.428201198577881,
      "learning_rate": 8.618218245595356e-06,
      "loss": 0.4669,
      "num_input_tokens_seen": 7056680,
      "step": 10795
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 3.411282777786255,
      "learning_rate": 8.616442146840427e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 7059432,
      "step": 10800
    },
    {
      "epoch": 6.370872641509434,
      "grad_norm": 1.8137084245681763,
      "learning_rate": 8.614665090616854e-06,
      "loss": 0.4099,
      "num_input_tokens_seen": 7062600,
      "step": 10805
    },
    {
      "epoch": 6.373820754716981,
      "grad_norm": 1.8732653856277466,
      "learning_rate": 8.61288707739512e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 7065832,
      "step": 10810
    },
    {
      "epoch": 6.376768867924528,
      "grad_norm": 1.4493833780288696,
      "learning_rate": 8.611108107645963e-06,
      "loss": 0.4861,
      "num_input_tokens_seen": 7068552,
      "step": 10815
    },
    {
      "epoch": 6.379716981132075,
      "grad_norm": 2.4043314456939697,
      "learning_rate": 8.609328181840368e-06,
      "loss": 0.5575,
      "num_input_tokens_seen": 7072552,
      "step": 10820
    },
    {
      "epoch": 6.382665094339623,
      "grad_norm": 2.9582250118255615,
      "learning_rate": 8.607547300449585e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 7075816,
      "step": 10825
    },
    {
      "epoch": 6.38561320754717,
      "grad_norm": 2.018313407897949,
      "learning_rate": 8.605765463945105e-06,
      "loss": 0.5315,
      "num_input_tokens_seen": 7079336,
      "step": 10830
    },
    {
      "epoch": 6.388561320754717,
      "grad_norm": 1.5002448558807373,
      "learning_rate": 8.603982672798678e-06,
      "loss": 0.4937,
      "num_input_tokens_seen": 7082824,
      "step": 10835
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 1.7112776041030884,
      "learning_rate": 8.602198927482309e-06,
      "loss": 0.3787,
      "num_input_tokens_seen": 7086088,
      "step": 10840
    },
    {
      "epoch": 6.394457547169811,
      "grad_norm": 4.001970291137695,
      "learning_rate": 8.600414228468245e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 7089320,
      "step": 10845
    },
    {
      "epoch": 6.397405660377358,
      "grad_norm": 1.8440855741500854,
      "learning_rate": 8.598628576229e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 7092424,
      "step": 10850
    },
    {
      "epoch": 6.400353773584905,
      "grad_norm": 1.7316794395446777,
      "learning_rate": 8.596841971237328e-06,
      "loss": 0.4365,
      "num_input_tokens_seen": 7095432,
      "step": 10855
    },
    {
      "epoch": 6.403301886792453,
      "grad_norm": 2.0240917205810547,
      "learning_rate": 8.595054413966246e-06,
      "loss": 0.5514,
      "num_input_tokens_seen": 7098632,
      "step": 10860
    },
    {
      "epoch": 6.40625,
      "grad_norm": 1.167580485343933,
      "learning_rate": 8.593265904889011e-06,
      "loss": 0.4741,
      "num_input_tokens_seen": 7101960,
      "step": 10865
    },
    {
      "epoch": 6.409198113207547,
      "grad_norm": 1.0853410959243774,
      "learning_rate": 8.591476444479141e-06,
      "loss": 0.3782,
      "num_input_tokens_seen": 7105960,
      "step": 10870
    },
    {
      "epoch": 6.412146226415095,
      "grad_norm": 1.6224892139434814,
      "learning_rate": 8.589686033210407e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 7108872,
      "step": 10875
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 1.4399867057800293,
      "learning_rate": 8.587894671556823e-06,
      "loss": 0.4091,
      "num_input_tokens_seen": 7112232,
      "step": 10880
    },
    {
      "epoch": 6.418042452830189,
      "grad_norm": 1.7170101404190063,
      "learning_rate": 8.586102359992663e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 7115656,
      "step": 10885
    },
    {
      "epoch": 6.4209905660377355,
      "grad_norm": 2.291621208190918,
      "learning_rate": 8.584309098992447e-06,
      "loss": 0.6804,
      "num_input_tokens_seen": 7118760,
      "step": 10890
    },
    {
      "epoch": 6.423938679245283,
      "grad_norm": 2.6894845962524414,
      "learning_rate": 8.58251488903095e-06,
      "loss": 0.5493,
      "num_input_tokens_seen": 7122024,
      "step": 10895
    },
    {
      "epoch": 6.42688679245283,
      "grad_norm": 2.023683786392212,
      "learning_rate": 8.580719730583196e-06,
      "loss": 0.3085,
      "num_input_tokens_seen": 7124744,
      "step": 10900
    },
    {
      "epoch": 6.429834905660377,
      "grad_norm": 1.2485380172729492,
      "learning_rate": 8.578923624124462e-06,
      "loss": 0.3609,
      "num_input_tokens_seen": 7128072,
      "step": 10905
    },
    {
      "epoch": 6.432783018867925,
      "grad_norm": 7.203639984130859,
      "learning_rate": 8.577126570130273e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 7131688,
      "step": 10910
    },
    {
      "epoch": 6.435731132075472,
      "grad_norm": 1.7754583358764648,
      "learning_rate": 8.575328569076408e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 7135336,
      "step": 10915
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 1.5462884902954102,
      "learning_rate": 8.573529621438896e-06,
      "loss": 0.3394,
      "num_input_tokens_seen": 7138280,
      "step": 10920
    },
    {
      "epoch": 6.441627358490566,
      "grad_norm": 1.5043601989746094,
      "learning_rate": 8.571729727694015e-06,
      "loss": 0.3786,
      "num_input_tokens_seen": 7141128,
      "step": 10925
    },
    {
      "epoch": 6.444575471698113,
      "grad_norm": 1.7990341186523438,
      "learning_rate": 8.569928888318298e-06,
      "loss": 0.3523,
      "num_input_tokens_seen": 7145032,
      "step": 10930
    },
    {
      "epoch": 6.44752358490566,
      "grad_norm": 1.5199345350265503,
      "learning_rate": 8.56812710378852e-06,
      "loss": 0.405,
      "num_input_tokens_seen": 7149736,
      "step": 10935
    },
    {
      "epoch": 6.4504716981132075,
      "grad_norm": 1.1617056131362915,
      "learning_rate": 8.566324374581714e-06,
      "loss": 0.3479,
      "num_input_tokens_seen": 7153192,
      "step": 10940
    },
    {
      "epoch": 6.453419811320755,
      "grad_norm": 1.3552199602127075,
      "learning_rate": 8.564520701175158e-06,
      "loss": 0.4648,
      "num_input_tokens_seen": 7156616,
      "step": 10945
    },
    {
      "epoch": 6.456367924528302,
      "grad_norm": 1.5121878385543823,
      "learning_rate": 8.562716084046387e-06,
      "loss": 0.375,
      "num_input_tokens_seen": 7159208,
      "step": 10950
    },
    {
      "epoch": 6.459316037735849,
      "grad_norm": 1.8055434226989746,
      "learning_rate": 8.560910523673177e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 7162664,
      "step": 10955
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 1.691249966621399,
      "learning_rate": 8.55910402053356e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 7165800,
      "step": 10960
    },
    {
      "epoch": 6.465212264150943,
      "grad_norm": 1.9174267053604126,
      "learning_rate": 8.557296575105814e-06,
      "loss": 0.4404,
      "num_input_tokens_seen": 7168808,
      "step": 10965
    },
    {
      "epoch": 6.46816037735849,
      "grad_norm": 1.965831995010376,
      "learning_rate": 8.555488187868469e-06,
      "loss": 0.4369,
      "num_input_tokens_seen": 7172200,
      "step": 10970
    },
    {
      "epoch": 6.471108490566038,
      "grad_norm": 2.3758835792541504,
      "learning_rate": 8.5536788593003e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 7174920,
      "step": 10975
    },
    {
      "epoch": 6.474056603773585,
      "grad_norm": 1.3522013425827026,
      "learning_rate": 8.55186858988034e-06,
      "loss": 0.41,
      "num_input_tokens_seen": 7177064,
      "step": 10980
    },
    {
      "epoch": 6.477004716981132,
      "grad_norm": 1.191308856010437,
      "learning_rate": 8.550057380087863e-06,
      "loss": 0.3758,
      "num_input_tokens_seen": 7180680,
      "step": 10985
    },
    {
      "epoch": 6.4799528301886795,
      "grad_norm": 1.1893686056137085,
      "learning_rate": 8.548245230402396e-06,
      "loss": 0.4056,
      "num_input_tokens_seen": 7183816,
      "step": 10990
    },
    {
      "epoch": 6.482900943396227,
      "grad_norm": 1.0004298686981201,
      "learning_rate": 8.546432141303711e-06,
      "loss": 0.4711,
      "num_input_tokens_seen": 7186632,
      "step": 10995
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 1.568371295928955,
      "learning_rate": 8.544618113271833e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 7189384,
      "step": 11000
    },
    {
      "epoch": 6.4887971698113205,
      "grad_norm": 4.266327857971191,
      "learning_rate": 8.542803146787032e-06,
      "loss": 0.4877,
      "num_input_tokens_seen": 7192232,
      "step": 11005
    },
    {
      "epoch": 6.491745283018868,
      "grad_norm": 1.1772425174713135,
      "learning_rate": 8.54098724232983e-06,
      "loss": 0.2747,
      "num_input_tokens_seen": 7196360,
      "step": 11010
    },
    {
      "epoch": 6.494693396226415,
      "grad_norm": 1.2568563222885132,
      "learning_rate": 8.539170400380994e-06,
      "loss": 0.463,
      "num_input_tokens_seen": 7198696,
      "step": 11015
    },
    {
      "epoch": 6.497641509433962,
      "grad_norm": 3.349360466003418,
      "learning_rate": 8.537352621421542e-06,
      "loss": 0.4941,
      "num_input_tokens_seen": 7202184,
      "step": 11020
    },
    {
      "epoch": 6.50058962264151,
      "grad_norm": 2.370978832244873,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.4942,
      "num_input_tokens_seen": 7205128,
      "step": 11025
    },
    {
      "epoch": 6.503537735849057,
      "grad_norm": 1.0790369510650635,
      "learning_rate": 8.533714254396096e-06,
      "loss": 0.4035,
      "num_input_tokens_seen": 7209576,
      "step": 11030
    },
    {
      "epoch": 6.506485849056604,
      "grad_norm": 1.1742953062057495,
      "learning_rate": 8.531893667293375e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 7212360,
      "step": 11035
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 1.2841202020645142,
      "learning_rate": 8.530072145106585e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 7215144,
      "step": 11040
    },
    {
      "epoch": 6.512382075471698,
      "grad_norm": 1.9063777923583984,
      "learning_rate": 8.528249688317978e-06,
      "loss": 0.4112,
      "num_input_tokens_seen": 7217896,
      "step": 11045
    },
    {
      "epoch": 6.515330188679245,
      "grad_norm": 2.101836681365967,
      "learning_rate": 8.526426297410062e-06,
      "loss": 0.4426,
      "num_input_tokens_seen": 7221032,
      "step": 11050
    },
    {
      "epoch": 6.5182783018867925,
      "grad_norm": 2.7056548595428467,
      "learning_rate": 8.524601972865586e-06,
      "loss": 0.4861,
      "num_input_tokens_seen": 7224488,
      "step": 11055
    },
    {
      "epoch": 6.52122641509434,
      "grad_norm": 2.0528292655944824,
      "learning_rate": 8.522776715167548e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 7227016,
      "step": 11060
    },
    {
      "epoch": 6.524174528301887,
      "grad_norm": 0.9749739170074463,
      "learning_rate": 8.520950524799192e-06,
      "loss": 0.4736,
      "num_input_tokens_seen": 7231176,
      "step": 11065
    },
    {
      "epoch": 6.527122641509434,
      "grad_norm": 1.6627600193023682,
      "learning_rate": 8.51912340224401e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 7234600,
      "step": 11070
    },
    {
      "epoch": 6.530070754716981,
      "grad_norm": 2.599240303039551,
      "learning_rate": 8.51729534798574e-06,
      "loss": 0.5577,
      "num_input_tokens_seen": 7237992,
      "step": 11075
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 1.3375779390335083,
      "learning_rate": 8.515466362508369e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 7241864,
      "step": 11080
    },
    {
      "epoch": 6.535966981132075,
      "grad_norm": 2.6731817722320557,
      "learning_rate": 8.513636446296125e-06,
      "loss": 0.6303,
      "num_input_tokens_seen": 7244200,
      "step": 11085
    },
    {
      "epoch": 6.538915094339623,
      "grad_norm": 1.0463314056396484,
      "learning_rate": 8.51180559983349e-06,
      "loss": 0.3218,
      "num_input_tokens_seen": 7247912,
      "step": 11090
    },
    {
      "epoch": 6.54186320754717,
      "grad_norm": 2.685417413711548,
      "learning_rate": 8.50997382360519e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 7251272,
      "step": 11095
    },
    {
      "epoch": 6.544811320754717,
      "grad_norm": 1.9272539615631104,
      "learning_rate": 8.508141118096191e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 7254664,
      "step": 11100
    },
    {
      "epoch": 6.5477594339622645,
      "grad_norm": 2.239264726638794,
      "learning_rate": 8.506307483791712e-06,
      "loss": 0.4387,
      "num_input_tokens_seen": 7257576,
      "step": 11105
    },
    {
      "epoch": 6.550707547169811,
      "grad_norm": 1.636662483215332,
      "learning_rate": 8.504472921177215e-06,
      "loss": 0.575,
      "num_input_tokens_seen": 7261128,
      "step": 11110
    },
    {
      "epoch": 6.553655660377358,
      "grad_norm": 3.189784049987793,
      "learning_rate": 8.502637430738409e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 7263784,
      "step": 11115
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 2.081143617630005,
      "learning_rate": 8.500801012961248e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 7266856,
      "step": 11120
    },
    {
      "epoch": 6.559551886792453,
      "grad_norm": 5.8629021644592285,
      "learning_rate": 8.49896366833193e-06,
      "loss": 0.4025,
      "num_input_tokens_seen": 7269864,
      "step": 11125
    },
    {
      "epoch": 6.5625,
      "grad_norm": 1.9503377676010132,
      "learning_rate": 8.497125397336903e-06,
      "loss": 0.4267,
      "num_input_tokens_seen": 7273192,
      "step": 11130
    },
    {
      "epoch": 6.565448113207547,
      "grad_norm": 2.2085728645324707,
      "learning_rate": 8.495286200462854e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 7276456,
      "step": 11135
    },
    {
      "epoch": 6.568396226415095,
      "grad_norm": 2.5417211055755615,
      "learning_rate": 8.49344607819672e-06,
      "loss": 0.3457,
      "num_input_tokens_seen": 7279240,
      "step": 11140
    },
    {
      "epoch": 6.571344339622642,
      "grad_norm": 1.474778652191162,
      "learning_rate": 8.49160503102568e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 7283368,
      "step": 11145
    },
    {
      "epoch": 6.574292452830189,
      "grad_norm": 1.7758393287658691,
      "learning_rate": 8.489763059437161e-06,
      "loss": 0.4785,
      "num_input_tokens_seen": 7287080,
      "step": 11150
    },
    {
      "epoch": 6.5772405660377355,
      "grad_norm": 3.1584277153015137,
      "learning_rate": 8.487920163918833e-06,
      "loss": 0.5711,
      "num_input_tokens_seen": 7290952,
      "step": 11155
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 1.3473479747772217,
      "learning_rate": 8.486076344958607e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 7294632,
      "step": 11160
    },
    {
      "epoch": 6.58313679245283,
      "grad_norm": 1.2120039463043213,
      "learning_rate": 8.484231603044647e-06,
      "loss": 0.2961,
      "num_input_tokens_seen": 7297672,
      "step": 11165
    },
    {
      "epoch": 6.586084905660377,
      "grad_norm": 1.8242223262786865,
      "learning_rate": 8.482385938665352e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 7300392,
      "step": 11170
    },
    {
      "epoch": 6.589033018867925,
      "grad_norm": 0.9736544489860535,
      "learning_rate": 8.480539352309373e-06,
      "loss": 0.4813,
      "num_input_tokens_seen": 7303912,
      "step": 11175
    },
    {
      "epoch": 6.591981132075472,
      "grad_norm": 2.4560675621032715,
      "learning_rate": 8.478691844465598e-06,
      "loss": 0.4264,
      "num_input_tokens_seen": 7306504,
      "step": 11180
    },
    {
      "epoch": 6.594929245283019,
      "grad_norm": 1.3973829746246338,
      "learning_rate": 8.476843415623168e-06,
      "loss": 0.3892,
      "num_input_tokens_seen": 7309192,
      "step": 11185
    },
    {
      "epoch": 6.597877358490566,
      "grad_norm": 2.853372812271118,
      "learning_rate": 8.474994066271458e-06,
      "loss": 0.4042,
      "num_input_tokens_seen": 7311976,
      "step": 11190
    },
    {
      "epoch": 6.600825471698113,
      "grad_norm": 2.061276912689209,
      "learning_rate": 8.473143796900089e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 7315336,
      "step": 11195
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 2.2931954860687256,
      "learning_rate": 8.471292607998936e-06,
      "loss": 0.3309,
      "num_input_tokens_seen": 7322024,
      "step": 11200
    },
    {
      "epoch": 6.6067216981132075,
      "grad_norm": 1.9210284948349,
      "learning_rate": 8.469440500058104e-06,
      "loss": 0.523,
      "num_input_tokens_seen": 7324648,
      "step": 11205
    },
    {
      "epoch": 6.609669811320755,
      "grad_norm": 0.9596117734909058,
      "learning_rate": 8.467587473567945e-06,
      "loss": 0.3766,
      "num_input_tokens_seen": 7327784,
      "step": 11210
    },
    {
      "epoch": 6.612617924528302,
      "grad_norm": 1.9294731616973877,
      "learning_rate": 8.46573352901906e-06,
      "loss": 0.4335,
      "num_input_tokens_seen": 7330152,
      "step": 11215
    },
    {
      "epoch": 6.615566037735849,
      "grad_norm": 0.9903777837753296,
      "learning_rate": 8.463878666902286e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 7333032,
      "step": 11220
    },
    {
      "epoch": 6.618514150943396,
      "grad_norm": 1.6641567945480347,
      "learning_rate": 8.462022887708706e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 7336712,
      "step": 11225
    },
    {
      "epoch": 6.621462264150943,
      "grad_norm": 1.9121224880218506,
      "learning_rate": 8.460166191929646e-06,
      "loss": 0.4248,
      "num_input_tokens_seen": 7339752,
      "step": 11230
    },
    {
      "epoch": 6.62441037735849,
      "grad_norm": 3.1584746837615967,
      "learning_rate": 8.458308580056675e-06,
      "loss": 0.5522,
      "num_input_tokens_seen": 7342984,
      "step": 11235
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 2.3062522411346436,
      "learning_rate": 8.456450052581602e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 7345768,
      "step": 11240
    },
    {
      "epoch": 6.630306603773585,
      "grad_norm": 2.303839921951294,
      "learning_rate": 8.45459060999648e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 7349128,
      "step": 11245
    },
    {
      "epoch": 6.633254716981132,
      "grad_norm": 3.4046735763549805,
      "learning_rate": 8.452730252793608e-06,
      "loss": 0.484,
      "num_input_tokens_seen": 7351368,
      "step": 11250
    },
    {
      "epoch": 6.6362028301886795,
      "grad_norm": 1.37340247631073,
      "learning_rate": 8.450868981465519e-06,
      "loss": 0.2877,
      "num_input_tokens_seen": 7354088,
      "step": 11255
    },
    {
      "epoch": 6.639150943396227,
      "grad_norm": 2.091463565826416,
      "learning_rate": 8.449006796504997e-06,
      "loss": 0.3339,
      "num_input_tokens_seen": 7356968,
      "step": 11260
    },
    {
      "epoch": 6.642099056603773,
      "grad_norm": 3.127164840698242,
      "learning_rate": 8.44714369840506e-06,
      "loss": 0.3465,
      "num_input_tokens_seen": 7362312,
      "step": 11265
    },
    {
      "epoch": 6.6450471698113205,
      "grad_norm": 2.1118664741516113,
      "learning_rate": 8.445279687658973e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 7365256,
      "step": 11270
    },
    {
      "epoch": 6.647995283018868,
      "grad_norm": 2.920006275177002,
      "learning_rate": 8.44341476476024e-06,
      "loss": 0.4706,
      "num_input_tokens_seen": 7367848,
      "step": 11275
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 1.6199023723602295,
      "learning_rate": 8.441548930202608e-06,
      "loss": 0.5126,
      "num_input_tokens_seen": 7370920,
      "step": 11280
    },
    {
      "epoch": 6.653891509433962,
      "grad_norm": 2.126906394958496,
      "learning_rate": 8.439682184480065e-06,
      "loss": 0.3169,
      "num_input_tokens_seen": 7374376,
      "step": 11285
    },
    {
      "epoch": 6.65683962264151,
      "grad_norm": 1.4357770681381226,
      "learning_rate": 8.437814528086837e-06,
      "loss": 0.3732,
      "num_input_tokens_seen": 7378344,
      "step": 11290
    },
    {
      "epoch": 6.659787735849057,
      "grad_norm": 1.6092534065246582,
      "learning_rate": 8.435945961517398e-06,
      "loss": 0.4224,
      "num_input_tokens_seen": 7381480,
      "step": 11295
    },
    {
      "epoch": 6.662735849056604,
      "grad_norm": 2.1627089977264404,
      "learning_rate": 8.434076485266458e-06,
      "loss": 0.5945,
      "num_input_tokens_seen": 7384520,
      "step": 11300
    },
    {
      "epoch": 6.665683962264151,
      "grad_norm": 1.7997546195983887,
      "learning_rate": 8.432206099828969e-06,
      "loss": 0.4496,
      "num_input_tokens_seen": 7387336,
      "step": 11305
    },
    {
      "epoch": 6.668632075471698,
      "grad_norm": 2.030839443206787,
      "learning_rate": 8.430334805700122e-06,
      "loss": 0.3263,
      "num_input_tokens_seen": 7389992,
      "step": 11310
    },
    {
      "epoch": 6.671580188679245,
      "grad_norm": 1.7795969247817993,
      "learning_rate": 8.428462603375351e-06,
      "loss": 0.4031,
      "num_input_tokens_seen": 7393800,
      "step": 11315
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 1.6212395429611206,
      "learning_rate": 8.426589493350332e-06,
      "loss": 0.4581,
      "num_input_tokens_seen": 7397224,
      "step": 11320
    },
    {
      "epoch": 6.67747641509434,
      "grad_norm": 2.507599353790283,
      "learning_rate": 8.424715476120976e-06,
      "loss": 0.4448,
      "num_input_tokens_seen": 7399912,
      "step": 11325
    },
    {
      "epoch": 6.680424528301887,
      "grad_norm": 1.2958760261535645,
      "learning_rate": 8.422840552183437e-06,
      "loss": 0.3609,
      "num_input_tokens_seen": 7402952,
      "step": 11330
    },
    {
      "epoch": 6.683372641509434,
      "grad_norm": 2.460819959640503,
      "learning_rate": 8.420964722034111e-06,
      "loss": 0.4721,
      "num_input_tokens_seen": 7405800,
      "step": 11335
    },
    {
      "epoch": 6.686320754716981,
      "grad_norm": 2.6678357124328613,
      "learning_rate": 8.41908798616963e-06,
      "loss": 0.4702,
      "num_input_tokens_seen": 7408584,
      "step": 11340
    },
    {
      "epoch": 6.689268867924528,
      "grad_norm": 1.6535897254943848,
      "learning_rate": 8.41721034508687e-06,
      "loss": 0.4134,
      "num_input_tokens_seen": 7412328,
      "step": 11345
    },
    {
      "epoch": 6.692216981132075,
      "grad_norm": 2.5728485584259033,
      "learning_rate": 8.415331799282942e-06,
      "loss": 0.5288,
      "num_input_tokens_seen": 7415912,
      "step": 11350
    },
    {
      "epoch": 6.695165094339623,
      "grad_norm": 1.361088514328003,
      "learning_rate": 8.413452349255205e-06,
      "loss": 0.5729,
      "num_input_tokens_seen": 7418760,
      "step": 11355
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 1.8883126974105835,
      "learning_rate": 8.411571995501245e-06,
      "loss": 0.4886,
      "num_input_tokens_seen": 7422216,
      "step": 11360
    },
    {
      "epoch": 6.701061320754717,
      "grad_norm": 1.226893424987793,
      "learning_rate": 8.409690738518895e-06,
      "loss": 0.4417,
      "num_input_tokens_seen": 7425928,
      "step": 11365
    },
    {
      "epoch": 6.7040094339622645,
      "grad_norm": 1.3609343767166138,
      "learning_rate": 8.407808578806229e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 7429384,
      "step": 11370
    },
    {
      "epoch": 6.706957547169811,
      "grad_norm": 2.4475836753845215,
      "learning_rate": 8.405925516861555e-06,
      "loss": 0.395,
      "num_input_tokens_seen": 7432424,
      "step": 11375
    },
    {
      "epoch": 6.709905660377358,
      "grad_norm": 1.6876314878463745,
      "learning_rate": 8.40404155318342e-06,
      "loss": 0.4555,
      "num_input_tokens_seen": 7435560,
      "step": 11380
    },
    {
      "epoch": 6.712853773584905,
      "grad_norm": 1.6346091032028198,
      "learning_rate": 8.402156688270613e-06,
      "loss": 0.4286,
      "num_input_tokens_seen": 7438568,
      "step": 11385
    },
    {
      "epoch": 6.715801886792453,
      "grad_norm": 2.394557237625122,
      "learning_rate": 8.400270922622162e-06,
      "loss": 0.5124,
      "num_input_tokens_seen": 7442216,
      "step": 11390
    },
    {
      "epoch": 6.71875,
      "grad_norm": 2.3807761669158936,
      "learning_rate": 8.398384256737328e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 7445704,
      "step": 11395
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 5.5305705070495605,
      "learning_rate": 8.396496691115619e-06,
      "loss": 0.5209,
      "num_input_tokens_seen": 7448360,
      "step": 11400
    },
    {
      "epoch": 6.724646226415095,
      "grad_norm": 1.3679087162017822,
      "learning_rate": 8.39460822625677e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 7452360,
      "step": 11405
    },
    {
      "epoch": 6.727594339622642,
      "grad_norm": 1.6806442737579346,
      "learning_rate": 8.392718862660765e-06,
      "loss": 0.447,
      "num_input_tokens_seen": 7455496,
      "step": 11410
    },
    {
      "epoch": 6.730542452830189,
      "grad_norm": 2.7627673149108887,
      "learning_rate": 8.390828600827818e-06,
      "loss": 0.3738,
      "num_input_tokens_seen": 7459304,
      "step": 11415
    },
    {
      "epoch": 6.7334905660377355,
      "grad_norm": 1.0391032695770264,
      "learning_rate": 8.388937441258385e-06,
      "loss": 0.4554,
      "num_input_tokens_seen": 7462568,
      "step": 11420
    },
    {
      "epoch": 6.736438679245283,
      "grad_norm": 2.113389253616333,
      "learning_rate": 8.387045384453162e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 7465768,
      "step": 11425
    },
    {
      "epoch": 6.73938679245283,
      "grad_norm": 1.153096079826355,
      "learning_rate": 8.385152430913073e-06,
      "loss": 0.5043,
      "num_input_tokens_seen": 7469800,
      "step": 11430
    },
    {
      "epoch": 6.742334905660377,
      "grad_norm": 1.3497940301895142,
      "learning_rate": 8.383258581139288e-06,
      "loss": 0.4387,
      "num_input_tokens_seen": 7472360,
      "step": 11435
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 2.911217212677002,
      "learning_rate": 8.381363835633213e-06,
      "loss": 0.4743,
      "num_input_tokens_seen": 7475944,
      "step": 11440
    },
    {
      "epoch": 6.748231132075472,
      "grad_norm": 2.3676581382751465,
      "learning_rate": 8.379468194896492e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 7478728,
      "step": 11445
    },
    {
      "epoch": 6.751179245283019,
      "grad_norm": 1.7134976387023926,
      "learning_rate": 8.377571659431e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 7481128,
      "step": 11450
    },
    {
      "epoch": 6.754127358490566,
      "grad_norm": 1.8084092140197754,
      "learning_rate": 8.375674229738855e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 7483976,
      "step": 11455
    },
    {
      "epoch": 6.757075471698113,
      "grad_norm": 1.478624701499939,
      "learning_rate": 8.37377590632241e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 7487240,
      "step": 11460
    },
    {
      "epoch": 6.76002358490566,
      "grad_norm": 1.5224944353103638,
      "learning_rate": 8.371876689684253e-06,
      "loss": 0.3842,
      "num_input_tokens_seen": 7490280,
      "step": 11465
    },
    {
      "epoch": 6.7629716981132075,
      "grad_norm": 1.210831642150879,
      "learning_rate": 8.369976580327211e-06,
      "loss": 0.386,
      "num_input_tokens_seen": 7493480,
      "step": 11470
    },
    {
      "epoch": 6.765919811320755,
      "grad_norm": 1.5500469207763672,
      "learning_rate": 8.368075578754345e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 7497192,
      "step": 11475
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 1.1962625980377197,
      "learning_rate": 8.366173685468952e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 7500168,
      "step": 11480
    },
    {
      "epoch": 6.771816037735849,
      "grad_norm": 1.3476800918579102,
      "learning_rate": 8.364270900974572e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 7502920,
      "step": 11485
    },
    {
      "epoch": 6.774764150943396,
      "grad_norm": 2.1063578128814697,
      "learning_rate": 8.362367225774968e-06,
      "loss": 0.3379,
      "num_input_tokens_seen": 7505992,
      "step": 11490
    },
    {
      "epoch": 6.777712264150943,
      "grad_norm": 2.601313591003418,
      "learning_rate": 8.360462660374153e-06,
      "loss": 0.4944,
      "num_input_tokens_seen": 7508552,
      "step": 11495
    },
    {
      "epoch": 6.78066037735849,
      "grad_norm": 1.369203805923462,
      "learning_rate": 8.358557205276365e-06,
      "loss": 0.5949,
      "num_input_tokens_seen": 7511976,
      "step": 11500
    },
    {
      "epoch": 6.783608490566038,
      "grad_norm": 1.3838844299316406,
      "learning_rate": 8.356650860986083e-06,
      "loss": 0.3886,
      "num_input_tokens_seen": 7515048,
      "step": 11505
    },
    {
      "epoch": 6.786556603773585,
      "grad_norm": 2.1440820693969727,
      "learning_rate": 8.354743628008017e-06,
      "loss": 0.4403,
      "num_input_tokens_seen": 7518152,
      "step": 11510
    },
    {
      "epoch": 6.789504716981132,
      "grad_norm": 1.8052012920379639,
      "learning_rate": 8.35283550684712e-06,
      "loss": 0.5578,
      "num_input_tokens_seen": 7521512,
      "step": 11515
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 2.3348770141601562,
      "learning_rate": 8.350926498008572e-06,
      "loss": 0.4765,
      "num_input_tokens_seen": 7524968,
      "step": 11520
    },
    {
      "epoch": 6.795400943396227,
      "grad_norm": 1.949662685394287,
      "learning_rate": 8.34901660199779e-06,
      "loss": 0.4032,
      "num_input_tokens_seen": 7529672,
      "step": 11525
    },
    {
      "epoch": 6.798349056603773,
      "grad_norm": 2.1654369831085205,
      "learning_rate": 8.347105819320432e-06,
      "loss": 0.5893,
      "num_input_tokens_seen": 7533064,
      "step": 11530
    },
    {
      "epoch": 6.8012971698113205,
      "grad_norm": 1.6288096904754639,
      "learning_rate": 8.34519415048238e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 7535816,
      "step": 11535
    },
    {
      "epoch": 6.804245283018868,
      "grad_norm": 1.5632507801055908,
      "learning_rate": 8.343281595989761e-06,
      "loss": 0.3537,
      "num_input_tokens_seen": 7539144,
      "step": 11540
    },
    {
      "epoch": 6.807193396226415,
      "grad_norm": 1.302585244178772,
      "learning_rate": 8.341368156348933e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 7543240,
      "step": 11545
    },
    {
      "epoch": 6.810141509433962,
      "grad_norm": 1.0490241050720215,
      "learning_rate": 8.339453832066482e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 7546120,
      "step": 11550
    },
    {
      "epoch": 6.81308962264151,
      "grad_norm": 1.5826739072799683,
      "learning_rate": 8.337538623649237e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 7548872,
      "step": 11555
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 1.513465166091919,
      "learning_rate": 8.33562253160426e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 7552104,
      "step": 11560
    },
    {
      "epoch": 6.818985849056604,
      "grad_norm": 1.185442328453064,
      "learning_rate": 8.33370555643884e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 7556328,
      "step": 11565
    },
    {
      "epoch": 6.821933962264151,
      "grad_norm": 6.602989196777344,
      "learning_rate": 8.331787698660507e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 7559752,
      "step": 11570
    },
    {
      "epoch": 6.824882075471698,
      "grad_norm": 1.6025782823562622,
      "learning_rate": 8.32986895877702e-06,
      "loss": 0.3271,
      "num_input_tokens_seen": 7563624,
      "step": 11575
    },
    {
      "epoch": 6.827830188679245,
      "grad_norm": 1.7179795503616333,
      "learning_rate": 8.327949337296378e-06,
      "loss": 0.2796,
      "num_input_tokens_seen": 7566824,
      "step": 11580
    },
    {
      "epoch": 6.8307783018867925,
      "grad_norm": 1.8874346017837524,
      "learning_rate": 8.326028834726803e-06,
      "loss": 0.5547,
      "num_input_tokens_seen": 7569160,
      "step": 11585
    },
    {
      "epoch": 6.83372641509434,
      "grad_norm": 1.5048376321792603,
      "learning_rate": 8.324107451576762e-06,
      "loss": 0.35,
      "num_input_tokens_seen": 7572456,
      "step": 11590
    },
    {
      "epoch": 6.836674528301887,
      "grad_norm": 1.5739561319351196,
      "learning_rate": 8.322185188354947e-06,
      "loss": 0.3889,
      "num_input_tokens_seen": 7575144,
      "step": 11595
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 2.0472874641418457,
      "learning_rate": 8.320262045570284e-06,
      "loss": 0.4852,
      "num_input_tokens_seen": 7578120,
      "step": 11600
    },
    {
      "epoch": 6.842570754716981,
      "grad_norm": 2.9681150913238525,
      "learning_rate": 8.318338023731937e-06,
      "loss": 0.6051,
      "num_input_tokens_seen": 7581672,
      "step": 11605
    },
    {
      "epoch": 6.845518867924528,
      "grad_norm": 1.490749716758728,
      "learning_rate": 8.316413123349296e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 7585512,
      "step": 11610
    },
    {
      "epoch": 6.848466981132075,
      "grad_norm": 2.601746082305908,
      "learning_rate": 8.314487344931987e-06,
      "loss": 0.5049,
      "num_input_tokens_seen": 7588968,
      "step": 11615
    },
    {
      "epoch": 6.851415094339623,
      "grad_norm": 2.7221179008483887,
      "learning_rate": 8.31256068898987e-06,
      "loss": 0.5921,
      "num_input_tokens_seen": 7592264,
      "step": 11620
    },
    {
      "epoch": 6.85436320754717,
      "grad_norm": 1.8810186386108398,
      "learning_rate": 8.310633156033032e-06,
      "loss": 0.4456,
      "num_input_tokens_seen": 7595272,
      "step": 11625
    },
    {
      "epoch": 6.857311320754717,
      "grad_norm": 2.42887544631958,
      "learning_rate": 8.3087047465718e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 7597800,
      "step": 11630
    },
    {
      "epoch": 6.8602594339622645,
      "grad_norm": 2.02432918548584,
      "learning_rate": 8.306775461116727e-06,
      "loss": 0.3986,
      "num_input_tokens_seen": 7600296,
      "step": 11635
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 2.261789083480835,
      "learning_rate": 8.304845300178597e-06,
      "loss": 0.3651,
      "num_input_tokens_seen": 7603592,
      "step": 11640
    },
    {
      "epoch": 6.866155660377358,
      "grad_norm": 3.4058144092559814,
      "learning_rate": 8.302914264268433e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 7606728,
      "step": 11645
    },
    {
      "epoch": 6.869103773584905,
      "grad_norm": 2.5369131565093994,
      "learning_rate": 8.300982353897482e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 7610056,
      "step": 11650
    },
    {
      "epoch": 6.872051886792453,
      "grad_norm": 2.689807653427124,
      "learning_rate": 8.299049569577226e-06,
      "loss": 0.4241,
      "num_input_tokens_seen": 7612744,
      "step": 11655
    },
    {
      "epoch": 6.875,
      "grad_norm": 2.104665517807007,
      "learning_rate": 8.297115911819379e-06,
      "loss": 0.5385,
      "num_input_tokens_seen": 7615592,
      "step": 11660
    },
    {
      "epoch": 6.877948113207547,
      "grad_norm": 2.5054311752319336,
      "learning_rate": 8.295181381135884e-06,
      "loss": 0.5906,
      "num_input_tokens_seen": 7618696,
      "step": 11665
    },
    {
      "epoch": 6.880896226415095,
      "grad_norm": 1.5638967752456665,
      "learning_rate": 8.293245978038917e-06,
      "loss": 0.4943,
      "num_input_tokens_seen": 7621384,
      "step": 11670
    },
    {
      "epoch": 6.883844339622642,
      "grad_norm": 1.9542137384414673,
      "learning_rate": 8.291309703040884e-06,
      "loss": 0.6039,
      "num_input_tokens_seen": 7625064,
      "step": 11675
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 1.2781684398651123,
      "learning_rate": 8.289372556654422e-06,
      "loss": 0.5588,
      "num_input_tokens_seen": 7628744,
      "step": 11680
    },
    {
      "epoch": 6.8897405660377355,
      "grad_norm": 1.2859834432601929,
      "learning_rate": 8.287434539392401e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 7631336,
      "step": 11685
    },
    {
      "epoch": 6.892688679245283,
      "grad_norm": 2.1392107009887695,
      "learning_rate": 8.285495651767916e-06,
      "loss": 0.4828,
      "num_input_tokens_seen": 7634216,
      "step": 11690
    },
    {
      "epoch": 6.89563679245283,
      "grad_norm": 1.810211420059204,
      "learning_rate": 8.283555894294297e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 7637288,
      "step": 11695
    },
    {
      "epoch": 6.898584905660377,
      "grad_norm": 1.2920650243759155,
      "learning_rate": 8.281615267485105e-06,
      "loss": 0.3845,
      "num_input_tokens_seen": 7641544,
      "step": 11700
    },
    {
      "epoch": 6.901533018867925,
      "grad_norm": 2.016890525817871,
      "learning_rate": 8.279673771854127e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 7644680,
      "step": 11705
    },
    {
      "epoch": 6.904481132075472,
      "grad_norm": 1.2575345039367676,
      "learning_rate": 8.277731407915386e-06,
      "loss": 0.4566,
      "num_input_tokens_seen": 7647400,
      "step": 11710
    },
    {
      "epoch": 6.907429245283019,
      "grad_norm": 1.9934018850326538,
      "learning_rate": 8.275788176183126e-06,
      "loss": 0.4829,
      "num_input_tokens_seen": 7650280,
      "step": 11715
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 1.4815058708190918,
      "learning_rate": 8.273844077171827e-06,
      "loss": 0.5041,
      "num_input_tokens_seen": 7654152,
      "step": 11720
    },
    {
      "epoch": 6.913325471698113,
      "grad_norm": 1.9020379781723022,
      "learning_rate": 8.271899111396202e-06,
      "loss": 0.3613,
      "num_input_tokens_seen": 7656904,
      "step": 11725
    },
    {
      "epoch": 6.91627358490566,
      "grad_norm": 1.1504908800125122,
      "learning_rate": 8.269953279371185e-06,
      "loss": 0.5236,
      "num_input_tokens_seen": 7661000,
      "step": 11730
    },
    {
      "epoch": 6.9192216981132075,
      "grad_norm": 2.1240503787994385,
      "learning_rate": 8.268006581611945e-06,
      "loss": 0.5245,
      "num_input_tokens_seen": 7665192,
      "step": 11735
    },
    {
      "epoch": 6.922169811320755,
      "grad_norm": 2.5187435150146484,
      "learning_rate": 8.266059018633878e-06,
      "loss": 0.3944,
      "num_input_tokens_seen": 7667976,
      "step": 11740
    },
    {
      "epoch": 6.925117924528302,
      "grad_norm": 2.538396120071411,
      "learning_rate": 8.264110590952609e-06,
      "loss": 0.5619,
      "num_input_tokens_seen": 7670568,
      "step": 11745
    },
    {
      "epoch": 6.928066037735849,
      "grad_norm": 2.443323850631714,
      "learning_rate": 8.262161299083993e-06,
      "loss": 0.5352,
      "num_input_tokens_seen": 7674760,
      "step": 11750
    },
    {
      "epoch": 6.931014150943396,
      "grad_norm": 1.5477584600448608,
      "learning_rate": 8.260211143544117e-06,
      "loss": 0.369,
      "num_input_tokens_seen": 7677992,
      "step": 11755
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 1.2205750942230225,
      "learning_rate": 8.258260124849288e-06,
      "loss": 0.6591,
      "num_input_tokens_seen": 7681192,
      "step": 11760
    },
    {
      "epoch": 6.93691037735849,
      "grad_norm": 1.3757009506225586,
      "learning_rate": 8.256308243516048e-06,
      "loss": 0.415,
      "num_input_tokens_seen": 7685096,
      "step": 11765
    },
    {
      "epoch": 6.939858490566038,
      "grad_norm": 2.847487688064575,
      "learning_rate": 8.254355500061168e-06,
      "loss": 0.4373,
      "num_input_tokens_seen": 7688264,
      "step": 11770
    },
    {
      "epoch": 6.942806603773585,
      "grad_norm": 2.1439907550811768,
      "learning_rate": 8.252401895001643e-06,
      "loss": 0.2863,
      "num_input_tokens_seen": 7691976,
      "step": 11775
    },
    {
      "epoch": 6.945754716981132,
      "grad_norm": 1.1395341157913208,
      "learning_rate": 8.2504474288547e-06,
      "loss": 0.4718,
      "num_input_tokens_seen": 7695208,
      "step": 11780
    },
    {
      "epoch": 6.9487028301886795,
      "grad_norm": 1.635796308517456,
      "learning_rate": 8.248492102137791e-06,
      "loss": 0.3566,
      "num_input_tokens_seen": 7698984,
      "step": 11785
    },
    {
      "epoch": 6.951650943396227,
      "grad_norm": 2.419865131378174,
      "learning_rate": 8.246535915368596e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 7702408,
      "step": 11790
    },
    {
      "epoch": 6.954599056603773,
      "grad_norm": 1.4682364463806152,
      "learning_rate": 8.24457886906503e-06,
      "loss": 0.5221,
      "num_input_tokens_seen": 7705384,
      "step": 11795
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 2.278494119644165,
      "learning_rate": 8.242620963745222e-06,
      "loss": 0.4663,
      "num_input_tokens_seen": 7708168,
      "step": 11800
    },
    {
      "epoch": 6.960495283018868,
      "grad_norm": 1.5297703742980957,
      "learning_rate": 8.240662199927538e-06,
      "loss": 0.5335,
      "num_input_tokens_seen": 7711368,
      "step": 11805
    },
    {
      "epoch": 6.963443396226415,
      "grad_norm": 2.251765727996826,
      "learning_rate": 8.238702578130573e-06,
      "loss": 0.3916,
      "num_input_tokens_seen": 7714568,
      "step": 11810
    },
    {
      "epoch": 6.966391509433962,
      "grad_norm": 2.0876946449279785,
      "learning_rate": 8.23674209887314e-06,
      "loss": 0.5447,
      "num_input_tokens_seen": 7718728,
      "step": 11815
    },
    {
      "epoch": 6.96933962264151,
      "grad_norm": 1.6396509408950806,
      "learning_rate": 8.234780762674288e-06,
      "loss": 0.3439,
      "num_input_tokens_seen": 7723784,
      "step": 11820
    },
    {
      "epoch": 6.972287735849057,
      "grad_norm": 2.6655702590942383,
      "learning_rate": 8.232818570053286e-06,
      "loss": 0.5039,
      "num_input_tokens_seen": 7726696,
      "step": 11825
    },
    {
      "epoch": 6.975235849056604,
      "grad_norm": 2.259575128555298,
      "learning_rate": 8.230855521529637e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 7729576,
      "step": 11830
    },
    {
      "epoch": 6.978183962264151,
      "grad_norm": 1.3702157735824585,
      "learning_rate": 8.228891617623064e-06,
      "loss": 0.5229,
      "num_input_tokens_seen": 7732456,
      "step": 11835
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 1.1039721965789795,
      "learning_rate": 8.22692685885352e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 7736040,
      "step": 11840
    },
    {
      "epoch": 6.984080188679245,
      "grad_norm": 2.0172183513641357,
      "learning_rate": 8.224961245741183e-06,
      "loss": 0.3673,
      "num_input_tokens_seen": 7739432,
      "step": 11845
    },
    {
      "epoch": 6.9870283018867925,
      "grad_norm": 1.936835527420044,
      "learning_rate": 8.222994778806457e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 7741768,
      "step": 11850
    },
    {
      "epoch": 6.98997641509434,
      "grad_norm": 1.9658669233322144,
      "learning_rate": 8.221027458569972e-06,
      "loss": 0.3849,
      "num_input_tokens_seen": 7745704,
      "step": 11855
    },
    {
      "epoch": 6.992924528301887,
      "grad_norm": 3.4912712574005127,
      "learning_rate": 8.219059285552586e-06,
      "loss": 0.5236,
      "num_input_tokens_seen": 7748776,
      "step": 11860
    },
    {
      "epoch": 6.995872641509434,
      "grad_norm": 1.779822826385498,
      "learning_rate": 8.21709026027538e-06,
      "loss": 0.5378,
      "num_input_tokens_seen": 7751720,
      "step": 11865
    },
    {
      "epoch": 6.998820754716981,
      "grad_norm": 1.9152437448501587,
      "learning_rate": 8.215120383259664e-06,
      "loss": 0.6356,
      "num_input_tokens_seen": 7754312,
      "step": 11870
    },
    {
      "epoch": 7.001768867924528,
      "grad_norm": 3.5339598655700684,
      "learning_rate": 8.21314965502697e-06,
      "loss": 0.3513,
      "num_input_tokens_seen": 7756704,
      "step": 11875
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 3.194509506225586,
      "learning_rate": 8.211178076099056e-06,
      "loss": 0.4098,
      "num_input_tokens_seen": 7759072,
      "step": 11880
    },
    {
      "epoch": 7.007665094339623,
      "grad_norm": 2.3303802013397217,
      "learning_rate": 8.209205646997909e-06,
      "loss": 0.3362,
      "num_input_tokens_seen": 7761792,
      "step": 11885
    },
    {
      "epoch": 7.01061320754717,
      "grad_norm": 1.2496417760849,
      "learning_rate": 8.207232368245735e-06,
      "loss": 0.3419,
      "num_input_tokens_seen": 7764800,
      "step": 11890
    },
    {
      "epoch": 7.013561320754717,
      "grad_norm": 2.0059468746185303,
      "learning_rate": 8.205258240364968e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 7767712,
      "step": 11895
    },
    {
      "epoch": 7.0165094339622645,
      "grad_norm": 1.4415944814682007,
      "learning_rate": 8.203283263878268e-06,
      "loss": 0.5767,
      "num_input_tokens_seen": 7771136,
      "step": 11900
    },
    {
      "epoch": 7.019457547169812,
      "grad_norm": 1.529421329498291,
      "learning_rate": 8.201307439308518e-06,
      "loss": 0.5236,
      "num_input_tokens_seen": 7775648,
      "step": 11905
    },
    {
      "epoch": 7.022405660377358,
      "grad_norm": 2.356029987335205,
      "learning_rate": 8.199330767178828e-06,
      "loss": 0.495,
      "num_input_tokens_seen": 7779136,
      "step": 11910
    },
    {
      "epoch": 7.025353773584905,
      "grad_norm": 1.8927726745605469,
      "learning_rate": 8.19735324801253e-06,
      "loss": 0.4635,
      "num_input_tokens_seen": 7782112,
      "step": 11915
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 1.3859190940856934,
      "learning_rate": 8.195374882333178e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 7784928,
      "step": 11920
    },
    {
      "epoch": 7.03125,
      "grad_norm": 2.154345750808716,
      "learning_rate": 8.193395670664555e-06,
      "loss": 0.5005,
      "num_input_tokens_seen": 7789952,
      "step": 11925
    },
    {
      "epoch": 7.034198113207547,
      "grad_norm": 1.2849678993225098,
      "learning_rate": 8.191415613530667e-06,
      "loss": 0.3992,
      "num_input_tokens_seen": 7793344,
      "step": 11930
    },
    {
      "epoch": 7.037146226415095,
      "grad_norm": 2.2049002647399902,
      "learning_rate": 8.189434711455739e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 7797216,
      "step": 11935
    },
    {
      "epoch": 7.040094339622642,
      "grad_norm": 1.7427544593811035,
      "learning_rate": 8.187452964964226e-06,
      "loss": 0.5558,
      "num_input_tokens_seen": 7799872,
      "step": 11940
    },
    {
      "epoch": 7.043042452830188,
      "grad_norm": 1.2393203973770142,
      "learning_rate": 8.185470374580805e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 7803936,
      "step": 11945
    },
    {
      "epoch": 7.0459905660377355,
      "grad_norm": 2.892695188522339,
      "learning_rate": 8.183486940830371e-06,
      "loss": 0.3584,
      "num_input_tokens_seen": 7808672,
      "step": 11950
    },
    {
      "epoch": 7.048938679245283,
      "grad_norm": 1.130380392074585,
      "learning_rate": 8.18150266423805e-06,
      "loss": 0.513,
      "num_input_tokens_seen": 7812480,
      "step": 11955
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 2.0773181915283203,
      "learning_rate": 8.179517545329188e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 7815776,
      "step": 11960
    },
    {
      "epoch": 7.054834905660377,
      "grad_norm": 1.8320262432098389,
      "learning_rate": 8.177531584629353e-06,
      "loss": 0.3004,
      "num_input_tokens_seen": 7818912,
      "step": 11965
    },
    {
      "epoch": 7.057783018867925,
      "grad_norm": 2.433553695678711,
      "learning_rate": 8.175544782664335e-06,
      "loss": 0.2993,
      "num_input_tokens_seen": 7821920,
      "step": 11970
    },
    {
      "epoch": 7.060731132075472,
      "grad_norm": 3.495915651321411,
      "learning_rate": 8.173557139960151e-06,
      "loss": 0.4424,
      "num_input_tokens_seen": 7825312,
      "step": 11975
    },
    {
      "epoch": 7.063679245283019,
      "grad_norm": 1.5818253755569458,
      "learning_rate": 8.17156865704304e-06,
      "loss": 0.3301,
      "num_input_tokens_seen": 7828928,
      "step": 11980
    },
    {
      "epoch": 7.066627358490566,
      "grad_norm": 1.9010298252105713,
      "learning_rate": 8.169579334439453e-06,
      "loss": 0.4302,
      "num_input_tokens_seen": 7832256,
      "step": 11985
    },
    {
      "epoch": 7.069575471698113,
      "grad_norm": 2.3041789531707764,
      "learning_rate": 8.16758917267608e-06,
      "loss": 0.3873,
      "num_input_tokens_seen": 7835264,
      "step": 11990
    },
    {
      "epoch": 7.07252358490566,
      "grad_norm": 2.420262575149536,
      "learning_rate": 8.165598172279822e-06,
      "loss": 0.5285,
      "num_input_tokens_seen": 7838080,
      "step": 11995
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 4.761005878448486,
      "learning_rate": 8.163606333777804e-06,
      "loss": 0.5415,
      "num_input_tokens_seen": 7840800,
      "step": 12000
    },
    {
      "epoch": 7.078419811320755,
      "grad_norm": 1.8994665145874023,
      "learning_rate": 8.161613657697374e-06,
      "loss": 0.329,
      "num_input_tokens_seen": 7844032,
      "step": 12005
    },
    {
      "epoch": 7.081367924528302,
      "grad_norm": 10.220967292785645,
      "learning_rate": 8.159620144566103e-06,
      "loss": 0.5029,
      "num_input_tokens_seen": 7846624,
      "step": 12010
    },
    {
      "epoch": 7.084316037735849,
      "grad_norm": 2.221285581588745,
      "learning_rate": 8.157625794911782e-06,
      "loss": 0.675,
      "num_input_tokens_seen": 7850656,
      "step": 12015
    },
    {
      "epoch": 7.087264150943396,
      "grad_norm": 1.529366374015808,
      "learning_rate": 8.155630609262424e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 7853632,
      "step": 12020
    },
    {
      "epoch": 7.090212264150943,
      "grad_norm": 2.22088885307312,
      "learning_rate": 8.153634588146262e-06,
      "loss": 0.3657,
      "num_input_tokens_seen": 7856704,
      "step": 12025
    },
    {
      "epoch": 7.09316037735849,
      "grad_norm": 1.7446376085281372,
      "learning_rate": 8.15163773209175e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 7860064,
      "step": 12030
    },
    {
      "epoch": 7.096108490566038,
      "grad_norm": 2.5884950160980225,
      "learning_rate": 8.149640041627566e-06,
      "loss": 0.4992,
      "num_input_tokens_seen": 7863456,
      "step": 12035
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 2.640911817550659,
      "learning_rate": 8.147641517282608e-06,
      "loss": 0.6397,
      "num_input_tokens_seen": 7866560,
      "step": 12040
    },
    {
      "epoch": 7.102004716981132,
      "grad_norm": 1.080689549446106,
      "learning_rate": 8.145642159585992e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 7870784,
      "step": 12045
    },
    {
      "epoch": 7.1049528301886795,
      "grad_norm": 1.7651944160461426,
      "learning_rate": 8.143641969067057e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 7873728,
      "step": 12050
    },
    {
      "epoch": 7.107900943396227,
      "grad_norm": 1.8429083824157715,
      "learning_rate": 8.141640946255362e-06,
      "loss": 0.3343,
      "num_input_tokens_seen": 7876480,
      "step": 12055
    },
    {
      "epoch": 7.110849056603773,
      "grad_norm": 2.1653099060058594,
      "learning_rate": 8.139639091680687e-06,
      "loss": 0.384,
      "num_input_tokens_seen": 7880224,
      "step": 12060
    },
    {
      "epoch": 7.1137971698113205,
      "grad_norm": 1.5170201063156128,
      "learning_rate": 8.137636405873031e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 7883680,
      "step": 12065
    },
    {
      "epoch": 7.116745283018868,
      "grad_norm": 2.324798345565796,
      "learning_rate": 8.135632889362614e-06,
      "loss": 0.4022,
      "num_input_tokens_seen": 7886976,
      "step": 12070
    },
    {
      "epoch": 7.119693396226415,
      "grad_norm": 1.5216901302337646,
      "learning_rate": 8.133628542679879e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 7889536,
      "step": 12075
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 1.9480828046798706,
      "learning_rate": 8.131623366355478e-06,
      "loss": 0.5026,
      "num_input_tokens_seen": 7892384,
      "step": 12080
    },
    {
      "epoch": 7.12558962264151,
      "grad_norm": 2.2878921031951904,
      "learning_rate": 8.129617360920297e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 7895488,
      "step": 12085
    },
    {
      "epoch": 7.128537735849057,
      "grad_norm": 4.224729061126709,
      "learning_rate": 8.12761052690543e-06,
      "loss": 0.4001,
      "num_input_tokens_seen": 7898720,
      "step": 12090
    },
    {
      "epoch": 7.131485849056604,
      "grad_norm": 2.8494932651519775,
      "learning_rate": 8.125602864842197e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 7901728,
      "step": 12095
    },
    {
      "epoch": 7.134433962264151,
      "grad_norm": 2.5959672927856445,
      "learning_rate": 8.123594375262135e-06,
      "loss": 0.3214,
      "num_input_tokens_seen": 7904448,
      "step": 12100
    },
    {
      "epoch": 7.137382075471698,
      "grad_norm": 2.5960066318511963,
      "learning_rate": 8.121585058697e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 7907680,
      "step": 12105
    },
    {
      "epoch": 7.140330188679245,
      "grad_norm": 1.3043678998947144,
      "learning_rate": 8.119574915678767e-06,
      "loss": 0.3146,
      "num_input_tokens_seen": 7910560,
      "step": 12110
    },
    {
      "epoch": 7.1432783018867925,
      "grad_norm": 1.9987351894378662,
      "learning_rate": 8.117563946739632e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 7913824,
      "step": 12115
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 1.7881391048431396,
      "learning_rate": 8.115552152412006e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 7916864,
      "step": 12120
    },
    {
      "epoch": 7.149174528301887,
      "grad_norm": 1.8918739557266235,
      "learning_rate": 8.11353953322852e-06,
      "loss": 0.299,
      "num_input_tokens_seen": 7919456,
      "step": 12125
    },
    {
      "epoch": 7.152122641509434,
      "grad_norm": 2.5359981060028076,
      "learning_rate": 8.111526089722024e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 7922592,
      "step": 12130
    },
    {
      "epoch": 7.155070754716981,
      "grad_norm": 1.628157377243042,
      "learning_rate": 8.109511822425586e-06,
      "loss": 0.4687,
      "num_input_tokens_seen": 7925344,
      "step": 12135
    },
    {
      "epoch": 7.158018867924528,
      "grad_norm": 2.7215464115142822,
      "learning_rate": 8.107496731872491e-06,
      "loss": 0.5015,
      "num_input_tokens_seen": 7928768,
      "step": 12140
    },
    {
      "epoch": 7.160966981132075,
      "grad_norm": 2.235278606414795,
      "learning_rate": 8.105480818596243e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 7931744,
      "step": 12145
    },
    {
      "epoch": 7.163915094339623,
      "grad_norm": 2.2407376766204834,
      "learning_rate": 8.103464083130566e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 7934560,
      "step": 12150
    },
    {
      "epoch": 7.16686320754717,
      "grad_norm": 2.6209843158721924,
      "learning_rate": 8.101446526009397e-06,
      "loss": 0.44,
      "num_input_tokens_seen": 7938432,
      "step": 12155
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 2.3419835567474365,
      "learning_rate": 8.099428147766894e-06,
      "loss": 0.4427,
      "num_input_tokens_seen": 7941408,
      "step": 12160
    },
    {
      "epoch": 7.1727594339622645,
      "grad_norm": 1.2462983131408691,
      "learning_rate": 8.097408948937431e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 7944608,
      "step": 12165
    },
    {
      "epoch": 7.175707547169812,
      "grad_norm": 1.8592244386672974,
      "learning_rate": 8.095388930055599e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 7947520,
      "step": 12170
    },
    {
      "epoch": 7.178655660377358,
      "grad_norm": 1.6499296426773071,
      "learning_rate": 8.093368091656209e-06,
      "loss": 0.3685,
      "num_input_tokens_seen": 7950176,
      "step": 12175
    },
    {
      "epoch": 7.181603773584905,
      "grad_norm": 2.2457525730133057,
      "learning_rate": 8.091346434274284e-06,
      "loss": 0.5016,
      "num_input_tokens_seen": 7953536,
      "step": 12180
    },
    {
      "epoch": 7.184551886792453,
      "grad_norm": 4.398120403289795,
      "learning_rate": 8.089323958445068e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 7956352,
      "step": 12185
    },
    {
      "epoch": 7.1875,
      "grad_norm": 2.2671868801116943,
      "learning_rate": 8.08730066470402e-06,
      "loss": 0.44,
      "num_input_tokens_seen": 7959264,
      "step": 12190
    },
    {
      "epoch": 7.190448113207547,
      "grad_norm": 3.5046403408050537,
      "learning_rate": 8.085276553586814e-06,
      "loss": 0.3755,
      "num_input_tokens_seen": 7961824,
      "step": 12195
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 1.523985743522644,
      "learning_rate": 8.083251625629345e-06,
      "loss": 0.4165,
      "num_input_tokens_seen": 7964896,
      "step": 12200
    },
    {
      "epoch": 7.196344339622642,
      "grad_norm": 4.03644323348999,
      "learning_rate": 8.08122588136772e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 7967392,
      "step": 12205
    },
    {
      "epoch": 7.199292452830188,
      "grad_norm": 2.884920358657837,
      "learning_rate": 8.079199321338262e-06,
      "loss": 0.3672,
      "num_input_tokens_seen": 7971424,
      "step": 12210
    },
    {
      "epoch": 7.2022405660377355,
      "grad_norm": 5.370350360870361,
      "learning_rate": 8.077171946077516e-06,
      "loss": 0.5021,
      "num_input_tokens_seen": 7974912,
      "step": 12215
    },
    {
      "epoch": 7.205188679245283,
      "grad_norm": 2.226308584213257,
      "learning_rate": 8.075143756122232e-06,
      "loss": 0.4371,
      "num_input_tokens_seen": 7977728,
      "step": 12220
    },
    {
      "epoch": 7.20813679245283,
      "grad_norm": 2.4273085594177246,
      "learning_rate": 8.073114752009388e-06,
      "loss": 0.3436,
      "num_input_tokens_seen": 7980352,
      "step": 12225
    },
    {
      "epoch": 7.211084905660377,
      "grad_norm": 1.8086092472076416,
      "learning_rate": 8.071084934276168e-06,
      "loss": 0.4196,
      "num_input_tokens_seen": 7983456,
      "step": 12230
    },
    {
      "epoch": 7.214033018867925,
      "grad_norm": 1.5103647708892822,
      "learning_rate": 8.069054303459976e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 7986656,
      "step": 12235
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 3.350346326828003,
      "learning_rate": 8.06702286009843e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 7990624,
      "step": 12240
    },
    {
      "epoch": 7.219929245283019,
      "grad_norm": 2.4687492847442627,
      "learning_rate": 8.064990604729363e-06,
      "loss": 0.4521,
      "num_input_tokens_seen": 7993760,
      "step": 12245
    },
    {
      "epoch": 7.222877358490566,
      "grad_norm": 2.9592607021331787,
      "learning_rate": 8.062957537890827e-06,
      "loss": 0.4764,
      "num_input_tokens_seen": 7996000,
      "step": 12250
    },
    {
      "epoch": 7.225825471698113,
      "grad_norm": 2.2037088871002197,
      "learning_rate": 8.060923660121081e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 8000352,
      "step": 12255
    },
    {
      "epoch": 7.22877358490566,
      "grad_norm": 1.8424619436264038,
      "learning_rate": 8.058888971958603e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 8004320,
      "step": 12260
    },
    {
      "epoch": 7.2317216981132075,
      "grad_norm": 1.953383445739746,
      "learning_rate": 8.056853473942085e-06,
      "loss": 0.3397,
      "num_input_tokens_seen": 8007872,
      "step": 12265
    },
    {
      "epoch": 7.234669811320755,
      "grad_norm": 1.4460313320159912,
      "learning_rate": 8.054817166610438e-06,
      "loss": 0.4495,
      "num_input_tokens_seen": 8011552,
      "step": 12270
    },
    {
      "epoch": 7.237617924528302,
      "grad_norm": 3.0899202823638916,
      "learning_rate": 8.052780050502781e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 8015168,
      "step": 12275
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 2.308255910873413,
      "learning_rate": 8.050742126158448e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 8017664,
      "step": 12280
    },
    {
      "epoch": 7.243514150943396,
      "grad_norm": 1.8830066919326782,
      "learning_rate": 8.04870339411699e-06,
      "loss": 0.2817,
      "num_input_tokens_seen": 8020896,
      "step": 12285
    },
    {
      "epoch": 7.246462264150943,
      "grad_norm": 2.279186964035034,
      "learning_rate": 8.046663854918166e-06,
      "loss": 0.5647,
      "num_input_tokens_seen": 8024512,
      "step": 12290
    },
    {
      "epoch": 7.24941037735849,
      "grad_norm": 1.9630522727966309,
      "learning_rate": 8.044623509101959e-06,
      "loss": 0.5489,
      "num_input_tokens_seen": 8027360,
      "step": 12295
    },
    {
      "epoch": 7.252358490566038,
      "grad_norm": 1.5792992115020752,
      "learning_rate": 8.042582357208557e-06,
      "loss": 0.4961,
      "num_input_tokens_seen": 8032128,
      "step": 12300
    },
    {
      "epoch": 7.255306603773585,
      "grad_norm": 1.0844857692718506,
      "learning_rate": 8.04054039977836e-06,
      "loss": 0.2579,
      "num_input_tokens_seen": 8035584,
      "step": 12305
    },
    {
      "epoch": 7.258254716981132,
      "grad_norm": 2.7664501667022705,
      "learning_rate": 8.038497637351992e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 8038784,
      "step": 12310
    },
    {
      "epoch": 7.2612028301886795,
      "grad_norm": 1.7518281936645508,
      "learning_rate": 8.036454070470276e-06,
      "loss": 0.3732,
      "num_input_tokens_seen": 8041344,
      "step": 12315
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 1.753612756729126,
      "learning_rate": 8.03440969967426e-06,
      "loss": 0.5332,
      "num_input_tokens_seen": 8044576,
      "step": 12320
    },
    {
      "epoch": 7.267099056603773,
      "grad_norm": 1.3995728492736816,
      "learning_rate": 8.032364525505198e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 8047648,
      "step": 12325
    },
    {
      "epoch": 7.2700471698113205,
      "grad_norm": 2.8808352947235107,
      "learning_rate": 8.030318548504561e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 8051040,
      "step": 12330
    },
    {
      "epoch": 7.272995283018868,
      "grad_norm": 1.8577425479888916,
      "learning_rate": 8.028271769214026e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 8054336,
      "step": 12335
    },
    {
      "epoch": 7.275943396226415,
      "grad_norm": 1.3512754440307617,
      "learning_rate": 8.02622418817549e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 8057920,
      "step": 12340
    },
    {
      "epoch": 7.278891509433962,
      "grad_norm": 1.8645884990692139,
      "learning_rate": 8.024175805931056e-06,
      "loss": 0.3804,
      "num_input_tokens_seen": 8061216,
      "step": 12345
    },
    {
      "epoch": 7.28183962264151,
      "grad_norm": 2.570061683654785,
      "learning_rate": 8.022126623023045e-06,
      "loss": 0.5239,
      "num_input_tokens_seen": 8065088,
      "step": 12350
    },
    {
      "epoch": 7.284787735849057,
      "grad_norm": 2.0161397457122803,
      "learning_rate": 8.020076639993987e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 8068416,
      "step": 12355
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 2.0487794876098633,
      "learning_rate": 8.018025857386622e-06,
      "loss": 0.3835,
      "num_input_tokens_seen": 8071328,
      "step": 12360
    },
    {
      "epoch": 7.290683962264151,
      "grad_norm": 2.080355167388916,
      "learning_rate": 8.015974275743905e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 8073952,
      "step": 12365
    },
    {
      "epoch": 7.293632075471698,
      "grad_norm": 2.274325132369995,
      "learning_rate": 8.013921895609e-06,
      "loss": 0.4223,
      "num_input_tokens_seen": 8076704,
      "step": 12370
    },
    {
      "epoch": 7.296580188679245,
      "grad_norm": 1.8114523887634277,
      "learning_rate": 8.011868717525283e-06,
      "loss": 0.3102,
      "num_input_tokens_seen": 8080480,
      "step": 12375
    },
    {
      "epoch": 7.2995283018867925,
      "grad_norm": 2.077540874481201,
      "learning_rate": 8.009814742036343e-06,
      "loss": 0.4929,
      "num_input_tokens_seen": 8083360,
      "step": 12380
    },
    {
      "epoch": 7.30247641509434,
      "grad_norm": 1.4760488271713257,
      "learning_rate": 8.007759969685979e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 8086880,
      "step": 12385
    },
    {
      "epoch": 7.305424528301887,
      "grad_norm": 4.747032642364502,
      "learning_rate": 8.005704401018199e-06,
      "loss": 0.4224,
      "num_input_tokens_seen": 8089088,
      "step": 12390
    },
    {
      "epoch": 7.308372641509434,
      "grad_norm": 3.4902751445770264,
      "learning_rate": 8.003648036577226e-06,
      "loss": 0.409,
      "num_input_tokens_seen": 8092736,
      "step": 12395
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 1.7141029834747314,
      "learning_rate": 8.00159087690749e-06,
      "loss": 0.3797,
      "num_input_tokens_seen": 8096064,
      "step": 12400
    },
    {
      "epoch": 7.314268867924528,
      "grad_norm": 1.9947317838668823,
      "learning_rate": 7.999532922553635e-06,
      "loss": 0.3876,
      "num_input_tokens_seen": 8099168,
      "step": 12405
    },
    {
      "epoch": 7.317216981132075,
      "grad_norm": 1.621710181236267,
      "learning_rate": 7.997474174060508e-06,
      "loss": 0.3146,
      "num_input_tokens_seen": 8103456,
      "step": 12410
    },
    {
      "epoch": 7.320165094339623,
      "grad_norm": 1.9514803886413574,
      "learning_rate": 7.995414631973179e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 8106560,
      "step": 12415
    },
    {
      "epoch": 7.32311320754717,
      "grad_norm": 1.9218391180038452,
      "learning_rate": 7.993354296836914e-06,
      "loss": 0.4568,
      "num_input_tokens_seen": 8109376,
      "step": 12420
    },
    {
      "epoch": 7.326061320754717,
      "grad_norm": 2.862060785293579,
      "learning_rate": 7.991293169197198e-06,
      "loss": 0.4614,
      "num_input_tokens_seen": 8113184,
      "step": 12425
    },
    {
      "epoch": 7.3290094339622645,
      "grad_norm": 2.180881977081299,
      "learning_rate": 7.989231249599725e-06,
      "loss": 0.4705,
      "num_input_tokens_seen": 8116512,
      "step": 12430
    },
    {
      "epoch": 7.331957547169811,
      "grad_norm": 2.0748820304870605,
      "learning_rate": 7.987168538590395e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 8119392,
      "step": 12435
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 1.9564100503921509,
      "learning_rate": 7.985105036715322e-06,
      "loss": 0.5489,
      "num_input_tokens_seen": 8122272,
      "step": 12440
    },
    {
      "epoch": 7.337853773584905,
      "grad_norm": 1.8913486003875732,
      "learning_rate": 7.983040744520823e-06,
      "loss": 0.4766,
      "num_input_tokens_seen": 8125664,
      "step": 12445
    },
    {
      "epoch": 7.340801886792453,
      "grad_norm": 2.1861531734466553,
      "learning_rate": 7.980975662553432e-06,
      "loss": 0.4766,
      "num_input_tokens_seen": 8128512,
      "step": 12450
    },
    {
      "epoch": 7.34375,
      "grad_norm": 1.673919439315796,
      "learning_rate": 7.978909791359888e-06,
      "loss": 0.381,
      "num_input_tokens_seen": 8131488,
      "step": 12455
    },
    {
      "epoch": 7.346698113207547,
      "grad_norm": 1.927797555923462,
      "learning_rate": 7.976843131487136e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 8135168,
      "step": 12460
    },
    {
      "epoch": 7.349646226415095,
      "grad_norm": 1.5589667558670044,
      "learning_rate": 7.974775683482337e-06,
      "loss": 0.419,
      "num_input_tokens_seen": 8138144,
      "step": 12465
    },
    {
      "epoch": 7.352594339622642,
      "grad_norm": 1.1726197004318237,
      "learning_rate": 7.972707447892855e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 8141376,
      "step": 12470
    },
    {
      "epoch": 7.355542452830189,
      "grad_norm": 4.077281475067139,
      "learning_rate": 7.970638425266264e-06,
      "loss": 0.4861,
      "num_input_tokens_seen": 8144416,
      "step": 12475
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 1.679072618484497,
      "learning_rate": 7.968568616150349e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 8146720,
      "step": 12480
    },
    {
      "epoch": 7.361438679245283,
      "grad_norm": 2.1428542137145996,
      "learning_rate": 7.966498021093096e-06,
      "loss": 0.3848,
      "num_input_tokens_seen": 8150496,
      "step": 12485
    },
    {
      "epoch": 7.36438679245283,
      "grad_norm": 1.7843437194824219,
      "learning_rate": 7.96442664064271e-06,
      "loss": 0.3761,
      "num_input_tokens_seen": 8153312,
      "step": 12490
    },
    {
      "epoch": 7.367334905660377,
      "grad_norm": 2.0392019748687744,
      "learning_rate": 7.962354475347593e-06,
      "loss": 0.426,
      "num_input_tokens_seen": 8156544,
      "step": 12495
    },
    {
      "epoch": 7.370283018867925,
      "grad_norm": 2.606898546218872,
      "learning_rate": 7.960281525756364e-06,
      "loss": 0.4223,
      "num_input_tokens_seen": 8158688,
      "step": 12500
    },
    {
      "epoch": 7.373231132075472,
      "grad_norm": 1.8667136430740356,
      "learning_rate": 7.95820779241784e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 8162592,
      "step": 12505
    },
    {
      "epoch": 7.376179245283019,
      "grad_norm": 2.3526113033294678,
      "learning_rate": 7.956133275881055e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 8165568,
      "step": 12510
    },
    {
      "epoch": 7.379127358490566,
      "grad_norm": 1.1191987991333008,
      "learning_rate": 7.954057976695244e-06,
      "loss": 0.2712,
      "num_input_tokens_seen": 8169184,
      "step": 12515
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 2.8422324657440186,
      "learning_rate": 7.951981895409854e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 8172160,
      "step": 12520
    },
    {
      "epoch": 7.38502358490566,
      "grad_norm": 3.6238527297973633,
      "learning_rate": 7.949905032574534e-06,
      "loss": 0.3948,
      "num_input_tokens_seen": 8174368,
      "step": 12525
    },
    {
      "epoch": 7.3879716981132075,
      "grad_norm": 1.7540556192398071,
      "learning_rate": 7.947827388739145e-06,
      "loss": 0.4525,
      "num_input_tokens_seen": 8177472,
      "step": 12530
    },
    {
      "epoch": 7.390919811320755,
      "grad_norm": 2.163388729095459,
      "learning_rate": 7.945748964453747e-06,
      "loss": 0.4632,
      "num_input_tokens_seen": 8181600,
      "step": 12535
    },
    {
      "epoch": 7.393867924528302,
      "grad_norm": 2.994271755218506,
      "learning_rate": 7.943669760268618e-06,
      "loss": 0.6428,
      "num_input_tokens_seen": 8185664,
      "step": 12540
    },
    {
      "epoch": 7.396816037735849,
      "grad_norm": 1.9568177461624146,
      "learning_rate": 7.941589776734232e-06,
      "loss": 0.3386,
      "num_input_tokens_seen": 8188736,
      "step": 12545
    },
    {
      "epoch": 7.399764150943396,
      "grad_norm": 3.4246387481689453,
      "learning_rate": 7.939509014401277e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 8191232,
      "step": 12550
    },
    {
      "epoch": 7.402712264150943,
      "grad_norm": 2.4000089168548584,
      "learning_rate": 7.93742747382064e-06,
      "loss": 0.3517,
      "num_input_tokens_seen": 8195328,
      "step": 12555
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 1.3705377578735352,
      "learning_rate": 7.93534515554342e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 8198688,
      "step": 12560
    },
    {
      "epoch": 7.408608490566038,
      "grad_norm": 1.7115085124969482,
      "learning_rate": 7.933262060120918e-06,
      "loss": 0.276,
      "num_input_tokens_seen": 8201824,
      "step": 12565
    },
    {
      "epoch": 7.411556603773585,
      "grad_norm": 2.026766300201416,
      "learning_rate": 7.931178188104646e-06,
      "loss": 0.3098,
      "num_input_tokens_seen": 8205248,
      "step": 12570
    },
    {
      "epoch": 7.414504716981132,
      "grad_norm": 2.8529162406921387,
      "learning_rate": 7.929093540046317e-06,
      "loss": 0.5164,
      "num_input_tokens_seen": 8208256,
      "step": 12575
    },
    {
      "epoch": 7.4174528301886795,
      "grad_norm": 2.6985702514648438,
      "learning_rate": 7.927008116497848e-06,
      "loss": 0.4581,
      "num_input_tokens_seen": 8210848,
      "step": 12580
    },
    {
      "epoch": 7.420400943396227,
      "grad_norm": 2.151388645172119,
      "learning_rate": 7.924921918011366e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 8213664,
      "step": 12585
    },
    {
      "epoch": 7.423349056603773,
      "grad_norm": 0.35098904371261597,
      "learning_rate": 7.9228349451392e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 8219168,
      "step": 12590
    },
    {
      "epoch": 7.4262971698113205,
      "grad_norm": 1.8342362642288208,
      "learning_rate": 7.920747198433884e-06,
      "loss": 0.419,
      "num_input_tokens_seen": 8221664,
      "step": 12595
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 1.7671475410461426,
      "learning_rate": 7.91865867844816e-06,
      "loss": 0.3728,
      "num_input_tokens_seen": 8225184,
      "step": 12600
    },
    {
      "epoch": 7.432193396226415,
      "grad_norm": 3.676335573196411,
      "learning_rate": 7.916569385734976e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 8228352,
      "step": 12605
    },
    {
      "epoch": 7.435141509433962,
      "grad_norm": 1.8697593212127686,
      "learning_rate": 7.914479320847474e-06,
      "loss": 0.4316,
      "num_input_tokens_seen": 8231072,
      "step": 12610
    },
    {
      "epoch": 7.43808962264151,
      "grad_norm": 2.0146260261535645,
      "learning_rate": 7.912388484339012e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 8234016,
      "step": 12615
    },
    {
      "epoch": 7.441037735849057,
      "grad_norm": 1.9886372089385986,
      "learning_rate": 7.910296876763147e-06,
      "loss": 0.4383,
      "num_input_tokens_seen": 8241920,
      "step": 12620
    },
    {
      "epoch": 7.443985849056604,
      "grad_norm": 1.385977029800415,
      "learning_rate": 7.90820449867364e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 8245760,
      "step": 12625
    },
    {
      "epoch": 7.446933962264151,
      "grad_norm": 1.5211071968078613,
      "learning_rate": 7.90611135062446e-06,
      "loss": 0.3443,
      "num_input_tokens_seen": 8248320,
      "step": 12630
    },
    {
      "epoch": 7.449882075471698,
      "grad_norm": 1.7209607362747192,
      "learning_rate": 7.904017433169775e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 8251744,
      "step": 12635
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 2.2223927974700928,
      "learning_rate": 7.901922746863957e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 8254400,
      "step": 12640
    },
    {
      "epoch": 7.4557783018867925,
      "grad_norm": 1.497497320175171,
      "learning_rate": 7.899827292261589e-06,
      "loss": 0.4075,
      "num_input_tokens_seen": 8257632,
      "step": 12645
    },
    {
      "epoch": 7.45872641509434,
      "grad_norm": 1.6228500604629517,
      "learning_rate": 7.897731069917444e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 8261184,
      "step": 12650
    },
    {
      "epoch": 7.461674528301887,
      "grad_norm": 1.6650136709213257,
      "learning_rate": 7.895634080386512e-06,
      "loss": 0.4241,
      "num_input_tokens_seen": 8264192,
      "step": 12655
    },
    {
      "epoch": 7.464622641509434,
      "grad_norm": 3.4806904792785645,
      "learning_rate": 7.893536324223977e-06,
      "loss": 0.5294,
      "num_input_tokens_seen": 8267680,
      "step": 12660
    },
    {
      "epoch": 7.467570754716981,
      "grad_norm": 3.2399303913116455,
      "learning_rate": 7.89143780198523e-06,
      "loss": 0.4371,
      "num_input_tokens_seen": 8270528,
      "step": 12665
    },
    {
      "epoch": 7.470518867924528,
      "grad_norm": 1.7679998874664307,
      "learning_rate": 7.889338514225862e-06,
      "loss": 0.3671,
      "num_input_tokens_seen": 8274016,
      "step": 12670
    },
    {
      "epoch": 7.473466981132075,
      "grad_norm": 4.800253868103027,
      "learning_rate": 7.887238461501671e-06,
      "loss": 0.4743,
      "num_input_tokens_seen": 8276832,
      "step": 12675
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 2.879516363143921,
      "learning_rate": 7.885137644368654e-06,
      "loss": 0.5329,
      "num_input_tokens_seen": 8281472,
      "step": 12680
    },
    {
      "epoch": 7.47936320754717,
      "grad_norm": 2.314478874206543,
      "learning_rate": 7.883036063383012e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 8285056,
      "step": 12685
    },
    {
      "epoch": 7.482311320754717,
      "grad_norm": 1.2154619693756104,
      "learning_rate": 7.880933719101148e-06,
      "loss": 0.4776,
      "num_input_tokens_seen": 8289152,
      "step": 12690
    },
    {
      "epoch": 7.4852594339622645,
      "grad_norm": 1.9537273645401,
      "learning_rate": 7.878830612079664e-06,
      "loss": 0.525,
      "num_input_tokens_seen": 8291680,
      "step": 12695
    },
    {
      "epoch": 7.488207547169811,
      "grad_norm": 2.777642011642456,
      "learning_rate": 7.876726742875369e-06,
      "loss": 0.5194,
      "num_input_tokens_seen": 8295040,
      "step": 12700
    },
    {
      "epoch": 7.491155660377358,
      "grad_norm": 1.4765856266021729,
      "learning_rate": 7.874622112045269e-06,
      "loss": 0.4806,
      "num_input_tokens_seen": 8298464,
      "step": 12705
    },
    {
      "epoch": 7.494103773584905,
      "grad_norm": 1.7408586740493774,
      "learning_rate": 7.872516720146578e-06,
      "loss": 0.3408,
      "num_input_tokens_seen": 8301248,
      "step": 12710
    },
    {
      "epoch": 7.497051886792453,
      "grad_norm": 2.2675421237945557,
      "learning_rate": 7.870410567736705e-06,
      "loss": 0.4959,
      "num_input_tokens_seen": 8305216,
      "step": 12715
    },
    {
      "epoch": 7.5,
      "grad_norm": 2.9936411380767822,
      "learning_rate": 7.868303655373264e-06,
      "loss": 0.489,
      "num_input_tokens_seen": 8309408,
      "step": 12720
    },
    {
      "epoch": 7.502948113207547,
      "grad_norm": 2.704676389694214,
      "learning_rate": 7.866195983614066e-06,
      "loss": 0.4774,
      "num_input_tokens_seen": 8313152,
      "step": 12725
    },
    {
      "epoch": 7.505896226415095,
      "grad_norm": 1.9438819885253906,
      "learning_rate": 7.864087553017133e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 8316000,
      "step": 12730
    },
    {
      "epoch": 7.508844339622642,
      "grad_norm": 1.4258575439453125,
      "learning_rate": 7.861978364140674e-06,
      "loss": 0.4859,
      "num_input_tokens_seen": 8319648,
      "step": 12735
    },
    {
      "epoch": 7.511792452830189,
      "grad_norm": 2.585756778717041,
      "learning_rate": 7.859868417543109e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 8323360,
      "step": 12740
    },
    {
      "epoch": 7.5147405660377355,
      "grad_norm": 2.237520217895508,
      "learning_rate": 7.857757713783055e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 8326144,
      "step": 12745
    },
    {
      "epoch": 7.517688679245283,
      "grad_norm": 3.7269821166992188,
      "learning_rate": 7.855646253419331e-06,
      "loss": 0.4049,
      "num_input_tokens_seen": 8331968,
      "step": 12750
    },
    {
      "epoch": 7.52063679245283,
      "grad_norm": 2.6455037593841553,
      "learning_rate": 7.853534037010952e-06,
      "loss": 0.3991,
      "num_input_tokens_seen": 8334496,
      "step": 12755
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 2.4723479747772217,
      "learning_rate": 7.851421065117142e-06,
      "loss": 0.5013,
      "num_input_tokens_seen": 8337376,
      "step": 12760
    },
    {
      "epoch": 7.526533018867925,
      "grad_norm": 4.94113826751709,
      "learning_rate": 7.849307338297314e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 8340768,
      "step": 12765
    },
    {
      "epoch": 7.529481132075472,
      "grad_norm": 1.1403754949569702,
      "learning_rate": 7.847192857111087e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 8344160,
      "step": 12770
    },
    {
      "epoch": 7.532429245283019,
      "grad_norm": 2.172560214996338,
      "learning_rate": 7.845077622118282e-06,
      "loss": 0.5216,
      "num_input_tokens_seen": 8347264,
      "step": 12775
    },
    {
      "epoch": 7.535377358490566,
      "grad_norm": 1.6669117212295532,
      "learning_rate": 7.842961633878916e-06,
      "loss": 0.3537,
      "num_input_tokens_seen": 8350400,
      "step": 12780
    },
    {
      "epoch": 7.538325471698113,
      "grad_norm": 2.1495351791381836,
      "learning_rate": 7.840844892953204e-06,
      "loss": 0.4784,
      "num_input_tokens_seen": 8352992,
      "step": 12785
    },
    {
      "epoch": 7.54127358490566,
      "grad_norm": 2.76818585395813,
      "learning_rate": 7.838727399901562e-06,
      "loss": 0.4667,
      "num_input_tokens_seen": 8355488,
      "step": 12790
    },
    {
      "epoch": 7.5442216981132075,
      "grad_norm": 7.441953659057617,
      "learning_rate": 7.836609155284607e-06,
      "loss": 0.5358,
      "num_input_tokens_seen": 8359008,
      "step": 12795
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 1.5105031728744507,
      "learning_rate": 7.834490159663154e-06,
      "loss": 0.5112,
      "num_input_tokens_seen": 8362304,
      "step": 12800
    },
    {
      "epoch": 7.550117924528302,
      "grad_norm": 1.9600744247436523,
      "learning_rate": 7.832370413598215e-06,
      "loss": 0.4039,
      "num_input_tokens_seen": 8365312,
      "step": 12805
    },
    {
      "epoch": 7.553066037735849,
      "grad_norm": 1.8072161674499512,
      "learning_rate": 7.830249917651003e-06,
      "loss": 0.3572,
      "num_input_tokens_seen": 8369120,
      "step": 12810
    },
    {
      "epoch": 7.556014150943396,
      "grad_norm": 2.0145347118377686,
      "learning_rate": 7.828128672382926e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 8372128,
      "step": 12815
    },
    {
      "epoch": 7.558962264150943,
      "grad_norm": 2.1070845127105713,
      "learning_rate": 7.826006678355596e-06,
      "loss": 0.43,
      "num_input_tokens_seen": 8375712,
      "step": 12820
    },
    {
      "epoch": 7.56191037735849,
      "grad_norm": 2.21620774269104,
      "learning_rate": 7.823883936130817e-06,
      "loss": 0.5394,
      "num_input_tokens_seen": 8378240,
      "step": 12825
    },
    {
      "epoch": 7.564858490566038,
      "grad_norm": 3.8304076194763184,
      "learning_rate": 7.821760446270597e-06,
      "loss": 0.4032,
      "num_input_tokens_seen": 8380992,
      "step": 12830
    },
    {
      "epoch": 7.567806603773585,
      "grad_norm": 1.8440310955047607,
      "learning_rate": 7.819636209337136e-06,
      "loss": 0.5093,
      "num_input_tokens_seen": 8384000,
      "step": 12835
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 2.327598810195923,
      "learning_rate": 7.817511225892838e-06,
      "loss": 0.5224,
      "num_input_tokens_seen": 8387200,
      "step": 12840
    },
    {
      "epoch": 7.5737028301886795,
      "grad_norm": 1.9870350360870361,
      "learning_rate": 7.8153854965003e-06,
      "loss": 0.4961,
      "num_input_tokens_seen": 8389984,
      "step": 12845
    },
    {
      "epoch": 7.576650943396227,
      "grad_norm": 2.839301109313965,
      "learning_rate": 7.813259021722319e-06,
      "loss": 0.426,
      "num_input_tokens_seen": 8392640,
      "step": 12850
    },
    {
      "epoch": 7.579599056603773,
      "grad_norm": 1.59361732006073,
      "learning_rate": 7.811131802121885e-06,
      "loss": 0.4758,
      "num_input_tokens_seen": 8395744,
      "step": 12855
    },
    {
      "epoch": 7.5825471698113205,
      "grad_norm": 3.2639541625976562,
      "learning_rate": 7.809003838262193e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 8398272,
      "step": 12860
    },
    {
      "epoch": 7.585495283018868,
      "grad_norm": 1.3946119546890259,
      "learning_rate": 7.806875130706628e-06,
      "loss": 0.3172,
      "num_input_tokens_seen": 8403264,
      "step": 12865
    },
    {
      "epoch": 7.588443396226415,
      "grad_norm": 1.6036895513534546,
      "learning_rate": 7.804745680018775e-06,
      "loss": 0.3946,
      "num_input_tokens_seen": 8406400,
      "step": 12870
    },
    {
      "epoch": 7.591391509433962,
      "grad_norm": 2.8741557598114014,
      "learning_rate": 7.802615486762418e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 8409024,
      "step": 12875
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 5.368066787719727,
      "learning_rate": 7.800484551501528e-06,
      "loss": 0.5098,
      "num_input_tokens_seen": 8412224,
      "step": 12880
    },
    {
      "epoch": 7.597287735849057,
      "grad_norm": 1.3189276456832886,
      "learning_rate": 7.798352874800285e-06,
      "loss": 0.2874,
      "num_input_tokens_seen": 8414944,
      "step": 12885
    },
    {
      "epoch": 7.600235849056604,
      "grad_norm": 1.333483099937439,
      "learning_rate": 7.79622045722306e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 8418432,
      "step": 12890
    },
    {
      "epoch": 7.603183962264151,
      "grad_norm": 1.0277557373046875,
      "learning_rate": 7.794087299334416e-06,
      "loss": 0.2895,
      "num_input_tokens_seen": 8420992,
      "step": 12895
    },
    {
      "epoch": 7.606132075471698,
      "grad_norm": 2.953380584716797,
      "learning_rate": 7.79195340169912e-06,
      "loss": 0.553,
      "num_input_tokens_seen": 8423968,
      "step": 12900
    },
    {
      "epoch": 7.609080188679245,
      "grad_norm": 1.8810580968856812,
      "learning_rate": 7.789818764882127e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 8427872,
      "step": 12905
    },
    {
      "epoch": 7.6120283018867925,
      "grad_norm": 2.841675281524658,
      "learning_rate": 7.78768338944859e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 8430464,
      "step": 12910
    },
    {
      "epoch": 7.61497641509434,
      "grad_norm": 2.060729742050171,
      "learning_rate": 7.785547275963865e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 8433088,
      "step": 12915
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 2.111762762069702,
      "learning_rate": 7.783410424993492e-06,
      "loss": 0.4823,
      "num_input_tokens_seen": 8435616,
      "step": 12920
    },
    {
      "epoch": 7.620872641509434,
      "grad_norm": 2.113705635070801,
      "learning_rate": 7.781272837103213e-06,
      "loss": 0.4319,
      "num_input_tokens_seen": 8438592,
      "step": 12925
    },
    {
      "epoch": 7.623820754716981,
      "grad_norm": 2.253535032272339,
      "learning_rate": 7.779134512858964e-06,
      "loss": 0.4764,
      "num_input_tokens_seen": 8442016,
      "step": 12930
    },
    {
      "epoch": 7.626768867924528,
      "grad_norm": 2.140029191970825,
      "learning_rate": 7.776995452826876e-06,
      "loss": 0.3972,
      "num_input_tokens_seen": 8444640,
      "step": 12935
    },
    {
      "epoch": 7.629716981132075,
      "grad_norm": 1.831196665763855,
      "learning_rate": 7.774855657573274e-06,
      "loss": 0.4029,
      "num_input_tokens_seen": 8448032,
      "step": 12940
    },
    {
      "epoch": 7.632665094339623,
      "grad_norm": 2.379000186920166,
      "learning_rate": 7.772715127664676e-06,
      "loss": 0.4368,
      "num_input_tokens_seen": 8451424,
      "step": 12945
    },
    {
      "epoch": 7.63561320754717,
      "grad_norm": 0.8232985734939575,
      "learning_rate": 7.7705738636678e-06,
      "loss": 0.3655,
      "num_input_tokens_seen": 8455456,
      "step": 12950
    },
    {
      "epoch": 7.638561320754717,
      "grad_norm": 3.6188085079193115,
      "learning_rate": 7.768431866149552e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 8458880,
      "step": 12955
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 4.133970737457275,
      "learning_rate": 7.766289135677035e-06,
      "loss": 0.5544,
      "num_input_tokens_seen": 8462208,
      "step": 12960
    },
    {
      "epoch": 7.644457547169811,
      "grad_norm": 2.3921561241149902,
      "learning_rate": 7.764145672817549e-06,
      "loss": 0.3443,
      "num_input_tokens_seen": 8465344,
      "step": 12965
    },
    {
      "epoch": 7.647405660377358,
      "grad_norm": 1.764865756034851,
      "learning_rate": 7.762001478138583e-06,
      "loss": 0.4881,
      "num_input_tokens_seen": 8469984,
      "step": 12970
    },
    {
      "epoch": 7.650353773584905,
      "grad_norm": 1.5439051389694214,
      "learning_rate": 7.759856552207822e-06,
      "loss": 0.5774,
      "num_input_tokens_seen": 8473056,
      "step": 12975
    },
    {
      "epoch": 7.653301886792453,
      "grad_norm": 3.051442861557007,
      "learning_rate": 7.757710895593144e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 8476800,
      "step": 12980
    },
    {
      "epoch": 7.65625,
      "grad_norm": 1.6308563947677612,
      "learning_rate": 7.755564508862623e-06,
      "loss": 0.4808,
      "num_input_tokens_seen": 8479712,
      "step": 12985
    },
    {
      "epoch": 7.659198113207547,
      "grad_norm": 2.2528645992279053,
      "learning_rate": 7.753417392584522e-06,
      "loss": 0.488,
      "num_input_tokens_seen": 8482432,
      "step": 12990
    },
    {
      "epoch": 7.662146226415095,
      "grad_norm": 6.912410259246826,
      "learning_rate": 7.751269547327298e-06,
      "loss": 0.3161,
      "num_input_tokens_seen": 8484960,
      "step": 12995
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 2.5700418949127197,
      "learning_rate": 7.749120973659606e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 8488288,
      "step": 13000
    },
    {
      "epoch": 7.668042452830189,
      "grad_norm": 1.7589749097824097,
      "learning_rate": 7.746971672150286e-06,
      "loss": 0.3011,
      "num_input_tokens_seen": 8491584,
      "step": 13005
    },
    {
      "epoch": 7.6709905660377355,
      "grad_norm": 2.172736883163452,
      "learning_rate": 7.74482164336838e-06,
      "loss": 0.5662,
      "num_input_tokens_seen": 8494656,
      "step": 13010
    },
    {
      "epoch": 7.673938679245283,
      "grad_norm": 1.9645341634750366,
      "learning_rate": 7.742670887883111e-06,
      "loss": 0.2838,
      "num_input_tokens_seen": 8496896,
      "step": 13015
    },
    {
      "epoch": 7.67688679245283,
      "grad_norm": 1.9462835788726807,
      "learning_rate": 7.740519406263905e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 8499424,
      "step": 13020
    },
    {
      "epoch": 7.679834905660377,
      "grad_norm": 1.393610954284668,
      "learning_rate": 7.738367199080376e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 8502784,
      "step": 13025
    },
    {
      "epoch": 7.682783018867925,
      "grad_norm": 0.8799817562103271,
      "learning_rate": 7.73621426690233e-06,
      "loss": 0.3568,
      "num_input_tokens_seen": 8506304,
      "step": 13030
    },
    {
      "epoch": 7.685731132075472,
      "grad_norm": 1.6788060665130615,
      "learning_rate": 7.734060610299764e-06,
      "loss": 0.3923,
      "num_input_tokens_seen": 8511488,
      "step": 13035
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 1.3899270296096802,
      "learning_rate": 7.731906229842869e-06,
      "loss": 0.4521,
      "num_input_tokens_seen": 8516192,
      "step": 13040
    },
    {
      "epoch": 7.691627358490566,
      "grad_norm": 1.2299786806106567,
      "learning_rate": 7.729751126102023e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 8519648,
      "step": 13045
    },
    {
      "epoch": 7.694575471698113,
      "grad_norm": 1.3812211751937866,
      "learning_rate": 7.727595299647805e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 8523808,
      "step": 13050
    },
    {
      "epoch": 7.69752358490566,
      "grad_norm": 1.9253334999084473,
      "learning_rate": 7.725438751050973e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 8527776,
      "step": 13055
    },
    {
      "epoch": 7.7004716981132075,
      "grad_norm": 2.042099714279175,
      "learning_rate": 7.723281480882489e-06,
      "loss": 0.5565,
      "num_input_tokens_seen": 8531072,
      "step": 13060
    },
    {
      "epoch": 7.703419811320755,
      "grad_norm": 3.6517539024353027,
      "learning_rate": 7.721123489713494e-06,
      "loss": 0.5542,
      "num_input_tokens_seen": 8534592,
      "step": 13065
    },
    {
      "epoch": 7.706367924528302,
      "grad_norm": 1.5470924377441406,
      "learning_rate": 7.718964778115328e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 8538624,
      "step": 13070
    },
    {
      "epoch": 7.709316037735849,
      "grad_norm": 2.0008859634399414,
      "learning_rate": 7.716805346659519e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 8541184,
      "step": 13075
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 1.31392240524292,
      "learning_rate": 7.714645195917788e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 8543680,
      "step": 13080
    },
    {
      "epoch": 7.715212264150943,
      "grad_norm": 1.7743488550186157,
      "learning_rate": 7.712484326462038e-06,
      "loss": 0.3328,
      "num_input_tokens_seen": 8547456,
      "step": 13085
    },
    {
      "epoch": 7.71816037735849,
      "grad_norm": 1.6390013694763184,
      "learning_rate": 7.710322738864375e-06,
      "loss": 0.5231,
      "num_input_tokens_seen": 8550528,
      "step": 13090
    },
    {
      "epoch": 7.721108490566038,
      "grad_norm": 1.85531485080719,
      "learning_rate": 7.708160433697085e-06,
      "loss": 0.518,
      "num_input_tokens_seen": 8553440,
      "step": 13095
    },
    {
      "epoch": 7.724056603773585,
      "grad_norm": 2.384963274002075,
      "learning_rate": 7.705997411532649e-06,
      "loss": 0.4927,
      "num_input_tokens_seen": 8556064,
      "step": 13100
    },
    {
      "epoch": 7.727004716981132,
      "grad_norm": 2.9405710697174072,
      "learning_rate": 7.703833672943735e-06,
      "loss": 0.4736,
      "num_input_tokens_seen": 8558624,
      "step": 13105
    },
    {
      "epoch": 7.7299528301886795,
      "grad_norm": 1.2509264945983887,
      "learning_rate": 7.701669218503206e-06,
      "loss": 0.398,
      "num_input_tokens_seen": 8562208,
      "step": 13110
    },
    {
      "epoch": 7.732900943396227,
      "grad_norm": 5.047940731048584,
      "learning_rate": 7.699504048784106e-06,
      "loss": 0.5974,
      "num_input_tokens_seen": 8565440,
      "step": 13115
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 2.7011373043060303,
      "learning_rate": 7.697338164359675e-06,
      "loss": 0.4728,
      "num_input_tokens_seen": 8567808,
      "step": 13120
    },
    {
      "epoch": 7.7387971698113205,
      "grad_norm": 1.810046672821045,
      "learning_rate": 7.69517156580334e-06,
      "loss": 0.3887,
      "num_input_tokens_seen": 8570432,
      "step": 13125
    },
    {
      "epoch": 7.741745283018868,
      "grad_norm": 4.012009620666504,
      "learning_rate": 7.693004253688716e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 8573824,
      "step": 13130
    },
    {
      "epoch": 7.744693396226415,
      "grad_norm": 1.257129192352295,
      "learning_rate": 7.690836228589613e-06,
      "loss": 0.4331,
      "num_input_tokens_seen": 8580384,
      "step": 13135
    },
    {
      "epoch": 7.747641509433962,
      "grad_norm": 1.8212661743164062,
      "learning_rate": 7.688667491080019e-06,
      "loss": 0.4069,
      "num_input_tokens_seen": 8583616,
      "step": 13140
    },
    {
      "epoch": 7.75058962264151,
      "grad_norm": 3.537367820739746,
      "learning_rate": 7.686498041734121e-06,
      "loss": 0.4604,
      "num_input_tokens_seen": 8586528,
      "step": 13145
    },
    {
      "epoch": 7.753537735849057,
      "grad_norm": 2.1566615104675293,
      "learning_rate": 7.684327881126285e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 8590912,
      "step": 13150
    },
    {
      "epoch": 7.756485849056604,
      "grad_norm": 3.5135486125946045,
      "learning_rate": 7.682157009831078e-06,
      "loss": 0.4522,
      "num_input_tokens_seen": 8594368,
      "step": 13155
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 1.6921586990356445,
      "learning_rate": 7.67998542842324e-06,
      "loss": 0.4106,
      "num_input_tokens_seen": 8597056,
      "step": 13160
    },
    {
      "epoch": 7.762382075471698,
      "grad_norm": 5.381982803344727,
      "learning_rate": 7.677813137477711e-06,
      "loss": 0.4687,
      "num_input_tokens_seen": 8600352,
      "step": 13165
    },
    {
      "epoch": 7.765330188679245,
      "grad_norm": 4.910697937011719,
      "learning_rate": 7.675640137569614e-06,
      "loss": 0.4392,
      "num_input_tokens_seen": 8602976,
      "step": 13170
    },
    {
      "epoch": 7.7682783018867925,
      "grad_norm": 4.213381767272949,
      "learning_rate": 7.673466429274257e-06,
      "loss": 0.3645,
      "num_input_tokens_seen": 8606336,
      "step": 13175
    },
    {
      "epoch": 7.77122641509434,
      "grad_norm": 1.380673885345459,
      "learning_rate": 7.671292013167143e-06,
      "loss": 0.4945,
      "num_input_tokens_seen": 8610400,
      "step": 13180
    },
    {
      "epoch": 7.774174528301887,
      "grad_norm": 2.5079081058502197,
      "learning_rate": 7.669116889823955e-06,
      "loss": 0.3468,
      "num_input_tokens_seen": 8613472,
      "step": 13185
    },
    {
      "epoch": 7.777122641509434,
      "grad_norm": 2.6769425868988037,
      "learning_rate": 7.666941059820567e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 8617280,
      "step": 13190
    },
    {
      "epoch": 7.780070754716981,
      "grad_norm": 2.1847896575927734,
      "learning_rate": 7.66476452373304e-06,
      "loss": 0.499,
      "num_input_tokens_seen": 8619584,
      "step": 13195
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 3.4186713695526123,
      "learning_rate": 7.66258728213762e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 8622624,
      "step": 13200
    },
    {
      "epoch": 7.785966981132075,
      "grad_norm": 1.9958317279815674,
      "learning_rate": 7.66040933561074e-06,
      "loss": 0.4815,
      "num_input_tokens_seen": 8625280,
      "step": 13205
    },
    {
      "epoch": 7.788915094339623,
      "grad_norm": 2.6439080238342285,
      "learning_rate": 7.658230684729027e-06,
      "loss": 0.3678,
      "num_input_tokens_seen": 8629696,
      "step": 13210
    },
    {
      "epoch": 7.79186320754717,
      "grad_norm": 2.09801983833313,
      "learning_rate": 7.656051330069282e-06,
      "loss": 0.6263,
      "num_input_tokens_seen": 8634080,
      "step": 13215
    },
    {
      "epoch": 7.794811320754717,
      "grad_norm": 2.2188522815704346,
      "learning_rate": 7.6538712722085e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 8637120,
      "step": 13220
    },
    {
      "epoch": 7.7977594339622645,
      "grad_norm": 1.7570682764053345,
      "learning_rate": 7.651690511723862e-06,
      "loss": 0.3376,
      "num_input_tokens_seen": 8639968,
      "step": 13225
    },
    {
      "epoch": 7.800707547169811,
      "grad_norm": 2.718661308288574,
      "learning_rate": 7.64950904919273e-06,
      "loss": 0.3352,
      "num_input_tokens_seen": 8643072,
      "step": 13230
    },
    {
      "epoch": 7.803655660377358,
      "grad_norm": 1.47187340259552,
      "learning_rate": 7.647326885192662e-06,
      "loss": 0.4097,
      "num_input_tokens_seen": 8646496,
      "step": 13235
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 2.5850493907928467,
      "learning_rate": 7.645144020301392e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 8649824,
      "step": 13240
    },
    {
      "epoch": 7.809551886792453,
      "grad_norm": 2.701981544494629,
      "learning_rate": 7.64296045509684e-06,
      "loss": 0.4419,
      "num_input_tokens_seen": 8652960,
      "step": 13245
    },
    {
      "epoch": 7.8125,
      "grad_norm": 1.7750754356384277,
      "learning_rate": 7.64077619015712e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 8655872,
      "step": 13250
    },
    {
      "epoch": 7.815448113207547,
      "grad_norm": 3.7290754318237305,
      "learning_rate": 7.638591226060519e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 8659168,
      "step": 13255
    },
    {
      "epoch": 7.818396226415095,
      "grad_norm": 4.366648197174072,
      "learning_rate": 7.636405563385522e-06,
      "loss": 0.5349,
      "num_input_tokens_seen": 8662336,
      "step": 13260
    },
    {
      "epoch": 7.821344339622642,
      "grad_norm": 3.6762187480926514,
      "learning_rate": 7.634219202710789e-06,
      "loss": 0.4487,
      "num_input_tokens_seen": 8666560,
      "step": 13265
    },
    {
      "epoch": 7.824292452830189,
      "grad_norm": 3.703329563140869,
      "learning_rate": 7.632032144615168e-06,
      "loss": 0.5006,
      "num_input_tokens_seen": 8669568,
      "step": 13270
    },
    {
      "epoch": 7.8272405660377355,
      "grad_norm": 1.3927212953567505,
      "learning_rate": 7.629844389677695e-06,
      "loss": 0.5345,
      "num_input_tokens_seen": 8672928,
      "step": 13275
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 1.392042636871338,
      "learning_rate": 7.627655938477586e-06,
      "loss": 0.384,
      "num_input_tokens_seen": 8676320,
      "step": 13280
    },
    {
      "epoch": 7.83313679245283,
      "grad_norm": 1.9699115753173828,
      "learning_rate": 7.6254667915942415e-06,
      "loss": 0.4809,
      "num_input_tokens_seen": 8679200,
      "step": 13285
    },
    {
      "epoch": 7.836084905660377,
      "grad_norm": 1.8889967203140259,
      "learning_rate": 7.62327694960725e-06,
      "loss": 0.3829,
      "num_input_tokens_seen": 8682144,
      "step": 13290
    },
    {
      "epoch": 7.839033018867925,
      "grad_norm": 2.4718267917633057,
      "learning_rate": 7.621086413096379e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 8685920,
      "step": 13295
    },
    {
      "epoch": 7.841981132075472,
      "grad_norm": 1.103031873703003,
      "learning_rate": 7.618895182641584e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 8689152,
      "step": 13300
    },
    {
      "epoch": 7.844929245283019,
      "grad_norm": 3.2554121017456055,
      "learning_rate": 7.6167032588230035e-06,
      "loss": 0.4811,
      "num_input_tokens_seen": 8692160,
      "step": 13305
    },
    {
      "epoch": 7.847877358490566,
      "grad_norm": 2.9319632053375244,
      "learning_rate": 7.614510642220958e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 8694880,
      "step": 13310
    },
    {
      "epoch": 7.850825471698113,
      "grad_norm": 1.452386736869812,
      "learning_rate": 7.612317333415951e-06,
      "loss": 0.3478,
      "num_input_tokens_seen": 8699232,
      "step": 13315
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 2.1855685710906982,
      "learning_rate": 7.610123332988673e-06,
      "loss": 0.4465,
      "num_input_tokens_seen": 8702752,
      "step": 13320
    },
    {
      "epoch": 7.8567216981132075,
      "grad_norm": 1.355332851409912,
      "learning_rate": 7.607928641519992e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 8705920,
      "step": 13325
    },
    {
      "epoch": 7.859669811320755,
      "grad_norm": 3.804752826690674,
      "learning_rate": 7.605733259590964e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 8709152,
      "step": 13330
    },
    {
      "epoch": 7.862617924528302,
      "grad_norm": 3.043949842453003,
      "learning_rate": 7.603537187782826e-06,
      "loss": 0.3568,
      "num_input_tokens_seen": 8711776,
      "step": 13335
    },
    {
      "epoch": 7.865566037735849,
      "grad_norm": 2.057572364807129,
      "learning_rate": 7.601340426676996e-06,
      "loss": 0.3324,
      "num_input_tokens_seen": 8714848,
      "step": 13340
    },
    {
      "epoch": 7.868514150943396,
      "grad_norm": 1.1065893173217773,
      "learning_rate": 7.599142976855077e-06,
      "loss": 0.3364,
      "num_input_tokens_seen": 8718048,
      "step": 13345
    },
    {
      "epoch": 7.871462264150943,
      "grad_norm": 2.151805877685547,
      "learning_rate": 7.596944838898854e-06,
      "loss": 0.4989,
      "num_input_tokens_seen": 8721472,
      "step": 13350
    },
    {
      "epoch": 7.87441037735849,
      "grad_norm": 1.353543996810913,
      "learning_rate": 7.594746013390293e-06,
      "loss": 0.381,
      "num_input_tokens_seen": 8725088,
      "step": 13355
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 1.658598780632019,
      "learning_rate": 7.59254650091154e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 8728448,
      "step": 13360
    },
    {
      "epoch": 7.880306603773585,
      "grad_norm": 2.0452380180358887,
      "learning_rate": 7.59034630204493e-06,
      "loss": 0.3036,
      "num_input_tokens_seen": 8731136,
      "step": 13365
    },
    {
      "epoch": 7.883254716981132,
      "grad_norm": 1.4282748699188232,
      "learning_rate": 7.588145417372972e-06,
      "loss": 0.4652,
      "num_input_tokens_seen": 8734784,
      "step": 13370
    },
    {
      "epoch": 7.8862028301886795,
      "grad_norm": 2.1600711345672607,
      "learning_rate": 7.585943847478361e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 8737984,
      "step": 13375
    },
    {
      "epoch": 7.889150943396227,
      "grad_norm": 2.0424246788024902,
      "learning_rate": 7.583741592943971e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 8741472,
      "step": 13380
    },
    {
      "epoch": 7.892099056603773,
      "grad_norm": 2.0373425483703613,
      "learning_rate": 7.581538654352859e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 8747776,
      "step": 13385
    },
    {
      "epoch": 7.8950471698113205,
      "grad_norm": 2.803098440170288,
      "learning_rate": 7.579335032288262e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 8752416,
      "step": 13390
    },
    {
      "epoch": 7.897995283018868,
      "grad_norm": 2.971954822540283,
      "learning_rate": 7.577130727333598e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 8755168,
      "step": 13395
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 1.6447069644927979,
      "learning_rate": 7.5749257400724695e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 8758560,
      "step": 13400
    },
    {
      "epoch": 7.903891509433962,
      "grad_norm": 2.004154920578003,
      "learning_rate": 7.572720071088653e-06,
      "loss": 0.5076,
      "num_input_tokens_seen": 8762208,
      "step": 13405
    },
    {
      "epoch": 7.90683962264151,
      "grad_norm": 1.1298751831054688,
      "learning_rate": 7.570513720966108e-06,
      "loss": 0.4435,
      "num_input_tokens_seen": 8765088,
      "step": 13410
    },
    {
      "epoch": 7.909787735849057,
      "grad_norm": 2.6579642295837402,
      "learning_rate": 7.56830669028898e-06,
      "loss": 0.5415,
      "num_input_tokens_seen": 8768160,
      "step": 13415
    },
    {
      "epoch": 7.912735849056604,
      "grad_norm": 3.0268828868865967,
      "learning_rate": 7.566098979641588e-06,
      "loss": 0.4829,
      "num_input_tokens_seen": 8775232,
      "step": 13420
    },
    {
      "epoch": 7.915683962264151,
      "grad_norm": 1.5259666442871094,
      "learning_rate": 7.563890589608427e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 8779264,
      "step": 13425
    },
    {
      "epoch": 7.918632075471698,
      "grad_norm": 2.330496311187744,
      "learning_rate": 7.561681520774187e-06,
      "loss": 0.2978,
      "num_input_tokens_seen": 8782464,
      "step": 13430
    },
    {
      "epoch": 7.921580188679245,
      "grad_norm": 1.8898367881774902,
      "learning_rate": 7.559471773723721e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 8785344,
      "step": 13435
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 1.280796766281128,
      "learning_rate": 7.557261349042073e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 8788640,
      "step": 13440
    },
    {
      "epoch": 7.92747641509434,
      "grad_norm": 1.4982537031173706,
      "learning_rate": 7.555050247314464e-06,
      "loss": 0.2891,
      "num_input_tokens_seen": 8792096,
      "step": 13445
    },
    {
      "epoch": 7.930424528301887,
      "grad_norm": 1.5618236064910889,
      "learning_rate": 7.552838469126289e-06,
      "loss": 0.4744,
      "num_input_tokens_seen": 8795712,
      "step": 13450
    },
    {
      "epoch": 7.933372641509434,
      "grad_norm": 2.2829606533050537,
      "learning_rate": 7.550626015063125e-06,
      "loss": 0.2809,
      "num_input_tokens_seen": 8798688,
      "step": 13455
    },
    {
      "epoch": 7.936320754716981,
      "grad_norm": 2.524898052215576,
      "learning_rate": 7.548412885710734e-06,
      "loss": 0.4522,
      "num_input_tokens_seen": 8801152,
      "step": 13460
    },
    {
      "epoch": 7.939268867924528,
      "grad_norm": 2.4906179904937744,
      "learning_rate": 7.546199081655048e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 8804032,
      "step": 13465
    },
    {
      "epoch": 7.942216981132075,
      "grad_norm": 3.076634168624878,
      "learning_rate": 7.54398460348218e-06,
      "loss": 0.397,
      "num_input_tokens_seen": 8806976,
      "step": 13470
    },
    {
      "epoch": 7.945165094339623,
      "grad_norm": 1.3567235469818115,
      "learning_rate": 7.541769451778425e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 8810144,
      "step": 13475
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 2.161062240600586,
      "learning_rate": 7.5395536271302536e-06,
      "loss": 0.4914,
      "num_input_tokens_seen": 8813184,
      "step": 13480
    },
    {
      "epoch": 7.951061320754717,
      "grad_norm": 1.9564601182937622,
      "learning_rate": 7.5373371301243136e-06,
      "loss": 0.4173,
      "num_input_tokens_seen": 8816448,
      "step": 13485
    },
    {
      "epoch": 7.9540094339622645,
      "grad_norm": 1.2747882604599,
      "learning_rate": 7.535119961347433e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 8819456,
      "step": 13490
    },
    {
      "epoch": 7.956957547169811,
      "grad_norm": 1.6545087099075317,
      "learning_rate": 7.532902121386618e-06,
      "loss": 0.3602,
      "num_input_tokens_seen": 8822432,
      "step": 13495
    },
    {
      "epoch": 7.959905660377358,
      "grad_norm": 1.7462671995162964,
      "learning_rate": 7.530683610829051e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 8827808,
      "step": 13500
    },
    {
      "epoch": 7.962853773584905,
      "grad_norm": 1.6405409574508667,
      "learning_rate": 7.5284644302620906e-06,
      "loss": 0.4956,
      "num_input_tokens_seen": 8830784,
      "step": 13505
    },
    {
      "epoch": 7.965801886792453,
      "grad_norm": 2.728203773498535,
      "learning_rate": 7.526244580273274e-06,
      "loss": 0.3822,
      "num_input_tokens_seen": 8833504,
      "step": 13510
    },
    {
      "epoch": 7.96875,
      "grad_norm": 1.9763630628585815,
      "learning_rate": 7.524024061450318e-06,
      "loss": 0.3951,
      "num_input_tokens_seen": 8836096,
      "step": 13515
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 2.535935401916504,
      "learning_rate": 7.521802874381115e-06,
      "loss": 0.4392,
      "num_input_tokens_seen": 8839200,
      "step": 13520
    },
    {
      "epoch": 7.974646226415095,
      "grad_norm": 1.614743709564209,
      "learning_rate": 7.519581019653731e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 8842176,
      "step": 13525
    },
    {
      "epoch": 7.977594339622642,
      "grad_norm": 3.1697046756744385,
      "learning_rate": 7.517358497856413e-06,
      "loss": 0.4952,
      "num_input_tokens_seen": 8845984,
      "step": 13530
    },
    {
      "epoch": 7.980542452830189,
      "grad_norm": 3.24099063873291,
      "learning_rate": 7.515135309577584e-06,
      "loss": 0.5388,
      "num_input_tokens_seen": 8848960,
      "step": 13535
    },
    {
      "epoch": 7.9834905660377355,
      "grad_norm": 1.6685577630996704,
      "learning_rate": 7.5129114554058425e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 8852448,
      "step": 13540
    },
    {
      "epoch": 7.986438679245283,
      "grad_norm": 2.333258867263794,
      "learning_rate": 7.510686935929963e-06,
      "loss": 0.3113,
      "num_input_tokens_seen": 8855648,
      "step": 13545
    },
    {
      "epoch": 7.98938679245283,
      "grad_norm": 1.116683006286621,
      "learning_rate": 7.5084617517388965e-06,
      "loss": 0.3627,
      "num_input_tokens_seen": 8858304,
      "step": 13550
    },
    {
      "epoch": 7.992334905660377,
      "grad_norm": 1.4402906894683838,
      "learning_rate": 7.506235903421771e-06,
      "loss": 0.5204,
      "num_input_tokens_seen": 8861024,
      "step": 13555
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 2.5634584426879883,
      "learning_rate": 7.504009391567889e-06,
      "loss": 0.3948,
      "num_input_tokens_seen": 8863296,
      "step": 13560
    },
    {
      "epoch": 7.998231132075472,
      "grad_norm": 2.3295438289642334,
      "learning_rate": 7.501782216766729e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 8865952,
      "step": 13565
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.49138548970222473,
      "eval_runtime": 19.1436,
      "eval_samples_per_second": 88.594,
      "eval_steps_per_second": 22.148,
      "num_input_tokens_seen": 8867608,
      "step": 13568
    },
    {
      "epoch": 8.00117924528302,
      "grad_norm": 1.6621562242507935,
      "learning_rate": 7.499554379607944e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 8869176,
      "step": 13570
    },
    {
      "epoch": 8.004127358490566,
      "grad_norm": 1.3259161710739136,
      "learning_rate": 7.497325880681365e-06,
      "loss": 0.4787,
      "num_input_tokens_seen": 8872920,
      "step": 13575
    },
    {
      "epoch": 8.007075471698114,
      "grad_norm": 1.8585065603256226,
      "learning_rate": 7.495096720576994e-06,
      "loss": 0.3805,
      "num_input_tokens_seen": 8876792,
      "step": 13580
    },
    {
      "epoch": 8.01002358490566,
      "grad_norm": 1.4864449501037598,
      "learning_rate": 7.492866899885017e-06,
      "loss": 0.3194,
      "num_input_tokens_seen": 8880120,
      "step": 13585
    },
    {
      "epoch": 8.012971698113208,
      "grad_norm": 3.2940778732299805,
      "learning_rate": 7.490636419195782e-06,
      "loss": 0.5012,
      "num_input_tokens_seen": 8883128,
      "step": 13590
    },
    {
      "epoch": 8.015919811320755,
      "grad_norm": 1.4467321634292603,
      "learning_rate": 7.488405279099821e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 8886520,
      "step": 13595
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 2.1244189739227295,
      "learning_rate": 7.48617348018784e-06,
      "loss": 0.4586,
      "num_input_tokens_seen": 8890104,
      "step": 13600
    },
    {
      "epoch": 8.02181603773585,
      "grad_norm": 3.129875898361206,
      "learning_rate": 7.4839410230507134e-06,
      "loss": 0.4729,
      "num_input_tokens_seen": 8892856,
      "step": 13605
    },
    {
      "epoch": 8.024764150943396,
      "grad_norm": 2.3229167461395264,
      "learning_rate": 7.481707908279496e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 8897048,
      "step": 13610
    },
    {
      "epoch": 8.027712264150944,
      "grad_norm": 3.378281593322754,
      "learning_rate": 7.4794741364654144e-06,
      "loss": 0.3617,
      "num_input_tokens_seen": 8900056,
      "step": 13615
    },
    {
      "epoch": 8.03066037735849,
      "grad_norm": 1.8339594602584839,
      "learning_rate": 7.477239708199871e-06,
      "loss": 0.5436,
      "num_input_tokens_seen": 8903224,
      "step": 13620
    },
    {
      "epoch": 8.033608490566039,
      "grad_norm": 3.5393452644348145,
      "learning_rate": 7.475004624074434e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 8906168,
      "step": 13625
    },
    {
      "epoch": 8.036556603773585,
      "grad_norm": 1.8796995878219604,
      "learning_rate": 7.4727688846808595e-06,
      "loss": 0.3406,
      "num_input_tokens_seen": 8908856,
      "step": 13630
    },
    {
      "epoch": 8.039504716981131,
      "grad_norm": 1.3400983810424805,
      "learning_rate": 7.4705324906110654e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 8913272,
      "step": 13635
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 3.562311887741089,
      "learning_rate": 7.4682954424571466e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 8916184,
      "step": 13640
    },
    {
      "epoch": 8.045400943396226,
      "grad_norm": 2.594835042953491,
      "learning_rate": 7.466057740811372e-06,
      "loss": 0.3832,
      "num_input_tokens_seen": 8919064,
      "step": 13645
    },
    {
      "epoch": 8.048349056603774,
      "grad_norm": 2.167931079864502,
      "learning_rate": 7.463819386266182e-06,
      "loss": 0.4352,
      "num_input_tokens_seen": 8922072,
      "step": 13650
    },
    {
      "epoch": 8.05129716981132,
      "grad_norm": 2.3030431270599365,
      "learning_rate": 7.461580379414191e-06,
      "loss": 0.7863,
      "num_input_tokens_seen": 8924408,
      "step": 13655
    },
    {
      "epoch": 8.054245283018869,
      "grad_norm": 2.10428524017334,
      "learning_rate": 7.459340720848187e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 8926840,
      "step": 13660
    },
    {
      "epoch": 8.057193396226415,
      "grad_norm": 3.4414076805114746,
      "learning_rate": 7.457100411161128e-06,
      "loss": 0.4432,
      "num_input_tokens_seen": 8929816,
      "step": 13665
    },
    {
      "epoch": 8.060141509433961,
      "grad_norm": 1.8417024612426758,
      "learning_rate": 7.454859450946144e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 8933656,
      "step": 13670
    },
    {
      "epoch": 8.06308962264151,
      "grad_norm": 1.0349085330963135,
      "learning_rate": 7.4526178407965396e-06,
      "loss": 0.5405,
      "num_input_tokens_seen": 8936568,
      "step": 13675
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 1.4251329898834229,
      "learning_rate": 7.450375581305794e-06,
      "loss": 0.3361,
      "num_input_tokens_seen": 8939544,
      "step": 13680
    },
    {
      "epoch": 8.068985849056604,
      "grad_norm": 1.7841787338256836,
      "learning_rate": 7.448132673067552e-06,
      "loss": 0.4291,
      "num_input_tokens_seen": 8943256,
      "step": 13685
    },
    {
      "epoch": 8.07193396226415,
      "grad_norm": 1.931891679763794,
      "learning_rate": 7.445889116675634e-06,
      "loss": 0.3215,
      "num_input_tokens_seen": 8946424,
      "step": 13690
    },
    {
      "epoch": 8.074882075471699,
      "grad_norm": 1.6666234731674194,
      "learning_rate": 7.443644912724031e-06,
      "loss": 0.3897,
      "num_input_tokens_seen": 8948984,
      "step": 13695
    },
    {
      "epoch": 8.077830188679245,
      "grad_norm": 2.5172975063323975,
      "learning_rate": 7.441400061806907e-06,
      "loss": 0.389,
      "num_input_tokens_seen": 8952088,
      "step": 13700
    },
    {
      "epoch": 8.080778301886792,
      "grad_norm": 2.0547380447387695,
      "learning_rate": 7.439154564518592e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 8954776,
      "step": 13705
    },
    {
      "epoch": 8.08372641509434,
      "grad_norm": 3.2828032970428467,
      "learning_rate": 7.436908421453597e-06,
      "loss": 0.5615,
      "num_input_tokens_seen": 8958008,
      "step": 13710
    },
    {
      "epoch": 8.086674528301886,
      "grad_norm": 1.6582406759262085,
      "learning_rate": 7.434661633206593e-06,
      "loss": 0.3704,
      "num_input_tokens_seen": 8961848,
      "step": 13715
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 1.960922360420227,
      "learning_rate": 7.4324142003724286e-06,
      "loss": 0.3805,
      "num_input_tokens_seen": 8965528,
      "step": 13720
    },
    {
      "epoch": 8.09257075471698,
      "grad_norm": 1.175669550895691,
      "learning_rate": 7.430166123546122e-06,
      "loss": 0.2851,
      "num_input_tokens_seen": 8968280,
      "step": 13725
    },
    {
      "epoch": 8.095518867924529,
      "grad_norm": 2.327145576477051,
      "learning_rate": 7.427917403322862e-06,
      "loss": 0.3152,
      "num_input_tokens_seen": 8971832,
      "step": 13730
    },
    {
      "epoch": 8.098466981132075,
      "grad_norm": 2.265737295150757,
      "learning_rate": 7.425668040298003e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 8975288,
      "step": 13735
    },
    {
      "epoch": 8.101415094339623,
      "grad_norm": 3.0419347286224365,
      "learning_rate": 7.4234180350670785e-06,
      "loss": 0.4495,
      "num_input_tokens_seen": 8977848,
      "step": 13740
    },
    {
      "epoch": 8.10436320754717,
      "grad_norm": 2.0738656520843506,
      "learning_rate": 7.421167388225785e-06,
      "loss": 0.4288,
      "num_input_tokens_seen": 8982968,
      "step": 13745
    },
    {
      "epoch": 8.107311320754716,
      "grad_norm": 1.5855865478515625,
      "learning_rate": 7.41891610036999e-06,
      "loss": 0.4241,
      "num_input_tokens_seen": 8986648,
      "step": 13750
    },
    {
      "epoch": 8.110259433962264,
      "grad_norm": 2.256459951400757,
      "learning_rate": 7.416664172095732e-06,
      "loss": 0.3385,
      "num_input_tokens_seen": 8989592,
      "step": 13755
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 1.8362722396850586,
      "learning_rate": 7.414411603999221e-06,
      "loss": 0.356,
      "num_input_tokens_seen": 8993272,
      "step": 13760
    },
    {
      "epoch": 8.116155660377359,
      "grad_norm": 2.322813034057617,
      "learning_rate": 7.4121583966768295e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 8996760,
      "step": 13765
    },
    {
      "epoch": 8.119103773584905,
      "grad_norm": 4.114555358886719,
      "learning_rate": 7.409904550725109e-06,
      "loss": 0.5273,
      "num_input_tokens_seen": 9004056,
      "step": 13770
    },
    {
      "epoch": 8.122051886792454,
      "grad_norm": 1.9996955394744873,
      "learning_rate": 7.407650066740771e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 9007192,
      "step": 13775
    },
    {
      "epoch": 8.125,
      "grad_norm": 3.493622064590454,
      "learning_rate": 7.405394945320702e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 9010648,
      "step": 13780
    },
    {
      "epoch": 8.127948113207546,
      "grad_norm": 2.0212185382843018,
      "learning_rate": 7.403139187061955e-06,
      "loss": 0.3583,
      "num_input_tokens_seen": 9013816,
      "step": 13785
    },
    {
      "epoch": 8.130896226415095,
      "grad_norm": 1.3997966051101685,
      "learning_rate": 7.400882792561752e-06,
      "loss": 0.3986,
      "num_input_tokens_seen": 9016792,
      "step": 13790
    },
    {
      "epoch": 8.133844339622641,
      "grad_norm": 1.8993357419967651,
      "learning_rate": 7.39862576241748e-06,
      "loss": 0.4524,
      "num_input_tokens_seen": 9019768,
      "step": 13795
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 2.1396842002868652,
      "learning_rate": 7.396368097226703e-06,
      "loss": 0.4438,
      "num_input_tokens_seen": 9022456,
      "step": 13800
    },
    {
      "epoch": 8.139740566037736,
      "grad_norm": 1.9632757902145386,
      "learning_rate": 7.394109797587144e-06,
      "loss": 0.4257,
      "num_input_tokens_seen": 9025880,
      "step": 13805
    },
    {
      "epoch": 8.142688679245284,
      "grad_norm": 1.7332069873809814,
      "learning_rate": 7.3918508640966956e-06,
      "loss": 0.3382,
      "num_input_tokens_seen": 9029304,
      "step": 13810
    },
    {
      "epoch": 8.14563679245283,
      "grad_norm": 2.4844751358032227,
      "learning_rate": 7.389591297353424e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 9032024,
      "step": 13815
    },
    {
      "epoch": 8.148584905660377,
      "grad_norm": 1.6667695045471191,
      "learning_rate": 7.3873310979555565e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 9035096,
      "step": 13820
    },
    {
      "epoch": 8.151533018867925,
      "grad_norm": 2.629072904586792,
      "learning_rate": 7.385070266501495e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 9038104,
      "step": 13825
    },
    {
      "epoch": 8.154481132075471,
      "grad_norm": 2.302917957305908,
      "learning_rate": 7.382808803589798e-06,
      "loss": 0.4787,
      "num_input_tokens_seen": 9040472,
      "step": 13830
    },
    {
      "epoch": 8.15742924528302,
      "grad_norm": 1.556125283241272,
      "learning_rate": 7.380546709819204e-06,
      "loss": 0.3541,
      "num_input_tokens_seen": 9043320,
      "step": 13835
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 2.9440553188323975,
      "learning_rate": 7.378283985788608e-06,
      "loss": 0.439,
      "num_input_tokens_seen": 9046744,
      "step": 13840
    },
    {
      "epoch": 8.163325471698114,
      "grad_norm": 4.965917110443115,
      "learning_rate": 7.376020632097076e-06,
      "loss": 0.6701,
      "num_input_tokens_seen": 9049624,
      "step": 13845
    },
    {
      "epoch": 8.16627358490566,
      "grad_norm": 2.8430848121643066,
      "learning_rate": 7.373756649343841e-06,
      "loss": 0.3875,
      "num_input_tokens_seen": 9051800,
      "step": 13850
    },
    {
      "epoch": 8.169221698113208,
      "grad_norm": 2.862823247909546,
      "learning_rate": 7.371492038128305e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 9055928,
      "step": 13855
    },
    {
      "epoch": 8.172169811320755,
      "grad_norm": 1.4085891246795654,
      "learning_rate": 7.36922679905003e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 9058328,
      "step": 13860
    },
    {
      "epoch": 8.175117924528301,
      "grad_norm": 2.986264944076538,
      "learning_rate": 7.366960932708749e-06,
      "loss": 0.3322,
      "num_input_tokens_seen": 9061848,
      "step": 13865
    },
    {
      "epoch": 8.17806603773585,
      "grad_norm": 1.1626472473144531,
      "learning_rate": 7.364694439704361e-06,
      "loss": 0.4164,
      "num_input_tokens_seen": 9065496,
      "step": 13870
    },
    {
      "epoch": 8.181014150943396,
      "grad_norm": 1.7250663042068481,
      "learning_rate": 7.3624273206369264e-06,
      "loss": 0.3822,
      "num_input_tokens_seen": 9068792,
      "step": 13875
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 2.0766329765319824,
      "learning_rate": 7.360159576106681e-06,
      "loss": 0.4589,
      "num_input_tokens_seen": 9073464,
      "step": 13880
    },
    {
      "epoch": 8.18691037735849,
      "grad_norm": 2.124788522720337,
      "learning_rate": 7.357891206714014e-06,
      "loss": 0.3762,
      "num_input_tokens_seen": 9076408,
      "step": 13885
    },
    {
      "epoch": 8.189858490566039,
      "grad_norm": 3.416912794113159,
      "learning_rate": 7.355622213059487e-06,
      "loss": 0.3291,
      "num_input_tokens_seen": 9079288,
      "step": 13890
    },
    {
      "epoch": 8.192806603773585,
      "grad_norm": 2.4355673789978027,
      "learning_rate": 7.353352595743829e-06,
      "loss": 0.4135,
      "num_input_tokens_seen": 9082328,
      "step": 13895
    },
    {
      "epoch": 8.195754716981131,
      "grad_norm": 2.2894399166107178,
      "learning_rate": 7.351082355367928e-06,
      "loss": 0.3193,
      "num_input_tokens_seen": 9084728,
      "step": 13900
    },
    {
      "epoch": 8.19870283018868,
      "grad_norm": 2.3049814701080322,
      "learning_rate": 7.34881149253284e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 9087544,
      "step": 13905
    },
    {
      "epoch": 8.201650943396226,
      "grad_norm": 1.6340093612670898,
      "learning_rate": 7.346540007839787e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 9091352,
      "step": 13910
    },
    {
      "epoch": 8.204599056603774,
      "grad_norm": 2.739292860031128,
      "learning_rate": 7.344267901890154e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 9094168,
      "step": 13915
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 1.4589697122573853,
      "learning_rate": 7.341995175285491e-06,
      "loss": 0.473,
      "num_input_tokens_seen": 9096920,
      "step": 13920
    },
    {
      "epoch": 8.210495283018869,
      "grad_norm": 3.6896467208862305,
      "learning_rate": 7.339721828627512e-06,
      "loss": 0.2621,
      "num_input_tokens_seen": 9099032,
      "step": 13925
    },
    {
      "epoch": 8.213443396226415,
      "grad_norm": 1.712661862373352,
      "learning_rate": 7.337447862518096e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 9101816,
      "step": 13930
    },
    {
      "epoch": 8.216391509433961,
      "grad_norm": 2.710340976715088,
      "learning_rate": 7.335173277559282e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 9105048,
      "step": 13935
    },
    {
      "epoch": 8.21933962264151,
      "grad_norm": 4.832684516906738,
      "learning_rate": 7.332898074353281e-06,
      "loss": 0.3807,
      "num_input_tokens_seen": 9108088,
      "step": 13940
    },
    {
      "epoch": 8.222287735849056,
      "grad_norm": 2.062472105026245,
      "learning_rate": 7.330622253502461e-06,
      "loss": 0.3265,
      "num_input_tokens_seen": 9111128,
      "step": 13945
    },
    {
      "epoch": 8.225235849056604,
      "grad_norm": 2.4425246715545654,
      "learning_rate": 7.3283458156093534e-06,
      "loss": 0.4493,
      "num_input_tokens_seen": 9113880,
      "step": 13950
    },
    {
      "epoch": 8.22818396226415,
      "grad_norm": 1.502716302871704,
      "learning_rate": 7.326068761276657e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 9119480,
      "step": 13955
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 2.6011204719543457,
      "learning_rate": 7.323791091107231e-06,
      "loss": 0.4316,
      "num_input_tokens_seen": 9123000,
      "step": 13960
    },
    {
      "epoch": 8.234080188679245,
      "grad_norm": 3.158933401107788,
      "learning_rate": 7.3215128057040986e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 9125976,
      "step": 13965
    },
    {
      "epoch": 8.237028301886792,
      "grad_norm": 3.5844616889953613,
      "learning_rate": 7.319233905670447e-06,
      "loss": 0.5116,
      "num_input_tokens_seen": 9128792,
      "step": 13970
    },
    {
      "epoch": 8.23997641509434,
      "grad_norm": 2.0982890129089355,
      "learning_rate": 7.316954391609622e-06,
      "loss": 0.5129,
      "num_input_tokens_seen": 9131864,
      "step": 13975
    },
    {
      "epoch": 8.242924528301886,
      "grad_norm": 2.433706283569336,
      "learning_rate": 7.314674264125137e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 9134552,
      "step": 13980
    },
    {
      "epoch": 8.245872641509434,
      "grad_norm": 1.8108760118484497,
      "learning_rate": 7.312393523820665e-06,
      "loss": 0.3319,
      "num_input_tokens_seen": 9137592,
      "step": 13985
    },
    {
      "epoch": 8.24882075471698,
      "grad_norm": 2.008455514907837,
      "learning_rate": 7.310112171300041e-06,
      "loss": 0.3748,
      "num_input_tokens_seen": 9140440,
      "step": 13990
    },
    {
      "epoch": 8.251768867924529,
      "grad_norm": 2.9995944499969482,
      "learning_rate": 7.307830207167263e-06,
      "loss": 0.3402,
      "num_input_tokens_seen": 9143640,
      "step": 13995
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 2.65132212638855,
      "learning_rate": 7.305547632026493e-06,
      "loss": 0.3918,
      "num_input_tokens_seen": 9147128,
      "step": 14000
    },
    {
      "epoch": 8.257665094339623,
      "grad_norm": 2.3154497146606445,
      "learning_rate": 7.3032644464820515e-06,
      "loss": 0.5516,
      "num_input_tokens_seen": 9150552,
      "step": 14005
    },
    {
      "epoch": 8.26061320754717,
      "grad_norm": 2.177722692489624,
      "learning_rate": 7.30098065113842e-06,
      "loss": 0.4412,
      "num_input_tokens_seen": 9153272,
      "step": 14010
    },
    {
      "epoch": 8.263561320754716,
      "grad_norm": 2.7271900177001953,
      "learning_rate": 7.298696246600244e-06,
      "loss": 0.4681,
      "num_input_tokens_seen": 9156536,
      "step": 14015
    },
    {
      "epoch": 8.266509433962264,
      "grad_norm": 2.105330467224121,
      "learning_rate": 7.2964112334723315e-06,
      "loss": 0.4606,
      "num_input_tokens_seen": 9159480,
      "step": 14020
    },
    {
      "epoch": 8.26945754716981,
      "grad_norm": 2.1727447509765625,
      "learning_rate": 7.294125612359647e-06,
      "loss": 0.409,
      "num_input_tokens_seen": 9163768,
      "step": 14025
    },
    {
      "epoch": 8.272405660377359,
      "grad_norm": 3.239534616470337,
      "learning_rate": 7.291839383867318e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 9166936,
      "step": 14030
    },
    {
      "epoch": 8.275353773584905,
      "grad_norm": 1.6477655172348022,
      "learning_rate": 7.289552548600638e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 9170872,
      "step": 14035
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 2.193758249282837,
      "learning_rate": 7.287265107165052e-06,
      "loss": 0.3696,
      "num_input_tokens_seen": 9173528,
      "step": 14040
    },
    {
      "epoch": 8.28125,
      "grad_norm": 1.3483048677444458,
      "learning_rate": 7.284977060166171e-06,
      "loss": 0.4108,
      "num_input_tokens_seen": 9177528,
      "step": 14045
    },
    {
      "epoch": 8.284198113207546,
      "grad_norm": 1.0243809223175049,
      "learning_rate": 7.282688408209766e-06,
      "loss": 0.504,
      "num_input_tokens_seen": 9181208,
      "step": 14050
    },
    {
      "epoch": 8.287146226415095,
      "grad_norm": 1.8333991765975952,
      "learning_rate": 7.2803991519017655e-06,
      "loss": 0.3085,
      "num_input_tokens_seen": 9183608,
      "step": 14055
    },
    {
      "epoch": 8.290094339622641,
      "grad_norm": 1.7030141353607178,
      "learning_rate": 7.2781092918482634e-06,
      "loss": 0.414,
      "num_input_tokens_seen": 9187224,
      "step": 14060
    },
    {
      "epoch": 8.29304245283019,
      "grad_norm": 3.065843105316162,
      "learning_rate": 7.275818828655508e-06,
      "loss": 0.3923,
      "num_input_tokens_seen": 9190840,
      "step": 14065
    },
    {
      "epoch": 8.295990566037736,
      "grad_norm": 1.2811897993087769,
      "learning_rate": 7.27352776292991e-06,
      "loss": 0.331,
      "num_input_tokens_seen": 9194584,
      "step": 14070
    },
    {
      "epoch": 8.298938679245284,
      "grad_norm": 2.101893901824951,
      "learning_rate": 7.271236095278036e-06,
      "loss": 0.3277,
      "num_input_tokens_seen": 9197496,
      "step": 14075
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 4.474690914154053,
      "learning_rate": 7.2689438263066195e-06,
      "loss": 0.4616,
      "num_input_tokens_seen": 9200120,
      "step": 14080
    },
    {
      "epoch": 8.304834905660377,
      "grad_norm": 1.8350647687911987,
      "learning_rate": 7.266650956622546e-06,
      "loss": 0.4524,
      "num_input_tokens_seen": 9203544,
      "step": 14085
    },
    {
      "epoch": 8.307783018867925,
      "grad_norm": 1.816725492477417,
      "learning_rate": 7.2643574868328625e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 9207480,
      "step": 14090
    },
    {
      "epoch": 8.310731132075471,
      "grad_norm": 1.5641651153564453,
      "learning_rate": 7.262063417544776e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 9210296,
      "step": 14095
    },
    {
      "epoch": 8.31367924528302,
      "grad_norm": 3.8366916179656982,
      "learning_rate": 7.25976874936565e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 9214712,
      "step": 14100
    },
    {
      "epoch": 8.316627358490566,
      "grad_norm": 1.6589109897613525,
      "learning_rate": 7.257473482903009e-06,
      "loss": 0.346,
      "num_input_tokens_seen": 9218168,
      "step": 14105
    },
    {
      "epoch": 8.319575471698114,
      "grad_norm": 2.310345411300659,
      "learning_rate": 7.255177618764534e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 9221048,
      "step": 14110
    },
    {
      "epoch": 8.32252358490566,
      "grad_norm": 2.0052247047424316,
      "learning_rate": 7.252881157558065e-06,
      "loss": 0.4253,
      "num_input_tokens_seen": 9223928,
      "step": 14115
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 2.991630792617798,
      "learning_rate": 7.250584099891602e-06,
      "loss": 0.4071,
      "num_input_tokens_seen": 9227384,
      "step": 14120
    },
    {
      "epoch": 8.328419811320755,
      "grad_norm": 2.206829071044922,
      "learning_rate": 7.248286446373296e-06,
      "loss": 0.4302,
      "num_input_tokens_seen": 9231288,
      "step": 14125
    },
    {
      "epoch": 8.331367924528301,
      "grad_norm": 3.045832872390747,
      "learning_rate": 7.245988197611466e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 9233912,
      "step": 14130
    },
    {
      "epoch": 8.33431603773585,
      "grad_norm": 3.773101568222046,
      "learning_rate": 7.2436893542145805e-06,
      "loss": 0.4188,
      "num_input_tokens_seen": 9236920,
      "step": 14135
    },
    {
      "epoch": 8.337264150943396,
      "grad_norm": 2.406062126159668,
      "learning_rate": 7.241389916791269e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 9239416,
      "step": 14140
    },
    {
      "epoch": 8.340212264150944,
      "grad_norm": 3.86338472366333,
      "learning_rate": 7.239089885950317e-06,
      "loss": 0.515,
      "num_input_tokens_seen": 9242616,
      "step": 14145
    },
    {
      "epoch": 8.34316037735849,
      "grad_norm": 2.9663383960723877,
      "learning_rate": 7.236789262300667e-06,
      "loss": 0.4211,
      "num_input_tokens_seen": 9245176,
      "step": 14150
    },
    {
      "epoch": 8.346108490566039,
      "grad_norm": 3.4104533195495605,
      "learning_rate": 7.23448804645142e-06,
      "loss": 0.5527,
      "num_input_tokens_seen": 9248312,
      "step": 14155
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 1.8138375282287598,
      "learning_rate": 7.232186239011834e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 9252152,
      "step": 14160
    },
    {
      "epoch": 8.352004716981131,
      "grad_norm": 1.5364997386932373,
      "learning_rate": 7.2298838405913195e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 9254456,
      "step": 14165
    },
    {
      "epoch": 8.35495283018868,
      "grad_norm": 1.9962011575698853,
      "learning_rate": 7.227580851799448e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 9257912,
      "step": 14170
    },
    {
      "epoch": 8.357900943396226,
      "grad_norm": 2.7346320152282715,
      "learning_rate": 7.2252772732459455e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 9261656,
      "step": 14175
    },
    {
      "epoch": 8.360849056603774,
      "grad_norm": 2.494647979736328,
      "learning_rate": 7.222973105540696e-06,
      "loss": 0.5646,
      "num_input_tokens_seen": 9265496,
      "step": 14180
    },
    {
      "epoch": 8.36379716981132,
      "grad_norm": 3.7523081302642822,
      "learning_rate": 7.2206683492937345e-06,
      "loss": 0.5314,
      "num_input_tokens_seen": 9268696,
      "step": 14185
    },
    {
      "epoch": 8.366745283018869,
      "grad_norm": 2.2180802822113037,
      "learning_rate": 7.218363005115259e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 9271704,
      "step": 14190
    },
    {
      "epoch": 8.369693396226415,
      "grad_norm": 2.1022439002990723,
      "learning_rate": 7.216057073615617e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 9274424,
      "step": 14195
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 1.6614010334014893,
      "learning_rate": 7.21375055540531e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 9277912,
      "step": 14200
    },
    {
      "epoch": 8.37558962264151,
      "grad_norm": 5.482511520385742,
      "learning_rate": 7.211443451095007e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 9281688,
      "step": 14205
    },
    {
      "epoch": 8.378537735849056,
      "grad_norm": 2.181302785873413,
      "learning_rate": 7.2091357612955185e-06,
      "loss": 0.5068,
      "num_input_tokens_seen": 9285144,
      "step": 14210
    },
    {
      "epoch": 8.381485849056604,
      "grad_norm": 1.876547932624817,
      "learning_rate": 7.206827486617816e-06,
      "loss": 0.3892,
      "num_input_tokens_seen": 9287992,
      "step": 14215
    },
    {
      "epoch": 8.38443396226415,
      "grad_norm": 1.6392223834991455,
      "learning_rate": 7.204518627673026e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 9290808,
      "step": 14220
    },
    {
      "epoch": 8.387382075471699,
      "grad_norm": 2.5411877632141113,
      "learning_rate": 7.202209185072428e-06,
      "loss": 0.4828,
      "num_input_tokens_seen": 9293496,
      "step": 14225
    },
    {
      "epoch": 8.390330188679245,
      "grad_norm": 3.1776459217071533,
      "learning_rate": 7.199899159427457e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 9297016,
      "step": 14230
    },
    {
      "epoch": 8.393278301886792,
      "grad_norm": 2.040161371231079,
      "learning_rate": 7.1975885513497035e-06,
      "loss": 0.3837,
      "num_input_tokens_seen": 9300504,
      "step": 14235
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 7.589986324310303,
      "learning_rate": 7.195277361450909e-06,
      "loss": 0.249,
      "num_input_tokens_seen": 9303640,
      "step": 14240
    },
    {
      "epoch": 8.399174528301886,
      "grad_norm": 2.920724391937256,
      "learning_rate": 7.192965590342973e-06,
      "loss": 0.4613,
      "num_input_tokens_seen": 9307000,
      "step": 14245
    },
    {
      "epoch": 8.402122641509434,
      "grad_norm": 2.1028432846069336,
      "learning_rate": 7.190653238637945e-06,
      "loss": 0.4077,
      "num_input_tokens_seen": 9309592,
      "step": 14250
    },
    {
      "epoch": 8.40507075471698,
      "grad_norm": 3.7883617877960205,
      "learning_rate": 7.18834030694803e-06,
      "loss": 0.5013,
      "num_input_tokens_seen": 9312696,
      "step": 14255
    },
    {
      "epoch": 8.408018867924529,
      "grad_norm": 2.4471495151519775,
      "learning_rate": 7.186026795885589e-06,
      "loss": 0.292,
      "num_input_tokens_seen": 9316120,
      "step": 14260
    },
    {
      "epoch": 8.410966981132075,
      "grad_norm": 1.8734937906265259,
      "learning_rate": 7.183712706063133e-06,
      "loss": 0.5021,
      "num_input_tokens_seen": 9320056,
      "step": 14265
    },
    {
      "epoch": 8.413915094339623,
      "grad_norm": 1.7022895812988281,
      "learning_rate": 7.1813980380933255e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 9323672,
      "step": 14270
    },
    {
      "epoch": 8.41686320754717,
      "grad_norm": 2.545719861984253,
      "learning_rate": 7.179082792588986e-06,
      "loss": 0.334,
      "num_input_tokens_seen": 9326616,
      "step": 14275
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 2.0917768478393555,
      "learning_rate": 7.176766970163087e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 9329656,
      "step": 14280
    },
    {
      "epoch": 8.422759433962264,
      "grad_norm": 2.5109658241271973,
      "learning_rate": 7.17445057142875e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 9332920,
      "step": 14285
    },
    {
      "epoch": 8.42570754716981,
      "grad_norm": 2.9947516918182373,
      "learning_rate": 7.172133596999253e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 9336728,
      "step": 14290
    },
    {
      "epoch": 8.428655660377359,
      "grad_norm": 3.720810890197754,
      "learning_rate": 7.1698160474880255e-06,
      "loss": 0.5027,
      "num_input_tokens_seen": 9339128,
      "step": 14295
    },
    {
      "epoch": 8.431603773584905,
      "grad_norm": 1.725210189819336,
      "learning_rate": 7.167497923508648e-06,
      "loss": 0.2423,
      "num_input_tokens_seen": 9342296,
      "step": 14300
    },
    {
      "epoch": 8.434551886792454,
      "grad_norm": 2.4579312801361084,
      "learning_rate": 7.165179225674854e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 9345144,
      "step": 14305
    },
    {
      "epoch": 8.4375,
      "grad_norm": 1.694855809211731,
      "learning_rate": 7.1628599546005276e-06,
      "loss": 0.4519,
      "num_input_tokens_seen": 9348056,
      "step": 14310
    },
    {
      "epoch": 8.440448113207546,
      "grad_norm": 1.7585989236831665,
      "learning_rate": 7.160540110899708e-06,
      "loss": 0.5295,
      "num_input_tokens_seen": 9351000,
      "step": 14315
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 1.563844919204712,
      "learning_rate": 7.158219695186582e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 9355224,
      "step": 14320
    },
    {
      "epoch": 8.446344339622641,
      "grad_norm": 1.8563148975372314,
      "learning_rate": 7.1558987080754905e-06,
      "loss": 0.3857,
      "num_input_tokens_seen": 9358776,
      "step": 14325
    },
    {
      "epoch": 8.44929245283019,
      "grad_norm": 3.490443706512451,
      "learning_rate": 7.1535771501809245e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 9361624,
      "step": 14330
    },
    {
      "epoch": 8.452240566037736,
      "grad_norm": 1.677544355392456,
      "learning_rate": 7.151255022117527e-06,
      "loss": 0.4343,
      "num_input_tokens_seen": 9364792,
      "step": 14335
    },
    {
      "epoch": 8.455188679245284,
      "grad_norm": 2.730604648590088,
      "learning_rate": 7.148932324500091e-06,
      "loss": 0.3782,
      "num_input_tokens_seen": 9367672,
      "step": 14340
    },
    {
      "epoch": 8.45813679245283,
      "grad_norm": 1.8237375020980835,
      "learning_rate": 7.146609057943559e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 9371928,
      "step": 14345
    },
    {
      "epoch": 8.461084905660377,
      "grad_norm": 1.964385747909546,
      "learning_rate": 7.14428522306303e-06,
      "loss": 0.4318,
      "num_input_tokens_seen": 9376024,
      "step": 14350
    },
    {
      "epoch": 8.464033018867925,
      "grad_norm": 2.487470865249634,
      "learning_rate": 7.141960820473745e-06,
      "loss": 0.4234,
      "num_input_tokens_seen": 9378872,
      "step": 14355
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 1.3759405612945557,
      "learning_rate": 7.139635850791102e-06,
      "loss": 0.616,
      "num_input_tokens_seen": 9382744,
      "step": 14360
    },
    {
      "epoch": 8.46992924528302,
      "grad_norm": 1.6250027418136597,
      "learning_rate": 7.137310314630647e-06,
      "loss": 0.3084,
      "num_input_tokens_seen": 9385784,
      "step": 14365
    },
    {
      "epoch": 8.472877358490566,
      "grad_norm": 2.631709575653076,
      "learning_rate": 7.134984212608074e-06,
      "loss": 0.5072,
      "num_input_tokens_seen": 9389336,
      "step": 14370
    },
    {
      "epoch": 8.475825471698114,
      "grad_norm": 2.13462233543396,
      "learning_rate": 7.13265754533923e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 9392024,
      "step": 14375
    },
    {
      "epoch": 8.47877358490566,
      "grad_norm": 1.930031418800354,
      "learning_rate": 7.130330313440109e-06,
      "loss": 0.3803,
      "num_input_tokens_seen": 9396216,
      "step": 14380
    },
    {
      "epoch": 8.481721698113208,
      "grad_norm": 4.915639400482178,
      "learning_rate": 7.128002517526856e-06,
      "loss": 0.3111,
      "num_input_tokens_seen": 9399288,
      "step": 14385
    },
    {
      "epoch": 8.484669811320755,
      "grad_norm": 2.127092123031616,
      "learning_rate": 7.1256741582157654e-06,
      "loss": 0.4458,
      "num_input_tokens_seen": 9402456,
      "step": 14390
    },
    {
      "epoch": 8.487617924528301,
      "grad_norm": 2.0153439044952393,
      "learning_rate": 7.123345236123282e-06,
      "loss": 0.3443,
      "num_input_tokens_seen": 9406392,
      "step": 14395
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 2.4597301483154297,
      "learning_rate": 7.121015751865994e-06,
      "loss": 0.553,
      "num_input_tokens_seen": 9408664,
      "step": 14400
    },
    {
      "epoch": 8.493514150943396,
      "grad_norm": 1.5532995462417603,
      "learning_rate": 7.118685706060645e-06,
      "loss": 0.4263,
      "num_input_tokens_seen": 9410648,
      "step": 14405
    },
    {
      "epoch": 8.496462264150944,
      "grad_norm": 2.505246877670288,
      "learning_rate": 7.116355099324126e-06,
      "loss": 0.4285,
      "num_input_tokens_seen": 9413432,
      "step": 14410
    },
    {
      "epoch": 8.49941037735849,
      "grad_norm": 1.6740667819976807,
      "learning_rate": 7.114023932273471e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 9416824,
      "step": 14415
    },
    {
      "epoch": 8.502358490566039,
      "grad_norm": 2.2221038341522217,
      "learning_rate": 7.111692205525871e-06,
      "loss": 0.3804,
      "num_input_tokens_seen": 9419928,
      "step": 14420
    },
    {
      "epoch": 8.505306603773585,
      "grad_norm": 1.7809739112854004,
      "learning_rate": 7.109359919698658e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 9424088,
      "step": 14425
    },
    {
      "epoch": 8.508254716981131,
      "grad_norm": 1.4886516332626343,
      "learning_rate": 7.107027075409316e-06,
      "loss": 0.4539,
      "num_input_tokens_seen": 9427128,
      "step": 14430
    },
    {
      "epoch": 8.51120283018868,
      "grad_norm": 2.514921188354492,
      "learning_rate": 7.104693673275475e-06,
      "loss": 0.4814,
      "num_input_tokens_seen": 9430168,
      "step": 14435
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 2.60798978805542,
      "learning_rate": 7.1023597139149115e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 9433592,
      "step": 14440
    },
    {
      "epoch": 8.517099056603774,
      "grad_norm": 3.3288321495056152,
      "learning_rate": 7.100025197945555e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 9436632,
      "step": 14445
    },
    {
      "epoch": 8.52004716981132,
      "grad_norm": 2.0994646549224854,
      "learning_rate": 7.097690125985476e-06,
      "loss": 0.5775,
      "num_input_tokens_seen": 9439960,
      "step": 14450
    },
    {
      "epoch": 8.522995283018869,
      "grad_norm": 1.8848986625671387,
      "learning_rate": 7.095354498652895e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 9443096,
      "step": 14455
    },
    {
      "epoch": 8.525943396226415,
      "grad_norm": 1.727368712425232,
      "learning_rate": 7.093018316566182e-06,
      "loss": 0.4475,
      "num_input_tokens_seen": 9445880,
      "step": 14460
    },
    {
      "epoch": 8.528891509433961,
      "grad_norm": 1.434965968132019,
      "learning_rate": 7.0906815803438465e-06,
      "loss": 0.3629,
      "num_input_tokens_seen": 9448952,
      "step": 14465
    },
    {
      "epoch": 8.53183962264151,
      "grad_norm": 3.493102788925171,
      "learning_rate": 7.088344290604554e-06,
      "loss": 0.3946,
      "num_input_tokens_seen": 9451768,
      "step": 14470
    },
    {
      "epoch": 8.534787735849056,
      "grad_norm": 1.9537174701690674,
      "learning_rate": 7.086006447967111e-06,
      "loss": 0.3855,
      "num_input_tokens_seen": 9455480,
      "step": 14475
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 1.9688647985458374,
      "learning_rate": 7.08366805305047e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 9459032,
      "step": 14480
    },
    {
      "epoch": 8.54068396226415,
      "grad_norm": 2.065587043762207,
      "learning_rate": 7.08132910647373e-06,
      "loss": 0.481,
      "num_input_tokens_seen": 9463256,
      "step": 14485
    },
    {
      "epoch": 8.543632075471699,
      "grad_norm": 2.6344969272613525,
      "learning_rate": 7.078989608856142e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 9467000,
      "step": 14490
    },
    {
      "epoch": 8.546580188679245,
      "grad_norm": 2.4032373428344727,
      "learning_rate": 7.076649560817092e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 9470680,
      "step": 14495
    },
    {
      "epoch": 8.549528301886792,
      "grad_norm": 2.502298593521118,
      "learning_rate": 7.0743089629761245e-06,
      "loss": 0.3472,
      "num_input_tokens_seen": 9473048,
      "step": 14500
    },
    {
      "epoch": 8.55247641509434,
      "grad_norm": 4.8303728103637695,
      "learning_rate": 7.071967815952917e-06,
      "loss": 0.5175,
      "num_input_tokens_seen": 9475800,
      "step": 14505
    },
    {
      "epoch": 8.555424528301886,
      "grad_norm": 2.0471091270446777,
      "learning_rate": 7.0696261203673e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 9479384,
      "step": 14510
    },
    {
      "epoch": 8.558372641509434,
      "grad_norm": 2.243032217025757,
      "learning_rate": 7.067283876839249e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 9482328,
      "step": 14515
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 2.6794323921203613,
      "learning_rate": 7.064941085988884e-06,
      "loss": 0.4335,
      "num_input_tokens_seen": 9485560,
      "step": 14520
    },
    {
      "epoch": 8.564268867924529,
      "grad_norm": 2.1054933071136475,
      "learning_rate": 7.062597748436464e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 9488216,
      "step": 14525
    },
    {
      "epoch": 8.567216981132075,
      "grad_norm": 2.1133108139038086,
      "learning_rate": 7.060253864802402e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 9491896,
      "step": 14530
    },
    {
      "epoch": 8.570165094339622,
      "grad_norm": 3.3617796897888184,
      "learning_rate": 7.05790943570725e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 9494520,
      "step": 14535
    },
    {
      "epoch": 8.57311320754717,
      "grad_norm": 2.126572847366333,
      "learning_rate": 7.055564461771704e-06,
      "loss": 0.309,
      "num_input_tokens_seen": 9498712,
      "step": 14540
    },
    {
      "epoch": 8.576061320754716,
      "grad_norm": 2.007002115249634,
      "learning_rate": 7.053218943616611e-06,
      "loss": 0.4337,
      "num_input_tokens_seen": 9501880,
      "step": 14545
    },
    {
      "epoch": 8.579009433962264,
      "grad_norm": 1.53638756275177,
      "learning_rate": 7.050872881862952e-06,
      "loss": 0.3947,
      "num_input_tokens_seen": 9505688,
      "step": 14550
    },
    {
      "epoch": 8.58195754716981,
      "grad_norm": 1.3224937915802002,
      "learning_rate": 7.04852627713186e-06,
      "loss": 0.4286,
      "num_input_tokens_seen": 9508600,
      "step": 14555
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 2.5079472064971924,
      "learning_rate": 7.0461791300446055e-06,
      "loss": 0.4641,
      "num_input_tokens_seen": 9511512,
      "step": 14560
    },
    {
      "epoch": 8.587853773584905,
      "grad_norm": 2.990499496459961,
      "learning_rate": 7.043831441222611e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 9514008,
      "step": 14565
    },
    {
      "epoch": 8.590801886792454,
      "grad_norm": 2.0288491249084473,
      "learning_rate": 7.04148321128743e-06,
      "loss": 0.4535,
      "num_input_tokens_seen": 9516760,
      "step": 14570
    },
    {
      "epoch": 8.59375,
      "grad_norm": 2.080861806869507,
      "learning_rate": 7.039134440860773e-06,
      "loss": 0.5917,
      "num_input_tokens_seen": 9520216,
      "step": 14575
    },
    {
      "epoch": 8.596698113207546,
      "grad_norm": 2.4236607551574707,
      "learning_rate": 7.036785130564484e-06,
      "loss": 0.3078,
      "num_input_tokens_seen": 9523544,
      "step": 14580
    },
    {
      "epoch": 8.599646226415095,
      "grad_norm": 2.4392125606536865,
      "learning_rate": 7.0344352810205544e-06,
      "loss": 0.3991,
      "num_input_tokens_seen": 9526968,
      "step": 14585
    },
    {
      "epoch": 8.602594339622641,
      "grad_norm": 2.386894464492798,
      "learning_rate": 7.032084892851115e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 9529912,
      "step": 14590
    },
    {
      "epoch": 8.60554245283019,
      "grad_norm": 3.8859059810638428,
      "learning_rate": 7.0297339666784425e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 9533272,
      "step": 14595
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 2.709977865219116,
      "learning_rate": 7.0273825031249556e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 9536600,
      "step": 14600
    },
    {
      "epoch": 8.611438679245284,
      "grad_norm": 2.1295385360717773,
      "learning_rate": 7.025030502813213e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 9539992,
      "step": 14605
    },
    {
      "epoch": 8.61438679245283,
      "grad_norm": 2.415579319000244,
      "learning_rate": 7.022677966365917e-06,
      "loss": 0.4831,
      "num_input_tokens_seen": 9542872,
      "step": 14610
    },
    {
      "epoch": 8.617334905660378,
      "grad_norm": 1.893093228340149,
      "learning_rate": 7.020324894405913e-06,
      "loss": 0.259,
      "num_input_tokens_seen": 9545944,
      "step": 14615
    },
    {
      "epoch": 8.620283018867925,
      "grad_norm": 2.899062395095825,
      "learning_rate": 7.017971287556185e-06,
      "loss": 0.4075,
      "num_input_tokens_seen": 9550136,
      "step": 14620
    },
    {
      "epoch": 8.623231132075471,
      "grad_norm": 3.261199712753296,
      "learning_rate": 7.015617146439863e-06,
      "loss": 0.3073,
      "num_input_tokens_seen": 9552504,
      "step": 14625
    },
    {
      "epoch": 8.62617924528302,
      "grad_norm": 4.335634231567383,
      "learning_rate": 7.0132624716802125e-06,
      "loss": 0.4452,
      "num_input_tokens_seen": 9555672,
      "step": 14630
    },
    {
      "epoch": 8.629127358490566,
      "grad_norm": 2.168668270111084,
      "learning_rate": 7.0109072639006474e-06,
      "loss": 0.3084,
      "num_input_tokens_seen": 9558392,
      "step": 14635
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 3.091409683227539,
      "learning_rate": 7.008551523724717e-06,
      "loss": 0.3454,
      "num_input_tokens_seen": 9564440,
      "step": 14640
    },
    {
      "epoch": 8.63502358490566,
      "grad_norm": 2.3820512294769287,
      "learning_rate": 7.006195251776116e-06,
      "loss": 0.4137,
      "num_input_tokens_seen": 9567480,
      "step": 14645
    },
    {
      "epoch": 8.637971698113208,
      "grad_norm": 3.3596670627593994,
      "learning_rate": 7.003838448678674e-06,
      "loss": 0.4579,
      "num_input_tokens_seen": 9570520,
      "step": 14650
    },
    {
      "epoch": 8.640919811320755,
      "grad_norm": 1.675384283065796,
      "learning_rate": 7.0014811150563675e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 9574744,
      "step": 14655
    },
    {
      "epoch": 8.643867924528301,
      "grad_norm": 3.0339221954345703,
      "learning_rate": 6.999123251533311e-06,
      "loss": 0.4121,
      "num_input_tokens_seen": 9578360,
      "step": 14660
    },
    {
      "epoch": 8.64681603773585,
      "grad_norm": 2.526143789291382,
      "learning_rate": 6.996764858733756e-06,
      "loss": 0.4831,
      "num_input_tokens_seen": 9581080,
      "step": 14665
    },
    {
      "epoch": 8.649764150943396,
      "grad_norm": 4.713304042816162,
      "learning_rate": 6.994405937282099e-06,
      "loss": 0.3017,
      "num_input_tokens_seen": 9583992,
      "step": 14670
    },
    {
      "epoch": 8.652712264150944,
      "grad_norm": 3.157762050628662,
      "learning_rate": 6.9920464878028745e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 9587416,
      "step": 14675
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 2.9744067192077637,
      "learning_rate": 6.989686510920758e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 9590520,
      "step": 14680
    },
    {
      "epoch": 8.658608490566039,
      "grad_norm": 2.0312788486480713,
      "learning_rate": 6.9873260072605634e-06,
      "loss": 0.4967,
      "num_input_tokens_seen": 9593272,
      "step": 14685
    },
    {
      "epoch": 8.661556603773585,
      "grad_norm": 2.00307035446167,
      "learning_rate": 6.984964977447243e-06,
      "loss": 0.3497,
      "num_input_tokens_seen": 9595736,
      "step": 14690
    },
    {
      "epoch": 8.664504716981131,
      "grad_norm": 2.57092022895813,
      "learning_rate": 6.982603422105889e-06,
      "loss": 0.4725,
      "num_input_tokens_seen": 9598744,
      "step": 14695
    },
    {
      "epoch": 8.66745283018868,
      "grad_norm": 3.3950090408325195,
      "learning_rate": 6.980241341861736e-06,
      "loss": 0.3113,
      "num_input_tokens_seen": 9602488,
      "step": 14700
    },
    {
      "epoch": 8.670400943396226,
      "grad_norm": 2.9208874702453613,
      "learning_rate": 6.977878737340153e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 9605784,
      "step": 14705
    },
    {
      "epoch": 8.673349056603774,
      "grad_norm": 1.779417634010315,
      "learning_rate": 6.97551560916665e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 9608568,
      "step": 14710
    },
    {
      "epoch": 8.67629716981132,
      "grad_norm": 2.8439300060272217,
      "learning_rate": 6.973151957966875e-06,
      "loss": 0.3687,
      "num_input_tokens_seen": 9611384,
      "step": 14715
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 2.7440383434295654,
      "learning_rate": 6.970787784366616e-06,
      "loss": 0.4987,
      "num_input_tokens_seen": 9614200,
      "step": 14720
    },
    {
      "epoch": 8.682193396226415,
      "grad_norm": 1.8343182802200317,
      "learning_rate": 6.968423088991797e-06,
      "loss": 0.3537,
      "num_input_tokens_seen": 9617016,
      "step": 14725
    },
    {
      "epoch": 8.685141509433961,
      "grad_norm": 1.9954990148544312,
      "learning_rate": 6.966057872468481e-06,
      "loss": 0.438,
      "num_input_tokens_seen": 9620376,
      "step": 14730
    },
    {
      "epoch": 8.68808962264151,
      "grad_norm": 5.010305404663086,
      "learning_rate": 6.963692135422872e-06,
      "loss": 0.3545,
      "num_input_tokens_seen": 9622968,
      "step": 14735
    },
    {
      "epoch": 8.691037735849056,
      "grad_norm": 1.8315011262893677,
      "learning_rate": 6.961325878481305e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 9626136,
      "step": 14740
    },
    {
      "epoch": 8.693985849056604,
      "grad_norm": 1.5511070489883423,
      "learning_rate": 6.958959102270259e-06,
      "loss": 0.4112,
      "num_input_tokens_seen": 9630456,
      "step": 14745
    },
    {
      "epoch": 8.69693396226415,
      "grad_norm": 1.396677851676941,
      "learning_rate": 6.95659180741635e-06,
      "loss": 0.3257,
      "num_input_tokens_seen": 9633976,
      "step": 14750
    },
    {
      "epoch": 8.699882075471699,
      "grad_norm": 1.6523910760879517,
      "learning_rate": 6.954223994546326e-06,
      "loss": 0.3461,
      "num_input_tokens_seen": 9637496,
      "step": 14755
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 2.5929982662200928,
      "learning_rate": 6.951855664287077e-06,
      "loss": 0.3902,
      "num_input_tokens_seen": 9640152,
      "step": 14760
    },
    {
      "epoch": 8.705778301886792,
      "grad_norm": 2.27486515045166,
      "learning_rate": 6.9494868172656304e-06,
      "loss": 0.6281,
      "num_input_tokens_seen": 9642808,
      "step": 14765
    },
    {
      "epoch": 8.70872641509434,
      "grad_norm": 3.1466143131256104,
      "learning_rate": 6.947117454109146e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 9645784,
      "step": 14770
    },
    {
      "epoch": 8.711674528301886,
      "grad_norm": 1.737451434135437,
      "learning_rate": 6.944747575444924e-06,
      "loss": 0.5281,
      "num_input_tokens_seen": 9648440,
      "step": 14775
    },
    {
      "epoch": 8.714622641509434,
      "grad_norm": 1.5840263366699219,
      "learning_rate": 6.942377181900399e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 9651608,
      "step": 14780
    },
    {
      "epoch": 8.71757075471698,
      "grad_norm": 1.9076385498046875,
      "learning_rate": 6.940006274103146e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 9654520,
      "step": 14785
    },
    {
      "epoch": 8.720518867924529,
      "grad_norm": 1.9383708238601685,
      "learning_rate": 6.93763485268087e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 9657016,
      "step": 14790
    },
    {
      "epoch": 8.723466981132075,
      "grad_norm": 1.414978265762329,
      "learning_rate": 6.935262918261416e-06,
      "loss": 0.3795,
      "num_input_tokens_seen": 9659736,
      "step": 14795
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 2.196211338043213,
      "learning_rate": 6.932890471472764e-06,
      "loss": 0.443,
      "num_input_tokens_seen": 9662424,
      "step": 14800
    },
    {
      "epoch": 8.72936320754717,
      "grad_norm": 1.6135673522949219,
      "learning_rate": 6.930517512943029e-06,
      "loss": 0.5096,
      "num_input_tokens_seen": 9665912,
      "step": 14805
    },
    {
      "epoch": 8.732311320754716,
      "grad_norm": 2.906925678253174,
      "learning_rate": 6.928144043300463e-06,
      "loss": 0.4652,
      "num_input_tokens_seen": 9670168,
      "step": 14810
    },
    {
      "epoch": 8.735259433962264,
      "grad_norm": 2.068936586380005,
      "learning_rate": 6.925770063173451e-06,
      "loss": 0.3021,
      "num_input_tokens_seen": 9673752,
      "step": 14815
    },
    {
      "epoch": 8.73820754716981,
      "grad_norm": 2.86344051361084,
      "learning_rate": 6.923395573190514e-06,
      "loss": 0.3525,
      "num_input_tokens_seen": 9677784,
      "step": 14820
    },
    {
      "epoch": 8.741155660377359,
      "grad_norm": 3.4262564182281494,
      "learning_rate": 6.921020573980313e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 9680024,
      "step": 14825
    },
    {
      "epoch": 8.744103773584905,
      "grad_norm": 1.8139543533325195,
      "learning_rate": 6.918645066171634e-06,
      "loss": 0.4844,
      "num_input_tokens_seen": 9683736,
      "step": 14830
    },
    {
      "epoch": 8.747051886792454,
      "grad_norm": 2.7502970695495605,
      "learning_rate": 6.916269050393404e-06,
      "loss": 0.4784,
      "num_input_tokens_seen": 9686776,
      "step": 14835
    },
    {
      "epoch": 8.75,
      "grad_norm": 3.1812684535980225,
      "learning_rate": 6.913892527274686e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 9689784,
      "step": 14840
    },
    {
      "epoch": 8.752948113207546,
      "grad_norm": 3.1909079551696777,
      "learning_rate": 6.9115154974446716e-06,
      "loss": 0.4718,
      "num_input_tokens_seen": 9693080,
      "step": 14845
    },
    {
      "epoch": 8.755896226415095,
      "grad_norm": 2.6671175956726074,
      "learning_rate": 6.909137961532692e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 9696184,
      "step": 14850
    },
    {
      "epoch": 8.758844339622641,
      "grad_norm": 1.6736253499984741,
      "learning_rate": 6.906759920168209e-06,
      "loss": 0.4979,
      "num_input_tokens_seen": 9700312,
      "step": 14855
    },
    {
      "epoch": 8.76179245283019,
      "grad_norm": 2.353541612625122,
      "learning_rate": 6.90438137398082e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 9703416,
      "step": 14860
    },
    {
      "epoch": 8.764740566037736,
      "grad_norm": 2.7418267726898193,
      "learning_rate": 6.902002323600252e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 9706360,
      "step": 14865
    },
    {
      "epoch": 8.767688679245284,
      "grad_norm": 1.8580416440963745,
      "learning_rate": 6.899622769656373e-06,
      "loss": 0.4364,
      "num_input_tokens_seen": 9708696,
      "step": 14870
    },
    {
      "epoch": 8.77063679245283,
      "grad_norm": 2.4957351684570312,
      "learning_rate": 6.897242712779179e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 9712024,
      "step": 14875
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 1.9319666624069214,
      "learning_rate": 6.894862153598802e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 9716152,
      "step": 14880
    },
    {
      "epoch": 8.776533018867925,
      "grad_norm": 1.8483657836914062,
      "learning_rate": 6.892481092745502e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 9719224,
      "step": 14885
    },
    {
      "epoch": 8.779481132075471,
      "grad_norm": 2.722543954849243,
      "learning_rate": 6.890099530849677e-06,
      "loss": 0.4105,
      "num_input_tokens_seen": 9723352,
      "step": 14890
    },
    {
      "epoch": 8.78242924528302,
      "grad_norm": 2.192185640335083,
      "learning_rate": 6.887717468541855e-06,
      "loss": 0.4984,
      "num_input_tokens_seen": 9726456,
      "step": 14895
    },
    {
      "epoch": 8.785377358490566,
      "grad_norm": 1.9870884418487549,
      "learning_rate": 6.885334906452696e-06,
      "loss": 0.4706,
      "num_input_tokens_seen": 9730104,
      "step": 14900
    },
    {
      "epoch": 8.788325471698114,
      "grad_norm": 4.403491497039795,
      "learning_rate": 6.882951845212997e-06,
      "loss": 0.3823,
      "num_input_tokens_seen": 9733016,
      "step": 14905
    },
    {
      "epoch": 8.79127358490566,
      "grad_norm": 2.5240285396575928,
      "learning_rate": 6.880568285453682e-06,
      "loss": 0.3168,
      "num_input_tokens_seen": 9736248,
      "step": 14910
    },
    {
      "epoch": 8.794221698113208,
      "grad_norm": 1.6845852136611938,
      "learning_rate": 6.878184227805807e-06,
      "loss": 0.3706,
      "num_input_tokens_seen": 9739928,
      "step": 14915
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 4.971837043762207,
      "learning_rate": 6.8757996729005645e-06,
      "loss": 0.3559,
      "num_input_tokens_seen": 9742456,
      "step": 14920
    },
    {
      "epoch": 8.800117924528301,
      "grad_norm": 2.684661865234375,
      "learning_rate": 6.8734146213692756e-06,
      "loss": 0.6257,
      "num_input_tokens_seen": 9745560,
      "step": 14925
    },
    {
      "epoch": 8.80306603773585,
      "grad_norm": 2.0261383056640625,
      "learning_rate": 6.87102907384339e-06,
      "loss": 0.3752,
      "num_input_tokens_seen": 9748824,
      "step": 14930
    },
    {
      "epoch": 8.806014150943396,
      "grad_norm": 1.0806154012680054,
      "learning_rate": 6.868643030954494e-06,
      "loss": 0.3596,
      "num_input_tokens_seen": 9752536,
      "step": 14935
    },
    {
      "epoch": 8.808962264150944,
      "grad_norm": 1.417434573173523,
      "learning_rate": 6.866256493334302e-06,
      "loss": 0.4135,
      "num_input_tokens_seen": 9760536,
      "step": 14940
    },
    {
      "epoch": 8.81191037735849,
      "grad_norm": 2.162376880645752,
      "learning_rate": 6.863869461614659e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 9763896,
      "step": 14945
    },
    {
      "epoch": 8.814858490566039,
      "grad_norm": 1.2979878187179565,
      "learning_rate": 6.861481936427545e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 9766680,
      "step": 14950
    },
    {
      "epoch": 8.817806603773585,
      "grad_norm": 3.0892908573150635,
      "learning_rate": 6.859093918405067e-06,
      "loss": 0.3763,
      "num_input_tokens_seen": 9769336,
      "step": 14955
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 1.8431278467178345,
      "learning_rate": 6.856705408179458e-06,
      "loss": 0.469,
      "num_input_tokens_seen": 9772440,
      "step": 14960
    },
    {
      "epoch": 8.82370283018868,
      "grad_norm": 3.80165958404541,
      "learning_rate": 6.854316406383093e-06,
      "loss": 0.3598,
      "num_input_tokens_seen": 9777304,
      "step": 14965
    },
    {
      "epoch": 8.826650943396226,
      "grad_norm": 3.5148727893829346,
      "learning_rate": 6.8519269136484665e-06,
      "loss": 0.4343,
      "num_input_tokens_seen": 9780696,
      "step": 14970
    },
    {
      "epoch": 8.829599056603774,
      "grad_norm": 2.234282970428467,
      "learning_rate": 6.849536930608208e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 9784344,
      "step": 14975
    },
    {
      "epoch": 8.83254716981132,
      "grad_norm": 4.30142879486084,
      "learning_rate": 6.847146457895078e-06,
      "loss": 0.3293,
      "num_input_tokens_seen": 9787800,
      "step": 14980
    },
    {
      "epoch": 8.835495283018869,
      "grad_norm": 2.248476266860962,
      "learning_rate": 6.8447554961419615e-06,
      "loss": 0.4482,
      "num_input_tokens_seen": 9790712,
      "step": 14985
    },
    {
      "epoch": 8.838443396226415,
      "grad_norm": 3.2127740383148193,
      "learning_rate": 6.842364045981876e-06,
      "loss": 0.6037,
      "num_input_tokens_seen": 9793720,
      "step": 14990
    },
    {
      "epoch": 8.841391509433961,
      "grad_norm": 2.464416027069092,
      "learning_rate": 6.83997210804797e-06,
      "loss": 0.5177,
      "num_input_tokens_seen": 9797816,
      "step": 14995
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 3.9430973529815674,
      "learning_rate": 6.837579682973519e-06,
      "loss": 0.4866,
      "num_input_tokens_seen": 9800600,
      "step": 15000
    },
    {
      "epoch": 8.847287735849056,
      "grad_norm": 3.8792178630828857,
      "learning_rate": 6.835186771391926e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 9804088,
      "step": 15005
    },
    {
      "epoch": 8.850235849056604,
      "grad_norm": 2.9914603233337402,
      "learning_rate": 6.8327933739367266e-06,
      "loss": 0.4833,
      "num_input_tokens_seen": 9807320,
      "step": 15010
    },
    {
      "epoch": 8.85318396226415,
      "grad_norm": 3.978513479232788,
      "learning_rate": 6.830399491241584e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 9810872,
      "step": 15015
    },
    {
      "epoch": 8.856132075471699,
      "grad_norm": 1.7317399978637695,
      "learning_rate": 6.828005123940287e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 9813976,
      "step": 15020
    },
    {
      "epoch": 8.859080188679245,
      "grad_norm": 4.423895835876465,
      "learning_rate": 6.825610272666754e-06,
      "loss": 0.3494,
      "num_input_tokens_seen": 9817976,
      "step": 15025
    },
    {
      "epoch": 8.862028301886792,
      "grad_norm": 2.1139910221099854,
      "learning_rate": 6.823214938055034e-06,
      "loss": 0.3694,
      "num_input_tokens_seen": 9820504,
      "step": 15030
    },
    {
      "epoch": 8.86497641509434,
      "grad_norm": 1.4282034635543823,
      "learning_rate": 6.8208191207393e-06,
      "loss": 0.405,
      "num_input_tokens_seen": 9823384,
      "step": 15035
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 1.0235377550125122,
      "learning_rate": 6.818422821353859e-06,
      "loss": 0.295,
      "num_input_tokens_seen": 9826648,
      "step": 15040
    },
    {
      "epoch": 8.870872641509434,
      "grad_norm": 2.4906466007232666,
      "learning_rate": 6.816026040533139e-06,
      "loss": 0.3552,
      "num_input_tokens_seen": 9830072,
      "step": 15045
    },
    {
      "epoch": 8.87382075471698,
      "grad_norm": 3.060526132583618,
      "learning_rate": 6.8136287789116966e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 9832536,
      "step": 15050
    },
    {
      "epoch": 8.876768867924529,
      "grad_norm": 2.645847797393799,
      "learning_rate": 6.81123103712422e-06,
      "loss": 0.494,
      "num_input_tokens_seen": 9835512,
      "step": 15055
    },
    {
      "epoch": 8.879716981132075,
      "grad_norm": 1.4168665409088135,
      "learning_rate": 6.808832815805518e-06,
      "loss": 0.4521,
      "num_input_tokens_seen": 9839320,
      "step": 15060
    },
    {
      "epoch": 8.882665094339622,
      "grad_norm": 1.9911712408065796,
      "learning_rate": 6.806434115590534e-06,
      "loss": 0.3892,
      "num_input_tokens_seen": 9842712,
      "step": 15065
    },
    {
      "epoch": 8.88561320754717,
      "grad_norm": 2.7948098182678223,
      "learning_rate": 6.804034937114332e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 9845720,
      "step": 15070
    },
    {
      "epoch": 8.888561320754716,
      "grad_norm": 1.5716015100479126,
      "learning_rate": 6.8016352810121064e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 9848248,
      "step": 15075
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 1.091664433479309,
      "learning_rate": 6.799235147919176e-06,
      "loss": 0.3059,
      "num_input_tokens_seen": 9852376,
      "step": 15080
    },
    {
      "epoch": 8.89445754716981,
      "grad_norm": 2.625401735305786,
      "learning_rate": 6.796834538470985e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 9855384,
      "step": 15085
    },
    {
      "epoch": 8.897405660377359,
      "grad_norm": 2.3745346069335938,
      "learning_rate": 6.794433453303106e-06,
      "loss": 0.4646,
      "num_input_tokens_seen": 9861464,
      "step": 15090
    },
    {
      "epoch": 8.900353773584905,
      "grad_norm": 3.165189027786255,
      "learning_rate": 6.792031893051238e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 9864024,
      "step": 15095
    },
    {
      "epoch": 8.903301886792454,
      "grad_norm": 1.9517658948898315,
      "learning_rate": 6.789629858351201e-06,
      "loss": 0.4157,
      "num_input_tokens_seen": 9866648,
      "step": 15100
    },
    {
      "epoch": 8.90625,
      "grad_norm": 1.5453767776489258,
      "learning_rate": 6.787227349838946e-06,
      "loss": 0.3579,
      "num_input_tokens_seen": 9869656,
      "step": 15105
    },
    {
      "epoch": 8.909198113207546,
      "grad_norm": 1.7292914390563965,
      "learning_rate": 6.784824368150548e-06,
      "loss": 0.3205,
      "num_input_tokens_seen": 9872792,
      "step": 15110
    },
    {
      "epoch": 8.912146226415095,
      "grad_norm": 1.4954750537872314,
      "learning_rate": 6.7824209139222076e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 9876952,
      "step": 15115
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 1.3489394187927246,
      "learning_rate": 6.780016987790248e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 9880952,
      "step": 15120
    },
    {
      "epoch": 8.91804245283019,
      "grad_norm": 2.641477584838867,
      "learning_rate": 6.7776125903911194e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 9885016,
      "step": 15125
    },
    {
      "epoch": 8.920990566037736,
      "grad_norm": 2.5882956981658936,
      "learning_rate": 6.775207722361396e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 9887768,
      "step": 15130
    },
    {
      "epoch": 8.923938679245284,
      "grad_norm": 1.6560536623001099,
      "learning_rate": 6.772802384337778e-06,
      "loss": 0.3462,
      "num_input_tokens_seen": 9890776,
      "step": 15135
    },
    {
      "epoch": 8.92688679245283,
      "grad_norm": 2.8866806030273438,
      "learning_rate": 6.770396576957088e-06,
      "loss": 0.4352,
      "num_input_tokens_seen": 9893688,
      "step": 15140
    },
    {
      "epoch": 8.929834905660378,
      "grad_norm": 1.9271970987319946,
      "learning_rate": 6.767990300856274e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 9896248,
      "step": 15145
    },
    {
      "epoch": 8.932783018867925,
      "grad_norm": 2.9706873893737793,
      "learning_rate": 6.765583556672408e-06,
      "loss": 0.4446,
      "num_input_tokens_seen": 9899608,
      "step": 15150
    },
    {
      "epoch": 8.935731132075471,
      "grad_norm": 1.878253698348999,
      "learning_rate": 6.763176345042687e-06,
      "loss": 0.6457,
      "num_input_tokens_seen": 9903448,
      "step": 15155
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 2.6108944416046143,
      "learning_rate": 6.760768666604429e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 9906936,
      "step": 15160
    },
    {
      "epoch": 8.941627358490566,
      "grad_norm": 2.2790870666503906,
      "learning_rate": 6.758360521995079e-06,
      "loss": 0.4164,
      "num_input_tokens_seen": 9910008,
      "step": 15165
    },
    {
      "epoch": 8.944575471698114,
      "grad_norm": 3.396894931793213,
      "learning_rate": 6.755951911852202e-06,
      "loss": 0.539,
      "num_input_tokens_seen": 9913336,
      "step": 15170
    },
    {
      "epoch": 8.94752358490566,
      "grad_norm": 1.3758920431137085,
      "learning_rate": 6.7535428368134885e-06,
      "loss": 0.3994,
      "num_input_tokens_seen": 9916120,
      "step": 15175
    },
    {
      "epoch": 8.950471698113208,
      "grad_norm": 1.8899223804473877,
      "learning_rate": 6.751133297516752e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 9919000,
      "step": 15180
    },
    {
      "epoch": 8.953419811320755,
      "grad_norm": 1.9505271911621094,
      "learning_rate": 6.748723294599928e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 9922008,
      "step": 15185
    },
    {
      "epoch": 8.956367924528301,
      "grad_norm": 1.8311269283294678,
      "learning_rate": 6.746312828701075e-06,
      "loss": 0.3945,
      "num_input_tokens_seen": 9924696,
      "step": 15190
    },
    {
      "epoch": 8.95931603773585,
      "grad_norm": 2.2431890964508057,
      "learning_rate": 6.743901900458374e-06,
      "loss": 0.2724,
      "num_input_tokens_seen": 9927864,
      "step": 15195
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 2.8137271404266357,
      "learning_rate": 6.741490510510129e-06,
      "loss": 0.4776,
      "num_input_tokens_seen": 9932504,
      "step": 15200
    },
    {
      "epoch": 8.965212264150944,
      "grad_norm": 2.260615348815918,
      "learning_rate": 6.7390786594947665e-06,
      "loss": 0.4803,
      "num_input_tokens_seen": 9937240,
      "step": 15205
    },
    {
      "epoch": 8.96816037735849,
      "grad_norm": 1.679549217224121,
      "learning_rate": 6.7366663480508335e-06,
      "loss": 0.4554,
      "num_input_tokens_seen": 9940888,
      "step": 15210
    },
    {
      "epoch": 8.971108490566039,
      "grad_norm": 2.7082278728485107,
      "learning_rate": 6.734253576817002e-06,
      "loss": 0.3445,
      "num_input_tokens_seen": 9943768,
      "step": 15215
    },
    {
      "epoch": 8.974056603773585,
      "grad_norm": 3.8038523197174072,
      "learning_rate": 6.731840346432061e-06,
      "loss": 0.3718,
      "num_input_tokens_seen": 9946264,
      "step": 15220
    },
    {
      "epoch": 8.977004716981131,
      "grad_norm": 3.385287046432495,
      "learning_rate": 6.729426657534922e-06,
      "loss": 0.5041,
      "num_input_tokens_seen": 9949496,
      "step": 15225
    },
    {
      "epoch": 8.97995283018868,
      "grad_norm": 1.9505316019058228,
      "learning_rate": 6.727012510764624e-06,
      "loss": 0.4231,
      "num_input_tokens_seen": 9952344,
      "step": 15230
    },
    {
      "epoch": 8.982900943396226,
      "grad_norm": 5.72893762588501,
      "learning_rate": 6.724597906760322e-06,
      "loss": 0.3757,
      "num_input_tokens_seen": 9955544,
      "step": 15235
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 3.8513402938842773,
      "learning_rate": 6.722182846161289e-06,
      "loss": 0.3581,
      "num_input_tokens_seen": 9959096,
      "step": 15240
    },
    {
      "epoch": 8.98879716981132,
      "grad_norm": 4.276385307312012,
      "learning_rate": 6.719767329606926e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 9961848,
      "step": 15245
    },
    {
      "epoch": 8.991745283018869,
      "grad_norm": 1.9907737970352173,
      "learning_rate": 6.717351357736751e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 9964696,
      "step": 15250
    },
    {
      "epoch": 8.994693396226415,
      "grad_norm": 3.0572190284729004,
      "learning_rate": 6.7149349311904025e-06,
      "loss": 0.4326,
      "num_input_tokens_seen": 9967832,
      "step": 15255
    },
    {
      "epoch": 8.997641509433961,
      "grad_norm": 1.4281933307647705,
      "learning_rate": 6.712518050607642e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 9970968,
      "step": 15260
    },
    {
      "epoch": 9.00058962264151,
      "grad_norm": 2.1415960788726807,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.3471,
      "num_input_tokens_seen": 9973376,
      "step": 15265
    },
    {
      "epoch": 9.003537735849056,
      "grad_norm": 1.331816554069519,
      "learning_rate": 6.707682929892513e-06,
      "loss": 0.4686,
      "num_input_tokens_seen": 9976320,
      "step": 15270
    },
    {
      "epoch": 9.006485849056604,
      "grad_norm": 3.080887794494629,
      "learning_rate": 6.705264691040266e-06,
      "loss": 0.4713,
      "num_input_tokens_seen": 9979008,
      "step": 15275
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 1.7653372287750244,
      "learning_rate": 6.7028460007118435e-06,
      "loss": 0.2584,
      "num_input_tokens_seen": 9981888,
      "step": 15280
    },
    {
      "epoch": 9.012382075471699,
      "grad_norm": 1.4327203035354614,
      "learning_rate": 6.700426859547602e-06,
      "loss": 0.3952,
      "num_input_tokens_seen": 9984608,
      "step": 15285
    },
    {
      "epoch": 9.015330188679245,
      "grad_norm": 4.05945348739624,
      "learning_rate": 6.6980072681880224e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 9988544,
      "step": 15290
    },
    {
      "epoch": 9.018278301886792,
      "grad_norm": 1.696528434753418,
      "learning_rate": 6.695587227273699e-06,
      "loss": 0.3549,
      "num_input_tokens_seen": 9991936,
      "step": 15295
    },
    {
      "epoch": 9.02122641509434,
      "grad_norm": 3.1099066734313965,
      "learning_rate": 6.69316673744535e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 9996096,
      "step": 15300
    },
    {
      "epoch": 9.024174528301886,
      "grad_norm": 3.3325438499450684,
      "learning_rate": 6.6907457993438115e-06,
      "loss": 0.4756,
      "num_input_tokens_seen": 9999904,
      "step": 15305
    },
    {
      "epoch": 9.027122641509434,
      "grad_norm": 1.9469714164733887,
      "learning_rate": 6.688324413610036e-06,
      "loss": 0.3491,
      "num_input_tokens_seen": 10002976,
      "step": 15310
    },
    {
      "epoch": 9.03007075471698,
      "grad_norm": 2.234377384185791,
      "learning_rate": 6.685902580885094e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 10005440,
      "step": 15315
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 2.4996230602264404,
      "learning_rate": 6.6834803018101794e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 10008832,
      "step": 15320
    },
    {
      "epoch": 9.035966981132075,
      "grad_norm": 1.981650948524475,
      "learning_rate": 6.681057577026599e-06,
      "loss": 0.3936,
      "num_input_tokens_seen": 10012320,
      "step": 15325
    },
    {
      "epoch": 9.038915094339623,
      "grad_norm": 2.3567802906036377,
      "learning_rate": 6.6786344071757795e-06,
      "loss": 0.2831,
      "num_input_tokens_seen": 10014400,
      "step": 15330
    },
    {
      "epoch": 9.04186320754717,
      "grad_norm": 2.6236753463745117,
      "learning_rate": 6.676210792899267e-06,
      "loss": 0.3635,
      "num_input_tokens_seen": 10018464,
      "step": 15335
    },
    {
      "epoch": 9.044811320754716,
      "grad_norm": 1.6384528875350952,
      "learning_rate": 6.6737867348387235e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 10022368,
      "step": 15340
    },
    {
      "epoch": 9.047759433962264,
      "grad_norm": 1.981227993965149,
      "learning_rate": 6.671362233635926e-06,
      "loss": 0.305,
      "num_input_tokens_seen": 10024992,
      "step": 15345
    },
    {
      "epoch": 9.05070754716981,
      "grad_norm": 2.756669282913208,
      "learning_rate": 6.668937289932775e-06,
      "loss": 0.4994,
      "num_input_tokens_seen": 10028800,
      "step": 15350
    },
    {
      "epoch": 9.053655660377359,
      "grad_norm": 1.4891854524612427,
      "learning_rate": 6.666511904371285e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 10031392,
      "step": 15355
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 2.9035892486572266,
      "learning_rate": 6.664086077593586e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 10034528,
      "step": 15360
    },
    {
      "epoch": 9.059551886792454,
      "grad_norm": 2.5645909309387207,
      "learning_rate": 6.661659810241924e-06,
      "loss": 0.478,
      "num_input_tokens_seen": 10038816,
      "step": 15365
    },
    {
      "epoch": 9.0625,
      "grad_norm": 2.1033129692077637,
      "learning_rate": 6.659233102958667e-06,
      "loss": 0.2947,
      "num_input_tokens_seen": 10042144,
      "step": 15370
    },
    {
      "epoch": 9.065448113207546,
      "grad_norm": 3.889389753341675,
      "learning_rate": 6.6568059563862965e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 10046144,
      "step": 15375
    },
    {
      "epoch": 9.068396226415095,
      "grad_norm": 1.6679675579071045,
      "learning_rate": 6.654378371167407e-06,
      "loss": 0.3953,
      "num_input_tokens_seen": 10049120,
      "step": 15380
    },
    {
      "epoch": 9.071344339622641,
      "grad_norm": 3.071485757827759,
      "learning_rate": 6.651950347944715e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 10052032,
      "step": 15385
    },
    {
      "epoch": 9.07429245283019,
      "grad_norm": 3.04103684425354,
      "learning_rate": 6.649521887361048e-06,
      "loss": 0.3856,
      "num_input_tokens_seen": 10055200,
      "step": 15390
    },
    {
      "epoch": 9.077240566037736,
      "grad_norm": 2.575634002685547,
      "learning_rate": 6.647092990059352e-06,
      "loss": 0.5212,
      "num_input_tokens_seen": 10058784,
      "step": 15395
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 3.5128839015960693,
      "learning_rate": 6.644663656682689e-06,
      "loss": 0.3869,
      "num_input_tokens_seen": 10061472,
      "step": 15400
    },
    {
      "epoch": 9.08313679245283,
      "grad_norm": 2.3795769214630127,
      "learning_rate": 6.642233887874234e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 10065216,
      "step": 15405
    },
    {
      "epoch": 9.086084905660377,
      "grad_norm": 2.605787515640259,
      "learning_rate": 6.639803684277279e-06,
      "loss": 0.3427,
      "num_input_tokens_seen": 10068032,
      "step": 15410
    },
    {
      "epoch": 9.089033018867925,
      "grad_norm": 1.4319419860839844,
      "learning_rate": 6.637373046535233e-06,
      "loss": 0.3743,
      "num_input_tokens_seen": 10072736,
      "step": 15415
    },
    {
      "epoch": 9.091981132075471,
      "grad_norm": 2.5210437774658203,
      "learning_rate": 6.634941975291617e-06,
      "loss": 0.5048,
      "num_input_tokens_seen": 10075232,
      "step": 15420
    },
    {
      "epoch": 9.09492924528302,
      "grad_norm": 4.4856390953063965,
      "learning_rate": 6.632510471190065e-06,
      "loss": 0.3688,
      "num_input_tokens_seen": 10078016,
      "step": 15425
    },
    {
      "epoch": 9.097877358490566,
      "grad_norm": 2.1707725524902344,
      "learning_rate": 6.630078534874332e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 10080672,
      "step": 15430
    },
    {
      "epoch": 9.100825471698114,
      "grad_norm": 1.9849909543991089,
      "learning_rate": 6.62764616698828e-06,
      "loss": 0.3361,
      "num_input_tokens_seen": 10083584,
      "step": 15435
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 1.614782691001892,
      "learning_rate": 6.625213368175889e-06,
      "loss": 0.5287,
      "num_input_tokens_seen": 10087552,
      "step": 15440
    },
    {
      "epoch": 9.106721698113208,
      "grad_norm": 1.361579179763794,
      "learning_rate": 6.622780139081256e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 10090272,
      "step": 15445
    },
    {
      "epoch": 9.109669811320755,
      "grad_norm": 3.4422192573547363,
      "learning_rate": 6.620346480348589e-06,
      "loss": 0.4221,
      "num_input_tokens_seen": 10093984,
      "step": 15450
    },
    {
      "epoch": 9.112617924528301,
      "grad_norm": 3.1677730083465576,
      "learning_rate": 6.617912392622206e-06,
      "loss": 0.4065,
      "num_input_tokens_seen": 10098016,
      "step": 15455
    },
    {
      "epoch": 9.11556603773585,
      "grad_norm": 2.346208095550537,
      "learning_rate": 6.615477876546544e-06,
      "loss": 0.3162,
      "num_input_tokens_seen": 10101248,
      "step": 15460
    },
    {
      "epoch": 9.118514150943396,
      "grad_norm": 1.8543267250061035,
      "learning_rate": 6.6130429327661535e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 10104864,
      "step": 15465
    },
    {
      "epoch": 9.121462264150944,
      "grad_norm": 2.716524839401245,
      "learning_rate": 6.610607561925694e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 10108416,
      "step": 15470
    },
    {
      "epoch": 9.12441037735849,
      "grad_norm": 3.1955411434173584,
      "learning_rate": 6.608171764669941e-06,
      "loss": 0.3626,
      "num_input_tokens_seen": 10111712,
      "step": 15475
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 2.9627437591552734,
      "learning_rate": 6.605735541643783e-06,
      "loss": 0.3336,
      "num_input_tokens_seen": 10114944,
      "step": 15480
    },
    {
      "epoch": 9.130306603773585,
      "grad_norm": 1.8272061347961426,
      "learning_rate": 6.603298893492219e-06,
      "loss": 0.2861,
      "num_input_tokens_seen": 10117632,
      "step": 15485
    },
    {
      "epoch": 9.133254716981131,
      "grad_norm": 1.8005974292755127,
      "learning_rate": 6.6008618208603625e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 10120992,
      "step": 15490
    },
    {
      "epoch": 9.13620283018868,
      "grad_norm": 3.366361618041992,
      "learning_rate": 6.598424324393442e-06,
      "loss": 0.4471,
      "num_input_tokens_seen": 10123840,
      "step": 15495
    },
    {
      "epoch": 9.139150943396226,
      "grad_norm": 1.0778284072875977,
      "learning_rate": 6.595986404736792e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 10127776,
      "step": 15500
    },
    {
      "epoch": 9.142099056603774,
      "grad_norm": 1.7818533182144165,
      "learning_rate": 6.5935480625358615e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 10132288,
      "step": 15505
    },
    {
      "epoch": 9.14504716981132,
      "grad_norm": 1.8036285638809204,
      "learning_rate": 6.591109298436216e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 10135680,
      "step": 15510
    },
    {
      "epoch": 9.147995283018869,
      "grad_norm": 2.523897409439087,
      "learning_rate": 6.5886701130835255e-06,
      "loss": 0.5248,
      "num_input_tokens_seen": 10139488,
      "step": 15515
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 2.1909806728363037,
      "learning_rate": 6.586230507123574e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 10143200,
      "step": 15520
    },
    {
      "epoch": 9.153891509433961,
      "grad_norm": 3.9399771690368652,
      "learning_rate": 6.583790481202261e-06,
      "loss": 0.4788,
      "num_input_tokens_seen": 10145376,
      "step": 15525
    },
    {
      "epoch": 9.15683962264151,
      "grad_norm": 2.587693452835083,
      "learning_rate": 6.5813500359655925e-06,
      "loss": 0.2898,
      "num_input_tokens_seen": 10148064,
      "step": 15530
    },
    {
      "epoch": 9.159787735849056,
      "grad_norm": 1.6846200227737427,
      "learning_rate": 6.578909172059687e-06,
      "loss": 0.3302,
      "num_input_tokens_seen": 10151488,
      "step": 15535
    },
    {
      "epoch": 9.162735849056604,
      "grad_norm": 1.7303704023361206,
      "learning_rate": 6.576467890130772e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 10154656,
      "step": 15540
    },
    {
      "epoch": 9.16568396226415,
      "grad_norm": 3.500183582305908,
      "learning_rate": 6.574026190825191e-06,
      "loss": 0.4383,
      "num_input_tokens_seen": 10157056,
      "step": 15545
    },
    {
      "epoch": 9.168632075471699,
      "grad_norm": 2.6772124767303467,
      "learning_rate": 6.57158407478939e-06,
      "loss": 0.4037,
      "num_input_tokens_seen": 10160992,
      "step": 15550
    },
    {
      "epoch": 9.171580188679245,
      "grad_norm": 1.660550832748413,
      "learning_rate": 6.569141542669935e-06,
      "loss": 0.6202,
      "num_input_tokens_seen": 10164448,
      "step": 15555
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 1.5178688764572144,
      "learning_rate": 6.566698595113492e-06,
      "loss": 0.3512,
      "num_input_tokens_seen": 10167328,
      "step": 15560
    },
    {
      "epoch": 9.17747641509434,
      "grad_norm": 2.2286252975463867,
      "learning_rate": 6.564255232766843e-06,
      "loss": 0.4526,
      "num_input_tokens_seen": 10170400,
      "step": 15565
    },
    {
      "epoch": 9.180424528301886,
      "grad_norm": 1.4649215936660767,
      "learning_rate": 6.561811456276881e-06,
      "loss": 0.4075,
      "num_input_tokens_seen": 10173472,
      "step": 15570
    },
    {
      "epoch": 9.183372641509434,
      "grad_norm": 2.10317325592041,
      "learning_rate": 6.559367266290605e-06,
      "loss": 0.2712,
      "num_input_tokens_seen": 10181312,
      "step": 15575
    },
    {
      "epoch": 9.18632075471698,
      "grad_norm": 2.1095528602600098,
      "learning_rate": 6.556922663455123e-06,
      "loss": 0.3653,
      "num_input_tokens_seen": 10184736,
      "step": 15580
    },
    {
      "epoch": 9.189268867924529,
      "grad_norm": 2.715970039367676,
      "learning_rate": 6.554477648417657e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 10187488,
      "step": 15585
    },
    {
      "epoch": 9.192216981132075,
      "grad_norm": 2.019785165786743,
      "learning_rate": 6.552032221825535e-06,
      "loss": 0.4857,
      "num_input_tokens_seen": 10190656,
      "step": 15590
    },
    {
      "epoch": 9.195165094339623,
      "grad_norm": 4.537968158721924,
      "learning_rate": 6.549586384326192e-06,
      "loss": 0.4397,
      "num_input_tokens_seen": 10194048,
      "step": 15595
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 1.6079845428466797,
      "learning_rate": 6.547140136567176e-06,
      "loss": 0.3052,
      "num_input_tokens_seen": 10197472,
      "step": 15600
    },
    {
      "epoch": 9.201061320754716,
      "grad_norm": 5.363461494445801,
      "learning_rate": 6.544693479196142e-06,
      "loss": 0.4841,
      "num_input_tokens_seen": 10200960,
      "step": 15605
    },
    {
      "epoch": 9.204009433962264,
      "grad_norm": 1.9231796264648438,
      "learning_rate": 6.542246412860851e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 10204320,
      "step": 15610
    },
    {
      "epoch": 9.20695754716981,
      "grad_norm": 4.801885604858398,
      "learning_rate": 6.5397989382091754e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 10207168,
      "step": 15615
    },
    {
      "epoch": 9.209905660377359,
      "grad_norm": 2.3074417114257812,
      "learning_rate": 6.537351055889096e-06,
      "loss": 0.3971,
      "num_input_tokens_seen": 10210784,
      "step": 15620
    },
    {
      "epoch": 9.212853773584905,
      "grad_norm": 2.8190464973449707,
      "learning_rate": 6.534902766548698e-06,
      "loss": 0.3085,
      "num_input_tokens_seen": 10213920,
      "step": 15625
    },
    {
      "epoch": 9.215801886792454,
      "grad_norm": 1.7802828550338745,
      "learning_rate": 6.532454070836176e-06,
      "loss": 0.4626,
      "num_input_tokens_seen": 10217184,
      "step": 15630
    },
    {
      "epoch": 9.21875,
      "grad_norm": 1.9891241788864136,
      "learning_rate": 6.530004969399836e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 10220640,
      "step": 15635
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 3.4818596839904785,
      "learning_rate": 6.5275554628880865e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 10223424,
      "step": 15640
    },
    {
      "epoch": 9.224646226415095,
      "grad_norm": 3.0463409423828125,
      "learning_rate": 6.525105551949444e-06,
      "loss": 0.5453,
      "num_input_tokens_seen": 10227072,
      "step": 15645
    },
    {
      "epoch": 9.227594339622641,
      "grad_norm": 4.407172203063965,
      "learning_rate": 6.522655237232535e-06,
      "loss": 0.639,
      "num_input_tokens_seen": 10230560,
      "step": 15650
    },
    {
      "epoch": 9.23054245283019,
      "grad_norm": 3.5959701538085938,
      "learning_rate": 6.5202045193860885e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 10233728,
      "step": 15655
    },
    {
      "epoch": 9.233490566037736,
      "grad_norm": 2.3856470584869385,
      "learning_rate": 6.517753399058944e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 10237152,
      "step": 15660
    },
    {
      "epoch": 9.236438679245284,
      "grad_norm": 2.3964321613311768,
      "learning_rate": 6.515301876900047e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 10239808,
      "step": 15665
    },
    {
      "epoch": 9.23938679245283,
      "grad_norm": 2.4412336349487305,
      "learning_rate": 6.512849953558445e-06,
      "loss": 0.4554,
      "num_input_tokens_seen": 10242496,
      "step": 15670
    },
    {
      "epoch": 9.242334905660377,
      "grad_norm": 6.7289276123046875,
      "learning_rate": 6.510397629683301e-06,
      "loss": 0.3891,
      "num_input_tokens_seen": 10245536,
      "step": 15675
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 1.498336672782898,
      "learning_rate": 6.507944905923872e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 10248544,
      "step": 15680
    },
    {
      "epoch": 9.248231132075471,
      "grad_norm": 2.884206533432007,
      "learning_rate": 6.505491782929531e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 10251648,
      "step": 15685
    },
    {
      "epoch": 9.25117924528302,
      "grad_norm": 5.707211494445801,
      "learning_rate": 6.5030382613497535e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 10253856,
      "step": 15690
    },
    {
      "epoch": 9.254127358490566,
      "grad_norm": 2.5112898349761963,
      "learning_rate": 6.500584341834119e-06,
      "loss": 0.3184,
      "num_input_tokens_seen": 10256992,
      "step": 15695
    },
    {
      "epoch": 9.257075471698114,
      "grad_norm": 3.376957893371582,
      "learning_rate": 6.498130025032312e-06,
      "loss": 0.3125,
      "num_input_tokens_seen": 10259680,
      "step": 15700
    },
    {
      "epoch": 9.26002358490566,
      "grad_norm": 1.8870755434036255,
      "learning_rate": 6.495675311594123e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 10262560,
      "step": 15705
    },
    {
      "epoch": 9.262971698113208,
      "grad_norm": 1.7279924154281616,
      "learning_rate": 6.493220202169452e-06,
      "loss": 0.2906,
      "num_input_tokens_seen": 10265856,
      "step": 15710
    },
    {
      "epoch": 9.265919811320755,
      "grad_norm": 1.9374191761016846,
      "learning_rate": 6.490764697408295e-06,
      "loss": 0.2836,
      "num_input_tokens_seen": 10269024,
      "step": 15715
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 2.205547332763672,
      "learning_rate": 6.48830879796076e-06,
      "loss": 0.3398,
      "num_input_tokens_seen": 10271776,
      "step": 15720
    },
    {
      "epoch": 9.27181603773585,
      "grad_norm": 2.7809596061706543,
      "learning_rate": 6.4858525044770546e-06,
      "loss": 0.3634,
      "num_input_tokens_seen": 10274976,
      "step": 15725
    },
    {
      "epoch": 9.274764150943396,
      "grad_norm": 4.34331750869751,
      "learning_rate": 6.483395817607497e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 10278080,
      "step": 15730
    },
    {
      "epoch": 9.277712264150944,
      "grad_norm": 5.8867597579956055,
      "learning_rate": 6.480938738002504e-06,
      "loss": 0.4265,
      "num_input_tokens_seen": 10281344,
      "step": 15735
    },
    {
      "epoch": 9.28066037735849,
      "grad_norm": 2.888827085494995,
      "learning_rate": 6.478481266312597e-06,
      "loss": 0.3957,
      "num_input_tokens_seen": 10285536,
      "step": 15740
    },
    {
      "epoch": 9.283608490566039,
      "grad_norm": 2.4943392276763916,
      "learning_rate": 6.476023403188403e-06,
      "loss": 0.5169,
      "num_input_tokens_seen": 10288416,
      "step": 15745
    },
    {
      "epoch": 9.286556603773585,
      "grad_norm": 3.7860286235809326,
      "learning_rate": 6.473565149280651e-06,
      "loss": 0.5996,
      "num_input_tokens_seen": 10291616,
      "step": 15750
    },
    {
      "epoch": 9.289504716981131,
      "grad_norm": 2.1486220359802246,
      "learning_rate": 6.471106505240175e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 10294144,
      "step": 15755
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 3.539700746536255,
      "learning_rate": 6.468647471717914e-06,
      "loss": 0.4203,
      "num_input_tokens_seen": 10297120,
      "step": 15760
    },
    {
      "epoch": 9.295400943396226,
      "grad_norm": 2.18477463722229,
      "learning_rate": 6.466188049364902e-06,
      "loss": 0.3335,
      "num_input_tokens_seen": 10299680,
      "step": 15765
    },
    {
      "epoch": 9.298349056603774,
      "grad_norm": 2.8212945461273193,
      "learning_rate": 6.463728238832288e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 10302848,
      "step": 15770
    },
    {
      "epoch": 9.30129716981132,
      "grad_norm": 1.390258550643921,
      "learning_rate": 6.461268040771311e-06,
      "loss": 0.3014,
      "num_input_tokens_seen": 10306272,
      "step": 15775
    },
    {
      "epoch": 9.304245283018869,
      "grad_norm": 2.182462215423584,
      "learning_rate": 6.458807455833326e-06,
      "loss": 0.3478,
      "num_input_tokens_seen": 10309728,
      "step": 15780
    },
    {
      "epoch": 9.307193396226415,
      "grad_norm": 1.9704664945602417,
      "learning_rate": 6.456346484669778e-06,
      "loss": 0.3882,
      "num_input_tokens_seen": 10313312,
      "step": 15785
    },
    {
      "epoch": 9.310141509433961,
      "grad_norm": 1.609020471572876,
      "learning_rate": 6.4538851279322225e-06,
      "loss": 0.3674,
      "num_input_tokens_seen": 10316192,
      "step": 15790
    },
    {
      "epoch": 9.31308962264151,
      "grad_norm": 3.488882064819336,
      "learning_rate": 6.451423386272312e-06,
      "loss": 0.5373,
      "num_input_tokens_seen": 10319616,
      "step": 15795
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 2.748314380645752,
      "learning_rate": 6.448961260341806e-06,
      "loss": 0.4031,
      "num_input_tokens_seen": 10322912,
      "step": 15800
    },
    {
      "epoch": 9.318985849056604,
      "grad_norm": 3.282627582550049,
      "learning_rate": 6.446498750792563e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 10325824,
      "step": 15805
    },
    {
      "epoch": 9.32193396226415,
      "grad_norm": 5.932957172393799,
      "learning_rate": 6.444035858276538e-06,
      "loss": 0.4956,
      "num_input_tokens_seen": 10328608,
      "step": 15810
    },
    {
      "epoch": 9.324882075471699,
      "grad_norm": 1.9018197059631348,
      "learning_rate": 6.441572583445799e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 10331712,
      "step": 15815
    },
    {
      "epoch": 9.327830188679245,
      "grad_norm": 3.3458659648895264,
      "learning_rate": 6.439108926952504e-06,
      "loss": 0.5112,
      "num_input_tokens_seen": 10334464,
      "step": 15820
    },
    {
      "epoch": 9.330778301886792,
      "grad_norm": 3.170395612716675,
      "learning_rate": 6.43664488944892e-06,
      "loss": 0.3103,
      "num_input_tokens_seen": 10337440,
      "step": 15825
    },
    {
      "epoch": 9.33372641509434,
      "grad_norm": 1.7229005098342896,
      "learning_rate": 6.434180471587409e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 10341120,
      "step": 15830
    },
    {
      "epoch": 9.336674528301886,
      "grad_norm": 1.8675578832626343,
      "learning_rate": 6.431715674020438e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 10344448,
      "step": 15835
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 3.0031251907348633,
      "learning_rate": 6.429250497400571e-06,
      "loss": 0.3024,
      "num_input_tokens_seen": 10348608,
      "step": 15840
    },
    {
      "epoch": 9.34257075471698,
      "grad_norm": 3.4657466411590576,
      "learning_rate": 6.426784942380475e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 10351584,
      "step": 15845
    },
    {
      "epoch": 9.345518867924529,
      "grad_norm": 1.7072380781173706,
      "learning_rate": 6.424319009612917e-06,
      "loss": 0.5521,
      "num_input_tokens_seen": 10354656,
      "step": 15850
    },
    {
      "epoch": 9.348466981132075,
      "grad_norm": 2.376465082168579,
      "learning_rate": 6.421852699750763e-06,
      "loss": 0.3913,
      "num_input_tokens_seen": 10357504,
      "step": 15855
    },
    {
      "epoch": 9.351415094339623,
      "grad_norm": 2.4121932983398438,
      "learning_rate": 6.4193860134469775e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 10360320,
      "step": 15860
    },
    {
      "epoch": 9.35436320754717,
      "grad_norm": 2.338000774383545,
      "learning_rate": 6.416918951354629e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 10362912,
      "step": 15865
    },
    {
      "epoch": 9.357311320754716,
      "grad_norm": 3.566300392150879,
      "learning_rate": 6.4144515141268816e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 10366048,
      "step": 15870
    },
    {
      "epoch": 9.360259433962264,
      "grad_norm": 3.6571462154388428,
      "learning_rate": 6.411983702416999e-06,
      "loss": 0.4644,
      "num_input_tokens_seen": 10369056,
      "step": 15875
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 1.7207053899765015,
      "learning_rate": 6.409515516878346e-06,
      "loss": 0.3383,
      "num_input_tokens_seen": 10372576,
      "step": 15880
    },
    {
      "epoch": 9.366155660377359,
      "grad_norm": 2.7244324684143066,
      "learning_rate": 6.407046958164387e-06,
      "loss": 0.4457,
      "num_input_tokens_seen": 10380032,
      "step": 15885
    },
    {
      "epoch": 9.369103773584905,
      "grad_norm": 3.127641439437866,
      "learning_rate": 6.404578026928679e-06,
      "loss": 0.3822,
      "num_input_tokens_seen": 10382592,
      "step": 15890
    },
    {
      "epoch": 9.372051886792454,
      "grad_norm": 1.755202293395996,
      "learning_rate": 6.402108723824887e-06,
      "loss": 0.3792,
      "num_input_tokens_seen": 10386048,
      "step": 15895
    },
    {
      "epoch": 9.375,
      "grad_norm": 4.480114459991455,
      "learning_rate": 6.399639049506767e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 10390240,
      "step": 15900
    },
    {
      "epoch": 9.377948113207546,
      "grad_norm": 3.6758601665496826,
      "learning_rate": 6.397169004628177e-06,
      "loss": 0.5759,
      "num_input_tokens_seen": 10394080,
      "step": 15905
    },
    {
      "epoch": 9.380896226415095,
      "grad_norm": 2.36454701423645,
      "learning_rate": 6.39469858984307e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 10396960,
      "step": 15910
    },
    {
      "epoch": 9.383844339622641,
      "grad_norm": 2.496279239654541,
      "learning_rate": 6.3922278058055024e-06,
      "loss": 0.3136,
      "num_input_tokens_seen": 10399648,
      "step": 15915
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 2.3780415058135986,
      "learning_rate": 6.389756653169622e-06,
      "loss": 0.4443,
      "num_input_tokens_seen": 10402336,
      "step": 15920
    },
    {
      "epoch": 9.389740566037736,
      "grad_norm": 2.4705562591552734,
      "learning_rate": 6.387285132589678e-06,
      "loss": 0.4462,
      "num_input_tokens_seen": 10405632,
      "step": 15925
    },
    {
      "epoch": 9.392688679245284,
      "grad_norm": 2.1138458251953125,
      "learning_rate": 6.3848132447200166e-06,
      "loss": 0.4073,
      "num_input_tokens_seen": 10408448,
      "step": 15930
    },
    {
      "epoch": 9.39563679245283,
      "grad_norm": 2.3524551391601562,
      "learning_rate": 6.38234099021508e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 10412000,
      "step": 15935
    },
    {
      "epoch": 9.398584905660377,
      "grad_norm": 3.1682446002960205,
      "learning_rate": 6.379868369729409e-06,
      "loss": 0.3805,
      "num_input_tokens_seen": 10415392,
      "step": 15940
    },
    {
      "epoch": 9.401533018867925,
      "grad_norm": 2.011284351348877,
      "learning_rate": 6.377395383917642e-06,
      "loss": 0.3233,
      "num_input_tokens_seen": 10418368,
      "step": 15945
    },
    {
      "epoch": 9.404481132075471,
      "grad_norm": 1.7249884605407715,
      "learning_rate": 6.374922033434507e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 10422592,
      "step": 15950
    },
    {
      "epoch": 9.40742924528302,
      "grad_norm": 3.0263454914093018,
      "learning_rate": 6.372448318934842e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 10426080,
      "step": 15955
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 1.9412637948989868,
      "learning_rate": 6.369974241073569e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 10429024,
      "step": 15960
    },
    {
      "epoch": 9.413325471698114,
      "grad_norm": 1.1896816492080688,
      "learning_rate": 6.367499800505709e-06,
      "loss": 0.516,
      "num_input_tokens_seen": 10432672,
      "step": 15965
    },
    {
      "epoch": 9.41627358490566,
      "grad_norm": 2.2452759742736816,
      "learning_rate": 6.365024997886384e-06,
      "loss": 0.3979,
      "num_input_tokens_seen": 10436000,
      "step": 15970
    },
    {
      "epoch": 9.419221698113208,
      "grad_norm": 3.163442850112915,
      "learning_rate": 6.362549833870808e-06,
      "loss": 0.4866,
      "num_input_tokens_seen": 10438624,
      "step": 15975
    },
    {
      "epoch": 9.422169811320755,
      "grad_norm": 1.1726953983306885,
      "learning_rate": 6.360074309114293e-06,
      "loss": 0.3059,
      "num_input_tokens_seen": 10442784,
      "step": 15980
    },
    {
      "epoch": 9.425117924528301,
      "grad_norm": 1.8725285530090332,
      "learning_rate": 6.357598424272241e-06,
      "loss": 0.3721,
      "num_input_tokens_seen": 10445536,
      "step": 15985
    },
    {
      "epoch": 9.42806603773585,
      "grad_norm": 1.75450599193573,
      "learning_rate": 6.355122180000156e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 10448320,
      "step": 15990
    },
    {
      "epoch": 9.431014150943396,
      "grad_norm": 2.3833396434783936,
      "learning_rate": 6.352645576953635e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 10451008,
      "step": 15995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 1.9738472700119019,
      "learning_rate": 6.350168615788366e-06,
      "loss": 0.3774,
      "num_input_tokens_seen": 10454176,
      "step": 16000
    },
    {
      "epoch": 9.43691037735849,
      "grad_norm": 2.683997392654419,
      "learning_rate": 6.347691297160137e-06,
      "loss": 0.5164,
      "num_input_tokens_seen": 10457216,
      "step": 16005
    },
    {
      "epoch": 9.439858490566039,
      "grad_norm": 1.3543111085891724,
      "learning_rate": 6.34521362172483e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 10459616,
      "step": 16010
    },
    {
      "epoch": 9.442806603773585,
      "grad_norm": 1.8764266967773438,
      "learning_rate": 6.342735590138417e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 10462048,
      "step": 16015
    },
    {
      "epoch": 9.445754716981131,
      "grad_norm": 3.595707416534424,
      "learning_rate": 6.340257203056972e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 10464352,
      "step": 16020
    },
    {
      "epoch": 9.44870283018868,
      "grad_norm": 2.844599962234497,
      "learning_rate": 6.3377784611366554e-06,
      "loss": 0.605,
      "num_input_tokens_seen": 10467712,
      "step": 16025
    },
    {
      "epoch": 9.451650943396226,
      "grad_norm": 2.6746275424957275,
      "learning_rate": 6.335299365033726e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 10470400,
      "step": 16030
    },
    {
      "epoch": 9.454599056603774,
      "grad_norm": 2.0208332538604736,
      "learning_rate": 6.3328199154045346e-06,
      "loss": 0.3799,
      "num_input_tokens_seen": 10473920,
      "step": 16035
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 1.7468231916427612,
      "learning_rate": 6.3303401129055265e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 10477408,
      "step": 16040
    },
    {
      "epoch": 9.460495283018869,
      "grad_norm": 2.3630073070526123,
      "learning_rate": 6.32785995819324e-06,
      "loss": 0.438,
      "num_input_tokens_seen": 10480096,
      "step": 16045
    },
    {
      "epoch": 9.463443396226415,
      "grad_norm": 2.4280097484588623,
      "learning_rate": 6.3253794519243075e-06,
      "loss": 0.3358,
      "num_input_tokens_seen": 10483008,
      "step": 16050
    },
    {
      "epoch": 9.466391509433961,
      "grad_norm": 2.2033209800720215,
      "learning_rate": 6.322898594755452e-06,
      "loss": 0.3934,
      "num_input_tokens_seen": 10485984,
      "step": 16055
    },
    {
      "epoch": 9.46933962264151,
      "grad_norm": 4.023126125335693,
      "learning_rate": 6.320417387343492e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 10488192,
      "step": 16060
    },
    {
      "epoch": 9.472287735849056,
      "grad_norm": 2.5232553482055664,
      "learning_rate": 6.3179358303453386e-06,
      "loss": 0.395,
      "num_input_tokens_seen": 10491232,
      "step": 16065
    },
    {
      "epoch": 9.475235849056604,
      "grad_norm": 4.92816162109375,
      "learning_rate": 6.315453924417995e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 10494336,
      "step": 16070
    },
    {
      "epoch": 9.47818396226415,
      "grad_norm": 2.6144378185272217,
      "learning_rate": 6.312971670218554e-06,
      "loss": 0.4041,
      "num_input_tokens_seen": 10497440,
      "step": 16075
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 1.9908008575439453,
      "learning_rate": 6.3104890684042055e-06,
      "loss": 0.3657,
      "num_input_tokens_seen": 10500224,
      "step": 16080
    },
    {
      "epoch": 9.484080188679245,
      "grad_norm": 1.6848758459091187,
      "learning_rate": 6.308006119632228e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 10504000,
      "step": 16085
    },
    {
      "epoch": 9.487028301886792,
      "grad_norm": 1.8162635564804077,
      "learning_rate": 6.305522824559993e-06,
      "loss": 0.3485,
      "num_input_tokens_seen": 10507296,
      "step": 16090
    },
    {
      "epoch": 9.48997641509434,
      "grad_norm": 2.775977373123169,
      "learning_rate": 6.303039183844965e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 10510784,
      "step": 16095
    },
    {
      "epoch": 9.492924528301886,
      "grad_norm": 1.9606175422668457,
      "learning_rate": 6.300555198144697e-06,
      "loss": 0.4554,
      "num_input_tokens_seen": 10514592,
      "step": 16100
    },
    {
      "epoch": 9.495872641509434,
      "grad_norm": 3.71420955657959,
      "learning_rate": 6.2980708681168335e-06,
      "loss": 0.3353,
      "num_input_tokens_seen": 10517248,
      "step": 16105
    },
    {
      "epoch": 9.49882075471698,
      "grad_norm": 3.606518507003784,
      "learning_rate": 6.2955861944191145e-06,
      "loss": 0.3436,
      "num_input_tokens_seen": 10520928,
      "step": 16110
    },
    {
      "epoch": 9.501768867924529,
      "grad_norm": 2.8836395740509033,
      "learning_rate": 6.293101177709367e-06,
      "loss": 0.432,
      "num_input_tokens_seen": 10527264,
      "step": 16115
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 1.9970805644989014,
      "learning_rate": 6.290615818645509e-06,
      "loss": 0.4275,
      "num_input_tokens_seen": 10530528,
      "step": 16120
    },
    {
      "epoch": 9.507665094339622,
      "grad_norm": 1.5784622430801392,
      "learning_rate": 6.288130117885552e-06,
      "loss": 0.2501,
      "num_input_tokens_seen": 10533088,
      "step": 16125
    },
    {
      "epoch": 9.51061320754717,
      "grad_norm": 3.0485551357269287,
      "learning_rate": 6.285644076087594e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 10535744,
      "step": 16130
    },
    {
      "epoch": 9.513561320754716,
      "grad_norm": 3.091919422149658,
      "learning_rate": 6.283157693909826e-06,
      "loss": 0.345,
      "num_input_tokens_seen": 10539488,
      "step": 16135
    },
    {
      "epoch": 9.516509433962264,
      "grad_norm": 4.458471775054932,
      "learning_rate": 6.280670972010528e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 10542176,
      "step": 16140
    },
    {
      "epoch": 9.51945754716981,
      "grad_norm": 2.239933729171753,
      "learning_rate": 6.278183911048072e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 10544928,
      "step": 16145
    },
    {
      "epoch": 9.522405660377359,
      "grad_norm": 2.526273488998413,
      "learning_rate": 6.275696511680915e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 10547648,
      "step": 16150
    },
    {
      "epoch": 9.525353773584905,
      "grad_norm": 1.8488178253173828,
      "learning_rate": 6.27320877456761e-06,
      "loss": 0.2481,
      "num_input_tokens_seen": 10551168,
      "step": 16155
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 4.2491912841796875,
      "learning_rate": 6.270720700366793e-06,
      "loss": 0.3234,
      "num_input_tokens_seen": 10555072,
      "step": 16160
    },
    {
      "epoch": 9.53125,
      "grad_norm": 2.362513303756714,
      "learning_rate": 6.2682322897371974e-06,
      "loss": 0.4341,
      "num_input_tokens_seen": 10557568,
      "step": 16165
    },
    {
      "epoch": 9.534198113207546,
      "grad_norm": 2.9463720321655273,
      "learning_rate": 6.265743543337634e-06,
      "loss": 0.3947,
      "num_input_tokens_seen": 10560608,
      "step": 16170
    },
    {
      "epoch": 9.537146226415095,
      "grad_norm": 2.8580873012542725,
      "learning_rate": 6.2632544618270156e-06,
      "loss": 0.3695,
      "num_input_tokens_seen": 10563904,
      "step": 16175
    },
    {
      "epoch": 9.540094339622641,
      "grad_norm": 2.2977294921875,
      "learning_rate": 6.260765045864332e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 10566720,
      "step": 16180
    },
    {
      "epoch": 9.54304245283019,
      "grad_norm": 1.798058032989502,
      "learning_rate": 6.258275296108669e-06,
      "loss": 0.2965,
      "num_input_tokens_seen": 10569408,
      "step": 16185
    },
    {
      "epoch": 9.545990566037736,
      "grad_norm": 1.6972498893737793,
      "learning_rate": 6.2557852132191985e-06,
      "loss": 0.3311,
      "num_input_tokens_seen": 10571712,
      "step": 16190
    },
    {
      "epoch": 9.548938679245284,
      "grad_norm": 2.3189690113067627,
      "learning_rate": 6.253294797855182e-06,
      "loss": 0.3536,
      "num_input_tokens_seen": 10574208,
      "step": 16195
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 1.6504532098770142,
      "learning_rate": 6.250804050675964e-06,
      "loss": 0.3973,
      "num_input_tokens_seen": 10578208,
      "step": 16200
    },
    {
      "epoch": 9.554834905660378,
      "grad_norm": 2.4056222438812256,
      "learning_rate": 6.248312972340984e-06,
      "loss": 0.3172,
      "num_input_tokens_seen": 10581152,
      "step": 16205
    },
    {
      "epoch": 9.557783018867925,
      "grad_norm": 4.411544322967529,
      "learning_rate": 6.2458215635097656e-06,
      "loss": 0.3862,
      "num_input_tokens_seen": 10584224,
      "step": 16210
    },
    {
      "epoch": 9.560731132075471,
      "grad_norm": 1.9530009031295776,
      "learning_rate": 6.2433298248419175e-06,
      "loss": 0.4331,
      "num_input_tokens_seen": 10588192,
      "step": 16215
    },
    {
      "epoch": 9.56367924528302,
      "grad_norm": 2.2928414344787598,
      "learning_rate": 6.2408377569971405e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 10591168,
      "step": 16220
    },
    {
      "epoch": 9.566627358490566,
      "grad_norm": 1.3964862823486328,
      "learning_rate": 6.238345360635221e-06,
      "loss": 0.3192,
      "num_input_tokens_seen": 10594240,
      "step": 16225
    },
    {
      "epoch": 9.569575471698114,
      "grad_norm": 3.570502996444702,
      "learning_rate": 6.2358526364160274e-06,
      "loss": 0.3946,
      "num_input_tokens_seen": 10596736,
      "step": 16230
    },
    {
      "epoch": 9.57252358490566,
      "grad_norm": 3.030200958251953,
      "learning_rate": 6.2333595849995245e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 10599296,
      "step": 16235
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 1.7806761264801025,
      "learning_rate": 6.230866207045756e-06,
      "loss": 0.3455,
      "num_input_tokens_seen": 10602976,
      "step": 16240
    },
    {
      "epoch": 9.578419811320755,
      "grad_norm": 3.6636722087860107,
      "learning_rate": 6.228372503214853e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 10605504,
      "step": 16245
    },
    {
      "epoch": 9.581367924528301,
      "grad_norm": 3.7286529541015625,
      "learning_rate": 6.225878474167035e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 10608352,
      "step": 16250
    },
    {
      "epoch": 9.58431603773585,
      "grad_norm": 2.06373929977417,
      "learning_rate": 6.22338412056261e-06,
      "loss": 0.2917,
      "num_input_tokens_seen": 10611072,
      "step": 16255
    },
    {
      "epoch": 9.587264150943396,
      "grad_norm": 3.1554698944091797,
      "learning_rate": 6.220889443061966e-06,
      "loss": 0.441,
      "num_input_tokens_seen": 10614272,
      "step": 16260
    },
    {
      "epoch": 9.590212264150944,
      "grad_norm": 2.2311768531799316,
      "learning_rate": 6.2183944423255796e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 10617184,
      "step": 16265
    },
    {
      "epoch": 9.59316037735849,
      "grad_norm": 1.7069822549819946,
      "learning_rate": 6.215899119014015e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 10620640,
      "step": 16270
    },
    {
      "epoch": 9.596108490566039,
      "grad_norm": 3.524672508239746,
      "learning_rate": 6.2134034737879175e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 10624832,
      "step": 16275
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 1.3416675329208374,
      "learning_rate": 6.2109075073080205e-06,
      "loss": 0.3693,
      "num_input_tokens_seen": 10628704,
      "step": 16280
    },
    {
      "epoch": 9.602004716981131,
      "grad_norm": 2.396172046661377,
      "learning_rate": 6.2084112202351425e-06,
      "loss": 0.3686,
      "num_input_tokens_seen": 10631776,
      "step": 16285
    },
    {
      "epoch": 9.60495283018868,
      "grad_norm": 1.1870280504226685,
      "learning_rate": 6.205914613230186e-06,
      "loss": 0.3139,
      "num_input_tokens_seen": 10634976,
      "step": 16290
    },
    {
      "epoch": 9.607900943396226,
      "grad_norm": 3.530200958251953,
      "learning_rate": 6.203417686954138e-06,
      "loss": 0.5008,
      "num_input_tokens_seen": 10639360,
      "step": 16295
    },
    {
      "epoch": 9.610849056603774,
      "grad_norm": 2.2217001914978027,
      "learning_rate": 6.2009204420680706e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 10645792,
      "step": 16300
    },
    {
      "epoch": 9.61379716981132,
      "grad_norm": 2.8113760948181152,
      "learning_rate": 6.198422879233141e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 10649440,
      "step": 16305
    },
    {
      "epoch": 9.616745283018869,
      "grad_norm": 2.6348636150360107,
      "learning_rate": 6.1959249991105895e-06,
      "loss": 0.3659,
      "num_input_tokens_seen": 10652192,
      "step": 16310
    },
    {
      "epoch": 9.619693396226415,
      "grad_norm": 2.958040475845337,
      "learning_rate": 6.19342680236174e-06,
      "loss": 0.3951,
      "num_input_tokens_seen": 10655104,
      "step": 16315
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 3.1591506004333496,
      "learning_rate": 6.190928289648003e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 10658880,
      "step": 16320
    },
    {
      "epoch": 9.62558962264151,
      "grad_norm": 2.290562391281128,
      "learning_rate": 6.188429461630866e-06,
      "loss": 0.3187,
      "num_input_tokens_seen": 10662464,
      "step": 16325
    },
    {
      "epoch": 9.628537735849056,
      "grad_norm": 2.415940761566162,
      "learning_rate": 6.185930318971909e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 10665824,
      "step": 16330
    },
    {
      "epoch": 9.631485849056604,
      "grad_norm": 5.0167365074157715,
      "learning_rate": 6.1834308623327885e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 10668704,
      "step": 16335
    },
    {
      "epoch": 9.63443396226415,
      "grad_norm": 4.060776710510254,
      "learning_rate": 6.180931092375247e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 10671648,
      "step": 16340
    },
    {
      "epoch": 9.637382075471699,
      "grad_norm": 4.225286960601807,
      "learning_rate": 6.1784310097611075e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 10674688,
      "step": 16345
    },
    {
      "epoch": 9.640330188679245,
      "grad_norm": 3.3126354217529297,
      "learning_rate": 6.1759306151522815e-06,
      "loss": 0.3501,
      "num_input_tokens_seen": 10677600,
      "step": 16350
    },
    {
      "epoch": 9.643278301886792,
      "grad_norm": 3.136636257171631,
      "learning_rate": 6.173429909210755e-06,
      "loss": 0.4741,
      "num_input_tokens_seen": 10681088,
      "step": 16355
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 3.1799864768981934,
      "learning_rate": 6.170928892598606e-06,
      "loss": 0.4729,
      "num_input_tokens_seen": 10684128,
      "step": 16360
    },
    {
      "epoch": 9.649174528301886,
      "grad_norm": 1.99306058883667,
      "learning_rate": 6.168427565977984e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 10687488,
      "step": 16365
    },
    {
      "epoch": 9.652122641509434,
      "grad_norm": 3.39628529548645,
      "learning_rate": 6.165925930011129e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 10690368,
      "step": 16370
    },
    {
      "epoch": 9.65507075471698,
      "grad_norm": 2.9993488788604736,
      "learning_rate": 6.163423985360359e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 10693504,
      "step": 16375
    },
    {
      "epoch": 9.658018867924529,
      "grad_norm": 2.750208854675293,
      "learning_rate": 6.160921732688076e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 10696256,
      "step": 16380
    },
    {
      "epoch": 9.660966981132075,
      "grad_norm": 1.7799794673919678,
      "learning_rate": 6.158419172656759e-06,
      "loss": 0.2819,
      "num_input_tokens_seen": 10699168,
      "step": 16385
    },
    {
      "epoch": 9.663915094339622,
      "grad_norm": 3.7818591594696045,
      "learning_rate": 6.155916305928974e-06,
      "loss": 0.4419,
      "num_input_tokens_seen": 10702528,
      "step": 16390
    },
    {
      "epoch": 9.66686320754717,
      "grad_norm": 3.0710859298706055,
      "learning_rate": 6.153413133167366e-06,
      "loss": 0.3336,
      "num_input_tokens_seen": 10708704,
      "step": 16395
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 2.094120502471924,
      "learning_rate": 6.1509096550346596e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 10712672,
      "step": 16400
    },
    {
      "epoch": 9.672759433962264,
      "grad_norm": 1.226839303970337,
      "learning_rate": 6.148405872193661e-06,
      "loss": 0.3532,
      "num_input_tokens_seen": 10715968,
      "step": 16405
    },
    {
      "epoch": 9.67570754716981,
      "grad_norm": 2.4415035247802734,
      "learning_rate": 6.14590178530726e-06,
      "loss": 0.3605,
      "num_input_tokens_seen": 10719328,
      "step": 16410
    },
    {
      "epoch": 9.678655660377359,
      "grad_norm": 2.1392781734466553,
      "learning_rate": 6.143397395038422e-06,
      "loss": 0.4813,
      "num_input_tokens_seen": 10722496,
      "step": 16415
    },
    {
      "epoch": 9.681603773584905,
      "grad_norm": 3.565587043762207,
      "learning_rate": 6.140892702050196e-06,
      "loss": 0.3453,
      "num_input_tokens_seen": 10725728,
      "step": 16420
    },
    {
      "epoch": 9.684551886792454,
      "grad_norm": 2.069631338119507,
      "learning_rate": 6.138387707005711e-06,
      "loss": 0.3117,
      "num_input_tokens_seen": 10728320,
      "step": 16425
    },
    {
      "epoch": 9.6875,
      "grad_norm": 1.9791226387023926,
      "learning_rate": 6.135882410568172e-06,
      "loss": 0.3449,
      "num_input_tokens_seen": 10732256,
      "step": 16430
    },
    {
      "epoch": 9.690448113207546,
      "grad_norm": 2.568251848220825,
      "learning_rate": 6.133376813400872e-06,
      "loss": 0.2879,
      "num_input_tokens_seen": 10734784,
      "step": 16435
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 2.4495906829833984,
      "learning_rate": 6.130870916167175e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 10738752,
      "step": 16440
    },
    {
      "epoch": 9.696344339622641,
      "grad_norm": 3.1261863708496094,
      "learning_rate": 6.128364719530528e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 10741888,
      "step": 16445
    },
    {
      "epoch": 9.69929245283019,
      "grad_norm": 3.011477470397949,
      "learning_rate": 6.125858224154459e-06,
      "loss": 0.4203,
      "num_input_tokens_seen": 10744992,
      "step": 16450
    },
    {
      "epoch": 9.702240566037736,
      "grad_norm": 6.573583126068115,
      "learning_rate": 6.123351430702576e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 10749120,
      "step": 16455
    },
    {
      "epoch": 9.705188679245284,
      "grad_norm": 1.6321626901626587,
      "learning_rate": 6.1208443398385575e-06,
      "loss": 0.3797,
      "num_input_tokens_seen": 10751872,
      "step": 16460
    },
    {
      "epoch": 9.70813679245283,
      "grad_norm": 2.0667264461517334,
      "learning_rate": 6.118336952226169e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 10755552,
      "step": 16465
    },
    {
      "epoch": 9.711084905660378,
      "grad_norm": 2.4497179985046387,
      "learning_rate": 6.115829268529254e-06,
      "loss": 0.4533,
      "num_input_tokens_seen": 10758176,
      "step": 16470
    },
    {
      "epoch": 9.714033018867925,
      "grad_norm": 2.031749725341797,
      "learning_rate": 6.1133212894117326e-06,
      "loss": 0.3463,
      "num_input_tokens_seen": 10761312,
      "step": 16475
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 1.4427075386047363,
      "learning_rate": 6.1108130155375986e-06,
      "loss": 0.3468,
      "num_input_tokens_seen": 10765152,
      "step": 16480
    },
    {
      "epoch": 9.71992924528302,
      "grad_norm": 3.0966665744781494,
      "learning_rate": 6.108304447570933e-06,
      "loss": 0.3598,
      "num_input_tokens_seen": 10768480,
      "step": 16485
    },
    {
      "epoch": 9.722877358490566,
      "grad_norm": 2.6871533393859863,
      "learning_rate": 6.105795586175888e-06,
      "loss": 0.3972,
      "num_input_tokens_seen": 10771232,
      "step": 16490
    },
    {
      "epoch": 9.725825471698114,
      "grad_norm": 1.946668267250061,
      "learning_rate": 6.1032864320166954e-06,
      "loss": 0.3784,
      "num_input_tokens_seen": 10774624,
      "step": 16495
    },
    {
      "epoch": 9.72877358490566,
      "grad_norm": 2.5856223106384277,
      "learning_rate": 6.100776985757666e-06,
      "loss": 0.3183,
      "num_input_tokens_seen": 10779360,
      "step": 16500
    },
    {
      "epoch": 9.731721698113208,
      "grad_norm": 2.0438878536224365,
      "learning_rate": 6.098267248063186e-06,
      "loss": 0.3288,
      "num_input_tokens_seen": 10782528,
      "step": 16505
    },
    {
      "epoch": 9.734669811320755,
      "grad_norm": 2.5015594959259033,
      "learning_rate": 6.0957572195977165e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 10786464,
      "step": 16510
    },
    {
      "epoch": 9.737617924528301,
      "grad_norm": 2.9551503658294678,
      "learning_rate": 6.0932469010258025e-06,
      "loss": 0.3568,
      "num_input_tokens_seen": 10789024,
      "step": 16515
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 2.0735690593719482,
      "learning_rate": 6.0907362930120594e-06,
      "loss": 0.528,
      "num_input_tokens_seen": 10792256,
      "step": 16520
    },
    {
      "epoch": 9.743514150943396,
      "grad_norm": 2.6473567485809326,
      "learning_rate": 6.088225396221181e-06,
      "loss": 0.3334,
      "num_input_tokens_seen": 10794784,
      "step": 16525
    },
    {
      "epoch": 9.746462264150944,
      "grad_norm": 2.4034645557403564,
      "learning_rate": 6.0857142113179415e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 10798208,
      "step": 16530
    },
    {
      "epoch": 9.74941037735849,
      "grad_norm": 3.59378719329834,
      "learning_rate": 6.083202738967182e-06,
      "loss": 0.4475,
      "num_input_tokens_seen": 10801056,
      "step": 16535
    },
    {
      "epoch": 9.752358490566039,
      "grad_norm": 2.1253271102905273,
      "learning_rate": 6.0806909798338324e-06,
      "loss": 0.3704,
      "num_input_tokens_seen": 10804128,
      "step": 16540
    },
    {
      "epoch": 9.755306603773585,
      "grad_norm": 1.687952995300293,
      "learning_rate": 6.0781789345828854e-06,
      "loss": 0.4834,
      "num_input_tokens_seen": 10806752,
      "step": 16545
    },
    {
      "epoch": 9.758254716981131,
      "grad_norm": 2.3955793380737305,
      "learning_rate": 6.0756666038794195e-06,
      "loss": 0.2965,
      "num_input_tokens_seen": 10810784,
      "step": 16550
    },
    {
      "epoch": 9.76120283018868,
      "grad_norm": 3.198855400085449,
      "learning_rate": 6.073153988388586e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 10814304,
      "step": 16555
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 2.076637029647827,
      "learning_rate": 6.070641088775608e-06,
      "loss": 0.3705,
      "num_input_tokens_seen": 10817184,
      "step": 16560
    },
    {
      "epoch": 9.767099056603774,
      "grad_norm": 2.2118148803710938,
      "learning_rate": 6.068127905705787e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 10820832,
      "step": 16565
    },
    {
      "epoch": 9.77004716981132,
      "grad_norm": 2.4881536960601807,
      "learning_rate": 6.065614439844501e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 10823360,
      "step": 16570
    },
    {
      "epoch": 9.772995283018869,
      "grad_norm": 1.8466612100601196,
      "learning_rate": 6.063100691857198e-06,
      "loss": 0.5066,
      "num_input_tokens_seen": 10826048,
      "step": 16575
    },
    {
      "epoch": 9.775943396226415,
      "grad_norm": 2.9641366004943848,
      "learning_rate": 6.060586662409407e-06,
      "loss": 0.467,
      "num_input_tokens_seen": 10828928,
      "step": 16580
    },
    {
      "epoch": 9.778891509433961,
      "grad_norm": 1.708082914352417,
      "learning_rate": 6.058072352166724e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 10832992,
      "step": 16585
    },
    {
      "epoch": 9.78183962264151,
      "grad_norm": 1.702005386352539,
      "learning_rate": 6.055557761794826e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 10836768,
      "step": 16590
    },
    {
      "epoch": 9.784787735849056,
      "grad_norm": 2.0201401710510254,
      "learning_rate": 6.053042891959462e-06,
      "loss": 0.5,
      "num_input_tokens_seen": 10840352,
      "step": 16595
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 1.7130177021026611,
      "learning_rate": 6.050527743326455e-06,
      "loss": 0.3592,
      "num_input_tokens_seen": 10843904,
      "step": 16600
    },
    {
      "epoch": 9.79068396226415,
      "grad_norm": 1.4081774950027466,
      "learning_rate": 6.048012316561699e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 10846976,
      "step": 16605
    },
    {
      "epoch": 9.793632075471699,
      "grad_norm": 2.825112819671631,
      "learning_rate": 6.045496612331166e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 10851040,
      "step": 16610
    },
    {
      "epoch": 9.796580188679245,
      "grad_norm": 2.7591233253479004,
      "learning_rate": 6.0429806313009e-06,
      "loss": 0.4033,
      "num_input_tokens_seen": 10853888,
      "step": 16615
    },
    {
      "epoch": 9.799528301886792,
      "grad_norm": 3.659418821334839,
      "learning_rate": 6.040464374137015e-06,
      "loss": 0.527,
      "num_input_tokens_seen": 10856896,
      "step": 16620
    },
    {
      "epoch": 9.80247641509434,
      "grad_norm": 2.3377346992492676,
      "learning_rate": 6.0379478415057045e-06,
      "loss": 0.4221,
      "num_input_tokens_seen": 10860224,
      "step": 16625
    },
    {
      "epoch": 9.805424528301886,
      "grad_norm": 2.246450901031494,
      "learning_rate": 6.035431034073228e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 10863552,
      "step": 16630
    },
    {
      "epoch": 9.808372641509434,
      "grad_norm": 5.673642158508301,
      "learning_rate": 6.0329139525059235e-06,
      "loss": 0.3483,
      "num_input_tokens_seen": 10866336,
      "step": 16635
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 2.305879592895508,
      "learning_rate": 6.030396597470198e-06,
      "loss": 0.3039,
      "num_input_tokens_seen": 10869952,
      "step": 16640
    },
    {
      "epoch": 9.814268867924529,
      "grad_norm": 2.4129884243011475,
      "learning_rate": 6.027878969632534e-06,
      "loss": 0.5478,
      "num_input_tokens_seen": 10872544,
      "step": 16645
    },
    {
      "epoch": 9.817216981132075,
      "grad_norm": 2.520533323287964,
      "learning_rate": 6.025361069659482e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 10875296,
      "step": 16650
    },
    {
      "epoch": 9.820165094339622,
      "grad_norm": 1.6545016765594482,
      "learning_rate": 6.022842898217668e-06,
      "loss": 0.4136,
      "num_input_tokens_seen": 10879232,
      "step": 16655
    },
    {
      "epoch": 9.82311320754717,
      "grad_norm": 2.255671262741089,
      "learning_rate": 6.020324455973788e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 10882272,
      "step": 16660
    },
    {
      "epoch": 9.826061320754716,
      "grad_norm": 1.8184921741485596,
      "learning_rate": 6.017805743594612e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 10886208,
      "step": 16665
    },
    {
      "epoch": 9.829009433962264,
      "grad_norm": 2.69875168800354,
      "learning_rate": 6.0152867617469776e-06,
      "loss": 0.5213,
      "num_input_tokens_seen": 10889440,
      "step": 16670
    },
    {
      "epoch": 9.83195754716981,
      "grad_norm": 3.985114336013794,
      "learning_rate": 6.012767511097799e-06,
      "loss": 0.4999,
      "num_input_tokens_seen": 10893792,
      "step": 16675
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 2.1120641231536865,
      "learning_rate": 6.010247992314055e-06,
      "loss": 0.3601,
      "num_input_tokens_seen": 10897824,
      "step": 16680
    },
    {
      "epoch": 9.837853773584905,
      "grad_norm": 3.0255351066589355,
      "learning_rate": 6.007728206062802e-06,
      "loss": 0.4326,
      "num_input_tokens_seen": 10900544,
      "step": 16685
    },
    {
      "epoch": 9.840801886792454,
      "grad_norm": 2.4413115978240967,
      "learning_rate": 6.005208153011163e-06,
      "loss": 0.3347,
      "num_input_tokens_seen": 10904576,
      "step": 16690
    },
    {
      "epoch": 9.84375,
      "grad_norm": 7.531404972076416,
      "learning_rate": 6.0026878338263335e-06,
      "loss": 0.4238,
      "num_input_tokens_seen": 10907424,
      "step": 16695
    },
    {
      "epoch": 9.846698113207546,
      "grad_norm": 2.256152629852295,
      "learning_rate": 6.000167249175579e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 10910848,
      "step": 16700
    },
    {
      "epoch": 9.849646226415095,
      "grad_norm": 5.964111328125,
      "learning_rate": 5.997646399726236e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 10913472,
      "step": 16705
    },
    {
      "epoch": 9.852594339622641,
      "grad_norm": 3.007678508758545,
      "learning_rate": 5.995125286145707e-06,
      "loss": 0.4038,
      "num_input_tokens_seen": 10916288,
      "step": 16710
    },
    {
      "epoch": 9.85554245283019,
      "grad_norm": 2.078181743621826,
      "learning_rate": 5.99260390910147e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 10919264,
      "step": 16715
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 2.9315991401672363,
      "learning_rate": 5.990082269261071e-06,
      "loss": 0.5835,
      "num_input_tokens_seen": 10922528,
      "step": 16720
    },
    {
      "epoch": 9.861438679245284,
      "grad_norm": 2.319767951965332,
      "learning_rate": 5.987560367292123e-06,
      "loss": 0.3941,
      "num_input_tokens_seen": 10925536,
      "step": 16725
    },
    {
      "epoch": 9.86438679245283,
      "grad_norm": 1.777725338935852,
      "learning_rate": 5.985038203862313e-06,
      "loss": 0.4076,
      "num_input_tokens_seen": 10928608,
      "step": 16730
    },
    {
      "epoch": 9.867334905660378,
      "grad_norm": 2.366572141647339,
      "learning_rate": 5.982515779639393e-06,
      "loss": 0.4887,
      "num_input_tokens_seen": 10931936,
      "step": 16735
    },
    {
      "epoch": 9.870283018867925,
      "grad_norm": 2.825746774673462,
      "learning_rate": 5.979993095291186e-06,
      "loss": 0.3371,
      "num_input_tokens_seen": 10935392,
      "step": 16740
    },
    {
      "epoch": 9.873231132075471,
      "grad_norm": 3.5361640453338623,
      "learning_rate": 5.977470151485582e-06,
      "loss": 0.4805,
      "num_input_tokens_seen": 10939488,
      "step": 16745
    },
    {
      "epoch": 9.87617924528302,
      "grad_norm": 3.1838505268096924,
      "learning_rate": 5.974946948890544e-06,
      "loss": 0.3753,
      "num_input_tokens_seen": 10943168,
      "step": 16750
    },
    {
      "epoch": 9.879127358490566,
      "grad_norm": 2.4686288833618164,
      "learning_rate": 5.9724234881740994e-06,
      "loss": 0.2782,
      "num_input_tokens_seen": 10946336,
      "step": 16755
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 2.1041171550750732,
      "learning_rate": 5.9698997700043445e-06,
      "loss": 0.3687,
      "num_input_tokens_seen": 10949984,
      "step": 16760
    },
    {
      "epoch": 9.88502358490566,
      "grad_norm": 3.844438314437866,
      "learning_rate": 5.9673757950494475e-06,
      "loss": 0.3952,
      "num_input_tokens_seen": 10952960,
      "step": 16765
    },
    {
      "epoch": 9.887971698113208,
      "grad_norm": 2.329035758972168,
      "learning_rate": 5.964851563977639e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 10955744,
      "step": 16770
    },
    {
      "epoch": 9.890919811320755,
      "grad_norm": 1.8626985549926758,
      "learning_rate": 5.962327077457219e-06,
      "loss": 0.342,
      "num_input_tokens_seen": 10958784,
      "step": 16775
    },
    {
      "epoch": 9.893867924528301,
      "grad_norm": 4.286154270172119,
      "learning_rate": 5.959802336156558e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 10961696,
      "step": 16780
    },
    {
      "epoch": 9.89681603773585,
      "grad_norm": 2.039259910583496,
      "learning_rate": 5.957277340744094e-06,
      "loss": 0.5501,
      "num_input_tokens_seen": 10964800,
      "step": 16785
    },
    {
      "epoch": 9.899764150943396,
      "grad_norm": 2.8672666549682617,
      "learning_rate": 5.954752091888326e-06,
      "loss": 0.4058,
      "num_input_tokens_seen": 10967936,
      "step": 16790
    },
    {
      "epoch": 9.902712264150944,
      "grad_norm": 3.277561664581299,
      "learning_rate": 5.952226590257829e-06,
      "loss": 0.5249,
      "num_input_tokens_seen": 10971104,
      "step": 16795
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 1.9575984477996826,
      "learning_rate": 5.949700836521239e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 10974592,
      "step": 16800
    },
    {
      "epoch": 9.908608490566039,
      "grad_norm": 1.7465177774429321,
      "learning_rate": 5.947174831347257e-06,
      "loss": 0.4029,
      "num_input_tokens_seen": 10977856,
      "step": 16805
    },
    {
      "epoch": 9.911556603773585,
      "grad_norm": 2.123689651489258,
      "learning_rate": 5.944648575404657e-06,
      "loss": 0.4739,
      "num_input_tokens_seen": 10980704,
      "step": 16810
    },
    {
      "epoch": 9.914504716981131,
      "grad_norm": 6.276925086975098,
      "learning_rate": 5.942122069362276e-06,
      "loss": 0.4076,
      "num_input_tokens_seen": 10983168,
      "step": 16815
    },
    {
      "epoch": 9.91745283018868,
      "grad_norm": 2.3685097694396973,
      "learning_rate": 5.939595313889016e-06,
      "loss": 0.3086,
      "num_input_tokens_seen": 10986976,
      "step": 16820
    },
    {
      "epoch": 9.920400943396226,
      "grad_norm": 2.092226266860962,
      "learning_rate": 5.937068309653848e-06,
      "loss": 0.36,
      "num_input_tokens_seen": 10989792,
      "step": 16825
    },
    {
      "epoch": 9.923349056603774,
      "grad_norm": 2.8103108406066895,
      "learning_rate": 5.934541057325807e-06,
      "loss": 0.439,
      "num_input_tokens_seen": 10992416,
      "step": 16830
    },
    {
      "epoch": 9.92629716981132,
      "grad_norm": 3.672915458679199,
      "learning_rate": 5.932013557573992e-06,
      "loss": 0.4833,
      "num_input_tokens_seen": 10996064,
      "step": 16835
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 1.9881372451782227,
      "learning_rate": 5.929485811067572e-06,
      "loss": 0.2981,
      "num_input_tokens_seen": 10999872,
      "step": 16840
    },
    {
      "epoch": 9.932193396226415,
      "grad_norm": 2.2922279834747314,
      "learning_rate": 5.926957818475778e-06,
      "loss": 0.5142,
      "num_input_tokens_seen": 11002656,
      "step": 16845
    },
    {
      "epoch": 9.935141509433961,
      "grad_norm": 1.5152888298034668,
      "learning_rate": 5.924429580467905e-06,
      "loss": 0.2648,
      "num_input_tokens_seen": 11006080,
      "step": 16850
    },
    {
      "epoch": 9.93808962264151,
      "grad_norm": 2.677478790283203,
      "learning_rate": 5.921901097713317e-06,
      "loss": 0.4115,
      "num_input_tokens_seen": 11009920,
      "step": 16855
    },
    {
      "epoch": 9.941037735849056,
      "grad_norm": 2.491976737976074,
      "learning_rate": 5.919372370881442e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 11014080,
      "step": 16860
    },
    {
      "epoch": 9.943985849056604,
      "grad_norm": 2.3317885398864746,
      "learning_rate": 5.916843400641768e-06,
      "loss": 0.4397,
      "num_input_tokens_seen": 11017920,
      "step": 16865
    },
    {
      "epoch": 9.94693396226415,
      "grad_norm": 3.8569834232330322,
      "learning_rate": 5.914314187663851e-06,
      "loss": 0.3484,
      "num_input_tokens_seen": 11020544,
      "step": 16870
    },
    {
      "epoch": 9.949882075471699,
      "grad_norm": 2.7212486267089844,
      "learning_rate": 5.911784732617314e-06,
      "loss": 0.4703,
      "num_input_tokens_seen": 11023360,
      "step": 16875
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 2.4578142166137695,
      "learning_rate": 5.90925503617184e-06,
      "loss": 0.4954,
      "num_input_tokens_seen": 11026112,
      "step": 16880
    },
    {
      "epoch": 9.955778301886792,
      "grad_norm": 2.0207157135009766,
      "learning_rate": 5.9067250989971745e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 11028864,
      "step": 16885
    },
    {
      "epoch": 9.95872641509434,
      "grad_norm": 2.4161627292633057,
      "learning_rate": 5.904194921763133e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 11031648,
      "step": 16890
    },
    {
      "epoch": 9.961674528301886,
      "grad_norm": 2.069093942642212,
      "learning_rate": 5.901664505139589e-06,
      "loss": 0.2938,
      "num_input_tokens_seen": 11034560,
      "step": 16895
    },
    {
      "epoch": 9.964622641509434,
      "grad_norm": 2.473487377166748,
      "learning_rate": 5.8991338497964814e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 11037888,
      "step": 16900
    },
    {
      "epoch": 9.96757075471698,
      "grad_norm": 2.2018752098083496,
      "learning_rate": 5.896602956403812e-06,
      "loss": 0.6396,
      "num_input_tokens_seen": 11042144,
      "step": 16905
    },
    {
      "epoch": 9.970518867924529,
      "grad_norm": 2.407668352127075,
      "learning_rate": 5.894071825631645e-06,
      "loss": 0.3088,
      "num_input_tokens_seen": 11045536,
      "step": 16910
    },
    {
      "epoch": 9.973466981132075,
      "grad_norm": 2.9631149768829346,
      "learning_rate": 5.891540458150109e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 11048416,
      "step": 16915
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 2.8481290340423584,
      "learning_rate": 5.889008854629395e-06,
      "loss": 0.5791,
      "num_input_tokens_seen": 11051616,
      "step": 16920
    },
    {
      "epoch": 9.97936320754717,
      "grad_norm": 2.373610734939575,
      "learning_rate": 5.886477015739754e-06,
      "loss": 0.4668,
      "num_input_tokens_seen": 11054048,
      "step": 16925
    },
    {
      "epoch": 9.982311320754716,
      "grad_norm": 2.3709583282470703,
      "learning_rate": 5.883944942151502e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 11056864,
      "step": 16930
    },
    {
      "epoch": 9.985259433962264,
      "grad_norm": 2.1882011890411377,
      "learning_rate": 5.88141263453502e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 11060192,
      "step": 16935
    },
    {
      "epoch": 9.98820754716981,
      "grad_norm": 2.7894251346588135,
      "learning_rate": 5.878880093560744e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 11063776,
      "step": 16940
    },
    {
      "epoch": 9.991155660377359,
      "grad_norm": 2.730959892272949,
      "learning_rate": 5.876347319899173e-06,
      "loss": 0.4676,
      "num_input_tokens_seen": 11067200,
      "step": 16945
    },
    {
      "epoch": 9.994103773584905,
      "grad_norm": 2.561547040939331,
      "learning_rate": 5.873814314220874e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 11072288,
      "step": 16950
    },
    {
      "epoch": 9.997051886792454,
      "grad_norm": 2.30741024017334,
      "learning_rate": 5.871281077196469e-06,
      "loss": 0.3449,
      "num_input_tokens_seen": 11074656,
      "step": 16955
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.574556350708008,
      "learning_rate": 5.868747609496643e-06,
      "loss": 0.356,
      "num_input_tokens_seen": 11077216,
      "step": 16960
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.5061902403831482,
      "eval_runtime": 19.1486,
      "eval_samples_per_second": 88.571,
      "eval_steps_per_second": 22.143,
      "num_input_tokens_seen": 11077216,
      "step": 16960
    },
    {
      "epoch": 10.002948113207546,
      "grad_norm": 2.112058162689209,
      "learning_rate": 5.8662139117921435e-06,
      "loss": 0.3574,
      "num_input_tokens_seen": 11080896,
      "step": 16965
    },
    {
      "epoch": 10.005896226415095,
      "grad_norm": 2.636873483657837,
      "learning_rate": 5.8636799847537785e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 11083776,
      "step": 16970
    },
    {
      "epoch": 10.008844339622641,
      "grad_norm": 2.790105104446411,
      "learning_rate": 5.861145829052415e-06,
      "loss": 0.3884,
      "num_input_tokens_seen": 11086560,
      "step": 16975
    },
    {
      "epoch": 10.01179245283019,
      "grad_norm": 2.0542263984680176,
      "learning_rate": 5.858611445358982e-06,
      "loss": 0.2895,
      "num_input_tokens_seen": 11090656,
      "step": 16980
    },
    {
      "epoch": 10.014740566037736,
      "grad_norm": 2.6753740310668945,
      "learning_rate": 5.856076834344468e-06,
      "loss": 0.302,
      "num_input_tokens_seen": 11094560,
      "step": 16985
    },
    {
      "epoch": 10.017688679245284,
      "grad_norm": 5.089707374572754,
      "learning_rate": 5.853541996679924e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 11097856,
      "step": 16990
    },
    {
      "epoch": 10.02063679245283,
      "grad_norm": 2.6079349517822266,
      "learning_rate": 5.851006933036456e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 11100512,
      "step": 16995
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 2.120274782180786,
      "learning_rate": 5.848471644085236e-06,
      "loss": 0.4089,
      "num_input_tokens_seen": 11104128,
      "step": 17000
    },
    {
      "epoch": 10.026533018867925,
      "grad_norm": 3.739114999771118,
      "learning_rate": 5.84593613049749e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 11106976,
      "step": 17005
    },
    {
      "epoch": 10.029481132075471,
      "grad_norm": 1.5719140768051147,
      "learning_rate": 5.843400392944509e-06,
      "loss": 0.4002,
      "num_input_tokens_seen": 11109952,
      "step": 17010
    },
    {
      "epoch": 10.03242924528302,
      "grad_norm": 1.5226027965545654,
      "learning_rate": 5.840864432097639e-06,
      "loss": 0.2778,
      "num_input_tokens_seen": 11113344,
      "step": 17015
    },
    {
      "epoch": 10.035377358490566,
      "grad_norm": 2.902195930480957,
      "learning_rate": 5.838328248628286e-06,
      "loss": 0.2949,
      "num_input_tokens_seen": 11115904,
      "step": 17020
    },
    {
      "epoch": 10.038325471698114,
      "grad_norm": 2.746558666229248,
      "learning_rate": 5.835791843207916e-06,
      "loss": 0.3781,
      "num_input_tokens_seen": 11119008,
      "step": 17025
    },
    {
      "epoch": 10.04127358490566,
      "grad_norm": 3.880944013595581,
      "learning_rate": 5.833255216508056e-06,
      "loss": 0.4532,
      "num_input_tokens_seen": 11122496,
      "step": 17030
    },
    {
      "epoch": 10.044221698113208,
      "grad_norm": 2.3490216732025146,
      "learning_rate": 5.830718369200284e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 11125088,
      "step": 17035
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 3.318497896194458,
      "learning_rate": 5.828181301956244e-06,
      "loss": 0.2967,
      "num_input_tokens_seen": 11128832,
      "step": 17040
    },
    {
      "epoch": 10.050117924528301,
      "grad_norm": 3.024944543838501,
      "learning_rate": 5.825644015447636e-06,
      "loss": 0.2963,
      "num_input_tokens_seen": 11131808,
      "step": 17045
    },
    {
      "epoch": 10.05306603773585,
      "grad_norm": 1.179841160774231,
      "learning_rate": 5.823106510346216e-06,
      "loss": 0.3159,
      "num_input_tokens_seen": 11135776,
      "step": 17050
    },
    {
      "epoch": 10.056014150943396,
      "grad_norm": 2.581219434738159,
      "learning_rate": 5.820568787323798e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 11139328,
      "step": 17055
    },
    {
      "epoch": 10.058962264150944,
      "grad_norm": 2.7380270957946777,
      "learning_rate": 5.818030847052258e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 11142464,
      "step": 17060
    },
    {
      "epoch": 10.06191037735849,
      "grad_norm": 2.3669400215148926,
      "learning_rate": 5.815492690203528e-06,
      "loss": 0.3693,
      "num_input_tokens_seen": 11144736,
      "step": 17065
    },
    {
      "epoch": 10.064858490566039,
      "grad_norm": 2.2337875366210938,
      "learning_rate": 5.812954317449591e-06,
      "loss": 0.3174,
      "num_input_tokens_seen": 11148448,
      "step": 17070
    },
    {
      "epoch": 10.067806603773585,
      "grad_norm": 3.1516125202178955,
      "learning_rate": 5.810415729462496e-06,
      "loss": 0.4081,
      "num_input_tokens_seen": 11152000,
      "step": 17075
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 3.072516918182373,
      "learning_rate": 5.807876926914344e-06,
      "loss": 0.2589,
      "num_input_tokens_seen": 11154592,
      "step": 17080
    },
    {
      "epoch": 10.07370283018868,
      "grad_norm": 2.4809658527374268,
      "learning_rate": 5.8053379104772935e-06,
      "loss": 0.4032,
      "num_input_tokens_seen": 11157024,
      "step": 17085
    },
    {
      "epoch": 10.076650943396226,
      "grad_norm": 2.0432913303375244,
      "learning_rate": 5.802798680823562e-06,
      "loss": 0.4259,
      "num_input_tokens_seen": 11159744,
      "step": 17090
    },
    {
      "epoch": 10.079599056603774,
      "grad_norm": 3.903622627258301,
      "learning_rate": 5.8002592386254186e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 11163136,
      "step": 17095
    },
    {
      "epoch": 10.08254716981132,
      "grad_norm": 2.140369176864624,
      "learning_rate": 5.797719584555192e-06,
      "loss": 0.3029,
      "num_input_tokens_seen": 11166432,
      "step": 17100
    },
    {
      "epoch": 10.085495283018869,
      "grad_norm": 3.921924352645874,
      "learning_rate": 5.795179719285269e-06,
      "loss": 0.4102,
      "num_input_tokens_seen": 11169312,
      "step": 17105
    },
    {
      "epoch": 10.088443396226415,
      "grad_norm": 3.969426393508911,
      "learning_rate": 5.792639643488086e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 11172768,
      "step": 17110
    },
    {
      "epoch": 10.091391509433961,
      "grad_norm": 2.207195281982422,
      "learning_rate": 5.7900993578361434e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 11176064,
      "step": 17115
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 2.6136326789855957,
      "learning_rate": 5.7875588630019895e-06,
      "loss": 0.3111,
      "num_input_tokens_seen": 11179680,
      "step": 17120
    },
    {
      "epoch": 10.097287735849056,
      "grad_norm": 1.9569365978240967,
      "learning_rate": 5.7850181596582335e-06,
      "loss": 0.3701,
      "num_input_tokens_seen": 11183840,
      "step": 17125
    },
    {
      "epoch": 10.100235849056604,
      "grad_norm": 2.7761807441711426,
      "learning_rate": 5.782477248477535e-06,
      "loss": 0.3907,
      "num_input_tokens_seen": 11186368,
      "step": 17130
    },
    {
      "epoch": 10.10318396226415,
      "grad_norm": 2.2603743076324463,
      "learning_rate": 5.779936130132614e-06,
      "loss": 0.3711,
      "num_input_tokens_seen": 11190048,
      "step": 17135
    },
    {
      "epoch": 10.106132075471699,
      "grad_norm": 3.287369966506958,
      "learning_rate": 5.777394805296242e-06,
      "loss": 0.5555,
      "num_input_tokens_seen": 11193632,
      "step": 17140
    },
    {
      "epoch": 10.109080188679245,
      "grad_norm": 2.975034236907959,
      "learning_rate": 5.774853274641243e-06,
      "loss": 0.2816,
      "num_input_tokens_seen": 11196128,
      "step": 17145
    },
    {
      "epoch": 10.112028301886792,
      "grad_norm": 4.992257595062256,
      "learning_rate": 5.772311538840501e-06,
      "loss": 0.4711,
      "num_input_tokens_seen": 11199680,
      "step": 17150
    },
    {
      "epoch": 10.11497641509434,
      "grad_norm": 3.449916124343872,
      "learning_rate": 5.76976959856695e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 11202048,
      "step": 17155
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 6.460193157196045,
      "learning_rate": 5.767227454493582e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 11204896,
      "step": 17160
    },
    {
      "epoch": 10.120872641509434,
      "grad_norm": 6.626892566680908,
      "learning_rate": 5.764685107293436e-06,
      "loss": 0.3886,
      "num_input_tokens_seen": 11207808,
      "step": 17165
    },
    {
      "epoch": 10.12382075471698,
      "grad_norm": 2.705941677093506,
      "learning_rate": 5.762142557639614e-06,
      "loss": 0.4677,
      "num_input_tokens_seen": 11211456,
      "step": 17170
    },
    {
      "epoch": 10.126768867924529,
      "grad_norm": 2.9977800846099854,
      "learning_rate": 5.759599806205266e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 11214528,
      "step": 17175
    },
    {
      "epoch": 10.129716981132075,
      "grad_norm": 1.5917412042617798,
      "learning_rate": 5.757056853663594e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 11217984,
      "step": 17180
    },
    {
      "epoch": 10.132665094339623,
      "grad_norm": 2.3645260334014893,
      "learning_rate": 5.754513700687858e-06,
      "loss": 0.336,
      "num_input_tokens_seen": 11220992,
      "step": 17185
    },
    {
      "epoch": 10.13561320754717,
      "grad_norm": 1.8643184900283813,
      "learning_rate": 5.751970347951369e-06,
      "loss": 0.4452,
      "num_input_tokens_seen": 11224160,
      "step": 17190
    },
    {
      "epoch": 10.138561320754716,
      "grad_norm": 2.853123188018799,
      "learning_rate": 5.7494267961274875e-06,
      "loss": 0.3822,
      "num_input_tokens_seen": 11227104,
      "step": 17195
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 2.9463248252868652,
      "learning_rate": 5.746883045889633e-06,
      "loss": 0.4974,
      "num_input_tokens_seen": 11229824,
      "step": 17200
    },
    {
      "epoch": 10.14445754716981,
      "grad_norm": 2.569958448410034,
      "learning_rate": 5.744339097911272e-06,
      "loss": 0.3804,
      "num_input_tokens_seen": 11233024,
      "step": 17205
    },
    {
      "epoch": 10.147405660377359,
      "grad_norm": 1.30960214138031,
      "learning_rate": 5.741794952865928e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 11235872,
      "step": 17210
    },
    {
      "epoch": 10.150353773584905,
      "grad_norm": 2.5726544857025146,
      "learning_rate": 5.739250611427173e-06,
      "loss": 0.3394,
      "num_input_tokens_seen": 11239520,
      "step": 17215
    },
    {
      "epoch": 10.153301886792454,
      "grad_norm": 3.0558876991271973,
      "learning_rate": 5.736706074268633e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 11243072,
      "step": 17220
    },
    {
      "epoch": 10.15625,
      "grad_norm": 2.491466522216797,
      "learning_rate": 5.734161342063984e-06,
      "loss": 0.3868,
      "num_input_tokens_seen": 11247552,
      "step": 17225
    },
    {
      "epoch": 10.159198113207546,
      "grad_norm": 3.145909070968628,
      "learning_rate": 5.731616415486957e-06,
      "loss": 0.4932,
      "num_input_tokens_seen": 11252000,
      "step": 17230
    },
    {
      "epoch": 10.162146226415095,
      "grad_norm": 2.2103474140167236,
      "learning_rate": 5.72907129521133e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 11254720,
      "step": 17235
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 2.1223642826080322,
      "learning_rate": 5.726525981910935e-06,
      "loss": 0.3929,
      "num_input_tokens_seen": 11258080,
      "step": 17240
    },
    {
      "epoch": 10.16804245283019,
      "grad_norm": 3.004373788833618,
      "learning_rate": 5.723980476259658e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 11261344,
      "step": 17245
    },
    {
      "epoch": 10.170990566037736,
      "grad_norm": 1.5305474996566772,
      "learning_rate": 5.721434778931426e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 11266112,
      "step": 17250
    },
    {
      "epoch": 10.173938679245284,
      "grad_norm": 2.7990121841430664,
      "learning_rate": 5.7188888906002284e-06,
      "loss": 0.4261,
      "num_input_tokens_seen": 11269440,
      "step": 17255
    },
    {
      "epoch": 10.17688679245283,
      "grad_norm": 1.9553160667419434,
      "learning_rate": 5.716342811940098e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 11272288,
      "step": 17260
    },
    {
      "epoch": 10.179834905660377,
      "grad_norm": 2.889065980911255,
      "learning_rate": 5.713796543625123e-06,
      "loss": 0.3277,
      "num_input_tokens_seen": 11275872,
      "step": 17265
    },
    {
      "epoch": 10.182783018867925,
      "grad_norm": 2.202157974243164,
      "learning_rate": 5.711250086329435e-06,
      "loss": 0.4769,
      "num_input_tokens_seen": 11279424,
      "step": 17270
    },
    {
      "epoch": 10.185731132075471,
      "grad_norm": 1.275183081626892,
      "learning_rate": 5.7087034407272225e-06,
      "loss": 0.3234,
      "num_input_tokens_seen": 11282880,
      "step": 17275
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 2.6807029247283936,
      "learning_rate": 5.70615660749272e-06,
      "loss": 0.2945,
      "num_input_tokens_seen": 11285600,
      "step": 17280
    },
    {
      "epoch": 10.191627358490566,
      "grad_norm": 6.062461853027344,
      "learning_rate": 5.7036095873002106e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 11288992,
      "step": 17285
    },
    {
      "epoch": 10.194575471698114,
      "grad_norm": 1.6470695734024048,
      "learning_rate": 5.701062380824032e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 11292704,
      "step": 17290
    },
    {
      "epoch": 10.19752358490566,
      "grad_norm": 3.893112897872925,
      "learning_rate": 5.698514988738566e-06,
      "loss": 0.3629,
      "num_input_tokens_seen": 11295520,
      "step": 17295
    },
    {
      "epoch": 10.200471698113208,
      "grad_norm": 3.8282268047332764,
      "learning_rate": 5.6959674117182465e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 11298080,
      "step": 17300
    },
    {
      "epoch": 10.203419811320755,
      "grad_norm": 2.9139583110809326,
      "learning_rate": 5.693419650437554e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 11300896,
      "step": 17305
    },
    {
      "epoch": 10.206367924528301,
      "grad_norm": 2.1333422660827637,
      "learning_rate": 5.690871705571022e-06,
      "loss": 0.3726,
      "num_input_tokens_seen": 11305856,
      "step": 17310
    },
    {
      "epoch": 10.20931603773585,
      "grad_norm": 2.225630521774292,
      "learning_rate": 5.688323577793229e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 11309568,
      "step": 17315
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 2.9951884746551514,
      "learning_rate": 5.685775267778801e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 11313408,
      "step": 17320
    },
    {
      "epoch": 10.215212264150944,
      "grad_norm": 1.6047955751419067,
      "learning_rate": 5.683226776202416e-06,
      "loss": 0.3031,
      "num_input_tokens_seen": 11316224,
      "step": 17325
    },
    {
      "epoch": 10.21816037735849,
      "grad_norm": 2.8794357776641846,
      "learning_rate": 5.680678103738798e-06,
      "loss": 0.3546,
      "num_input_tokens_seen": 11321184,
      "step": 17330
    },
    {
      "epoch": 10.221108490566039,
      "grad_norm": 2.953395366668701,
      "learning_rate": 5.678129251062717e-06,
      "loss": 0.3986,
      "num_input_tokens_seen": 11323808,
      "step": 17335
    },
    {
      "epoch": 10.224056603773585,
      "grad_norm": 2.47658371925354,
      "learning_rate": 5.675580218848995e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 11326752,
      "step": 17340
    },
    {
      "epoch": 10.227004716981131,
      "grad_norm": 4.680934906005859,
      "learning_rate": 5.673031007772498e-06,
      "loss": 0.3892,
      "num_input_tokens_seen": 11331104,
      "step": 17345
    },
    {
      "epoch": 10.22995283018868,
      "grad_norm": 1.577209234237671,
      "learning_rate": 5.670481618508141e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 11333920,
      "step": 17350
    },
    {
      "epoch": 10.232900943396226,
      "grad_norm": 3.425941228866577,
      "learning_rate": 5.667932051730887e-06,
      "loss": 0.4227,
      "num_input_tokens_seen": 11337024,
      "step": 17355
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 2.968404769897461,
      "learning_rate": 5.6653823081157434e-06,
      "loss": 0.4326,
      "num_input_tokens_seen": 11340128,
      "step": 17360
    },
    {
      "epoch": 10.23879716981132,
      "grad_norm": 2.5554916858673096,
      "learning_rate": 5.662832388337766e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 11342912,
      "step": 17365
    },
    {
      "epoch": 10.241745283018869,
      "grad_norm": 2.1316909790039062,
      "learning_rate": 5.660282293072057e-06,
      "loss": 0.4375,
      "num_input_tokens_seen": 11345792,
      "step": 17370
    },
    {
      "epoch": 10.244693396226415,
      "grad_norm": 2.1705164909362793,
      "learning_rate": 5.657732022993765e-06,
      "loss": 0.3082,
      "num_input_tokens_seen": 11348672,
      "step": 17375
    },
    {
      "epoch": 10.247641509433961,
      "grad_norm": 1.7026745080947876,
      "learning_rate": 5.655181578778085e-06,
      "loss": 0.3155,
      "num_input_tokens_seen": 11351264,
      "step": 17380
    },
    {
      "epoch": 10.25058962264151,
      "grad_norm": 3.691415548324585,
      "learning_rate": 5.65263096110026e-06,
      "loss": 0.4029,
      "num_input_tokens_seen": 11354016,
      "step": 17385
    },
    {
      "epoch": 10.253537735849056,
      "grad_norm": 2.0694663524627686,
      "learning_rate": 5.650080170635573e-06,
      "loss": 0.2934,
      "num_input_tokens_seen": 11356416,
      "step": 17390
    },
    {
      "epoch": 10.256485849056604,
      "grad_norm": 1.7829653024673462,
      "learning_rate": 5.647529208059359e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 11359392,
      "step": 17395
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 2.7029449939727783,
      "learning_rate": 5.6449780740469985e-06,
      "loss": 0.3419,
      "num_input_tokens_seen": 11362560,
      "step": 17400
    },
    {
      "epoch": 10.262382075471699,
      "grad_norm": 2.124021291732788,
      "learning_rate": 5.642426769273912e-06,
      "loss": 0.3306,
      "num_input_tokens_seen": 11365248,
      "step": 17405
    },
    {
      "epoch": 10.265330188679245,
      "grad_norm": 2.851135492324829,
      "learning_rate": 5.63987529441557e-06,
      "loss": 0.4161,
      "num_input_tokens_seen": 11368736,
      "step": 17410
    },
    {
      "epoch": 10.268278301886792,
      "grad_norm": 2.077836513519287,
      "learning_rate": 5.637323650147487e-06,
      "loss": 0.4255,
      "num_input_tokens_seen": 11373376,
      "step": 17415
    },
    {
      "epoch": 10.27122641509434,
      "grad_norm": 1.9708411693572998,
      "learning_rate": 5.63477183714522e-06,
      "loss": 0.3329,
      "num_input_tokens_seen": 11377152,
      "step": 17420
    },
    {
      "epoch": 10.274174528301886,
      "grad_norm": 1.4919012784957886,
      "learning_rate": 5.632219856084373e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 11380544,
      "step": 17425
    },
    {
      "epoch": 10.277122641509434,
      "grad_norm": 4.491013526916504,
      "learning_rate": 5.6296677076405944e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 11383296,
      "step": 17430
    },
    {
      "epoch": 10.28007075471698,
      "grad_norm": 1.860515832901001,
      "learning_rate": 5.627115392489578e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 11386496,
      "step": 17435
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 3.0695648193359375,
      "learning_rate": 5.624562911307058e-06,
      "loss": 0.2833,
      "num_input_tokens_seen": 11390272,
      "step": 17440
    },
    {
      "epoch": 10.285966981132075,
      "grad_norm": 2.617431402206421,
      "learning_rate": 5.622010264768813e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 11394432,
      "step": 17445
    },
    {
      "epoch": 10.288915094339623,
      "grad_norm": 2.888895273208618,
      "learning_rate": 5.619457453550673e-06,
      "loss": 0.3954,
      "num_input_tokens_seen": 11397056,
      "step": 17450
    },
    {
      "epoch": 10.29186320754717,
      "grad_norm": 2.528139352798462,
      "learning_rate": 5.616904478328502e-06,
      "loss": 0.2974,
      "num_input_tokens_seen": 11400768,
      "step": 17455
    },
    {
      "epoch": 10.294811320754716,
      "grad_norm": 7.799339294433594,
      "learning_rate": 5.6143513397782104e-06,
      "loss": 0.3821,
      "num_input_tokens_seen": 11403712,
      "step": 17460
    },
    {
      "epoch": 10.297759433962264,
      "grad_norm": 1.9704792499542236,
      "learning_rate": 5.611798038575755e-06,
      "loss": 0.3394,
      "num_input_tokens_seen": 11406944,
      "step": 17465
    },
    {
      "epoch": 10.30070754716981,
      "grad_norm": 4.176156520843506,
      "learning_rate": 5.609244575397131e-06,
      "loss": 0.439,
      "num_input_tokens_seen": 11409888,
      "step": 17470
    },
    {
      "epoch": 10.303655660377359,
      "grad_norm": 3.3911826610565186,
      "learning_rate": 5.606690950918381e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 11412896,
      "step": 17475
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 2.7131316661834717,
      "learning_rate": 5.604137165815586e-06,
      "loss": 0.4509,
      "num_input_tokens_seen": 11415712,
      "step": 17480
    },
    {
      "epoch": 10.309551886792454,
      "grad_norm": 2.284796714782715,
      "learning_rate": 5.601583220764874e-06,
      "loss": 0.2941,
      "num_input_tokens_seen": 11418112,
      "step": 17485
    },
    {
      "epoch": 10.3125,
      "grad_norm": 2.916926383972168,
      "learning_rate": 5.599029116442409e-06,
      "loss": 0.3057,
      "num_input_tokens_seen": 11423616,
      "step": 17490
    },
    {
      "epoch": 10.315448113207546,
      "grad_norm": 1.6925725936889648,
      "learning_rate": 5.596474853524406e-06,
      "loss": 0.4496,
      "num_input_tokens_seen": 11427936,
      "step": 17495
    },
    {
      "epoch": 10.318396226415095,
      "grad_norm": 2.4235377311706543,
      "learning_rate": 5.593920432687115e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 11431008,
      "step": 17500
    },
    {
      "epoch": 10.321344339622641,
      "grad_norm": 2.645493984222412,
      "learning_rate": 5.591365854606829e-06,
      "loss": 0.3765,
      "num_input_tokens_seen": 11434656,
      "step": 17505
    },
    {
      "epoch": 10.32429245283019,
      "grad_norm": 2.0462758541107178,
      "learning_rate": 5.588811119959885e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 11437984,
      "step": 17510
    },
    {
      "epoch": 10.327240566037736,
      "grad_norm": 2.2767951488494873,
      "learning_rate": 5.58625622942266e-06,
      "loss": 0.3743,
      "num_input_tokens_seen": 11441696,
      "step": 17515
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 6.34493350982666,
      "learning_rate": 5.58370118367157e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 11446144,
      "step": 17520
    },
    {
      "epoch": 10.33313679245283,
      "grad_norm": 2.2103374004364014,
      "learning_rate": 5.581145983383077e-06,
      "loss": 0.3197,
      "num_input_tokens_seen": 11449152,
      "step": 17525
    },
    {
      "epoch": 10.336084905660377,
      "grad_norm": 1.6146049499511719,
      "learning_rate": 5.57859062923368e-06,
      "loss": 0.3039,
      "num_input_tokens_seen": 11452288,
      "step": 17530
    },
    {
      "epoch": 10.339033018867925,
      "grad_norm": 1.6563220024108887,
      "learning_rate": 5.5760351218999194e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 11455968,
      "step": 17535
    },
    {
      "epoch": 10.341981132075471,
      "grad_norm": 3.2218542098999023,
      "learning_rate": 5.573479462058379e-06,
      "loss": 0.3066,
      "num_input_tokens_seen": 11459520,
      "step": 17540
    },
    {
      "epoch": 10.34492924528302,
      "grad_norm": 1.95578134059906,
      "learning_rate": 5.570923650385679e-06,
      "loss": 0.3152,
      "num_input_tokens_seen": 11463008,
      "step": 17545
    },
    {
      "epoch": 10.347877358490566,
      "grad_norm": 5.442994117736816,
      "learning_rate": 5.568367687558481e-06,
      "loss": 0.4677,
      "num_input_tokens_seen": 11465504,
      "step": 17550
    },
    {
      "epoch": 10.350825471698114,
      "grad_norm": 3.237537145614624,
      "learning_rate": 5.56581157425349e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 11468960,
      "step": 17555
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 3.6067051887512207,
      "learning_rate": 5.563255311147446e-06,
      "loss": 0.4129,
      "num_input_tokens_seen": 11471552,
      "step": 17560
    },
    {
      "epoch": 10.356721698113208,
      "grad_norm": 2.684098482131958,
      "learning_rate": 5.560698898917129e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 11475680,
      "step": 17565
    },
    {
      "epoch": 10.359669811320755,
      "grad_norm": 3.790410280227661,
      "learning_rate": 5.558142338239365e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 11478048,
      "step": 17570
    },
    {
      "epoch": 10.362617924528301,
      "grad_norm": 1.210429072380066,
      "learning_rate": 5.55558562979101e-06,
      "loss": 0.4038,
      "num_input_tokens_seen": 11482272,
      "step": 17575
    },
    {
      "epoch": 10.36556603773585,
      "grad_norm": 2.6478431224823,
      "learning_rate": 5.553028774248964e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 11485952,
      "step": 17580
    },
    {
      "epoch": 10.368514150943396,
      "grad_norm": 1.9827935695648193,
      "learning_rate": 5.5504717722901665e-06,
      "loss": 0.3876,
      "num_input_tokens_seen": 11489312,
      "step": 17585
    },
    {
      "epoch": 10.371462264150944,
      "grad_norm": 1.5696144104003906,
      "learning_rate": 5.547914624591597e-06,
      "loss": 0.4528,
      "num_input_tokens_seen": 11493248,
      "step": 17590
    },
    {
      "epoch": 10.37441037735849,
      "grad_norm": 2.735292911529541,
      "learning_rate": 5.545357331830269e-06,
      "loss": 0.3015,
      "num_input_tokens_seen": 11496192,
      "step": 17595
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 1.7629132270812988,
      "learning_rate": 5.542799894683235e-06,
      "loss": 0.339,
      "num_input_tokens_seen": 11500800,
      "step": 17600
    },
    {
      "epoch": 10.380306603773585,
      "grad_norm": 2.619920015335083,
      "learning_rate": 5.540242313827591e-06,
      "loss": 0.3299,
      "num_input_tokens_seen": 11503936,
      "step": 17605
    },
    {
      "epoch": 10.383254716981131,
      "grad_norm": 3.4268085956573486,
      "learning_rate": 5.537684589940466e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 11506848,
      "step": 17610
    },
    {
      "epoch": 10.38620283018868,
      "grad_norm": 1.8338828086853027,
      "learning_rate": 5.535126723699029e-06,
      "loss": 0.3466,
      "num_input_tokens_seen": 11510144,
      "step": 17615
    },
    {
      "epoch": 10.389150943396226,
      "grad_norm": 2.513178586959839,
      "learning_rate": 5.532568715780485e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 11513920,
      "step": 17620
    },
    {
      "epoch": 10.392099056603774,
      "grad_norm": 2.357240915298462,
      "learning_rate": 5.530010566862077e-06,
      "loss": 0.3706,
      "num_input_tokens_seen": 11517344,
      "step": 17625
    },
    {
      "epoch": 10.39504716981132,
      "grad_norm": 3.289005994796753,
      "learning_rate": 5.527452277621089e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 11520032,
      "step": 17630
    },
    {
      "epoch": 10.397995283018869,
      "grad_norm": 3.377303123474121,
      "learning_rate": 5.524893848734837e-06,
      "loss": 0.3725,
      "num_input_tokens_seen": 11522976,
      "step": 17635
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 2.51676082611084,
      "learning_rate": 5.522335280880676e-06,
      "loss": 0.4172,
      "num_input_tokens_seen": 11527680,
      "step": 17640
    },
    {
      "epoch": 10.403891509433961,
      "grad_norm": 1.8994343280792236,
      "learning_rate": 5.519776574735999e-06,
      "loss": 0.4275,
      "num_input_tokens_seen": 11530688,
      "step": 17645
    },
    {
      "epoch": 10.40683962264151,
      "grad_norm": 1.898848295211792,
      "learning_rate": 5.5172177309782325e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 11534848,
      "step": 17650
    },
    {
      "epoch": 10.409787735849056,
      "grad_norm": 2.26517391204834,
      "learning_rate": 5.514658750284844e-06,
      "loss": 0.479,
      "num_input_tokens_seen": 11538176,
      "step": 17655
    },
    {
      "epoch": 10.412735849056604,
      "grad_norm": 2.0276670455932617,
      "learning_rate": 5.512099633333332e-06,
      "loss": 0.4913,
      "num_input_tokens_seen": 11540800,
      "step": 17660
    },
    {
      "epoch": 10.41568396226415,
      "grad_norm": 2.0876083374023438,
      "learning_rate": 5.509540380801236e-06,
      "loss": 0.399,
      "num_input_tokens_seen": 11544960,
      "step": 17665
    },
    {
      "epoch": 10.418632075471699,
      "grad_norm": 2.947951078414917,
      "learning_rate": 5.506980993366129e-06,
      "loss": 0.4711,
      "num_input_tokens_seen": 11548608,
      "step": 17670
    },
    {
      "epoch": 10.421580188679245,
      "grad_norm": 5.269805431365967,
      "learning_rate": 5.504421471705616e-06,
      "loss": 0.3158,
      "num_input_tokens_seen": 11552000,
      "step": 17675
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 3.058739423751831,
      "learning_rate": 5.501861816497346e-06,
      "loss": 0.4708,
      "num_input_tokens_seen": 11554688,
      "step": 17680
    },
    {
      "epoch": 10.42747641509434,
      "grad_norm": 3.068485975265503,
      "learning_rate": 5.499302028418998e-06,
      "loss": 0.4061,
      "num_input_tokens_seen": 11557344,
      "step": 17685
    },
    {
      "epoch": 10.430424528301886,
      "grad_norm": 3.190065622329712,
      "learning_rate": 5.496742108148285e-06,
      "loss": 0.4366,
      "num_input_tokens_seen": 11560032,
      "step": 17690
    },
    {
      "epoch": 10.433372641509434,
      "grad_norm": 2.980930805206299,
      "learning_rate": 5.494182056362959e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 11563424,
      "step": 17695
    },
    {
      "epoch": 10.43632075471698,
      "grad_norm": 1.9980535507202148,
      "learning_rate": 5.491621873740804e-06,
      "loss": 0.3177,
      "num_input_tokens_seen": 11566112,
      "step": 17700
    },
    {
      "epoch": 10.439268867924529,
      "grad_norm": 3.9577462673187256,
      "learning_rate": 5.4890615609596375e-06,
      "loss": 0.4035,
      "num_input_tokens_seen": 11568960,
      "step": 17705
    },
    {
      "epoch": 10.442216981132075,
      "grad_norm": 2.0909650325775146,
      "learning_rate": 5.486501118697317e-06,
      "loss": 0.4087,
      "num_input_tokens_seen": 11572864,
      "step": 17710
    },
    {
      "epoch": 10.445165094339623,
      "grad_norm": 2.294403553009033,
      "learning_rate": 5.483940547631727e-06,
      "loss": 0.5016,
      "num_input_tokens_seen": 11577216,
      "step": 17715
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 3.668940544128418,
      "learning_rate": 5.481379848440792e-06,
      "loss": 0.3044,
      "num_input_tokens_seen": 11580192,
      "step": 17720
    },
    {
      "epoch": 10.451061320754716,
      "grad_norm": 1.6960254907608032,
      "learning_rate": 5.478819021802468e-06,
      "loss": 0.2856,
      "num_input_tokens_seen": 11583520,
      "step": 17725
    },
    {
      "epoch": 10.454009433962264,
      "grad_norm": 5.286823272705078,
      "learning_rate": 5.476258068394743e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 11586336,
      "step": 17730
    },
    {
      "epoch": 10.45695754716981,
      "grad_norm": 1.801348090171814,
      "learning_rate": 5.473696988895644e-06,
      "loss": 0.3059,
      "num_input_tokens_seen": 11589568,
      "step": 17735
    },
    {
      "epoch": 10.459905660377359,
      "grad_norm": 1.7994965314865112,
      "learning_rate": 5.471135783983224e-06,
      "loss": 0.3091,
      "num_input_tokens_seen": 11592832,
      "step": 17740
    },
    {
      "epoch": 10.462853773584905,
      "grad_norm": 3.709779739379883,
      "learning_rate": 5.4685744543355745e-06,
      "loss": 0.4638,
      "num_input_tokens_seen": 11595648,
      "step": 17745
    },
    {
      "epoch": 10.465801886792454,
      "grad_norm": 1.5172752141952515,
      "learning_rate": 5.466013000630819e-06,
      "loss": 0.3601,
      "num_input_tokens_seen": 11598976,
      "step": 17750
    },
    {
      "epoch": 10.46875,
      "grad_norm": 2.1951160430908203,
      "learning_rate": 5.463451423547114e-06,
      "loss": 0.436,
      "num_input_tokens_seen": 11603360,
      "step": 17755
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 2.853409767150879,
      "learning_rate": 5.460889723762647e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 11605632,
      "step": 17760
    },
    {
      "epoch": 10.474646226415095,
      "grad_norm": 1.7093557119369507,
      "learning_rate": 5.458327901955639e-06,
      "loss": 0.3566,
      "num_input_tokens_seen": 11608128,
      "step": 17765
    },
    {
      "epoch": 10.477594339622641,
      "grad_norm": 3.0885679721832275,
      "learning_rate": 5.455765958804344e-06,
      "loss": 0.2541,
      "num_input_tokens_seen": 11611072,
      "step": 17770
    },
    {
      "epoch": 10.48054245283019,
      "grad_norm": 4.776382923126221,
      "learning_rate": 5.4532038949870455e-06,
      "loss": 0.4862,
      "num_input_tokens_seen": 11614080,
      "step": 17775
    },
    {
      "epoch": 10.483490566037736,
      "grad_norm": 1.4229613542556763,
      "learning_rate": 5.450641711182066e-06,
      "loss": 0.4691,
      "num_input_tokens_seen": 11616544,
      "step": 17780
    },
    {
      "epoch": 10.486438679245284,
      "grad_norm": 1.5716538429260254,
      "learning_rate": 5.448079408067748e-06,
      "loss": 0.4008,
      "num_input_tokens_seen": 11620352,
      "step": 17785
    },
    {
      "epoch": 10.48938679245283,
      "grad_norm": 2.6080310344696045,
      "learning_rate": 5.4455169863224775e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 11623168,
      "step": 17790
    },
    {
      "epoch": 10.492334905660377,
      "grad_norm": 3.276071071624756,
      "learning_rate": 5.442954446624664e-06,
      "loss": 0.4973,
      "num_input_tokens_seen": 11626112,
      "step": 17795
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 3.4211113452911377,
      "learning_rate": 5.440391789652752e-06,
      "loss": 0.3493,
      "num_input_tokens_seen": 11629632,
      "step": 17800
    },
    {
      "epoch": 10.498231132075471,
      "grad_norm": 2.7757272720336914,
      "learning_rate": 5.437829016085216e-06,
      "loss": 0.3499,
      "num_input_tokens_seen": 11632480,
      "step": 17805
    },
    {
      "epoch": 10.50117924528302,
      "grad_norm": 2.525200366973877,
      "learning_rate": 5.435266126600561e-06,
      "loss": 0.3999,
      "num_input_tokens_seen": 11635584,
      "step": 17810
    },
    {
      "epoch": 10.504127358490566,
      "grad_norm": 2.159573554992676,
      "learning_rate": 5.4327031218773215e-06,
      "loss": 0.3443,
      "num_input_tokens_seen": 11638432,
      "step": 17815
    },
    {
      "epoch": 10.507075471698114,
      "grad_norm": 4.052877426147461,
      "learning_rate": 5.430140002594067e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 11642144,
      "step": 17820
    },
    {
      "epoch": 10.51002358490566,
      "grad_norm": 2.274242401123047,
      "learning_rate": 5.4275767694293934e-06,
      "loss": 0.3507,
      "num_input_tokens_seen": 11645568,
      "step": 17825
    },
    {
      "epoch": 10.512971698113208,
      "grad_norm": 4.130377292633057,
      "learning_rate": 5.425013423061926e-06,
      "loss": 0.5165,
      "num_input_tokens_seen": 11649056,
      "step": 17830
    },
    {
      "epoch": 10.515919811320755,
      "grad_norm": 3.9317069053649902,
      "learning_rate": 5.422449964170324e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 11653120,
      "step": 17835
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 4.330565929412842,
      "learning_rate": 5.419886393433275e-06,
      "loss": 0.3058,
      "num_input_tokens_seen": 11657632,
      "step": 17840
    },
    {
      "epoch": 10.52181603773585,
      "grad_norm": 5.769122123718262,
      "learning_rate": 5.417322711529491e-06,
      "loss": 0.3361,
      "num_input_tokens_seen": 11660992,
      "step": 17845
    },
    {
      "epoch": 10.524764150943396,
      "grad_norm": 2.3091564178466797,
      "learning_rate": 5.4147589191377224e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 11664672,
      "step": 17850
    },
    {
      "epoch": 10.527712264150944,
      "grad_norm": 1.7396917343139648,
      "learning_rate": 5.412195016936742e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 11668672,
      "step": 17855
    },
    {
      "epoch": 10.53066037735849,
      "grad_norm": 4.496405124664307,
      "learning_rate": 5.409631005605354e-06,
      "loss": 0.6391,
      "num_input_tokens_seen": 11671392,
      "step": 17860
    },
    {
      "epoch": 10.533608490566039,
      "grad_norm": 3.3741204738616943,
      "learning_rate": 5.407066885822391e-06,
      "loss": 0.498,
      "num_input_tokens_seen": 11674560,
      "step": 17865
    },
    {
      "epoch": 10.536556603773585,
      "grad_norm": 2.8106727600097656,
      "learning_rate": 5.404502658266717e-06,
      "loss": 0.4049,
      "num_input_tokens_seen": 11677696,
      "step": 17870
    },
    {
      "epoch": 10.539504716981131,
      "grad_norm": 2.331644058227539,
      "learning_rate": 5.4019383236172195e-06,
      "loss": 0.4972,
      "num_input_tokens_seen": 11681504,
      "step": 17875
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 3.853442668914795,
      "learning_rate": 5.39937388255282e-06,
      "loss": 0.4049,
      "num_input_tokens_seen": 11684672,
      "step": 17880
    },
    {
      "epoch": 10.545400943396226,
      "grad_norm": 2.4221765995025635,
      "learning_rate": 5.3968093357524645e-06,
      "loss": 0.4579,
      "num_input_tokens_seen": 11688032,
      "step": 17885
    },
    {
      "epoch": 10.548349056603774,
      "grad_norm": 1.4227325916290283,
      "learning_rate": 5.3942446838951245e-06,
      "loss": 0.2653,
      "num_input_tokens_seen": 11690400,
      "step": 17890
    },
    {
      "epoch": 10.55129716981132,
      "grad_norm": 3.2254207134246826,
      "learning_rate": 5.3916799276598074e-06,
      "loss": 0.5792,
      "num_input_tokens_seen": 11693664,
      "step": 17895
    },
    {
      "epoch": 10.554245283018869,
      "grad_norm": 1.7772433757781982,
      "learning_rate": 5.3891150677255425e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 11696320,
      "step": 17900
    },
    {
      "epoch": 10.557193396226415,
      "grad_norm": 2.60695481300354,
      "learning_rate": 5.386550104771384e-06,
      "loss": 0.4233,
      "num_input_tokens_seen": 11699872,
      "step": 17905
    },
    {
      "epoch": 10.560141509433961,
      "grad_norm": 2.7642438411712646,
      "learning_rate": 5.3839850394764205e-06,
      "loss": 0.2066,
      "num_input_tokens_seen": 11702912,
      "step": 17910
    },
    {
      "epoch": 10.56308962264151,
      "grad_norm": 3.271554708480835,
      "learning_rate": 5.381419872519763e-06,
      "loss": 0.3491,
      "num_input_tokens_seen": 11705664,
      "step": 17915
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 1.6263501644134521,
      "learning_rate": 5.378854604580549e-06,
      "loss": 0.3225,
      "num_input_tokens_seen": 11709792,
      "step": 17920
    },
    {
      "epoch": 10.568985849056604,
      "grad_norm": 2.964728593826294,
      "learning_rate": 5.376289236337946e-06,
      "loss": 0.5011,
      "num_input_tokens_seen": 11713024,
      "step": 17925
    },
    {
      "epoch": 10.57193396226415,
      "grad_norm": 3.215033531188965,
      "learning_rate": 5.373723768471147e-06,
      "loss": 0.5421,
      "num_input_tokens_seen": 11715904,
      "step": 17930
    },
    {
      "epoch": 10.574882075471699,
      "grad_norm": 8.855051040649414,
      "learning_rate": 5.37115820165937e-06,
      "loss": 0.313,
      "num_input_tokens_seen": 11718720,
      "step": 17935
    },
    {
      "epoch": 10.577830188679245,
      "grad_norm": 2.347040891647339,
      "learning_rate": 5.368592536581858e-06,
      "loss": 0.4501,
      "num_input_tokens_seen": 11721856,
      "step": 17940
    },
    {
      "epoch": 10.580778301886792,
      "grad_norm": 4.751888751983643,
      "learning_rate": 5.366026773917885e-06,
      "loss": 0.3368,
      "num_input_tokens_seen": 11724544,
      "step": 17945
    },
    {
      "epoch": 10.58372641509434,
      "grad_norm": 2.483494997024536,
      "learning_rate": 5.363460914346746e-06,
      "loss": 0.3003,
      "num_input_tokens_seen": 11727648,
      "step": 17950
    },
    {
      "epoch": 10.586674528301886,
      "grad_norm": 2.1051361560821533,
      "learning_rate": 5.360894958547762e-06,
      "loss": 0.4458,
      "num_input_tokens_seen": 11730464,
      "step": 17955
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 1.8857148885726929,
      "learning_rate": 5.358328907200284e-06,
      "loss": 0.2751,
      "num_input_tokens_seen": 11733600,
      "step": 17960
    },
    {
      "epoch": 10.59257075471698,
      "grad_norm": 1.6844147443771362,
      "learning_rate": 5.355762760983682e-06,
      "loss": 0.2988,
      "num_input_tokens_seen": 11736544,
      "step": 17965
    },
    {
      "epoch": 10.595518867924529,
      "grad_norm": 2.696889877319336,
      "learning_rate": 5.353196520577356e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 11739424,
      "step": 17970
    },
    {
      "epoch": 10.598466981132075,
      "grad_norm": 2.7280452251434326,
      "learning_rate": 5.35063018666073e-06,
      "loss": 0.4453,
      "num_input_tokens_seen": 11742592,
      "step": 17975
    },
    {
      "epoch": 10.601415094339622,
      "grad_norm": 2.7200305461883545,
      "learning_rate": 5.3480637599132515e-06,
      "loss": 0.367,
      "num_input_tokens_seen": 11745536,
      "step": 17980
    },
    {
      "epoch": 10.60436320754717,
      "grad_norm": 4.168461799621582,
      "learning_rate": 5.34549724101439e-06,
      "loss": 0.4476,
      "num_input_tokens_seen": 11749280,
      "step": 17985
    },
    {
      "epoch": 10.607311320754716,
      "grad_norm": 1.5932660102844238,
      "learning_rate": 5.342930630643646e-06,
      "loss": 0.5224,
      "num_input_tokens_seen": 11751840,
      "step": 17990
    },
    {
      "epoch": 10.610259433962264,
      "grad_norm": 2.4405980110168457,
      "learning_rate": 5.340363929480541e-06,
      "loss": 0.3563,
      "num_input_tokens_seen": 11754880,
      "step": 17995
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 1.572044849395752,
      "learning_rate": 5.3377971382046164e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 11758144,
      "step": 18000
    },
    {
      "epoch": 10.616155660377359,
      "grad_norm": 3.1554408073425293,
      "learning_rate": 5.335230257495446e-06,
      "loss": 0.3706,
      "num_input_tokens_seen": 11760800,
      "step": 18005
    },
    {
      "epoch": 10.619103773584905,
      "grad_norm": 2.918224573135376,
      "learning_rate": 5.3326632880326205e-06,
      "loss": 0.4275,
      "num_input_tokens_seen": 11763584,
      "step": 18010
    },
    {
      "epoch": 10.622051886792454,
      "grad_norm": 2.8243789672851562,
      "learning_rate": 5.3300962304957515e-06,
      "loss": 0.6045,
      "num_input_tokens_seen": 11766016,
      "step": 18015
    },
    {
      "epoch": 10.625,
      "grad_norm": 3.4763824939727783,
      "learning_rate": 5.327529085564487e-06,
      "loss": 0.4725,
      "num_input_tokens_seen": 11769344,
      "step": 18020
    },
    {
      "epoch": 10.627948113207546,
      "grad_norm": 2.389759063720703,
      "learning_rate": 5.324961853918485e-06,
      "loss": 0.3658,
      "num_input_tokens_seen": 11773536,
      "step": 18025
    },
    {
      "epoch": 10.630896226415095,
      "grad_norm": 3.1256370544433594,
      "learning_rate": 5.32239453623743e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 11776800,
      "step": 18030
    },
    {
      "epoch": 10.633844339622641,
      "grad_norm": 2.493598699569702,
      "learning_rate": 5.3198271332010335e-06,
      "loss": 0.3886,
      "num_input_tokens_seen": 11780544,
      "step": 18035
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 2.4832937717437744,
      "learning_rate": 5.317259645489024e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 11783680,
      "step": 18040
    },
    {
      "epoch": 10.639740566037736,
      "grad_norm": 2.4133715629577637,
      "learning_rate": 5.314692073781157e-06,
      "loss": 0.2992,
      "num_input_tokens_seen": 11786240,
      "step": 18045
    },
    {
      "epoch": 10.642688679245284,
      "grad_norm": 2.0527801513671875,
      "learning_rate": 5.312124418757207e-06,
      "loss": 0.3378,
      "num_input_tokens_seen": 11790240,
      "step": 18050
    },
    {
      "epoch": 10.64563679245283,
      "grad_norm": 5.2826619148254395,
      "learning_rate": 5.309556681096972e-06,
      "loss": 0.3169,
      "num_input_tokens_seen": 11792704,
      "step": 18055
    },
    {
      "epoch": 10.648584905660378,
      "grad_norm": 2.517688035964966,
      "learning_rate": 5.306988861480271e-06,
      "loss": 0.3248,
      "num_input_tokens_seen": 11796032,
      "step": 18060
    },
    {
      "epoch": 10.651533018867925,
      "grad_norm": 2.6958463191986084,
      "learning_rate": 5.304420960586946e-06,
      "loss": 0.376,
      "num_input_tokens_seen": 11798368,
      "step": 18065
    },
    {
      "epoch": 10.654481132075471,
      "grad_norm": 2.0004935264587402,
      "learning_rate": 5.3018529790968606e-06,
      "loss": 0.3711,
      "num_input_tokens_seen": 11802016,
      "step": 18070
    },
    {
      "epoch": 10.65742924528302,
      "grad_norm": 1.664948582649231,
      "learning_rate": 5.299284917689898e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 11806240,
      "step": 18075
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 2.158820629119873,
      "learning_rate": 5.296716777045962e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 11809184,
      "step": 18080
    },
    {
      "epoch": 10.663325471698114,
      "grad_norm": 2.7022690773010254,
      "learning_rate": 5.294148557844983e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 11813248,
      "step": 18085
    },
    {
      "epoch": 10.66627358490566,
      "grad_norm": 2.4976425170898438,
      "learning_rate": 5.291580260766904e-06,
      "loss": 0.4568,
      "num_input_tokens_seen": 11815968,
      "step": 18090
    },
    {
      "epoch": 10.669221698113208,
      "grad_norm": 3.9949796199798584,
      "learning_rate": 5.289011886491694e-06,
      "loss": 0.4753,
      "num_input_tokens_seen": 11818624,
      "step": 18095
    },
    {
      "epoch": 10.672169811320755,
      "grad_norm": 2.560960531234741,
      "learning_rate": 5.286443435699342e-06,
      "loss": 0.3905,
      "num_input_tokens_seen": 11822240,
      "step": 18100
    },
    {
      "epoch": 10.675117924528301,
      "grad_norm": 2.4166297912597656,
      "learning_rate": 5.283874909069855e-06,
      "loss": 0.3418,
      "num_input_tokens_seen": 11824992,
      "step": 18105
    },
    {
      "epoch": 10.67806603773585,
      "grad_norm": 2.1950271129608154,
      "learning_rate": 5.281306307283263e-06,
      "loss": 0.4188,
      "num_input_tokens_seen": 11828032,
      "step": 18110
    },
    {
      "epoch": 10.681014150943396,
      "grad_norm": 4.421349048614502,
      "learning_rate": 5.2787376310196145e-06,
      "loss": 0.4302,
      "num_input_tokens_seen": 11830944,
      "step": 18115
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 3.616283893585205,
      "learning_rate": 5.276168880958977e-06,
      "loss": 0.3369,
      "num_input_tokens_seen": 11834496,
      "step": 18120
    },
    {
      "epoch": 10.68691037735849,
      "grad_norm": 1.867903232574463,
      "learning_rate": 5.273600057781437e-06,
      "loss": 0.3173,
      "num_input_tokens_seen": 11837312,
      "step": 18125
    },
    {
      "epoch": 10.689858490566039,
      "grad_norm": 2.358996629714966,
      "learning_rate": 5.271031162167103e-06,
      "loss": 0.3304,
      "num_input_tokens_seen": 11840928,
      "step": 18130
    },
    {
      "epoch": 10.692806603773585,
      "grad_norm": 2.0943846702575684,
      "learning_rate": 5.268462194796101e-06,
      "loss": 0.2916,
      "num_input_tokens_seen": 11844704,
      "step": 18135
    },
    {
      "epoch": 10.695754716981131,
      "grad_norm": 2.1024558544158936,
      "learning_rate": 5.265893156348576e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 11847744,
      "step": 18140
    },
    {
      "epoch": 10.69870283018868,
      "grad_norm": 3.050257444381714,
      "learning_rate": 5.2633240475046925e-06,
      "loss": 0.3511,
      "num_input_tokens_seen": 11850912,
      "step": 18145
    },
    {
      "epoch": 10.701650943396226,
      "grad_norm": 2.6693036556243896,
      "learning_rate": 5.2607548689446305e-06,
      "loss": 0.3053,
      "num_input_tokens_seen": 11854624,
      "step": 18150
    },
    {
      "epoch": 10.704599056603774,
      "grad_norm": 2.5336668491363525,
      "learning_rate": 5.258185621348595e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 11857824,
      "step": 18155
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 3.371170997619629,
      "learning_rate": 5.255616305396801e-06,
      "loss": 0.3831,
      "num_input_tokens_seen": 11860320,
      "step": 18160
    },
    {
      "epoch": 10.710495283018869,
      "grad_norm": 2.2351510524749756,
      "learning_rate": 5.253046921769491e-06,
      "loss": 0.3461,
      "num_input_tokens_seen": 11865024,
      "step": 18165
    },
    {
      "epoch": 10.713443396226415,
      "grad_norm": 3.533430337905884,
      "learning_rate": 5.250477471146916e-06,
      "loss": 0.564,
      "num_input_tokens_seen": 11867648,
      "step": 18170
    },
    {
      "epoch": 10.716391509433961,
      "grad_norm": 2.2561612129211426,
      "learning_rate": 5.2479079542093535e-06,
      "loss": 0.3044,
      "num_input_tokens_seen": 11870688,
      "step": 18175
    },
    {
      "epoch": 10.71933962264151,
      "grad_norm": 1.7384933233261108,
      "learning_rate": 5.245338371637091e-06,
      "loss": 0.3201,
      "num_input_tokens_seen": 11873856,
      "step": 18180
    },
    {
      "epoch": 10.722287735849056,
      "grad_norm": 4.056912899017334,
      "learning_rate": 5.242768724110437e-06,
      "loss": 0.5499,
      "num_input_tokens_seen": 11876896,
      "step": 18185
    },
    {
      "epoch": 10.725235849056604,
      "grad_norm": 1.9095032215118408,
      "learning_rate": 5.240199012309717e-06,
      "loss": 0.2746,
      "num_input_tokens_seen": 11883872,
      "step": 18190
    },
    {
      "epoch": 10.72818396226415,
      "grad_norm": 8.1639404296875,
      "learning_rate": 5.237629236915273e-06,
      "loss": 0.4156,
      "num_input_tokens_seen": 11886816,
      "step": 18195
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 1.5733028650283813,
      "learning_rate": 5.235059398607464e-06,
      "loss": 0.2979,
      "num_input_tokens_seen": 11889888,
      "step": 18200
    },
    {
      "epoch": 10.734080188679245,
      "grad_norm": 2.6612954139709473,
      "learning_rate": 5.232489498066665e-06,
      "loss": 0.3706,
      "num_input_tokens_seen": 11893216,
      "step": 18205
    },
    {
      "epoch": 10.737028301886792,
      "grad_norm": 3.0885848999023438,
      "learning_rate": 5.229919535973272e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 11896704,
      "step": 18210
    },
    {
      "epoch": 10.73997641509434,
      "grad_norm": 2.502556562423706,
      "learning_rate": 5.2273495130076905e-06,
      "loss": 0.464,
      "num_input_tokens_seen": 11899968,
      "step": 18215
    },
    {
      "epoch": 10.742924528301886,
      "grad_norm": 6.081595420837402,
      "learning_rate": 5.224779429850344e-06,
      "loss": 0.5001,
      "num_input_tokens_seen": 11903072,
      "step": 18220
    },
    {
      "epoch": 10.745872641509434,
      "grad_norm": 3.075894355773926,
      "learning_rate": 5.222209287181677e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 11905728,
      "step": 18225
    },
    {
      "epoch": 10.74882075471698,
      "grad_norm": 3.149172306060791,
      "learning_rate": 5.219639085682142e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 11907936,
      "step": 18230
    },
    {
      "epoch": 10.751768867924529,
      "grad_norm": 4.814205646514893,
      "learning_rate": 5.2170688260322124e-06,
      "loss": 0.3658,
      "num_input_tokens_seen": 11910240,
      "step": 18235
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 3.566417932510376,
      "learning_rate": 5.214498508912376e-06,
      "loss": 0.3872,
      "num_input_tokens_seen": 11912960,
      "step": 18240
    },
    {
      "epoch": 10.757665094339622,
      "grad_norm": 3.653684377670288,
      "learning_rate": 5.211928135003135e-06,
      "loss": 0.2908,
      "num_input_tokens_seen": 11915808,
      "step": 18245
    },
    {
      "epoch": 10.76061320754717,
      "grad_norm": 1.9069368839263916,
      "learning_rate": 5.209357704985007e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 11918368,
      "step": 18250
    },
    {
      "epoch": 10.763561320754716,
      "grad_norm": 2.686373710632324,
      "learning_rate": 5.206787219538524e-06,
      "loss": 0.3734,
      "num_input_tokens_seen": 11920832,
      "step": 18255
    },
    {
      "epoch": 10.766509433962264,
      "grad_norm": 2.0669944286346436,
      "learning_rate": 5.204216679344234e-06,
      "loss": 0.392,
      "num_input_tokens_seen": 11924864,
      "step": 18260
    },
    {
      "epoch": 10.76945754716981,
      "grad_norm": 2.3509879112243652,
      "learning_rate": 5.201646085082696e-06,
      "loss": 0.3009,
      "num_input_tokens_seen": 11927616,
      "step": 18265
    },
    {
      "epoch": 10.772405660377359,
      "grad_norm": 3.999713659286499,
      "learning_rate": 5.199075437434491e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 11930784,
      "step": 18270
    },
    {
      "epoch": 10.775353773584905,
      "grad_norm": 3.640404224395752,
      "learning_rate": 5.1965047370802046e-06,
      "loss": 0.4012,
      "num_input_tokens_seen": 11934016,
      "step": 18275
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 3.176971197128296,
      "learning_rate": 5.193933984700441e-06,
      "loss": 0.3895,
      "num_input_tokens_seen": 11936800,
      "step": 18280
    },
    {
      "epoch": 10.78125,
      "grad_norm": 3.334949016571045,
      "learning_rate": 5.1913631809758216e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 11939648,
      "step": 18285
    },
    {
      "epoch": 10.784198113207546,
      "grad_norm": 2.41296648979187,
      "learning_rate": 5.188792326586973e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 11942912,
      "step": 18290
    },
    {
      "epoch": 10.787146226415095,
      "grad_norm": 2.599114418029785,
      "learning_rate": 5.186221422214544e-06,
      "loss": 0.3263,
      "num_input_tokens_seen": 11945568,
      "step": 18295
    },
    {
      "epoch": 10.790094339622641,
      "grad_norm": 3.2736313343048096,
      "learning_rate": 5.1836504685391885e-06,
      "loss": 0.443,
      "num_input_tokens_seen": 11948672,
      "step": 18300
    },
    {
      "epoch": 10.79304245283019,
      "grad_norm": 3.2548632621765137,
      "learning_rate": 5.181079466241582e-06,
      "loss": 0.4675,
      "num_input_tokens_seen": 11951648,
      "step": 18305
    },
    {
      "epoch": 10.795990566037736,
      "grad_norm": 4.639283657073975,
      "learning_rate": 5.178508416002406e-06,
      "loss": 0.5134,
      "num_input_tokens_seen": 11955392,
      "step": 18310
    },
    {
      "epoch": 10.798938679245284,
      "grad_norm": 2.2992923259735107,
      "learning_rate": 5.175937318502357e-06,
      "loss": 0.4946,
      "num_input_tokens_seen": 11958880,
      "step": 18315
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 1.747592806816101,
      "learning_rate": 5.173366174422147e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 11961792,
      "step": 18320
    },
    {
      "epoch": 10.804834905660378,
      "grad_norm": 3.9357683658599854,
      "learning_rate": 5.170794984442492e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 11964704,
      "step": 18325
    },
    {
      "epoch": 10.807783018867925,
      "grad_norm": 2.109396457672119,
      "learning_rate": 5.16822374924413e-06,
      "loss": 0.3439,
      "num_input_tokens_seen": 11967040,
      "step": 18330
    },
    {
      "epoch": 10.810731132075471,
      "grad_norm": 11.139080047607422,
      "learning_rate": 5.165652469507806e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 11969632,
      "step": 18335
    },
    {
      "epoch": 10.81367924528302,
      "grad_norm": 3.1703648567199707,
      "learning_rate": 5.163081145914276e-06,
      "loss": 0.4107,
      "num_input_tokens_seen": 11972320,
      "step": 18340
    },
    {
      "epoch": 10.816627358490566,
      "grad_norm": 2.343620538711548,
      "learning_rate": 5.160509779144311e-06,
      "loss": 0.3141,
      "num_input_tokens_seen": 11976256,
      "step": 18345
    },
    {
      "epoch": 10.819575471698114,
      "grad_norm": 2.559418201446533,
      "learning_rate": 5.157938369878688e-06,
      "loss": 0.4253,
      "num_input_tokens_seen": 11979264,
      "step": 18350
    },
    {
      "epoch": 10.82252358490566,
      "grad_norm": 2.525541067123413,
      "learning_rate": 5.155366918798203e-06,
      "loss": 0.3071,
      "num_input_tokens_seen": 11982048,
      "step": 18355
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 2.4824178218841553,
      "learning_rate": 5.152795426583654e-06,
      "loss": 0.4022,
      "num_input_tokens_seen": 11985600,
      "step": 18360
    },
    {
      "epoch": 10.828419811320755,
      "grad_norm": 1.7608798742294312,
      "learning_rate": 5.15022389391586e-06,
      "loss": 0.455,
      "num_input_tokens_seen": 11988640,
      "step": 18365
    },
    {
      "epoch": 10.831367924528301,
      "grad_norm": 2.961484909057617,
      "learning_rate": 5.147652321475642e-06,
      "loss": 0.3716,
      "num_input_tokens_seen": 11992320,
      "step": 18370
    },
    {
      "epoch": 10.83431603773585,
      "grad_norm": 5.424807548522949,
      "learning_rate": 5.145080709943835e-06,
      "loss": 0.3271,
      "num_input_tokens_seen": 11995424,
      "step": 18375
    },
    {
      "epoch": 10.837264150943396,
      "grad_norm": 1.9254889488220215,
      "learning_rate": 5.142509060001285e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 11999360,
      "step": 18380
    },
    {
      "epoch": 10.840212264150944,
      "grad_norm": 3.195554733276367,
      "learning_rate": 5.139937372328847e-06,
      "loss": 0.4276,
      "num_input_tokens_seen": 12003360,
      "step": 18385
    },
    {
      "epoch": 10.84316037735849,
      "grad_norm": 3.019355058670044,
      "learning_rate": 5.1373656476073876e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 12006176,
      "step": 18390
    },
    {
      "epoch": 10.846108490566039,
      "grad_norm": 2.383373260498047,
      "learning_rate": 5.134793886517779e-06,
      "loss": 0.2887,
      "num_input_tokens_seen": 12009920,
      "step": 18395
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 2.029245138168335,
      "learning_rate": 5.1322220897409105e-06,
      "loss": 0.3938,
      "num_input_tokens_seen": 12012736,
      "step": 18400
    },
    {
      "epoch": 10.852004716981131,
      "grad_norm": 2.5142924785614014,
      "learning_rate": 5.129650257957671e-06,
      "loss": 0.3869,
      "num_input_tokens_seen": 12016576,
      "step": 18405
    },
    {
      "epoch": 10.85495283018868,
      "grad_norm": 1.6941502094268799,
      "learning_rate": 5.12707839184897e-06,
      "loss": 0.3267,
      "num_input_tokens_seen": 12019712,
      "step": 18410
    },
    {
      "epoch": 10.857900943396226,
      "grad_norm": 3.540437936782837,
      "learning_rate": 5.124506492095716e-06,
      "loss": 0.3793,
      "num_input_tokens_seen": 12022432,
      "step": 18415
    },
    {
      "epoch": 10.860849056603774,
      "grad_norm": 2.84694766998291,
      "learning_rate": 5.121934559378831e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 12025408,
      "step": 18420
    },
    {
      "epoch": 10.86379716981132,
      "grad_norm": 2.3983306884765625,
      "learning_rate": 5.1193625943792456e-06,
      "loss": 0.5443,
      "num_input_tokens_seen": 12028768,
      "step": 18425
    },
    {
      "epoch": 10.866745283018869,
      "grad_norm": 3.3345470428466797,
      "learning_rate": 5.116790597777901e-06,
      "loss": 0.5353,
      "num_input_tokens_seen": 12031488,
      "step": 18430
    },
    {
      "epoch": 10.869693396226415,
      "grad_norm": 1.9027084112167358,
      "learning_rate": 5.11421857025574e-06,
      "loss": 0.245,
      "num_input_tokens_seen": 12036192,
      "step": 18435
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 1.5639606714248657,
      "learning_rate": 5.111646512493721e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 12040512,
      "step": 18440
    },
    {
      "epoch": 10.87558962264151,
      "grad_norm": 2.640469789505005,
      "learning_rate": 5.109074425172806e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 12043680,
      "step": 18445
    },
    {
      "epoch": 10.878537735849056,
      "grad_norm": 2.4510624408721924,
      "learning_rate": 5.106502308973967e-06,
      "loss": 0.4755,
      "num_input_tokens_seen": 12051296,
      "step": 18450
    },
    {
      "epoch": 10.881485849056604,
      "grad_norm": 1.8619440793991089,
      "learning_rate": 5.103930164578184e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 12054688,
      "step": 18455
    },
    {
      "epoch": 10.88443396226415,
      "grad_norm": 1.9570364952087402,
      "learning_rate": 5.101357992666441e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 12057984,
      "step": 18460
    },
    {
      "epoch": 10.887382075471699,
      "grad_norm": 2.2112526893615723,
      "learning_rate": 5.098785793919733e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 12061792,
      "step": 18465
    },
    {
      "epoch": 10.890330188679245,
      "grad_norm": 3.1938085556030273,
      "learning_rate": 5.096213569019061e-06,
      "loss": 0.2566,
      "num_input_tokens_seen": 12066112,
      "step": 18470
    },
    {
      "epoch": 10.893278301886792,
      "grad_norm": 2.5193662643432617,
      "learning_rate": 5.0936413186454315e-06,
      "loss": 0.4194,
      "num_input_tokens_seen": 12069024,
      "step": 18475
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 3.5767180919647217,
      "learning_rate": 5.0910690434798584e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 12071616,
      "step": 18480
    },
    {
      "epoch": 10.899174528301886,
      "grad_norm": 1.890755295753479,
      "learning_rate": 5.088496744203364e-06,
      "loss": 0.3391,
      "num_input_tokens_seen": 12074752,
      "step": 18485
    },
    {
      "epoch": 10.902122641509434,
      "grad_norm": 2.243572235107422,
      "learning_rate": 5.085924421496976e-06,
      "loss": 0.3797,
      "num_input_tokens_seen": 12078784,
      "step": 18490
    },
    {
      "epoch": 10.90507075471698,
      "grad_norm": 2.966371536254883,
      "learning_rate": 5.083352076041725e-06,
      "loss": 0.4371,
      "num_input_tokens_seen": 12081504,
      "step": 18495
    },
    {
      "epoch": 10.908018867924529,
      "grad_norm": 2.7092092037200928,
      "learning_rate": 5.080779708518654e-06,
      "loss": 0.5117,
      "num_input_tokens_seen": 12084480,
      "step": 18500
    },
    {
      "epoch": 10.910966981132075,
      "grad_norm": 5.167614459991455,
      "learning_rate": 5.078207319608807e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 12086816,
      "step": 18505
    },
    {
      "epoch": 10.913915094339622,
      "grad_norm": 3.379915714263916,
      "learning_rate": 5.075634909993235e-06,
      "loss": 0.3692,
      "num_input_tokens_seen": 12089696,
      "step": 18510
    },
    {
      "epoch": 10.91686320754717,
      "grad_norm": 2.802248954772949,
      "learning_rate": 5.073062480352995e-06,
      "loss": 0.3132,
      "num_input_tokens_seen": 12092064,
      "step": 18515
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 5.269838333129883,
      "learning_rate": 5.070490031369149e-06,
      "loss": 0.3767,
      "num_input_tokens_seen": 12094080,
      "step": 18520
    },
    {
      "epoch": 10.922759433962264,
      "grad_norm": 3.403148889541626,
      "learning_rate": 5.067917563722762e-06,
      "loss": 0.3236,
      "num_input_tokens_seen": 12098240,
      "step": 18525
    },
    {
      "epoch": 10.92570754716981,
      "grad_norm": 2.975417137145996,
      "learning_rate": 5.065345078094907e-06,
      "loss": 0.5274,
      "num_input_tokens_seen": 12101056,
      "step": 18530
    },
    {
      "epoch": 10.928655660377359,
      "grad_norm": 6.292437553405762,
      "learning_rate": 5.062772575166663e-06,
      "loss": 0.3657,
      "num_input_tokens_seen": 12103584,
      "step": 18535
    },
    {
      "epoch": 10.931603773584905,
      "grad_norm": 2.543532609939575,
      "learning_rate": 5.0602000556191075e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 12110304,
      "step": 18540
    },
    {
      "epoch": 10.934551886792454,
      "grad_norm": 1.6381995677947998,
      "learning_rate": 5.0576275201333284e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 12113888,
      "step": 18545
    },
    {
      "epoch": 10.9375,
      "grad_norm": 2.395709276199341,
      "learning_rate": 5.055054969390415e-06,
      "loss": 0.387,
      "num_input_tokens_seen": 12116512,
      "step": 18550
    },
    {
      "epoch": 10.940448113207546,
      "grad_norm": 2.0614936351776123,
      "learning_rate": 5.052482404071461e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 12121024,
      "step": 18555
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 2.9037232398986816,
      "learning_rate": 5.049909824857564e-06,
      "loss": 0.288,
      "num_input_tokens_seen": 12123904,
      "step": 18560
    },
    {
      "epoch": 10.946344339622641,
      "grad_norm": 4.0091776847839355,
      "learning_rate": 5.047337232429827e-06,
      "loss": 0.3515,
      "num_input_tokens_seen": 12127648,
      "step": 18565
    },
    {
      "epoch": 10.94929245283019,
      "grad_norm": 3.5112311840057373,
      "learning_rate": 5.044764627469354e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 12130784,
      "step": 18570
    },
    {
      "epoch": 10.952240566037736,
      "grad_norm": 1.4793812036514282,
      "learning_rate": 5.042192010657251e-06,
      "loss": 0.2962,
      "num_input_tokens_seen": 12133728,
      "step": 18575
    },
    {
      "epoch": 10.955188679245284,
      "grad_norm": 5.3206024169921875,
      "learning_rate": 5.039619382674632e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 12137152,
      "step": 18580
    },
    {
      "epoch": 10.95813679245283,
      "grad_norm": 3.300123453140259,
      "learning_rate": 5.0370467442026115e-06,
      "loss": 0.4988,
      "num_input_tokens_seen": 12140672,
      "step": 18585
    },
    {
      "epoch": 10.961084905660378,
      "grad_norm": 1.8992621898651123,
      "learning_rate": 5.034474095922304e-06,
      "loss": 0.382,
      "num_input_tokens_seen": 12143936,
      "step": 18590
    },
    {
      "epoch": 10.964033018867925,
      "grad_norm": 4.129718780517578,
      "learning_rate": 5.031901438514832e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 12146368,
      "step": 18595
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 2.596972942352295,
      "learning_rate": 5.0293287726613185e-06,
      "loss": 0.323,
      "num_input_tokens_seen": 12149216,
      "step": 18600
    },
    {
      "epoch": 10.96992924528302,
      "grad_norm": 2.948406457901001,
      "learning_rate": 5.0267560990428836e-06,
      "loss": 0.4001,
      "num_input_tokens_seen": 12152128,
      "step": 18605
    },
    {
      "epoch": 10.972877358490566,
      "grad_norm": 2.712996244430542,
      "learning_rate": 5.024183418340657e-06,
      "loss": 0.483,
      "num_input_tokens_seen": 12159136,
      "step": 18610
    },
    {
      "epoch": 10.975825471698114,
      "grad_norm": 2.478860378265381,
      "learning_rate": 5.021610731235766e-06,
      "loss": 0.3727,
      "num_input_tokens_seen": 12161856,
      "step": 18615
    },
    {
      "epoch": 10.97877358490566,
      "grad_norm": 2.585320234298706,
      "learning_rate": 5.01903803840934e-06,
      "loss": 0.4808,
      "num_input_tokens_seen": 12164896,
      "step": 18620
    },
    {
      "epoch": 10.981721698113208,
      "grad_norm": 2.1764211654663086,
      "learning_rate": 5.016465340542514e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 12167744,
      "step": 18625
    },
    {
      "epoch": 10.984669811320755,
      "grad_norm": 3.206021308898926,
      "learning_rate": 5.013892638316417e-06,
      "loss": 0.3248,
      "num_input_tokens_seen": 12170624,
      "step": 18630
    },
    {
      "epoch": 10.987617924528301,
      "grad_norm": 2.208869457244873,
      "learning_rate": 5.011319932412182e-06,
      "loss": 0.3363,
      "num_input_tokens_seen": 12174016,
      "step": 18635
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 3.025620698928833,
      "learning_rate": 5.008747223510947e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 12176704,
      "step": 18640
    },
    {
      "epoch": 10.993514150943396,
      "grad_norm": 3.119352340698242,
      "learning_rate": 5.006174512293849e-06,
      "loss": 0.3919,
      "num_input_tokens_seen": 12180832,
      "step": 18645
    },
    {
      "epoch": 10.996462264150944,
      "grad_norm": 2.5751779079437256,
      "learning_rate": 5.003601799442019e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 12183552,
      "step": 18650
    },
    {
      "epoch": 10.99941037735849,
      "grad_norm": 1.6569411754608154,
      "learning_rate": 5.0010290856366e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 12186688,
      "step": 18655
    },
    {
      "epoch": 11.002358490566039,
      "grad_norm": 2.07817006111145,
      "learning_rate": 4.998456371558726e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 12189792,
      "step": 18660
    },
    {
      "epoch": 11.005306603773585,
      "grad_norm": 1.6101347208023071,
      "learning_rate": 4.995883657889531e-06,
      "loss": 0.2374,
      "num_input_tokens_seen": 12193440,
      "step": 18665
    },
    {
      "epoch": 11.008254716981131,
      "grad_norm": 2.6193971633911133,
      "learning_rate": 4.993310945310158e-06,
      "loss": 0.5764,
      "num_input_tokens_seen": 12197440,
      "step": 18670
    },
    {
      "epoch": 11.01120283018868,
      "grad_norm": 2.8443214893341064,
      "learning_rate": 4.9907382345017416e-06,
      "loss": 0.3863,
      "num_input_tokens_seen": 12199936,
      "step": 18675
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 3.610927104949951,
      "learning_rate": 4.988165526145416e-06,
      "loss": 0.3153,
      "num_input_tokens_seen": 12203008,
      "step": 18680
    },
    {
      "epoch": 11.017099056603774,
      "grad_norm": 2.3691766262054443,
      "learning_rate": 4.985592820922319e-06,
      "loss": 0.3065,
      "num_input_tokens_seen": 12205696,
      "step": 18685
    },
    {
      "epoch": 11.02004716981132,
      "grad_norm": 2.7915852069854736,
      "learning_rate": 4.983020119513586e-06,
      "loss": 0.3196,
      "num_input_tokens_seen": 12208480,
      "step": 18690
    },
    {
      "epoch": 11.022995283018869,
      "grad_norm": 2.748513698577881,
      "learning_rate": 4.9804474226003465e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 12210912,
      "step": 18695
    },
    {
      "epoch": 11.025943396226415,
      "grad_norm": 2.5785045623779297,
      "learning_rate": 4.977874730863739e-06,
      "loss": 0.376,
      "num_input_tokens_seen": 12214816,
      "step": 18700
    },
    {
      "epoch": 11.028891509433961,
      "grad_norm": 2.1167213916778564,
      "learning_rate": 4.975302044984889e-06,
      "loss": 0.2598,
      "num_input_tokens_seen": 12219616,
      "step": 18705
    },
    {
      "epoch": 11.03183962264151,
      "grad_norm": 2.484483003616333,
      "learning_rate": 4.972729365644931e-06,
      "loss": 0.2626,
      "num_input_tokens_seen": 12222400,
      "step": 18710
    },
    {
      "epoch": 11.034787735849056,
      "grad_norm": 2.920567274093628,
      "learning_rate": 4.97015669352499e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 12225024,
      "step": 18715
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 5.715683460235596,
      "learning_rate": 4.967584029306194e-06,
      "loss": 0.3993,
      "num_input_tokens_seen": 12231584,
      "step": 18720
    },
    {
      "epoch": 11.04068396226415,
      "grad_norm": 4.764193534851074,
      "learning_rate": 4.965011373669666e-06,
      "loss": 0.3204,
      "num_input_tokens_seen": 12234016,
      "step": 18725
    },
    {
      "epoch": 11.043632075471699,
      "grad_norm": 3.43125581741333,
      "learning_rate": 4.962438727296527e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 12237088,
      "step": 18730
    },
    {
      "epoch": 11.046580188679245,
      "grad_norm": 2.5600507259368896,
      "learning_rate": 4.959866090867897e-06,
      "loss": 0.3487,
      "num_input_tokens_seen": 12239616,
      "step": 18735
    },
    {
      "epoch": 11.049528301886792,
      "grad_norm": 2.2332050800323486,
      "learning_rate": 4.957293465064893e-06,
      "loss": 0.4369,
      "num_input_tokens_seen": 12242976,
      "step": 18740
    },
    {
      "epoch": 11.05247641509434,
      "grad_norm": 2.26415753364563,
      "learning_rate": 4.954720850568627e-06,
      "loss": 0.263,
      "num_input_tokens_seen": 12245984,
      "step": 18745
    },
    {
      "epoch": 11.055424528301886,
      "grad_norm": 2.660088300704956,
      "learning_rate": 4.952148248060212e-06,
      "loss": 0.4279,
      "num_input_tokens_seen": 12249376,
      "step": 18750
    },
    {
      "epoch": 11.058372641509434,
      "grad_norm": 1.383358120918274,
      "learning_rate": 4.949575658220755e-06,
      "loss": 0.2534,
      "num_input_tokens_seen": 12251744,
      "step": 18755
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 2.8924977779388428,
      "learning_rate": 4.947003081731359e-06,
      "loss": 0.2873,
      "num_input_tokens_seen": 12255360,
      "step": 18760
    },
    {
      "epoch": 11.064268867924529,
      "grad_norm": 2.9372799396514893,
      "learning_rate": 4.944430519273126e-06,
      "loss": 0.3862,
      "num_input_tokens_seen": 12258080,
      "step": 18765
    },
    {
      "epoch": 11.067216981132075,
      "grad_norm": 3.862826347351074,
      "learning_rate": 4.941857971527152e-06,
      "loss": 0.2884,
      "num_input_tokens_seen": 12260768,
      "step": 18770
    },
    {
      "epoch": 11.070165094339623,
      "grad_norm": 2.8268673419952393,
      "learning_rate": 4.93928543917453e-06,
      "loss": 0.3127,
      "num_input_tokens_seen": 12263232,
      "step": 18775
    },
    {
      "epoch": 11.07311320754717,
      "grad_norm": 2.020620346069336,
      "learning_rate": 4.93671292289635e-06,
      "loss": 0.2591,
      "num_input_tokens_seen": 12266336,
      "step": 18780
    },
    {
      "epoch": 11.076061320754716,
      "grad_norm": 2.624994993209839,
      "learning_rate": 4.934140423373698e-06,
      "loss": 0.3251,
      "num_input_tokens_seen": 12269632,
      "step": 18785
    },
    {
      "epoch": 11.079009433962264,
      "grad_norm": 4.177948951721191,
      "learning_rate": 4.931567941287651e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 12272480,
      "step": 18790
    },
    {
      "epoch": 11.08195754716981,
      "grad_norm": 3.717465877532959,
      "learning_rate": 4.9289954773192875e-06,
      "loss": 0.5082,
      "num_input_tokens_seen": 12274816,
      "step": 18795
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 2.3130388259887695,
      "learning_rate": 4.926423032149677e-06,
      "loss": 0.2893,
      "num_input_tokens_seen": 12277536,
      "step": 18800
    },
    {
      "epoch": 11.087853773584905,
      "grad_norm": 3.3192992210388184,
      "learning_rate": 4.923850606459883e-06,
      "loss": 0.5905,
      "num_input_tokens_seen": 12280192,
      "step": 18805
    },
    {
      "epoch": 11.090801886792454,
      "grad_norm": 4.04664421081543,
      "learning_rate": 4.921278200930972e-06,
      "loss": 0.3559,
      "num_input_tokens_seen": 12282656,
      "step": 18810
    },
    {
      "epoch": 11.09375,
      "grad_norm": 2.7599549293518066,
      "learning_rate": 4.918705816243996e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 12285952,
      "step": 18815
    },
    {
      "epoch": 11.096698113207546,
      "grad_norm": 3.0351381301879883,
      "learning_rate": 4.916133453080007e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 12289888,
      "step": 18820
    },
    {
      "epoch": 11.099646226415095,
      "grad_norm": 3.059849739074707,
      "learning_rate": 4.913561112120046e-06,
      "loss": 0.3049,
      "num_input_tokens_seen": 12292288,
      "step": 18825
    },
    {
      "epoch": 11.102594339622641,
      "grad_norm": 2.9580507278442383,
      "learning_rate": 4.910988794045154e-06,
      "loss": 0.4316,
      "num_input_tokens_seen": 12294688,
      "step": 18830
    },
    {
      "epoch": 11.10554245283019,
      "grad_norm": 2.651153326034546,
      "learning_rate": 4.9084164995363626e-06,
      "loss": 0.4217,
      "num_input_tokens_seen": 12297952,
      "step": 18835
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 3.077687978744507,
      "learning_rate": 4.905844229274697e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 12303808,
      "step": 18840
    },
    {
      "epoch": 11.111438679245284,
      "grad_norm": 3.0277867317199707,
      "learning_rate": 4.903271983941177e-06,
      "loss": 0.3354,
      "num_input_tokens_seen": 12306464,
      "step": 18845
    },
    {
      "epoch": 11.11438679245283,
      "grad_norm": 3.1911988258361816,
      "learning_rate": 4.900699764216818e-06,
      "loss": 0.3238,
      "num_input_tokens_seen": 12310304,
      "step": 18850
    },
    {
      "epoch": 11.117334905660377,
      "grad_norm": 3.3639252185821533,
      "learning_rate": 4.898127570782622e-06,
      "loss": 0.4372,
      "num_input_tokens_seen": 12314528,
      "step": 18855
    },
    {
      "epoch": 11.120283018867925,
      "grad_norm": 2.862252950668335,
      "learning_rate": 4.895555404319592e-06,
      "loss": 0.4883,
      "num_input_tokens_seen": 12318816,
      "step": 18860
    },
    {
      "epoch": 11.123231132075471,
      "grad_norm": 3.4091687202453613,
      "learning_rate": 4.89298326550872e-06,
      "loss": 0.4165,
      "num_input_tokens_seen": 12321504,
      "step": 18865
    },
    {
      "epoch": 11.12617924528302,
      "grad_norm": 2.546569347381592,
      "learning_rate": 4.8904111550309876e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 12325760,
      "step": 18870
    },
    {
      "epoch": 11.129127358490566,
      "grad_norm": 3.3297479152679443,
      "learning_rate": 4.8878390735673755e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 12328320,
      "step": 18875
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 3.8290083408355713,
      "learning_rate": 4.8852670217988505e-06,
      "loss": 0.4418,
      "num_input_tokens_seen": 12331488,
      "step": 18880
    },
    {
      "epoch": 11.13502358490566,
      "grad_norm": 3.857517957687378,
      "learning_rate": 4.882695000406377e-06,
      "loss": 0.3697,
      "num_input_tokens_seen": 12333888,
      "step": 18885
    },
    {
      "epoch": 11.137971698113208,
      "grad_norm": 2.149351119995117,
      "learning_rate": 4.880123010070909e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 12336640,
      "step": 18890
    },
    {
      "epoch": 11.140919811320755,
      "grad_norm": 1.9243743419647217,
      "learning_rate": 4.877551051473388e-06,
      "loss": 0.39,
      "num_input_tokens_seen": 12340896,
      "step": 18895
    },
    {
      "epoch": 11.143867924528301,
      "grad_norm": 3.900275707244873,
      "learning_rate": 4.874979125294755e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 12344736,
      "step": 18900
    },
    {
      "epoch": 11.14681603773585,
      "grad_norm": 1.9680038690567017,
      "learning_rate": 4.872407232215937e-06,
      "loss": 0.324,
      "num_input_tokens_seen": 12348288,
      "step": 18905
    },
    {
      "epoch": 11.149764150943396,
      "grad_norm": 2.5996077060699463,
      "learning_rate": 4.8698353729178546e-06,
      "loss": 0.4421,
      "num_input_tokens_seen": 12351424,
      "step": 18910
    },
    {
      "epoch": 11.152712264150944,
      "grad_norm": 1.8923604488372803,
      "learning_rate": 4.867263548081418e-06,
      "loss": 0.2962,
      "num_input_tokens_seen": 12355360,
      "step": 18915
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 2.211007833480835,
      "learning_rate": 4.8646917583875304e-06,
      "loss": 0.4701,
      "num_input_tokens_seen": 12358272,
      "step": 18920
    },
    {
      "epoch": 11.158608490566039,
      "grad_norm": 3.6852946281433105,
      "learning_rate": 4.862120004517082e-06,
      "loss": 0.433,
      "num_input_tokens_seen": 12361152,
      "step": 18925
    },
    {
      "epoch": 11.161556603773585,
      "grad_norm": 3.2426340579986572,
      "learning_rate": 4.859548287150956e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 12364480,
      "step": 18930
    },
    {
      "epoch": 11.164504716981131,
      "grad_norm": 1.8590847253799438,
      "learning_rate": 4.8569766069700275e-06,
      "loss": 0.3973,
      "num_input_tokens_seen": 12367296,
      "step": 18935
    },
    {
      "epoch": 11.16745283018868,
      "grad_norm": 2.6552910804748535,
      "learning_rate": 4.854404964655158e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 12370400,
      "step": 18940
    },
    {
      "epoch": 11.170400943396226,
      "grad_norm": 2.037202835083008,
      "learning_rate": 4.8518333608872015e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 12373632,
      "step": 18945
    },
    {
      "epoch": 11.173349056603774,
      "grad_norm": 3.639559030532837,
      "learning_rate": 4.849261796347002e-06,
      "loss": 0.3505,
      "num_input_tokens_seen": 12376384,
      "step": 18950
    },
    {
      "epoch": 11.17629716981132,
      "grad_norm": 2.8348350524902344,
      "learning_rate": 4.846690271715391e-06,
      "loss": 0.3116,
      "num_input_tokens_seen": 12379520,
      "step": 18955
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 2.341925859451294,
      "learning_rate": 4.844118787673191e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 12383040,
      "step": 18960
    },
    {
      "epoch": 11.182193396226415,
      "grad_norm": 2.2823874950408936,
      "learning_rate": 4.841547344901214e-06,
      "loss": 0.3337,
      "num_input_tokens_seen": 12385696,
      "step": 18965
    },
    {
      "epoch": 11.185141509433961,
      "grad_norm": 2.3264060020446777,
      "learning_rate": 4.838975944080261e-06,
      "loss": 0.3995,
      "num_input_tokens_seen": 12388736,
      "step": 18970
    },
    {
      "epoch": 11.18808962264151,
      "grad_norm": 1.9796124696731567,
      "learning_rate": 4.83640458589112e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 12392288,
      "step": 18975
    },
    {
      "epoch": 11.191037735849056,
      "grad_norm": 3.560814142227173,
      "learning_rate": 4.833833271014571e-06,
      "loss": 0.3305,
      "num_input_tokens_seen": 12397088,
      "step": 18980
    },
    {
      "epoch": 11.193985849056604,
      "grad_norm": 5.6418304443359375,
      "learning_rate": 4.831262000131379e-06,
      "loss": 0.3069,
      "num_input_tokens_seen": 12399424,
      "step": 18985
    },
    {
      "epoch": 11.19693396226415,
      "grad_norm": 4.770077705383301,
      "learning_rate": 4.828690773922299e-06,
      "loss": 0.314,
      "num_input_tokens_seen": 12404736,
      "step": 18990
    },
    {
      "epoch": 11.199882075471699,
      "grad_norm": 2.0901732444763184,
      "learning_rate": 4.826119593068074e-06,
      "loss": 0.3493,
      "num_input_tokens_seen": 12408960,
      "step": 18995
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 1.654471755027771,
      "learning_rate": 4.8235484582494375e-06,
      "loss": 0.5402,
      "num_input_tokens_seen": 12412256,
      "step": 19000
    },
    {
      "epoch": 11.205778301886792,
      "grad_norm": 2.906632661819458,
      "learning_rate": 4.8209773701471076e-06,
      "loss": 0.4032,
      "num_input_tokens_seen": 12415136,
      "step": 19005
    },
    {
      "epoch": 11.20872641509434,
      "grad_norm": 5.0455403327941895,
      "learning_rate": 4.818406329441789e-06,
      "loss": 0.4383,
      "num_input_tokens_seen": 12417312,
      "step": 19010
    },
    {
      "epoch": 11.211674528301886,
      "grad_norm": 3.264449119567871,
      "learning_rate": 4.815835336814179e-06,
      "loss": 0.2906,
      "num_input_tokens_seen": 12420064,
      "step": 19015
    },
    {
      "epoch": 11.214622641509434,
      "grad_norm": 4.146888732910156,
      "learning_rate": 4.813264392944957e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 12423200,
      "step": 19020
    },
    {
      "epoch": 11.21757075471698,
      "grad_norm": 4.022525310516357,
      "learning_rate": 4.8106934985147905e-06,
      "loss": 0.4519,
      "num_input_tokens_seen": 12426528,
      "step": 19025
    },
    {
      "epoch": 11.220518867924529,
      "grad_norm": 2.0552022457122803,
      "learning_rate": 4.808122654204338e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 12429376,
      "step": 19030
    },
    {
      "epoch": 11.223466981132075,
      "grad_norm": 2.082984447479248,
      "learning_rate": 4.805551860694239e-06,
      "loss": 0.2862,
      "num_input_tokens_seen": 12432320,
      "step": 19035
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 2.4792544841766357,
      "learning_rate": 4.802981118665122e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 12435744,
      "step": 19040
    },
    {
      "epoch": 11.22936320754717,
      "grad_norm": 3.216367244720459,
      "learning_rate": 4.800410428797604e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 12439712,
      "step": 19045
    },
    {
      "epoch": 11.232311320754716,
      "grad_norm": 2.659485340118408,
      "learning_rate": 4.797839791772286e-06,
      "loss": 0.2293,
      "num_input_tokens_seen": 12443328,
      "step": 19050
    },
    {
      "epoch": 11.235259433962264,
      "grad_norm": 5.107675075531006,
      "learning_rate": 4.795269208269752e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 12446432,
      "step": 19055
    },
    {
      "epoch": 11.23820754716981,
      "grad_norm": 2.060307741165161,
      "learning_rate": 4.792698678970579e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 12449152,
      "step": 19060
    },
    {
      "epoch": 11.241155660377359,
      "grad_norm": 5.250658988952637,
      "learning_rate": 4.7901282045553245e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 12452480,
      "step": 19065
    },
    {
      "epoch": 11.244103773584905,
      "grad_norm": 2.7702136039733887,
      "learning_rate": 4.787557785704531e-06,
      "loss": 0.2804,
      "num_input_tokens_seen": 12456672,
      "step": 19070
    },
    {
      "epoch": 11.247051886792454,
      "grad_norm": 3.0735023021698,
      "learning_rate": 4.784987423098731e-06,
      "loss": 0.3515,
      "num_input_tokens_seen": 12460928,
      "step": 19075
    },
    {
      "epoch": 11.25,
      "grad_norm": 2.261589765548706,
      "learning_rate": 4.7824171174184354e-06,
      "loss": 0.3526,
      "num_input_tokens_seen": 12466592,
      "step": 19080
    },
    {
      "epoch": 11.252948113207546,
      "grad_norm": 3.120527744293213,
      "learning_rate": 4.779846869344146e-06,
      "loss": 0.3079,
      "num_input_tokens_seen": 12469632,
      "step": 19085
    },
    {
      "epoch": 11.255896226415095,
      "grad_norm": 3.361133575439453,
      "learning_rate": 4.777276679556346e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 12472960,
      "step": 19090
    },
    {
      "epoch": 11.258844339622641,
      "grad_norm": 3.0226731300354004,
      "learning_rate": 4.774706548735507e-06,
      "loss": 0.2772,
      "num_input_tokens_seen": 12476256,
      "step": 19095
    },
    {
      "epoch": 11.26179245283019,
      "grad_norm": 3.0785770416259766,
      "learning_rate": 4.77213647756208e-06,
      "loss": 0.3069,
      "num_input_tokens_seen": 12479328,
      "step": 19100
    },
    {
      "epoch": 11.264740566037736,
      "grad_norm": 2.519599437713623,
      "learning_rate": 4.769566466716501e-06,
      "loss": 0.4354,
      "num_input_tokens_seen": 12483104,
      "step": 19105
    },
    {
      "epoch": 11.267688679245284,
      "grad_norm": 2.848919153213501,
      "learning_rate": 4.766996516879195e-06,
      "loss": 0.3889,
      "num_input_tokens_seen": 12486272,
      "step": 19110
    },
    {
      "epoch": 11.27063679245283,
      "grad_norm": 3.8280861377716064,
      "learning_rate": 4.764426628730564e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 12489408,
      "step": 19115
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 3.5057053565979004,
      "learning_rate": 4.761856802950999e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 12492864,
      "step": 19120
    },
    {
      "epoch": 11.276533018867925,
      "grad_norm": 2.2636024951934814,
      "learning_rate": 4.759287040220872e-06,
      "loss": 0.2686,
      "num_input_tokens_seen": 12495936,
      "step": 19125
    },
    {
      "epoch": 11.279481132075471,
      "grad_norm": 4.037961959838867,
      "learning_rate": 4.756717341220538e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 12499552,
      "step": 19130
    },
    {
      "epoch": 11.28242924528302,
      "grad_norm": 4.6358513832092285,
      "learning_rate": 4.7541477066303365e-06,
      "loss": 0.3343,
      "num_input_tokens_seen": 12502624,
      "step": 19135
    },
    {
      "epoch": 11.285377358490566,
      "grad_norm": 3.334432363510132,
      "learning_rate": 4.75157813713059e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 12505408,
      "step": 19140
    },
    {
      "epoch": 11.288325471698114,
      "grad_norm": 3.137725353240967,
      "learning_rate": 4.7490086334016e-06,
      "loss": 0.3324,
      "num_input_tokens_seen": 12508576,
      "step": 19145
    },
    {
      "epoch": 11.29127358490566,
      "grad_norm": 2.839376211166382,
      "learning_rate": 4.746439196123659e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 12510944,
      "step": 19150
    },
    {
      "epoch": 11.294221698113208,
      "grad_norm": 1.9199877977371216,
      "learning_rate": 4.743869825977032e-06,
      "loss": 0.3677,
      "num_input_tokens_seen": 12513952,
      "step": 19155
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 2.1961135864257812,
      "learning_rate": 4.741300523641972e-06,
      "loss": 0.3649,
      "num_input_tokens_seen": 12521344,
      "step": 19160
    },
    {
      "epoch": 11.300117924528301,
      "grad_norm": 3.245575428009033,
      "learning_rate": 4.738731289798715e-06,
      "loss": 0.2986,
      "num_input_tokens_seen": 12524672,
      "step": 19165
    },
    {
      "epoch": 11.30306603773585,
      "grad_norm": 3.4407589435577393,
      "learning_rate": 4.7361621251274744e-06,
      "loss": 0.4089,
      "num_input_tokens_seen": 12528320,
      "step": 19170
    },
    {
      "epoch": 11.306014150943396,
      "grad_norm": 11.975054740905762,
      "learning_rate": 4.733593030308446e-06,
      "loss": 0.3371,
      "num_input_tokens_seen": 12531168,
      "step": 19175
    },
    {
      "epoch": 11.308962264150944,
      "grad_norm": 2.186952590942383,
      "learning_rate": 4.731024006021814e-06,
      "loss": 0.4,
      "num_input_tokens_seen": 12535296,
      "step": 19180
    },
    {
      "epoch": 11.31191037735849,
      "grad_norm": 3.096693277359009,
      "learning_rate": 4.728455052947732e-06,
      "loss": 0.3422,
      "num_input_tokens_seen": 12540256,
      "step": 19185
    },
    {
      "epoch": 11.314858490566039,
      "grad_norm": 2.0854339599609375,
      "learning_rate": 4.725886171766349e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 12543936,
      "step": 19190
    },
    {
      "epoch": 11.317806603773585,
      "grad_norm": 2.489769458770752,
      "learning_rate": 4.723317363157781e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 12546240,
      "step": 19195
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 2.3422746658325195,
      "learning_rate": 4.720748627802135e-06,
      "loss": 0.3414,
      "num_input_tokens_seen": 12549888,
      "step": 19200
    },
    {
      "epoch": 11.32370283018868,
      "grad_norm": 2.0614945888519287,
      "learning_rate": 4.718179966379492e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 12552896,
      "step": 19205
    },
    {
      "epoch": 11.326650943396226,
      "grad_norm": 3.04833722114563,
      "learning_rate": 4.715611379569919e-06,
      "loss": 0.2778,
      "num_input_tokens_seen": 12556576,
      "step": 19210
    },
    {
      "epoch": 11.329599056603774,
      "grad_norm": 2.4084465503692627,
      "learning_rate": 4.713042868053458e-06,
      "loss": 0.3843,
      "num_input_tokens_seen": 12560128,
      "step": 19215
    },
    {
      "epoch": 11.33254716981132,
      "grad_norm": 2.109391689300537,
      "learning_rate": 4.7104744325101345e-06,
      "loss": 0.4855,
      "num_input_tokens_seen": 12563008,
      "step": 19220
    },
    {
      "epoch": 11.335495283018869,
      "grad_norm": 3.1284005641937256,
      "learning_rate": 4.7079060736199525e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 12566688,
      "step": 19225
    },
    {
      "epoch": 11.338443396226415,
      "grad_norm": 3.7895541191101074,
      "learning_rate": 4.705337792062897e-06,
      "loss": 0.5731,
      "num_input_tokens_seen": 12569120,
      "step": 19230
    },
    {
      "epoch": 11.341391509433961,
      "grad_norm": 3.45467472076416,
      "learning_rate": 4.702769588518931e-06,
      "loss": 0.3121,
      "num_input_tokens_seen": 12575712,
      "step": 19235
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 6.743478298187256,
      "learning_rate": 4.700201463667996e-06,
      "loss": 0.3426,
      "num_input_tokens_seen": 12578656,
      "step": 19240
    },
    {
      "epoch": 11.347287735849056,
      "grad_norm": 3.7938477993011475,
      "learning_rate": 4.697633418190017e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 12582432,
      "step": 19245
    },
    {
      "epoch": 11.350235849056604,
      "grad_norm": 4.066020488739014,
      "learning_rate": 4.695065452764893e-06,
      "loss": 0.1931,
      "num_input_tokens_seen": 12586144,
      "step": 19250
    },
    {
      "epoch": 11.35318396226415,
      "grad_norm": 1.9117176532745361,
      "learning_rate": 4.692497568072505e-06,
      "loss": 0.3398,
      "num_input_tokens_seen": 12589632,
      "step": 19255
    },
    {
      "epoch": 11.356132075471699,
      "grad_norm": 2.040886640548706,
      "learning_rate": 4.689929764792711e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 12593984,
      "step": 19260
    },
    {
      "epoch": 11.359080188679245,
      "grad_norm": 2.146101474761963,
      "learning_rate": 4.687362043605349e-06,
      "loss": 0.4099,
      "num_input_tokens_seen": 12596800,
      "step": 19265
    },
    {
      "epoch": 11.362028301886792,
      "grad_norm": 1.7401902675628662,
      "learning_rate": 4.684794405190231e-06,
      "loss": 0.3916,
      "num_input_tokens_seen": 12600736,
      "step": 19270
    },
    {
      "epoch": 11.36497641509434,
      "grad_norm": 2.249783515930176,
      "learning_rate": 4.682226850227155e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 12604064,
      "step": 19275
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 2.4771080017089844,
      "learning_rate": 4.6796593793958875e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 12607584,
      "step": 19280
    },
    {
      "epoch": 11.370872641509434,
      "grad_norm": 5.536300182342529,
      "learning_rate": 4.6770919933761815e-06,
      "loss": 0.2999,
      "num_input_tokens_seen": 12610112,
      "step": 19285
    },
    {
      "epoch": 11.37382075471698,
      "grad_norm": 2.0653223991394043,
      "learning_rate": 4.6745246928477615e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 12614624,
      "step": 19290
    },
    {
      "epoch": 11.376768867924529,
      "grad_norm": 2.810990333557129,
      "learning_rate": 4.671957478490332e-06,
      "loss": 0.3676,
      "num_input_tokens_seen": 12617376,
      "step": 19295
    },
    {
      "epoch": 11.379716981132075,
      "grad_norm": 3.3452584743499756,
      "learning_rate": 4.669390350983574e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 12620736,
      "step": 19300
    },
    {
      "epoch": 11.382665094339623,
      "grad_norm": 3.6998629570007324,
      "learning_rate": 4.666823311007145e-06,
      "loss": 0.335,
      "num_input_tokens_seen": 12623776,
      "step": 19305
    },
    {
      "epoch": 11.38561320754717,
      "grad_norm": 2.3542568683624268,
      "learning_rate": 4.664256359240681e-06,
      "loss": 0.3878,
      "num_input_tokens_seen": 12627264,
      "step": 19310
    },
    {
      "epoch": 11.388561320754716,
      "grad_norm": 1.79622483253479,
      "learning_rate": 4.661689496363793e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 12631008,
      "step": 19315
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 3.9413645267486572,
      "learning_rate": 4.659122723056068e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 12634400,
      "step": 19320
    },
    {
      "epoch": 11.39445754716981,
      "grad_norm": 2.983200788497925,
      "learning_rate": 4.656556039997072e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 12637632,
      "step": 19325
    },
    {
      "epoch": 11.397405660377359,
      "grad_norm": 2.901372194290161,
      "learning_rate": 4.653989447866345e-06,
      "loss": 0.4548,
      "num_input_tokens_seen": 12640384,
      "step": 19330
    },
    {
      "epoch": 11.400353773584905,
      "grad_norm": 2.3919856548309326,
      "learning_rate": 4.651422947343401e-06,
      "loss": 0.3804,
      "num_input_tokens_seen": 12643040,
      "step": 19335
    },
    {
      "epoch": 11.403301886792454,
      "grad_norm": 2.917827606201172,
      "learning_rate": 4.6488565391077355e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 12645632,
      "step": 19340
    },
    {
      "epoch": 11.40625,
      "grad_norm": 2.710474967956543,
      "learning_rate": 4.646290223838815e-06,
      "loss": 0.4279,
      "num_input_tokens_seen": 12648832,
      "step": 19345
    },
    {
      "epoch": 11.409198113207546,
      "grad_norm": 3.9233789443969727,
      "learning_rate": 4.64372400221608e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 12651712,
      "step": 19350
    },
    {
      "epoch": 11.412146226415095,
      "grad_norm": 1.7593247890472412,
      "learning_rate": 4.641157874918953e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 12654496,
      "step": 19355
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 3.4506092071533203,
      "learning_rate": 4.6385918426268245e-06,
      "loss": 0.4062,
      "num_input_tokens_seen": 12659040,
      "step": 19360
    },
    {
      "epoch": 11.41804245283019,
      "grad_norm": 2.654039144515991,
      "learning_rate": 4.636025906019062e-06,
      "loss": 0.3707,
      "num_input_tokens_seen": 12662176,
      "step": 19365
    },
    {
      "epoch": 11.420990566037736,
      "grad_norm": 2.5804386138916016,
      "learning_rate": 4.6334600657750115e-06,
      "loss": 0.5156,
      "num_input_tokens_seen": 12664992,
      "step": 19370
    },
    {
      "epoch": 11.423938679245284,
      "grad_norm": 1.8582115173339844,
      "learning_rate": 4.6308943225739855e-06,
      "loss": 0.3758,
      "num_input_tokens_seen": 12668224,
      "step": 19375
    },
    {
      "epoch": 11.42688679245283,
      "grad_norm": 2.5318286418914795,
      "learning_rate": 4.628328677095281e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 12670592,
      "step": 19380
    },
    {
      "epoch": 11.429834905660377,
      "grad_norm": 2.782984495162964,
      "learning_rate": 4.625763130018159e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 12673344,
      "step": 19385
    },
    {
      "epoch": 11.432783018867925,
      "grad_norm": 3.8476438522338867,
      "learning_rate": 4.6231976820218635e-06,
      "loss": 0.3471,
      "num_input_tokens_seen": 12676704,
      "step": 19390
    },
    {
      "epoch": 11.435731132075471,
      "grad_norm": 1.9571491479873657,
      "learning_rate": 4.620632333785605e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 12679872,
      "step": 19395
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 3.494539260864258,
      "learning_rate": 4.618067085988569e-06,
      "loss": 0.3551,
      "num_input_tokens_seen": 12683200,
      "step": 19400
    },
    {
      "epoch": 11.441627358490566,
      "grad_norm": 3.3328499794006348,
      "learning_rate": 4.615501939309921e-06,
      "loss": 0.5507,
      "num_input_tokens_seen": 12685792,
      "step": 19405
    },
    {
      "epoch": 11.444575471698114,
      "grad_norm": 3.395228862762451,
      "learning_rate": 4.612936894428791e-06,
      "loss": 0.3556,
      "num_input_tokens_seen": 12689728,
      "step": 19410
    },
    {
      "epoch": 11.44752358490566,
      "grad_norm": 3.2028725147247314,
      "learning_rate": 4.610371952024285e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 12693056,
      "step": 19415
    },
    {
      "epoch": 11.450471698113208,
      "grad_norm": 4.085222244262695,
      "learning_rate": 4.607807112775485e-06,
      "loss": 0.4284,
      "num_input_tokens_seen": 12696160,
      "step": 19420
    },
    {
      "epoch": 11.453419811320755,
      "grad_norm": 2.4255104064941406,
      "learning_rate": 4.605242377361441e-06,
      "loss": 0.2618,
      "num_input_tokens_seen": 12698816,
      "step": 19425
    },
    {
      "epoch": 11.456367924528301,
      "grad_norm": 4.056238174438477,
      "learning_rate": 4.60267774646118e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 12702528,
      "step": 19430
    },
    {
      "epoch": 11.45931603773585,
      "grad_norm": 1.9658514261245728,
      "learning_rate": 4.600113220753698e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 12705568,
      "step": 19435
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 2.670004367828369,
      "learning_rate": 4.597548800917964e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 12708288,
      "step": 19440
    },
    {
      "epoch": 11.465212264150944,
      "grad_norm": 3.9363582134246826,
      "learning_rate": 4.594984487632919e-06,
      "loss": 0.499,
      "num_input_tokens_seen": 12711680,
      "step": 19445
    },
    {
      "epoch": 11.46816037735849,
      "grad_norm": 4.023828506469727,
      "learning_rate": 4.592420281577478e-06,
      "loss": 0.5202,
      "num_input_tokens_seen": 12714368,
      "step": 19450
    },
    {
      "epoch": 11.471108490566039,
      "grad_norm": 5.754276752471924,
      "learning_rate": 4.589856183430521e-06,
      "loss": 0.4488,
      "num_input_tokens_seen": 12718272,
      "step": 19455
    },
    {
      "epoch": 11.474056603773585,
      "grad_norm": 5.895110130310059,
      "learning_rate": 4.58729219387091e-06,
      "loss": 0.3508,
      "num_input_tokens_seen": 12722336,
      "step": 19460
    },
    {
      "epoch": 11.477004716981131,
      "grad_norm": 4.148027420043945,
      "learning_rate": 4.584728313577468e-06,
      "loss": 0.6101,
      "num_input_tokens_seen": 12724928,
      "step": 19465
    },
    {
      "epoch": 11.47995283018868,
      "grad_norm": 2.7084238529205322,
      "learning_rate": 4.582164543228993e-06,
      "loss": 0.3869,
      "num_input_tokens_seen": 12727296,
      "step": 19470
    },
    {
      "epoch": 11.482900943396226,
      "grad_norm": 2.168855667114258,
      "learning_rate": 4.5796008835042574e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 12730528,
      "step": 19475
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 2.821993350982666,
      "learning_rate": 4.577037335082e-06,
      "loss": 0.3807,
      "num_input_tokens_seen": 12733568,
      "step": 19480
    },
    {
      "epoch": 11.48879716981132,
      "grad_norm": 2.6510560512542725,
      "learning_rate": 4.574473898640933e-06,
      "loss": 0.2784,
      "num_input_tokens_seen": 12736576,
      "step": 19485
    },
    {
      "epoch": 11.491745283018869,
      "grad_norm": 2.324392557144165,
      "learning_rate": 4.571910574859732e-06,
      "loss": 0.3425,
      "num_input_tokens_seen": 12740160,
      "step": 19490
    },
    {
      "epoch": 11.494693396226415,
      "grad_norm": 2.2524232864379883,
      "learning_rate": 4.5693473644170535e-06,
      "loss": 0.2843,
      "num_input_tokens_seen": 12742880,
      "step": 19495
    },
    {
      "epoch": 11.497641509433961,
      "grad_norm": 4.198735237121582,
      "learning_rate": 4.566784267991516e-06,
      "loss": 0.4119,
      "num_input_tokens_seen": 12746656,
      "step": 19500
    },
    {
      "epoch": 11.50058962264151,
      "grad_norm": 4.64422607421875,
      "learning_rate": 4.564221286261709e-06,
      "loss": 0.3071,
      "num_input_tokens_seen": 12749088,
      "step": 19505
    },
    {
      "epoch": 11.503537735849056,
      "grad_norm": 2.1209192276000977,
      "learning_rate": 4.5616584199061964e-06,
      "loss": 0.4455,
      "num_input_tokens_seen": 12752736,
      "step": 19510
    },
    {
      "epoch": 11.506485849056604,
      "grad_norm": 1.8846454620361328,
      "learning_rate": 4.559095669603506e-06,
      "loss": 0.4204,
      "num_input_tokens_seen": 12755808,
      "step": 19515
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 3.646672248840332,
      "learning_rate": 4.556533036032136e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 12758976,
      "step": 19520
    },
    {
      "epoch": 11.512382075471699,
      "grad_norm": 3.5801799297332764,
      "learning_rate": 4.553970519870557e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 12761824,
      "step": 19525
    },
    {
      "epoch": 11.515330188679245,
      "grad_norm": 2.0536041259765625,
      "learning_rate": 4.551408121797205e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 12764736,
      "step": 19530
    },
    {
      "epoch": 11.518278301886792,
      "grad_norm": 4.442261219024658,
      "learning_rate": 4.548845842490486e-06,
      "loss": 0.3319,
      "num_input_tokens_seen": 12767552,
      "step": 19535
    },
    {
      "epoch": 11.52122641509434,
      "grad_norm": 3.1297552585601807,
      "learning_rate": 4.5462836826287745e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 12770656,
      "step": 19540
    },
    {
      "epoch": 11.524174528301886,
      "grad_norm": 2.2051851749420166,
      "learning_rate": 4.543721642890414e-06,
      "loss": 0.3194,
      "num_input_tokens_seen": 12773280,
      "step": 19545
    },
    {
      "epoch": 11.527122641509434,
      "grad_norm": 1.3019304275512695,
      "learning_rate": 4.541159723953714e-06,
      "loss": 0.5599,
      "num_input_tokens_seen": 12777120,
      "step": 19550
    },
    {
      "epoch": 11.53007075471698,
      "grad_norm": 3.7880823612213135,
      "learning_rate": 4.538597926496955e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 12781248,
      "step": 19555
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 2.672839641571045,
      "learning_rate": 4.536036251198384e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 12784384,
      "step": 19560
    },
    {
      "epoch": 11.535966981132075,
      "grad_norm": 3.640153408050537,
      "learning_rate": 4.5334746987362124e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 12786880,
      "step": 19565
    },
    {
      "epoch": 11.538915094339622,
      "grad_norm": 3.2116458415985107,
      "learning_rate": 4.530913269788627e-06,
      "loss": 0.4056,
      "num_input_tokens_seen": 12790272,
      "step": 19570
    },
    {
      "epoch": 11.54186320754717,
      "grad_norm": 2.736027240753174,
      "learning_rate": 4.528351965033775e-06,
      "loss": 0.3894,
      "num_input_tokens_seen": 12793184,
      "step": 19575
    },
    {
      "epoch": 11.544811320754716,
      "grad_norm": 4.28379487991333,
      "learning_rate": 4.525790785149774e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 12796384,
      "step": 19580
    },
    {
      "epoch": 11.547759433962264,
      "grad_norm": 1.4380395412445068,
      "learning_rate": 4.523229730814705e-06,
      "loss": 0.4525,
      "num_input_tokens_seen": 12800064,
      "step": 19585
    },
    {
      "epoch": 11.55070754716981,
      "grad_norm": 2.0794293880462646,
      "learning_rate": 4.52066880270662e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 12803520,
      "step": 19590
    },
    {
      "epoch": 11.553655660377359,
      "grad_norm": 4.123380184173584,
      "learning_rate": 4.518108001503536e-06,
      "loss": 0.3774,
      "num_input_tokens_seen": 12806304,
      "step": 19595
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 2.7307510375976562,
      "learning_rate": 4.515547327883434e-06,
      "loss": 0.3563,
      "num_input_tokens_seen": 12809728,
      "step": 19600
    },
    {
      "epoch": 11.559551886792454,
      "grad_norm": 3.443293333053589,
      "learning_rate": 4.512986782524266e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 12813024,
      "step": 19605
    },
    {
      "epoch": 11.5625,
      "grad_norm": 2.774116039276123,
      "learning_rate": 4.510426366103946e-06,
      "loss": 0.4777,
      "num_input_tokens_seen": 12816000,
      "step": 19610
    },
    {
      "epoch": 11.565448113207546,
      "grad_norm": 3.624239444732666,
      "learning_rate": 4.5078660793003544e-06,
      "loss": 0.3255,
      "num_input_tokens_seen": 12818240,
      "step": 19615
    },
    {
      "epoch": 11.568396226415095,
      "grad_norm": 1.6827666759490967,
      "learning_rate": 4.505305922791341e-06,
      "loss": 0.3216,
      "num_input_tokens_seen": 12822272,
      "step": 19620
    },
    {
      "epoch": 11.571344339622641,
      "grad_norm": 1.5956286191940308,
      "learning_rate": 4.502745897254716e-06,
      "loss": 0.3007,
      "num_input_tokens_seen": 12825792,
      "step": 19625
    },
    {
      "epoch": 11.57429245283019,
      "grad_norm": 5.229098320007324,
      "learning_rate": 4.500186003368257e-06,
      "loss": 0.3627,
      "num_input_tokens_seen": 12828704,
      "step": 19630
    },
    {
      "epoch": 11.577240566037736,
      "grad_norm": 4.724757194519043,
      "learning_rate": 4.497626241809709e-06,
      "loss": 0.494,
      "num_input_tokens_seen": 12832032,
      "step": 19635
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 2.5149569511413574,
      "learning_rate": 4.4950666132567775e-06,
      "loss": 0.3158,
      "num_input_tokens_seen": 12835200,
      "step": 19640
    },
    {
      "epoch": 11.58313679245283,
      "grad_norm": 2.945343017578125,
      "learning_rate": 4.492507118387136e-06,
      "loss": 0.4259,
      "num_input_tokens_seen": 12838784,
      "step": 19645
    },
    {
      "epoch": 11.586084905660378,
      "grad_norm": 4.359920978546143,
      "learning_rate": 4.4899477578784235e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 12841536,
      "step": 19650
    },
    {
      "epoch": 11.589033018867925,
      "grad_norm": 3.5148541927337646,
      "learning_rate": 4.487388532408239e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 12845696,
      "step": 19655
    },
    {
      "epoch": 11.591981132075471,
      "grad_norm": 3.201462745666504,
      "learning_rate": 4.48482944265415e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 12848512,
      "step": 19660
    },
    {
      "epoch": 11.59492924528302,
      "grad_norm": 2.495408535003662,
      "learning_rate": 4.482270489293685e-06,
      "loss": 0.3145,
      "num_input_tokens_seen": 12851008,
      "step": 19665
    },
    {
      "epoch": 11.597877358490566,
      "grad_norm": 4.492279052734375,
      "learning_rate": 4.479711673004341e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 12854976,
      "step": 19670
    },
    {
      "epoch": 11.600825471698114,
      "grad_norm": 2.877711534500122,
      "learning_rate": 4.477152994463575e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 12858208,
      "step": 19675
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 2.4451868534088135,
      "learning_rate": 4.474594454348805e-06,
      "loss": 0.3444,
      "num_input_tokens_seen": 12861024,
      "step": 19680
    },
    {
      "epoch": 11.606721698113208,
      "grad_norm": 3.511920690536499,
      "learning_rate": 4.472036053337419e-06,
      "loss": 0.4039,
      "num_input_tokens_seen": 12863840,
      "step": 19685
    },
    {
      "epoch": 11.609669811320755,
      "grad_norm": 3.4084725379943848,
      "learning_rate": 4.469477792106761e-06,
      "loss": 0.3385,
      "num_input_tokens_seen": 12866560,
      "step": 19690
    },
    {
      "epoch": 11.612617924528301,
      "grad_norm": 5.191800117492676,
      "learning_rate": 4.466919671334146e-06,
      "loss": 0.3304,
      "num_input_tokens_seen": 12870400,
      "step": 19695
    },
    {
      "epoch": 11.61556603773585,
      "grad_norm": 3.8291516304016113,
      "learning_rate": 4.4643616916968456e-06,
      "loss": 0.4265,
      "num_input_tokens_seen": 12872992,
      "step": 19700
    },
    {
      "epoch": 11.618514150943396,
      "grad_norm": 4.297688961029053,
      "learning_rate": 4.461803853872095e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 12876224,
      "step": 19705
    },
    {
      "epoch": 11.621462264150944,
      "grad_norm": 2.2558701038360596,
      "learning_rate": 4.459246158537094e-06,
      "loss": 0.2329,
      "num_input_tokens_seen": 12880800,
      "step": 19710
    },
    {
      "epoch": 11.62441037735849,
      "grad_norm": 2.0393576622009277,
      "learning_rate": 4.456688606369003e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 12883776,
      "step": 19715
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 3.4001266956329346,
      "learning_rate": 4.4541311980449446e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 12887552,
      "step": 19720
    },
    {
      "epoch": 11.630306603773585,
      "grad_norm": 4.881402015686035,
      "learning_rate": 4.4515739342420045e-06,
      "loss": 0.3418,
      "num_input_tokens_seen": 12890016,
      "step": 19725
    },
    {
      "epoch": 11.633254716981131,
      "grad_norm": 3.279690980911255,
      "learning_rate": 4.4490168156372294e-06,
      "loss": 0.3791,
      "num_input_tokens_seen": 12893344,
      "step": 19730
    },
    {
      "epoch": 11.63620283018868,
      "grad_norm": 3.0450403690338135,
      "learning_rate": 4.446459842907626e-06,
      "loss": 0.3136,
      "num_input_tokens_seen": 12897344,
      "step": 19735
    },
    {
      "epoch": 11.639150943396226,
      "grad_norm": 2.718522310256958,
      "learning_rate": 4.443903016730165e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 12899712,
      "step": 19740
    },
    {
      "epoch": 11.642099056603774,
      "grad_norm": 3.835625648498535,
      "learning_rate": 4.4413463377817775e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 12903072,
      "step": 19745
    },
    {
      "epoch": 11.64504716981132,
      "grad_norm": 2.8474748134613037,
      "learning_rate": 4.438789806739353e-06,
      "loss": 0.2533,
      "num_input_tokens_seen": 12906080,
      "step": 19750
    },
    {
      "epoch": 11.647995283018869,
      "grad_norm": 2.2236063480377197,
      "learning_rate": 4.436233424279746e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 12909856,
      "step": 19755
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 3.613675355911255,
      "learning_rate": 4.433677191079771e-06,
      "loss": 0.3432,
      "num_input_tokens_seen": 12912864,
      "step": 19760
    },
    {
      "epoch": 11.653891509433961,
      "grad_norm": 4.901577949523926,
      "learning_rate": 4.431121107816201e-06,
      "loss": 0.3223,
      "num_input_tokens_seen": 12915488,
      "step": 19765
    },
    {
      "epoch": 11.65683962264151,
      "grad_norm": 2.188823938369751,
      "learning_rate": 4.4285651751657676e-06,
      "loss": 0.2682,
      "num_input_tokens_seen": 12918432,
      "step": 19770
    },
    {
      "epoch": 11.659787735849056,
      "grad_norm": 3.2545464038848877,
      "learning_rate": 4.4260093938051685e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 12921312,
      "step": 19775
    },
    {
      "epoch": 11.662735849056604,
      "grad_norm": 4.343674182891846,
      "learning_rate": 4.423453764411056e-06,
      "loss": 0.2302,
      "num_input_tokens_seen": 12924832,
      "step": 19780
    },
    {
      "epoch": 11.66568396226415,
      "grad_norm": 1.9771088361740112,
      "learning_rate": 4.4208982876600425e-06,
      "loss": 0.2796,
      "num_input_tokens_seen": 12928032,
      "step": 19785
    },
    {
      "epoch": 11.668632075471699,
      "grad_norm": 3.6045823097229004,
      "learning_rate": 4.418342964228705e-06,
      "loss": 0.39,
      "num_input_tokens_seen": 12930912,
      "step": 19790
    },
    {
      "epoch": 11.671580188679245,
      "grad_norm": 11.580513000488281,
      "learning_rate": 4.415787794793574e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 12933440,
      "step": 19795
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 4.959413051605225,
      "learning_rate": 4.4132327800311414e-06,
      "loss": 0.3127,
      "num_input_tokens_seen": 12936352,
      "step": 19800
    },
    {
      "epoch": 11.67747641509434,
      "grad_norm": 4.020618438720703,
      "learning_rate": 4.41067792061786e-06,
      "loss": 0.3051,
      "num_input_tokens_seen": 12940224,
      "step": 19805
    },
    {
      "epoch": 11.680424528301886,
      "grad_norm": 4.4778852462768555,
      "learning_rate": 4.408123217230139e-06,
      "loss": 0.2825,
      "num_input_tokens_seen": 12942784,
      "step": 19810
    },
    {
      "epoch": 11.683372641509434,
      "grad_norm": 2.738088369369507,
      "learning_rate": 4.4055686705443464e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 12946080,
      "step": 19815
    },
    {
      "epoch": 11.68632075471698,
      "grad_norm": 4.104277610778809,
      "learning_rate": 4.4030142812368106e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 12949088,
      "step": 19820
    },
    {
      "epoch": 11.689268867924529,
      "grad_norm": 5.421760082244873,
      "learning_rate": 4.400460049983817e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 12952352,
      "step": 19825
    },
    {
      "epoch": 11.692216981132075,
      "grad_norm": 2.43270206451416,
      "learning_rate": 4.397905977461608e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 12955200,
      "step": 19830
    },
    {
      "epoch": 11.695165094339622,
      "grad_norm": 2.16408634185791,
      "learning_rate": 4.395352064346387e-06,
      "loss": 0.3118,
      "num_input_tokens_seen": 12959200,
      "step": 19835
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 4.651562213897705,
      "learning_rate": 4.392798311314314e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 12962816,
      "step": 19840
    },
    {
      "epoch": 11.701061320754716,
      "grad_norm": 5.96450662612915,
      "learning_rate": 4.390244719041502e-06,
      "loss": 0.305,
      "num_input_tokens_seen": 12966496,
      "step": 19845
    },
    {
      "epoch": 11.704009433962264,
      "grad_norm": 2.8367936611175537,
      "learning_rate": 4.387691288204029e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 12969408,
      "step": 19850
    },
    {
      "epoch": 11.70695754716981,
      "grad_norm": 3.019251585006714,
      "learning_rate": 4.385138019477928e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 12972352,
      "step": 19855
    },
    {
      "epoch": 11.709905660377359,
      "grad_norm": 3.4007418155670166,
      "learning_rate": 4.3825849135391864e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 12975104,
      "step": 19860
    },
    {
      "epoch": 11.712853773584905,
      "grad_norm": 3.8827619552612305,
      "learning_rate": 4.380031971063748e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 12978400,
      "step": 19865
    },
    {
      "epoch": 11.715801886792454,
      "grad_norm": 5.611135005950928,
      "learning_rate": 4.37747919272752e-06,
      "loss": 0.3349,
      "num_input_tokens_seen": 12982720,
      "step": 19870
    },
    {
      "epoch": 11.71875,
      "grad_norm": 3.560457468032837,
      "learning_rate": 4.374926579206357e-06,
      "loss": 0.3886,
      "num_input_tokens_seen": 12985952,
      "step": 19875
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 2.2686843872070312,
      "learning_rate": 4.372374131176075e-06,
      "loss": 0.3037,
      "num_input_tokens_seen": 12989600,
      "step": 19880
    },
    {
      "epoch": 11.724646226415095,
      "grad_norm": 2.819779634475708,
      "learning_rate": 4.369821849312449e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 12992960,
      "step": 19885
    },
    {
      "epoch": 11.727594339622641,
      "grad_norm": 2.166043281555176,
      "learning_rate": 4.367269734291203e-06,
      "loss": 0.3345,
      "num_input_tokens_seen": 12996224,
      "step": 19890
    },
    {
      "epoch": 11.73054245283019,
      "grad_norm": 3.522712230682373,
      "learning_rate": 4.364717786788022e-06,
      "loss": 0.3426,
      "num_input_tokens_seen": 12999200,
      "step": 19895
    },
    {
      "epoch": 11.733490566037736,
      "grad_norm": 3.254236936569214,
      "learning_rate": 4.362166007478545e-06,
      "loss": 0.3335,
      "num_input_tokens_seen": 13002496,
      "step": 19900
    },
    {
      "epoch": 11.736438679245284,
      "grad_norm": 2.5725560188293457,
      "learning_rate": 4.3596143970383665e-06,
      "loss": 0.3673,
      "num_input_tokens_seen": 13005760,
      "step": 19905
    },
    {
      "epoch": 11.73938679245283,
      "grad_norm": 8.310726165771484,
      "learning_rate": 4.357062956143035e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 13010016,
      "step": 19910
    },
    {
      "epoch": 11.742334905660378,
      "grad_norm": 2.6428146362304688,
      "learning_rate": 4.354511685468059e-06,
      "loss": 0.411,
      "num_input_tokens_seen": 13013760,
      "step": 19915
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 3.390658378601074,
      "learning_rate": 4.351960585688894e-06,
      "loss": 0.4495,
      "num_input_tokens_seen": 13017824,
      "step": 19920
    },
    {
      "epoch": 11.748231132075471,
      "grad_norm": 2.9086384773254395,
      "learning_rate": 4.349409657480959e-06,
      "loss": 0.3723,
      "num_input_tokens_seen": 13020736,
      "step": 19925
    },
    {
      "epoch": 11.75117924528302,
      "grad_norm": 3.1141345500946045,
      "learning_rate": 4.34685890151962e-06,
      "loss": 0.32,
      "num_input_tokens_seen": 13024192,
      "step": 19930
    },
    {
      "epoch": 11.754127358490566,
      "grad_norm": 3.504436492919922,
      "learning_rate": 4.344308318480201e-06,
      "loss": 0.5421,
      "num_input_tokens_seen": 13026592,
      "step": 19935
    },
    {
      "epoch": 11.757075471698114,
      "grad_norm": 3.108581781387329,
      "learning_rate": 4.341757909037981e-06,
      "loss": 0.2954,
      "num_input_tokens_seen": 13029120,
      "step": 19940
    },
    {
      "epoch": 11.76002358490566,
      "grad_norm": 2.399872303009033,
      "learning_rate": 4.33920767386819e-06,
      "loss": 0.303,
      "num_input_tokens_seen": 13032512,
      "step": 19945
    },
    {
      "epoch": 11.762971698113208,
      "grad_norm": 2.4232728481292725,
      "learning_rate": 4.336657613646017e-06,
      "loss": 0.3917,
      "num_input_tokens_seen": 13035552,
      "step": 19950
    },
    {
      "epoch": 11.765919811320755,
      "grad_norm": 3.1901907920837402,
      "learning_rate": 4.3341077290466e-06,
      "loss": 0.3741,
      "num_input_tokens_seen": 13038624,
      "step": 19955
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 2.363543748855591,
      "learning_rate": 4.331558020745031e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 13041280,
      "step": 19960
    },
    {
      "epoch": 11.77181603773585,
      "grad_norm": 2.732208490371704,
      "learning_rate": 4.329008489416357e-06,
      "loss": 0.2819,
      "num_input_tokens_seen": 13044224,
      "step": 19965
    },
    {
      "epoch": 11.774764150943396,
      "grad_norm": 2.97947096824646,
      "learning_rate": 4.326459135735576e-06,
      "loss": 0.3809,
      "num_input_tokens_seen": 13046944,
      "step": 19970
    },
    {
      "epoch": 11.777712264150944,
      "grad_norm": 4.316673278808594,
      "learning_rate": 4.323909960377644e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 13049856,
      "step": 19975
    },
    {
      "epoch": 11.78066037735849,
      "grad_norm": 8.164325714111328,
      "learning_rate": 4.3213609640174625e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 13054080,
      "step": 19980
    },
    {
      "epoch": 11.783608490566039,
      "grad_norm": 2.7751338481903076,
      "learning_rate": 4.318812147329889e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 13056992,
      "step": 19985
    },
    {
      "epoch": 11.786556603773585,
      "grad_norm": 2.8392651081085205,
      "learning_rate": 4.316263510989737e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 13060512,
      "step": 19990
    },
    {
      "epoch": 11.789504716981131,
      "grad_norm": 4.087166786193848,
      "learning_rate": 4.313715055671768e-06,
      "loss": 0.3838,
      "num_input_tokens_seen": 13062816,
      "step": 19995
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 4.625625133514404,
      "learning_rate": 4.311166782050694e-06,
      "loss": 0.391,
      "num_input_tokens_seen": 13066080,
      "step": 20000
    },
    {
      "epoch": 11.795400943396226,
      "grad_norm": 2.522242784500122,
      "learning_rate": 4.308618690801184e-06,
      "loss": 0.3218,
      "num_input_tokens_seen": 13071040,
      "step": 20005
    },
    {
      "epoch": 11.798349056603774,
      "grad_norm": 2.150648832321167,
      "learning_rate": 4.3060707825978564e-06,
      "loss": 0.2983,
      "num_input_tokens_seen": 13073760,
      "step": 20010
    },
    {
      "epoch": 11.80129716981132,
      "grad_norm": 3.8908722400665283,
      "learning_rate": 4.303523058115278e-06,
      "loss": 0.4907,
      "num_input_tokens_seen": 13076736,
      "step": 20015
    },
    {
      "epoch": 11.804245283018869,
      "grad_norm": 7.16663932800293,
      "learning_rate": 4.300975518027972e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 13080000,
      "step": 20020
    },
    {
      "epoch": 11.807193396226415,
      "grad_norm": 2.136683702468872,
      "learning_rate": 4.298428163010411e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 13083808,
      "step": 20025
    },
    {
      "epoch": 11.810141509433961,
      "grad_norm": 1.5894784927368164,
      "learning_rate": 4.295880993737016e-06,
      "loss": 0.3343,
      "num_input_tokens_seen": 13087040,
      "step": 20030
    },
    {
      "epoch": 11.81308962264151,
      "grad_norm": 4.039005756378174,
      "learning_rate": 4.293334010882164e-06,
      "loss": 0.2976,
      "num_input_tokens_seen": 13089056,
      "step": 20035
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 4.293148994445801,
      "learning_rate": 4.290787215120178e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 13091776,
      "step": 20040
    },
    {
      "epoch": 11.818985849056604,
      "grad_norm": 2.4038197994232178,
      "learning_rate": 4.2882406071253315e-06,
      "loss": 0.3442,
      "num_input_tokens_seen": 13094848,
      "step": 20045
    },
    {
      "epoch": 11.82193396226415,
      "grad_norm": 3.7143845558166504,
      "learning_rate": 4.285694187571852e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 13097632,
      "step": 20050
    },
    {
      "epoch": 11.824882075471699,
      "grad_norm": 3.118560791015625,
      "learning_rate": 4.283147957133917e-06,
      "loss": 0.3417,
      "num_input_tokens_seen": 13100320,
      "step": 20055
    },
    {
      "epoch": 11.827830188679245,
      "grad_norm": 4.0494065284729,
      "learning_rate": 4.280601916485648e-06,
      "loss": 0.4076,
      "num_input_tokens_seen": 13103680,
      "step": 20060
    },
    {
      "epoch": 11.830778301886792,
      "grad_norm": 2.045502185821533,
      "learning_rate": 4.278056066301123e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 13106912,
      "step": 20065
    },
    {
      "epoch": 11.83372641509434,
      "grad_norm": 2.190617322921753,
      "learning_rate": 4.275510407254366e-06,
      "loss": 0.2423,
      "num_input_tokens_seen": 13110048,
      "step": 20070
    },
    {
      "epoch": 11.836674528301886,
      "grad_norm": 1.8312909603118896,
      "learning_rate": 4.27296494001935e-06,
      "loss": 0.2791,
      "num_input_tokens_seen": 13113568,
      "step": 20075
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 3.1066839694976807,
      "learning_rate": 4.2704196652700005e-06,
      "loss": 0.3916,
      "num_input_tokens_seen": 13117312,
      "step": 20080
    },
    {
      "epoch": 11.84257075471698,
      "grad_norm": 2.279017686843872,
      "learning_rate": 4.2678745836801895e-06,
      "loss": 0.3814,
      "num_input_tokens_seen": 13120512,
      "step": 20085
    },
    {
      "epoch": 11.845518867924529,
      "grad_norm": 1.9354933500289917,
      "learning_rate": 4.265329695923736e-06,
      "loss": 0.3863,
      "num_input_tokens_seen": 13124192,
      "step": 20090
    },
    {
      "epoch": 11.848466981132075,
      "grad_norm": 2.4930012226104736,
      "learning_rate": 4.262785002674412e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 13127200,
      "step": 20095
    },
    {
      "epoch": 11.851415094339622,
      "grad_norm": 3.1927742958068848,
      "learning_rate": 4.260240504605938e-06,
      "loss": 0.3886,
      "num_input_tokens_seen": 13130336,
      "step": 20100
    },
    {
      "epoch": 11.85436320754717,
      "grad_norm": 4.484797954559326,
      "learning_rate": 4.2576962023919775e-06,
      "loss": 0.3219,
      "num_input_tokens_seen": 13133312,
      "step": 20105
    },
    {
      "epoch": 11.857311320754716,
      "grad_norm": 3.507286548614502,
      "learning_rate": 4.255152096706145e-06,
      "loss": 0.3567,
      "num_input_tokens_seen": 13136736,
      "step": 20110
    },
    {
      "epoch": 11.860259433962264,
      "grad_norm": 3.785550594329834,
      "learning_rate": 4.252608188222007e-06,
      "loss": 0.4474,
      "num_input_tokens_seen": 13139840,
      "step": 20115
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 2.161871910095215,
      "learning_rate": 4.250064477613071e-06,
      "loss": 0.426,
      "num_input_tokens_seen": 13143232,
      "step": 20120
    },
    {
      "epoch": 11.866155660377359,
      "grad_norm": 3.2125909328460693,
      "learning_rate": 4.247520965552797e-06,
      "loss": 0.2909,
      "num_input_tokens_seen": 13145856,
      "step": 20125
    },
    {
      "epoch": 11.869103773584905,
      "grad_norm": 3.93833065032959,
      "learning_rate": 4.24497765271459e-06,
      "loss": 0.3159,
      "num_input_tokens_seen": 13149280,
      "step": 20130
    },
    {
      "epoch": 11.872051886792454,
      "grad_norm": 4.176203727722168,
      "learning_rate": 4.242434539771804e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 13151488,
      "step": 20135
    },
    {
      "epoch": 11.875,
      "grad_norm": 4.075594902038574,
      "learning_rate": 4.239891627397737e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 13155200,
      "step": 20140
    },
    {
      "epoch": 11.877948113207546,
      "grad_norm": 1.6404215097427368,
      "learning_rate": 4.237348916265638e-06,
      "loss": 0.3821,
      "num_input_tokens_seen": 13159008,
      "step": 20145
    },
    {
      "epoch": 11.880896226415095,
      "grad_norm": 2.389227867126465,
      "learning_rate": 4.234806407048702e-06,
      "loss": 0.5086,
      "num_input_tokens_seen": 13161728,
      "step": 20150
    },
    {
      "epoch": 11.883844339622641,
      "grad_norm": 2.9551029205322266,
      "learning_rate": 4.232264100420066e-06,
      "loss": 0.4263,
      "num_input_tokens_seen": 13164928,
      "step": 20155
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 2.9525225162506104,
      "learning_rate": 4.229721997052819e-06,
      "loss": 0.4446,
      "num_input_tokens_seen": 13168256,
      "step": 20160
    },
    {
      "epoch": 11.889740566037736,
      "grad_norm": 3.969352960586548,
      "learning_rate": 4.227180097619993e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 13171328,
      "step": 20165
    },
    {
      "epoch": 11.892688679245284,
      "grad_norm": 2.8390090465545654,
      "learning_rate": 4.224638402794566e-06,
      "loss": 0.2958,
      "num_input_tokens_seen": 13174528,
      "step": 20170
    },
    {
      "epoch": 11.89563679245283,
      "grad_norm": 1.8571231365203857,
      "learning_rate": 4.222096913249464e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 13177440,
      "step": 20175
    },
    {
      "epoch": 11.898584905660378,
      "grad_norm": 1.4821473360061646,
      "learning_rate": 4.219555629657558e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 13181152,
      "step": 20180
    },
    {
      "epoch": 11.901533018867925,
      "grad_norm": 2.376145124435425,
      "learning_rate": 4.21701455269166e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 13184256,
      "step": 20185
    },
    {
      "epoch": 11.904481132075471,
      "grad_norm": 2.1521339416503906,
      "learning_rate": 4.2144736830245356e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 13187712,
      "step": 20190
    },
    {
      "epoch": 11.90742924528302,
      "grad_norm": 3.1701760292053223,
      "learning_rate": 4.211933021328889e-06,
      "loss": 0.288,
      "num_input_tokens_seen": 13190528,
      "step": 20195
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 3.694705009460449,
      "learning_rate": 4.20939256827737e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 13193056,
      "step": 20200
    },
    {
      "epoch": 11.913325471698114,
      "grad_norm": 3.256448745727539,
      "learning_rate": 4.206852324542578e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 13196800,
      "step": 20205
    },
    {
      "epoch": 11.91627358490566,
      "grad_norm": 2.2158236503601074,
      "learning_rate": 4.2043122907970496e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 13201152,
      "step": 20210
    },
    {
      "epoch": 11.919221698113208,
      "grad_norm": 5.059680938720703,
      "learning_rate": 4.2017724677132715e-06,
      "loss": 0.2979,
      "num_input_tokens_seen": 13204192,
      "step": 20215
    },
    {
      "epoch": 11.922169811320755,
      "grad_norm": 1.7997976541519165,
      "learning_rate": 4.1992328559636734e-06,
      "loss": 0.2924,
      "num_input_tokens_seen": 13207232,
      "step": 20220
    },
    {
      "epoch": 11.925117924528301,
      "grad_norm": 4.4880218505859375,
      "learning_rate": 4.196693456220628e-06,
      "loss": 0.4304,
      "num_input_tokens_seen": 13210432,
      "step": 20225
    },
    {
      "epoch": 11.92806603773585,
      "grad_norm": 2.9171552658081055,
      "learning_rate": 4.194154269156452e-06,
      "loss": 0.3938,
      "num_input_tokens_seen": 13213536,
      "step": 20230
    },
    {
      "epoch": 11.931014150943396,
      "grad_norm": 2.3881304264068604,
      "learning_rate": 4.191615295443404e-06,
      "loss": 0.3694,
      "num_input_tokens_seen": 13216672,
      "step": 20235
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 2.420511484146118,
      "learning_rate": 4.189076535753692e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 13219552,
      "step": 20240
    },
    {
      "epoch": 11.93691037735849,
      "grad_norm": 1.6572902202606201,
      "learning_rate": 4.186537990759464e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 13223520,
      "step": 20245
    },
    {
      "epoch": 11.939858490566039,
      "grad_norm": 3.220273017883301,
      "learning_rate": 4.183999661132806e-06,
      "loss": 0.4379,
      "num_input_tokens_seen": 13226432,
      "step": 20250
    },
    {
      "epoch": 11.942806603773585,
      "grad_norm": 2.45338773727417,
      "learning_rate": 4.181461547545756e-06,
      "loss": 0.4421,
      "num_input_tokens_seen": 13230528,
      "step": 20255
    },
    {
      "epoch": 11.945754716981131,
      "grad_norm": 3.3699920177459717,
      "learning_rate": 4.178923650670289e-06,
      "loss": 0.2956,
      "num_input_tokens_seen": 13233248,
      "step": 20260
    },
    {
      "epoch": 11.94870283018868,
      "grad_norm": 2.404857635498047,
      "learning_rate": 4.176385971178324e-06,
      "loss": 0.3106,
      "num_input_tokens_seen": 13236800,
      "step": 20265
    },
    {
      "epoch": 11.951650943396226,
      "grad_norm": 2.115370750427246,
      "learning_rate": 4.1738485097417225e-06,
      "loss": 0.2573,
      "num_input_tokens_seen": 13239488,
      "step": 20270
    },
    {
      "epoch": 11.954599056603774,
      "grad_norm": 4.094048023223877,
      "learning_rate": 4.1713112670322886e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 13242784,
      "step": 20275
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 1.0294504165649414,
      "learning_rate": 4.168774243721768e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 13245696,
      "step": 20280
    },
    {
      "epoch": 11.960495283018869,
      "grad_norm": 3.927948236465454,
      "learning_rate": 4.166237440481849e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 13248928,
      "step": 20285
    },
    {
      "epoch": 11.963443396226415,
      "grad_norm": 3.8922832012176514,
      "learning_rate": 4.163700857984162e-06,
      "loss": 0.3486,
      "num_input_tokens_seen": 13251840,
      "step": 20290
    },
    {
      "epoch": 11.966391509433961,
      "grad_norm": 3.368195056915283,
      "learning_rate": 4.161164496900275e-06,
      "loss": 0.3388,
      "num_input_tokens_seen": 13255584,
      "step": 20295
    },
    {
      "epoch": 11.96933962264151,
      "grad_norm": 5.870173454284668,
      "learning_rate": 4.1586283579017036e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 13258784,
      "step": 20300
    },
    {
      "epoch": 11.972287735849056,
      "grad_norm": 4.158247470855713,
      "learning_rate": 4.156092441659901e-06,
      "loss": 0.2997,
      "num_input_tokens_seen": 13261440,
      "step": 20305
    },
    {
      "epoch": 11.975235849056604,
      "grad_norm": 8.09614372253418,
      "learning_rate": 4.153556748846261e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 13267968,
      "step": 20310
    },
    {
      "epoch": 11.97818396226415,
      "grad_norm": 3.2984607219696045,
      "learning_rate": 4.15102128013212e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 13272032,
      "step": 20315
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 5.433045387268066,
      "learning_rate": 4.1484860361887544e-06,
      "loss": 0.3755,
      "num_input_tokens_seen": 13275264,
      "step": 20320
    },
    {
      "epoch": 11.984080188679245,
      "grad_norm": 3.446495532989502,
      "learning_rate": 4.145951017687379e-06,
      "loss": 0.4719,
      "num_input_tokens_seen": 13277632,
      "step": 20325
    },
    {
      "epoch": 11.987028301886792,
      "grad_norm": 2.5376334190368652,
      "learning_rate": 4.1434162252991524e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 13280608,
      "step": 20330
    },
    {
      "epoch": 11.98997641509434,
      "grad_norm": 2.535027503967285,
      "learning_rate": 4.140881659695173e-06,
      "loss": 0.2923,
      "num_input_tokens_seen": 13283488,
      "step": 20335
    },
    {
      "epoch": 11.992924528301886,
      "grad_norm": 2.684250593185425,
      "learning_rate": 4.138347321546477e-06,
      "loss": 0.4455,
      "num_input_tokens_seen": 13287040,
      "step": 20340
    },
    {
      "epoch": 11.995872641509434,
      "grad_norm": 3.2501649856567383,
      "learning_rate": 4.13581321152404e-06,
      "loss": 0.4907,
      "num_input_tokens_seen": 13290432,
      "step": 20345
    },
    {
      "epoch": 11.99882075471698,
      "grad_norm": 3.1122658252716064,
      "learning_rate": 4.133279330298781e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 13293792,
      "step": 20350
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.5285955667495728,
      "eval_runtime": 19.0582,
      "eval_samples_per_second": 88.991,
      "eval_steps_per_second": 22.248,
      "num_input_tokens_seen": 13294344,
      "step": 20352
    },
    {
      "epoch": 12.001768867924529,
      "grad_norm": 2.995269775390625,
      "learning_rate": 4.130745678541555e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 13296264,
      "step": 20355
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 1.8506131172180176,
      "learning_rate": 4.128212256923155e-06,
      "loss": 0.2497,
      "num_input_tokens_seen": 13298824,
      "step": 20360
    },
    {
      "epoch": 12.007665094339623,
      "grad_norm": 2.377072811126709,
      "learning_rate": 4.125679066114318e-06,
      "loss": 0.2607,
      "num_input_tokens_seen": 13305448,
      "step": 20365
    },
    {
      "epoch": 12.01061320754717,
      "grad_norm": 2.0905508995056152,
      "learning_rate": 4.123146106785717e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 13307880,
      "step": 20370
    },
    {
      "epoch": 12.013561320754716,
      "grad_norm": 2.9763216972351074,
      "learning_rate": 4.1206133796079625e-06,
      "loss": 0.3014,
      "num_input_tokens_seen": 13310312,
      "step": 20375
    },
    {
      "epoch": 12.016509433962264,
      "grad_norm": 3.53157901763916,
      "learning_rate": 4.1180808852516065e-06,
      "loss": 0.2879,
      "num_input_tokens_seen": 13313096,
      "step": 20380
    },
    {
      "epoch": 12.01945754716981,
      "grad_norm": 3.346940040588379,
      "learning_rate": 4.115548624387136e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 13316008,
      "step": 20385
    },
    {
      "epoch": 12.022405660377359,
      "grad_norm": 2.980398178100586,
      "learning_rate": 4.113016597684979e-06,
      "loss": 0.3695,
      "num_input_tokens_seen": 13319176,
      "step": 20390
    },
    {
      "epoch": 12.025353773584905,
      "grad_norm": 2.063508987426758,
      "learning_rate": 4.110484805815502e-06,
      "loss": 0.4062,
      "num_input_tokens_seen": 13323208,
      "step": 20395
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 3.293639898300171,
      "learning_rate": 4.107953249449005e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 13326760,
      "step": 20400
    },
    {
      "epoch": 12.03125,
      "grad_norm": 3.1202104091644287,
      "learning_rate": 4.105421929255729e-06,
      "loss": 0.3712,
      "num_input_tokens_seen": 13329832,
      "step": 20405
    },
    {
      "epoch": 12.034198113207546,
      "grad_norm": 3.460252046585083,
      "learning_rate": 4.102890845905854e-06,
      "loss": 0.5248,
      "num_input_tokens_seen": 13333288,
      "step": 20410
    },
    {
      "epoch": 12.037146226415095,
      "grad_norm": 4.947984218597412,
      "learning_rate": 4.1003600000694935e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 13335848,
      "step": 20415
    },
    {
      "epoch": 12.040094339622641,
      "grad_norm": 2.148728370666504,
      "learning_rate": 4.0978293924167e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 13339944,
      "step": 20420
    },
    {
      "epoch": 12.04304245283019,
      "grad_norm": 2.9058282375335693,
      "learning_rate": 4.095299023617461e-06,
      "loss": 0.3034,
      "num_input_tokens_seen": 13342664,
      "step": 20425
    },
    {
      "epoch": 12.045990566037736,
      "grad_norm": 2.174978494644165,
      "learning_rate": 4.092768894341707e-06,
      "loss": 0.5513,
      "num_input_tokens_seen": 13345608,
      "step": 20430
    },
    {
      "epoch": 12.048938679245284,
      "grad_norm": 5.255092144012451,
      "learning_rate": 4.090239005259298e-06,
      "loss": 0.3537,
      "num_input_tokens_seen": 13348072,
      "step": 20435
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 2.394832134246826,
      "learning_rate": 4.087709357040033e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 13351688,
      "step": 20440
    },
    {
      "epoch": 12.054834905660377,
      "grad_norm": 3.0504393577575684,
      "learning_rate": 4.085179950353648e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 13354792,
      "step": 20445
    },
    {
      "epoch": 12.057783018867925,
      "grad_norm": 4.826006889343262,
      "learning_rate": 4.0826507858698135e-06,
      "loss": 0.3425,
      "num_input_tokens_seen": 13357864,
      "step": 20450
    },
    {
      "epoch": 12.060731132075471,
      "grad_norm": 4.504231929779053,
      "learning_rate": 4.080121864258136e-06,
      "loss": 0.3255,
      "num_input_tokens_seen": 13361320,
      "step": 20455
    },
    {
      "epoch": 12.06367924528302,
      "grad_norm": 3.6780197620391846,
      "learning_rate": 4.077593186188161e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 13364232,
      "step": 20460
    },
    {
      "epoch": 12.066627358490566,
      "grad_norm": 2.183443546295166,
      "learning_rate": 4.075064752329364e-06,
      "loss": 0.2921,
      "num_input_tokens_seen": 13366824,
      "step": 20465
    },
    {
      "epoch": 12.069575471698114,
      "grad_norm": 2.8168795108795166,
      "learning_rate": 4.0725365633511605e-06,
      "loss": 0.5119,
      "num_input_tokens_seen": 13369864,
      "step": 20470
    },
    {
      "epoch": 12.07252358490566,
      "grad_norm": 2.6497018337249756,
      "learning_rate": 4.070008619922899e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 13373544,
      "step": 20475
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 2.3086867332458496,
      "learning_rate": 4.067480922713864e-06,
      "loss": 0.3346,
      "num_input_tokens_seen": 13376872,
      "step": 20480
    },
    {
      "epoch": 12.078419811320755,
      "grad_norm": 2.9749228954315186,
      "learning_rate": 4.064953472393273e-06,
      "loss": 0.3441,
      "num_input_tokens_seen": 13379880,
      "step": 20485
    },
    {
      "epoch": 12.081367924528301,
      "grad_norm": 2.28490948677063,
      "learning_rate": 4.06242626963028e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 13382696,
      "step": 20490
    },
    {
      "epoch": 12.08431603773585,
      "grad_norm": 2.7174410820007324,
      "learning_rate": 4.059899315093972e-06,
      "loss": 0.3402,
      "num_input_tokens_seen": 13386536,
      "step": 20495
    },
    {
      "epoch": 12.087264150943396,
      "grad_norm": 1.899407148361206,
      "learning_rate": 4.057372609453374e-06,
      "loss": 0.2807,
      "num_input_tokens_seen": 13389288,
      "step": 20500
    },
    {
      "epoch": 12.090212264150944,
      "grad_norm": 2.117680549621582,
      "learning_rate": 4.054846153377439e-06,
      "loss": 0.3114,
      "num_input_tokens_seen": 13393352,
      "step": 20505
    },
    {
      "epoch": 12.09316037735849,
      "grad_norm": 3.383793354034424,
      "learning_rate": 4.052319947535058e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 13396200,
      "step": 20510
    },
    {
      "epoch": 12.096108490566039,
      "grad_norm": 3.5340113639831543,
      "learning_rate": 4.049793992595056e-06,
      "loss": 0.4038,
      "num_input_tokens_seen": 13399208,
      "step": 20515
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 4.834295272827148,
      "learning_rate": 4.047268289226187e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 13402856,
      "step": 20520
    },
    {
      "epoch": 12.102004716981131,
      "grad_norm": 2.4631476402282715,
      "learning_rate": 4.044742838097147e-06,
      "loss": 0.2905,
      "num_input_tokens_seen": 13406600,
      "step": 20525
    },
    {
      "epoch": 12.10495283018868,
      "grad_norm": 3.2188820838928223,
      "learning_rate": 4.0422176398765564e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 13410344,
      "step": 20530
    },
    {
      "epoch": 12.107900943396226,
      "grad_norm": 2.3243026733398438,
      "learning_rate": 4.039692695232975e-06,
      "loss": 0.2923,
      "num_input_tokens_seen": 13413384,
      "step": 20535
    },
    {
      "epoch": 12.110849056603774,
      "grad_norm": 2.1043717861175537,
      "learning_rate": 4.03716800483489e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 13416072,
      "step": 20540
    },
    {
      "epoch": 12.11379716981132,
      "grad_norm": 2.1728835105895996,
      "learning_rate": 4.034643569350726e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 13418824,
      "step": 20545
    },
    {
      "epoch": 12.116745283018869,
      "grad_norm": 1.847430944442749,
      "learning_rate": 4.032119389448837e-06,
      "loss": 0.35,
      "num_input_tokens_seen": 13422216,
      "step": 20550
    },
    {
      "epoch": 12.119693396226415,
      "grad_norm": 1.7857210636138916,
      "learning_rate": 4.0295954657975115e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 13426792,
      "step": 20555
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 9.512849807739258,
      "learning_rate": 4.027071799064968e-06,
      "loss": 0.4108,
      "num_input_tokens_seen": 13430216,
      "step": 20560
    },
    {
      "epoch": 12.12558962264151,
      "grad_norm": 19.448076248168945,
      "learning_rate": 4.02454838991936e-06,
      "loss": 0.3385,
      "num_input_tokens_seen": 13434152,
      "step": 20565
    },
    {
      "epoch": 12.128537735849056,
      "grad_norm": 2.0390233993530273,
      "learning_rate": 4.022025239028768e-06,
      "loss": 0.3097,
      "num_input_tokens_seen": 13437448,
      "step": 20570
    },
    {
      "epoch": 12.131485849056604,
      "grad_norm": 2.1481804847717285,
      "learning_rate": 4.0195023470612095e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 13440776,
      "step": 20575
    },
    {
      "epoch": 12.13443396226415,
      "grad_norm": 4.149892807006836,
      "learning_rate": 4.016979714684631e-06,
      "loss": 0.4545,
      "num_input_tokens_seen": 13442984,
      "step": 20580
    },
    {
      "epoch": 12.137382075471699,
      "grad_norm": 2.5265004634857178,
      "learning_rate": 4.014457342566909e-06,
      "loss": 0.3431,
      "num_input_tokens_seen": 13445704,
      "step": 20585
    },
    {
      "epoch": 12.140330188679245,
      "grad_norm": 3.832061290740967,
      "learning_rate": 4.011935231375853e-06,
      "loss": 0.2277,
      "num_input_tokens_seen": 13448872,
      "step": 20590
    },
    {
      "epoch": 12.143278301886792,
      "grad_norm": 2.831943988800049,
      "learning_rate": 4.009413381779203e-06,
      "loss": 0.3401,
      "num_input_tokens_seen": 13451912,
      "step": 20595
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 5.006485939025879,
      "learning_rate": 4.00689179444463e-06,
      "loss": 0.4031,
      "num_input_tokens_seen": 13454760,
      "step": 20600
    },
    {
      "epoch": 12.149174528301886,
      "grad_norm": 4.691247463226318,
      "learning_rate": 4.004370470039733e-06,
      "loss": 0.3611,
      "num_input_tokens_seen": 13457960,
      "step": 20605
    },
    {
      "epoch": 12.152122641509434,
      "grad_norm": 4.993732452392578,
      "learning_rate": 4.001849409232046e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 13461864,
      "step": 20610
    },
    {
      "epoch": 12.15507075471698,
      "grad_norm": 2.706949472427368,
      "learning_rate": 3.9993286126890274e-06,
      "loss": 0.3062,
      "num_input_tokens_seen": 13464808,
      "step": 20615
    },
    {
      "epoch": 12.158018867924529,
      "grad_norm": 4.822876453399658,
      "learning_rate": 3.996808081078074e-06,
      "loss": 0.3069,
      "num_input_tokens_seen": 13467976,
      "step": 20620
    },
    {
      "epoch": 12.160966981132075,
      "grad_norm": 2.3310904502868652,
      "learning_rate": 3.9942878150665035e-06,
      "loss": 0.3487,
      "num_input_tokens_seen": 13470920,
      "step": 20625
    },
    {
      "epoch": 12.163915094339623,
      "grad_norm": 2.31691575050354,
      "learning_rate": 3.991767815321569e-06,
      "loss": 0.3402,
      "num_input_tokens_seen": 13473768,
      "step": 20630
    },
    {
      "epoch": 12.16686320754717,
      "grad_norm": 3.8995275497436523,
      "learning_rate": 3.9892480825104504e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 13476232,
      "step": 20635
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 6.5006608963012695,
      "learning_rate": 3.986728617300257e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 13479400,
      "step": 20640
    },
    {
      "epoch": 12.172759433962264,
      "grad_norm": 6.510169506072998,
      "learning_rate": 3.984209420358031e-06,
      "loss": 0.3649,
      "num_input_tokens_seen": 13482248,
      "step": 20645
    },
    {
      "epoch": 12.17570754716981,
      "grad_norm": 2.992431402206421,
      "learning_rate": 3.981690492350738e-06,
      "loss": 0.2975,
      "num_input_tokens_seen": 13484648,
      "step": 20650
    },
    {
      "epoch": 12.178655660377359,
      "grad_norm": 2.7950279712677,
      "learning_rate": 3.979171833945276e-06,
      "loss": 0.2502,
      "num_input_tokens_seen": 13487688,
      "step": 20655
    },
    {
      "epoch": 12.181603773584905,
      "grad_norm": 5.115423679351807,
      "learning_rate": 3.97665344580847e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 13491016,
      "step": 20660
    },
    {
      "epoch": 12.184551886792454,
      "grad_norm": 4.521049976348877,
      "learning_rate": 3.974135328607075e-06,
      "loss": 0.4165,
      "num_input_tokens_seen": 13493800,
      "step": 20665
    },
    {
      "epoch": 12.1875,
      "grad_norm": 3.3234031200408936,
      "learning_rate": 3.971617483007773e-06,
      "loss": 0.3658,
      "num_input_tokens_seen": 13496648,
      "step": 20670
    },
    {
      "epoch": 12.190448113207546,
      "grad_norm": 3.1856353282928467,
      "learning_rate": 3.969099909677174e-06,
      "loss": 0.3026,
      "num_input_tokens_seen": 13500392,
      "step": 20675
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 5.189203262329102,
      "learning_rate": 3.9665826092818165e-06,
      "loss": 0.4524,
      "num_input_tokens_seen": 13503464,
      "step": 20680
    },
    {
      "epoch": 12.196344339622641,
      "grad_norm": 3.7492592334747314,
      "learning_rate": 3.964065582488168e-06,
      "loss": 0.4332,
      "num_input_tokens_seen": 13505864,
      "step": 20685
    },
    {
      "epoch": 12.19929245283019,
      "grad_norm": 1.7605834007263184,
      "learning_rate": 3.961548829962622e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 13509480,
      "step": 20690
    },
    {
      "epoch": 12.202240566037736,
      "grad_norm": 2.529864549636841,
      "learning_rate": 3.959032352371496e-06,
      "loss": 0.3373,
      "num_input_tokens_seen": 13512456,
      "step": 20695
    },
    {
      "epoch": 12.205188679245284,
      "grad_norm": 2.8706319332122803,
      "learning_rate": 3.956516150381043e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 13515976,
      "step": 20700
    },
    {
      "epoch": 12.20813679245283,
      "grad_norm": 2.315016508102417,
      "learning_rate": 3.954000224657436e-06,
      "loss": 0.334,
      "num_input_tokens_seen": 13518952,
      "step": 20705
    },
    {
      "epoch": 12.211084905660377,
      "grad_norm": 5.482354164123535,
      "learning_rate": 3.951484575866776e-06,
      "loss": 0.3664,
      "num_input_tokens_seen": 13522152,
      "step": 20710
    },
    {
      "epoch": 12.214033018867925,
      "grad_norm": 2.564690351486206,
      "learning_rate": 3.948969204675096e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 13526024,
      "step": 20715
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 1.744099497795105,
      "learning_rate": 3.946454111748346e-06,
      "loss": 0.2735,
      "num_input_tokens_seen": 13528968,
      "step": 20720
    },
    {
      "epoch": 12.21992924528302,
      "grad_norm": 3.045351028442383,
      "learning_rate": 3.943939297752413e-06,
      "loss": 0.2866,
      "num_input_tokens_seen": 13532616,
      "step": 20725
    },
    {
      "epoch": 12.222877358490566,
      "grad_norm": 1.4625201225280762,
      "learning_rate": 3.9414247633531e-06,
      "loss": 0.2635,
      "num_input_tokens_seen": 13535688,
      "step": 20730
    },
    {
      "epoch": 12.225825471698114,
      "grad_norm": 2.557241201400757,
      "learning_rate": 3.9389105092161454e-06,
      "loss": 0.2693,
      "num_input_tokens_seen": 13538024,
      "step": 20735
    },
    {
      "epoch": 12.22877358490566,
      "grad_norm": 4.044541835784912,
      "learning_rate": 3.936396536007205e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 13540936,
      "step": 20740
    },
    {
      "epoch": 12.231721698113208,
      "grad_norm": 2.7472991943359375,
      "learning_rate": 3.933882844391866e-06,
      "loss": 0.3369,
      "num_input_tokens_seen": 13545032,
      "step": 20745
    },
    {
      "epoch": 12.234669811320755,
      "grad_norm": 2.1390974521636963,
      "learning_rate": 3.931369435035639e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 13548072,
      "step": 20750
    },
    {
      "epoch": 12.237617924528301,
      "grad_norm": 2.2768681049346924,
      "learning_rate": 3.92885630860396e-06,
      "loss": 0.2826,
      "num_input_tokens_seen": 13551304,
      "step": 20755
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 2.3447816371917725,
      "learning_rate": 3.926343465762189e-06,
      "loss": 0.4699,
      "num_input_tokens_seen": 13553992,
      "step": 20760
    },
    {
      "epoch": 12.243514150943396,
      "grad_norm": 3.5453333854675293,
      "learning_rate": 3.923830907175613e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 13557000,
      "step": 20765
    },
    {
      "epoch": 12.246462264150944,
      "grad_norm": 2.715397357940674,
      "learning_rate": 3.921318633509442e-06,
      "loss": 0.2681,
      "num_input_tokens_seen": 13560520,
      "step": 20770
    },
    {
      "epoch": 12.24941037735849,
      "grad_norm": 9.032553672790527,
      "learning_rate": 3.918806645428811e-06,
      "loss": 0.3263,
      "num_input_tokens_seen": 13563848,
      "step": 20775
    },
    {
      "epoch": 12.252358490566039,
      "grad_norm": 3.0020182132720947,
      "learning_rate": 3.916294943598781e-06,
      "loss": 0.3951,
      "num_input_tokens_seen": 13567240,
      "step": 20780
    },
    {
      "epoch": 12.255306603773585,
      "grad_norm": 1.9228100776672363,
      "learning_rate": 3.913783528684336e-06,
      "loss": 0.3665,
      "num_input_tokens_seen": 13570984,
      "step": 20785
    },
    {
      "epoch": 12.258254716981131,
      "grad_norm": 4.466412544250488,
      "learning_rate": 3.911272401350381e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 13574344,
      "step": 20790
    },
    {
      "epoch": 12.26120283018868,
      "grad_norm": 2.349611520767212,
      "learning_rate": 3.9087615622617525e-06,
      "loss": 0.2633,
      "num_input_tokens_seen": 13577160,
      "step": 20795
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 3.4292752742767334,
      "learning_rate": 3.906251012083202e-06,
      "loss": 0.277,
      "num_input_tokens_seen": 13579816,
      "step": 20800
    },
    {
      "epoch": 12.267099056603774,
      "grad_norm": 4.5215630531311035,
      "learning_rate": 3.9037407514794085e-06,
      "loss": 0.3569,
      "num_input_tokens_seen": 13583208,
      "step": 20805
    },
    {
      "epoch": 12.27004716981132,
      "grad_norm": 3.7275662422180176,
      "learning_rate": 3.901230781114976e-06,
      "loss": 0.3705,
      "num_input_tokens_seen": 13586664,
      "step": 20810
    },
    {
      "epoch": 12.272995283018869,
      "grad_norm": 4.544991493225098,
      "learning_rate": 3.898721101654431e-06,
      "loss": 0.4013,
      "num_input_tokens_seen": 13589576,
      "step": 20815
    },
    {
      "epoch": 12.275943396226415,
      "grad_norm": 3.4388785362243652,
      "learning_rate": 3.896211713762221e-06,
      "loss": 0.4176,
      "num_input_tokens_seen": 13592520,
      "step": 20820
    },
    {
      "epoch": 12.278891509433961,
      "grad_norm": 2.700488805770874,
      "learning_rate": 3.893702618102715e-06,
      "loss": 0.2927,
      "num_input_tokens_seen": 13595880,
      "step": 20825
    },
    {
      "epoch": 12.28183962264151,
      "grad_norm": 3.167090654373169,
      "learning_rate": 3.891193815340211e-06,
      "loss": 0.3386,
      "num_input_tokens_seen": 13599432,
      "step": 20830
    },
    {
      "epoch": 12.284787735849056,
      "grad_norm": 2.197781801223755,
      "learning_rate": 3.888685306138922e-06,
      "loss": 0.3737,
      "num_input_tokens_seen": 13603304,
      "step": 20835
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 3.288383960723877,
      "learning_rate": 3.886177091162987e-06,
      "loss": 0.4645,
      "num_input_tokens_seen": 13608168,
      "step": 20840
    },
    {
      "epoch": 12.29068396226415,
      "grad_norm": 2.750422477722168,
      "learning_rate": 3.883669171076468e-06,
      "loss": 0.3549,
      "num_input_tokens_seen": 13611432,
      "step": 20845
    },
    {
      "epoch": 12.293632075471699,
      "grad_norm": 3.9897148609161377,
      "learning_rate": 3.881161546543348e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 13614952,
      "step": 20850
    },
    {
      "epoch": 12.296580188679245,
      "grad_norm": 3.860426425933838,
      "learning_rate": 3.8786542182275295e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 13617800,
      "step": 20855
    },
    {
      "epoch": 12.299528301886792,
      "grad_norm": 1.9516464471817017,
      "learning_rate": 3.87614718679284e-06,
      "loss": 0.4817,
      "num_input_tokens_seen": 13620392,
      "step": 20860
    },
    {
      "epoch": 12.30247641509434,
      "grad_norm": 2.693713426589966,
      "learning_rate": 3.8736404529030255e-06,
      "loss": 0.2961,
      "num_input_tokens_seen": 13623240,
      "step": 20865
    },
    {
      "epoch": 12.305424528301886,
      "grad_norm": 3.691365957260132,
      "learning_rate": 3.871134017221756e-06,
      "loss": 0.3494,
      "num_input_tokens_seen": 13626280,
      "step": 20870
    },
    {
      "epoch": 12.308372641509434,
      "grad_norm": 2.652278423309326,
      "learning_rate": 3.868627880412621e-06,
      "loss": 0.3129,
      "num_input_tokens_seen": 13629032,
      "step": 20875
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 7.2572102546691895,
      "learning_rate": 3.86612204313913e-06,
      "loss": 0.5072,
      "num_input_tokens_seen": 13631272,
      "step": 20880
    },
    {
      "epoch": 12.314268867924529,
      "grad_norm": 2.2897253036499023,
      "learning_rate": 3.863616506064714e-06,
      "loss": 0.3025,
      "num_input_tokens_seen": 13634088,
      "step": 20885
    },
    {
      "epoch": 12.317216981132075,
      "grad_norm": 3.4662582874298096,
      "learning_rate": 3.861111269852727e-06,
      "loss": 0.28,
      "num_input_tokens_seen": 13637384,
      "step": 20890
    },
    {
      "epoch": 12.320165094339623,
      "grad_norm": 2.384341239929199,
      "learning_rate": 3.858606335166439e-06,
      "loss": 0.3375,
      "num_input_tokens_seen": 13639880,
      "step": 20895
    },
    {
      "epoch": 12.32311320754717,
      "grad_norm": 2.494283437728882,
      "learning_rate": 3.8561017026690415e-06,
      "loss": 0.3023,
      "num_input_tokens_seen": 13643720,
      "step": 20900
    },
    {
      "epoch": 12.326061320754716,
      "grad_norm": 3.1592018604278564,
      "learning_rate": 3.8535973730236495e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 13647496,
      "step": 20905
    },
    {
      "epoch": 12.329009433962264,
      "grad_norm": 2.509104013442993,
      "learning_rate": 3.851093346893294e-06,
      "loss": 0.4627,
      "num_input_tokens_seen": 13651496,
      "step": 20910
    },
    {
      "epoch": 12.33195754716981,
      "grad_norm": 2.404008150100708,
      "learning_rate": 3.848589624940927e-06,
      "loss": 0.455,
      "num_input_tokens_seen": 13654728,
      "step": 20915
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 3.1260712146759033,
      "learning_rate": 3.846086207829417e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 13660456,
      "step": 20920
    },
    {
      "epoch": 12.337853773584905,
      "grad_norm": 2.6710073947906494,
      "learning_rate": 3.843583096221559e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 13663336,
      "step": 20925
    },
    {
      "epoch": 12.340801886792454,
      "grad_norm": 3.337674140930176,
      "learning_rate": 3.8410802907800596e-06,
      "loss": 0.4104,
      "num_input_tokens_seen": 13666216,
      "step": 20930
    },
    {
      "epoch": 12.34375,
      "grad_norm": 2.2680962085723877,
      "learning_rate": 3.838577792167546e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 13669512,
      "step": 20935
    },
    {
      "epoch": 12.346698113207546,
      "grad_norm": 3.8270130157470703,
      "learning_rate": 3.836075601046569e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 13672040,
      "step": 20940
    },
    {
      "epoch": 12.349646226415095,
      "grad_norm": 5.3278679847717285,
      "learning_rate": 3.833573718079594e-06,
      "loss": 0.4245,
      "num_input_tokens_seen": 13674600,
      "step": 20945
    },
    {
      "epoch": 12.352594339622641,
      "grad_norm": 3.143900156021118,
      "learning_rate": 3.831072143929002e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 13678024,
      "step": 20950
    },
    {
      "epoch": 12.35554245283019,
      "grad_norm": 3.8135550022125244,
      "learning_rate": 3.828570879257098e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 13680808,
      "step": 20955
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 4.760499954223633,
      "learning_rate": 3.826069924726102e-06,
      "loss": 0.49,
      "num_input_tokens_seen": 13684552,
      "step": 20960
    },
    {
      "epoch": 12.361438679245284,
      "grad_norm": 3.8336374759674072,
      "learning_rate": 3.823569280998154e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 13687464,
      "step": 20965
    },
    {
      "epoch": 12.36438679245283,
      "grad_norm": 2.6034443378448486,
      "learning_rate": 3.8210689487353095e-06,
      "loss": 0.4059,
      "num_input_tokens_seen": 13690152,
      "step": 20970
    },
    {
      "epoch": 12.367334905660377,
      "grad_norm": 4.756405353546143,
      "learning_rate": 3.818568928599539e-06,
      "loss": 0.4247,
      "num_input_tokens_seen": 13692872,
      "step": 20975
    },
    {
      "epoch": 12.370283018867925,
      "grad_norm": 6.018263816833496,
      "learning_rate": 3.816069221252739e-06,
      "loss": 0.3441,
      "num_input_tokens_seen": 13695368,
      "step": 20980
    },
    {
      "epoch": 12.373231132075471,
      "grad_norm": 2.6549811363220215,
      "learning_rate": 3.813569827356715e-06,
      "loss": 0.2419,
      "num_input_tokens_seen": 13697992,
      "step": 20985
    },
    {
      "epoch": 12.37617924528302,
      "grad_norm": 3.284407615661621,
      "learning_rate": 3.811070747573191e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 13700520,
      "step": 20990
    },
    {
      "epoch": 12.379127358490566,
      "grad_norm": 6.835721969604492,
      "learning_rate": 3.8085719825638098e-06,
      "loss": 0.3638,
      "num_input_tokens_seen": 13703624,
      "step": 20995
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 10.055959701538086,
      "learning_rate": 3.806073532990132e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 13707528,
      "step": 21000
    },
    {
      "epoch": 12.38502358490566,
      "grad_norm": 1.8716892004013062,
      "learning_rate": 3.8035753995136333e-06,
      "loss": 0.3184,
      "num_input_tokens_seen": 13710408,
      "step": 21005
    },
    {
      "epoch": 12.387971698113208,
      "grad_norm": 2.4773008823394775,
      "learning_rate": 3.801077582795702e-06,
      "loss": 0.2369,
      "num_input_tokens_seen": 13713608,
      "step": 21010
    },
    {
      "epoch": 12.390919811320755,
      "grad_norm": 3.0310864448547363,
      "learning_rate": 3.7985800834976494e-06,
      "loss": 0.2307,
      "num_input_tokens_seen": 13717864,
      "step": 21015
    },
    {
      "epoch": 12.393867924528301,
      "grad_norm": 3.7160608768463135,
      "learning_rate": 3.7960829022806965e-06,
      "loss": 0.3376,
      "num_input_tokens_seen": 13720744,
      "step": 21020
    },
    {
      "epoch": 12.39681603773585,
      "grad_norm": 5.075348377227783,
      "learning_rate": 3.793586039805984e-06,
      "loss": 0.2812,
      "num_input_tokens_seen": 13724360,
      "step": 21025
    },
    {
      "epoch": 12.399764150943396,
      "grad_norm": 3.7574923038482666,
      "learning_rate": 3.791089496734567e-06,
      "loss": 0.3601,
      "num_input_tokens_seen": 13727816,
      "step": 21030
    },
    {
      "epoch": 12.402712264150944,
      "grad_norm": 4.410711288452148,
      "learning_rate": 3.7885932737274163e-06,
      "loss": 0.2956,
      "num_input_tokens_seen": 13730824,
      "step": 21035
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 1.7549787759780884,
      "learning_rate": 3.7860973714454156e-06,
      "loss": 0.3247,
      "num_input_tokens_seen": 13734568,
      "step": 21040
    },
    {
      "epoch": 12.408608490566039,
      "grad_norm": 2.128951072692871,
      "learning_rate": 3.7836017905493695e-06,
      "loss": 0.4539,
      "num_input_tokens_seen": 13737736,
      "step": 21045
    },
    {
      "epoch": 12.411556603773585,
      "grad_norm": 5.231126308441162,
      "learning_rate": 3.7811065316999908e-06,
      "loss": 0.3057,
      "num_input_tokens_seen": 13741832,
      "step": 21050
    },
    {
      "epoch": 12.414504716981131,
      "grad_norm": 5.285285472869873,
      "learning_rate": 3.7786115955579105e-06,
      "loss": 0.2682,
      "num_input_tokens_seen": 13744040,
      "step": 21055
    },
    {
      "epoch": 12.41745283018868,
      "grad_norm": 3.159050941467285,
      "learning_rate": 3.7761169827836746e-06,
      "loss": 0.3746,
      "num_input_tokens_seen": 13747080,
      "step": 21060
    },
    {
      "epoch": 12.420400943396226,
      "grad_norm": 1.7390600442886353,
      "learning_rate": 3.773622694037743e-06,
      "loss": 0.2916,
      "num_input_tokens_seen": 13751496,
      "step": 21065
    },
    {
      "epoch": 12.423349056603774,
      "grad_norm": 2.6362318992614746,
      "learning_rate": 3.7711287299804865e-06,
      "loss": 0.419,
      "num_input_tokens_seen": 13754792,
      "step": 21070
    },
    {
      "epoch": 12.42629716981132,
      "grad_norm": 4.261575698852539,
      "learning_rate": 3.768635091272197e-06,
      "loss": 0.3942,
      "num_input_tokens_seen": 13758120,
      "step": 21075
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 4.064494609832764,
      "learning_rate": 3.7661417785730732e-06,
      "loss": 0.2827,
      "num_input_tokens_seen": 13761096,
      "step": 21080
    },
    {
      "epoch": 12.432193396226415,
      "grad_norm": 3.165087938308716,
      "learning_rate": 3.7636487925432304e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 13764808,
      "step": 21085
    },
    {
      "epoch": 12.435141509433961,
      "grad_norm": 2.7327182292938232,
      "learning_rate": 3.761156133842697e-06,
      "loss": 0.4338,
      "num_input_tokens_seen": 13768168,
      "step": 21090
    },
    {
      "epoch": 12.43808962264151,
      "grad_norm": 3.362520217895508,
      "learning_rate": 3.7586638031314182e-06,
      "loss": 0.3107,
      "num_input_tokens_seen": 13771560,
      "step": 21095
    },
    {
      "epoch": 12.441037735849056,
      "grad_norm": 2.6189818382263184,
      "learning_rate": 3.7561718010692477e-06,
      "loss": 0.3855,
      "num_input_tokens_seen": 13775432,
      "step": 21100
    },
    {
      "epoch": 12.443985849056604,
      "grad_norm": 3.5197579860687256,
      "learning_rate": 3.7536801283159523e-06,
      "loss": 0.2918,
      "num_input_tokens_seen": 13778856,
      "step": 21105
    },
    {
      "epoch": 12.44693396226415,
      "grad_norm": 2.631495952606201,
      "learning_rate": 3.7511887855312155e-06,
      "loss": 0.3363,
      "num_input_tokens_seen": 13781576,
      "step": 21110
    },
    {
      "epoch": 12.449882075471699,
      "grad_norm": 3.284714460372925,
      "learning_rate": 3.74869777337463e-06,
      "loss": 0.3237,
      "num_input_tokens_seen": 13784488,
      "step": 21115
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 3.011319398880005,
      "learning_rate": 3.7462070925057004e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 13787272,
      "step": 21120
    },
    {
      "epoch": 12.455778301886792,
      "grad_norm": 2.7078869342803955,
      "learning_rate": 3.7437167435838472e-06,
      "loss": 0.2825,
      "num_input_tokens_seen": 13791080,
      "step": 21125
    },
    {
      "epoch": 12.45872641509434,
      "grad_norm": 2.282879114151001,
      "learning_rate": 3.741226727268401e-06,
      "loss": 0.2616,
      "num_input_tokens_seen": 13794760,
      "step": 21130
    },
    {
      "epoch": 12.461674528301886,
      "grad_norm": 4.962332725524902,
      "learning_rate": 3.7387370442186027e-06,
      "loss": 0.4339,
      "num_input_tokens_seen": 13797576,
      "step": 21135
    },
    {
      "epoch": 12.464622641509434,
      "grad_norm": 2.7526962757110596,
      "learning_rate": 3.736247695093609e-06,
      "loss": 0.2881,
      "num_input_tokens_seen": 13802056,
      "step": 21140
    },
    {
      "epoch": 12.46757075471698,
      "grad_norm": 2.9923222064971924,
      "learning_rate": 3.7337586805524838e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 13805960,
      "step": 21145
    },
    {
      "epoch": 12.470518867924529,
      "grad_norm": 3.5290675163269043,
      "learning_rate": 3.731270001254205e-06,
      "loss": 0.3051,
      "num_input_tokens_seen": 13809032,
      "step": 21150
    },
    {
      "epoch": 12.473466981132075,
      "grad_norm": 5.629026412963867,
      "learning_rate": 3.728781657857661e-06,
      "loss": 0.2605,
      "num_input_tokens_seen": 13812360,
      "step": 21155
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 3.2082128524780273,
      "learning_rate": 3.726293651021653e-06,
      "loss": 0.3095,
      "num_input_tokens_seen": 13815144,
      "step": 21160
    },
    {
      "epoch": 12.47936320754717,
      "grad_norm": 3.3161277770996094,
      "learning_rate": 3.7238059814048888e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 13817864,
      "step": 21165
    },
    {
      "epoch": 12.482311320754716,
      "grad_norm": 2.120596170425415,
      "learning_rate": 3.7213186496659916e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 13820776,
      "step": 21170
    },
    {
      "epoch": 12.485259433962264,
      "grad_norm": 3.1434383392333984,
      "learning_rate": 3.718831656463493e-06,
      "loss": 0.3418,
      "num_input_tokens_seen": 13825000,
      "step": 21175
    },
    {
      "epoch": 12.48820754716981,
      "grad_norm": 4.214330196380615,
      "learning_rate": 3.7163450024558345e-06,
      "loss": 0.3124,
      "num_input_tokens_seen": 13827336,
      "step": 21180
    },
    {
      "epoch": 12.491155660377359,
      "grad_norm": 3.908937931060791,
      "learning_rate": 3.7138586883013694e-06,
      "loss": 0.4462,
      "num_input_tokens_seen": 13829928,
      "step": 21185
    },
    {
      "epoch": 12.494103773584905,
      "grad_norm": 3.2783925533294678,
      "learning_rate": 3.7113727146583573e-06,
      "loss": 0.3776,
      "num_input_tokens_seen": 13833448,
      "step": 21190
    },
    {
      "epoch": 12.497051886792454,
      "grad_norm": 4.030472278594971,
      "learning_rate": 3.7088870821849748e-06,
      "loss": 0.3747,
      "num_input_tokens_seen": 13835976,
      "step": 21195
    },
    {
      "epoch": 12.5,
      "grad_norm": 3.095993995666504,
      "learning_rate": 3.706401791539301e-06,
      "loss": 0.3271,
      "num_input_tokens_seen": 13842440,
      "step": 21200
    },
    {
      "epoch": 12.502948113207546,
      "grad_norm": 2.764364719390869,
      "learning_rate": 3.703916843379328e-06,
      "loss": 0.3427,
      "num_input_tokens_seen": 13845928,
      "step": 21205
    },
    {
      "epoch": 12.505896226415095,
      "grad_norm": 2.484802484512329,
      "learning_rate": 3.7014322383629575e-06,
      "loss": 0.3785,
      "num_input_tokens_seen": 13848808,
      "step": 21210
    },
    {
      "epoch": 12.508844339622641,
      "grad_norm": 4.204075336456299,
      "learning_rate": 3.6989479771479976e-06,
      "loss": 0.3769,
      "num_input_tokens_seen": 13851944,
      "step": 21215
    },
    {
      "epoch": 12.51179245283019,
      "grad_norm": 3.847012996673584,
      "learning_rate": 3.696464060392169e-06,
      "loss": 0.3037,
      "num_input_tokens_seen": 13854440,
      "step": 21220
    },
    {
      "epoch": 12.514740566037736,
      "grad_norm": 3.306812286376953,
      "learning_rate": 3.6939804887530962e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 13857064,
      "step": 21225
    },
    {
      "epoch": 12.517688679245284,
      "grad_norm": 2.6109683513641357,
      "learning_rate": 3.6914972628883196e-06,
      "loss": 0.3156,
      "num_input_tokens_seen": 13859720,
      "step": 21230
    },
    {
      "epoch": 12.52063679245283,
      "grad_norm": 2.273571729660034,
      "learning_rate": 3.6890143834552814e-06,
      "loss": 0.2943,
      "num_input_tokens_seen": 13862824,
      "step": 21235
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 3.1061275005340576,
      "learning_rate": 3.6865318511113347e-06,
      "loss": 0.3033,
      "num_input_tokens_seen": 13866824,
      "step": 21240
    },
    {
      "epoch": 12.526533018867925,
      "grad_norm": 3.150359869003296,
      "learning_rate": 3.684049666513742e-06,
      "loss": 0.4844,
      "num_input_tokens_seen": 13869864,
      "step": 21245
    },
    {
      "epoch": 12.529481132075471,
      "grad_norm": 2.6697885990142822,
      "learning_rate": 3.6815678303196715e-06,
      "loss": 0.3299,
      "num_input_tokens_seen": 13873384,
      "step": 21250
    },
    {
      "epoch": 12.53242924528302,
      "grad_norm": 3.050293207168579,
      "learning_rate": 3.6790863431861988e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 13876552,
      "step": 21255
    },
    {
      "epoch": 12.535377358490566,
      "grad_norm": 3.487201690673828,
      "learning_rate": 3.676605205770311e-06,
      "loss": 0.5546,
      "num_input_tokens_seen": 13879368,
      "step": 21260
    },
    {
      "epoch": 12.538325471698114,
      "grad_norm": 4.3891119956970215,
      "learning_rate": 3.674124418728898e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 13882088,
      "step": 21265
    },
    {
      "epoch": 12.54127358490566,
      "grad_norm": 7.5927205085754395,
      "learning_rate": 3.671643982718759e-06,
      "loss": 0.408,
      "num_input_tokens_seen": 13885480,
      "step": 21270
    },
    {
      "epoch": 12.544221698113208,
      "grad_norm": 2.9569761753082275,
      "learning_rate": 3.6691638983966017e-06,
      "loss": 0.2455,
      "num_input_tokens_seen": 13888360,
      "step": 21275
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 2.747605562210083,
      "learning_rate": 3.666684166419037e-06,
      "loss": 0.3078,
      "num_input_tokens_seen": 13892104,
      "step": 21280
    },
    {
      "epoch": 12.550117924528301,
      "grad_norm": 3.0418810844421387,
      "learning_rate": 3.6642047874425833e-06,
      "loss": 0.3669,
      "num_input_tokens_seen": 13898824,
      "step": 21285
    },
    {
      "epoch": 12.55306603773585,
      "grad_norm": 2.267571210861206,
      "learning_rate": 3.661725762123671e-06,
      "loss": 0.6203,
      "num_input_tokens_seen": 13901512,
      "step": 21290
    },
    {
      "epoch": 12.556014150943396,
      "grad_norm": 2.5589938163757324,
      "learning_rate": 3.659247091118631e-06,
      "loss": 0.4028,
      "num_input_tokens_seen": 13905480,
      "step": 21295
    },
    {
      "epoch": 12.558962264150944,
      "grad_norm": 2.3733720779418945,
      "learning_rate": 3.6567687750837027e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 13908648,
      "step": 21300
    },
    {
      "epoch": 12.56191037735849,
      "grad_norm": 3.881669044494629,
      "learning_rate": 3.6542908146750287e-06,
      "loss": 0.3133,
      "num_input_tokens_seen": 13911048,
      "step": 21305
    },
    {
      "epoch": 12.564858490566039,
      "grad_norm": 2.383603096008301,
      "learning_rate": 3.6518132105486624e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 13914088,
      "step": 21310
    },
    {
      "epoch": 12.567806603773585,
      "grad_norm": 3.1022331714630127,
      "learning_rate": 3.649335963360559e-06,
      "loss": 0.3018,
      "num_input_tokens_seen": 13916904,
      "step": 21315
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 2.505497455596924,
      "learning_rate": 3.6468590737665795e-06,
      "loss": 0.3971,
      "num_input_tokens_seen": 13921160,
      "step": 21320
    },
    {
      "epoch": 12.57370283018868,
      "grad_norm": 2.6197571754455566,
      "learning_rate": 3.6443825424224926e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 13923624,
      "step": 21325
    },
    {
      "epoch": 12.576650943396226,
      "grad_norm": 1.9581712484359741,
      "learning_rate": 3.64190636998397e-06,
      "loss": 0.3281,
      "num_input_tokens_seen": 13927080,
      "step": 21330
    },
    {
      "epoch": 12.579599056603774,
      "grad_norm": 3.0885415077209473,
      "learning_rate": 3.639430557106588e-06,
      "loss": 0.3663,
      "num_input_tokens_seen": 13929736,
      "step": 21335
    },
    {
      "epoch": 12.58254716981132,
      "grad_norm": 3.5131583213806152,
      "learning_rate": 3.6369551044458314e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 13934376,
      "step": 21340
    },
    {
      "epoch": 12.585495283018869,
      "grad_norm": 3.603179931640625,
      "learning_rate": 3.6344800126570846e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 13937544,
      "step": 21345
    },
    {
      "epoch": 12.588443396226415,
      "grad_norm": 3.7797067165374756,
      "learning_rate": 3.6320052823956385e-06,
      "loss": 0.3013,
      "num_input_tokens_seen": 13941256,
      "step": 21350
    },
    {
      "epoch": 12.591391509433961,
      "grad_norm": 2.212301015853882,
      "learning_rate": 3.6295309143166906e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 13944424,
      "step": 21355
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 4.5279765129089355,
      "learning_rate": 3.62705690907534e-06,
      "loss": 0.3375,
      "num_input_tokens_seen": 13947112,
      "step": 21360
    },
    {
      "epoch": 12.597287735849056,
      "grad_norm": 2.2268099784851074,
      "learning_rate": 3.624583267326588e-06,
      "loss": 0.3581,
      "num_input_tokens_seen": 13950280,
      "step": 21365
    },
    {
      "epoch": 12.600235849056604,
      "grad_norm": 2.970632791519165,
      "learning_rate": 3.6221099897253454e-06,
      "loss": 0.4591,
      "num_input_tokens_seen": 13953288,
      "step": 21370
    },
    {
      "epoch": 12.60318396226415,
      "grad_norm": 3.822704315185547,
      "learning_rate": 3.619637076926421e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 13957832,
      "step": 21375
    },
    {
      "epoch": 12.606132075471699,
      "grad_norm": 2.8128480911254883,
      "learning_rate": 3.617164529584528e-06,
      "loss": 0.2834,
      "num_input_tokens_seen": 13960904,
      "step": 21380
    },
    {
      "epoch": 12.609080188679245,
      "grad_norm": 3.724604845046997,
      "learning_rate": 3.614692348354286e-06,
      "loss": 0.3356,
      "num_input_tokens_seen": 13964136,
      "step": 21385
    },
    {
      "epoch": 12.612028301886792,
      "grad_norm": 2.0798704624176025,
      "learning_rate": 3.612220533890216e-06,
      "loss": 0.3243,
      "num_input_tokens_seen": 13967656,
      "step": 21390
    },
    {
      "epoch": 12.61497641509434,
      "grad_norm": 3.9214749336242676,
      "learning_rate": 3.609749086846741e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 13971144,
      "step": 21395
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 2.851613759994507,
      "learning_rate": 3.607278007878186e-06,
      "loss": 0.3513,
      "num_input_tokens_seen": 13975304,
      "step": 21400
    },
    {
      "epoch": 12.620872641509434,
      "grad_norm": 3.303439140319824,
      "learning_rate": 3.6048072976387817e-06,
      "loss": 0.3931,
      "num_input_tokens_seen": 13978696,
      "step": 21405
    },
    {
      "epoch": 12.62382075471698,
      "grad_norm": 7.3056769371032715,
      "learning_rate": 3.6023369567826585e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 13981544,
      "step": 21410
    },
    {
      "epoch": 12.626768867924529,
      "grad_norm": 2.8467283248901367,
      "learning_rate": 3.599866985963848e-06,
      "loss": 0.2721,
      "num_input_tokens_seen": 13984232,
      "step": 21415
    },
    {
      "epoch": 12.629716981132075,
      "grad_norm": 10.393388748168945,
      "learning_rate": 3.5973973858362885e-06,
      "loss": 0.4557,
      "num_input_tokens_seen": 13987464,
      "step": 21420
    },
    {
      "epoch": 12.632665094339622,
      "grad_norm": 4.37252950668335,
      "learning_rate": 3.594928157053816e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 13990792,
      "step": 21425
    },
    {
      "epoch": 12.63561320754717,
      "grad_norm": 2.7383339405059814,
      "learning_rate": 3.592459300270168e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 13994920,
      "step": 21430
    },
    {
      "epoch": 12.638561320754716,
      "grad_norm": 2.8082518577575684,
      "learning_rate": 3.589990816138988e-06,
      "loss": 0.2842,
      "num_input_tokens_seen": 13997352,
      "step": 21435
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 2.6382718086242676,
      "learning_rate": 3.587522705313816e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 14000584,
      "step": 21440
    },
    {
      "epoch": 12.64445754716981,
      "grad_norm": 5.086875915527344,
      "learning_rate": 3.585054968448094e-06,
      "loss": 0.3869,
      "num_input_tokens_seen": 14004104,
      "step": 21445
    },
    {
      "epoch": 12.647405660377359,
      "grad_norm": 2.2660017013549805,
      "learning_rate": 3.5825876061951686e-06,
      "loss": 0.3794,
      "num_input_tokens_seen": 14007400,
      "step": 21450
    },
    {
      "epoch": 12.650353773584905,
      "grad_norm": 3.876828670501709,
      "learning_rate": 3.5801206192082818e-06,
      "loss": 0.2692,
      "num_input_tokens_seen": 14010248,
      "step": 21455
    },
    {
      "epoch": 12.653301886792454,
      "grad_norm": 2.3132376670837402,
      "learning_rate": 3.577654008140582e-06,
      "loss": 0.3109,
      "num_input_tokens_seen": 14013640,
      "step": 21460
    },
    {
      "epoch": 12.65625,
      "grad_norm": 2.5358951091766357,
      "learning_rate": 3.5751877736451123e-06,
      "loss": 0.2891,
      "num_input_tokens_seen": 14016232,
      "step": 21465
    },
    {
      "epoch": 12.659198113207546,
      "grad_norm": 2.573917865753174,
      "learning_rate": 3.5727219163748205e-06,
      "loss": 0.4635,
      "num_input_tokens_seen": 14019496,
      "step": 21470
    },
    {
      "epoch": 12.662146226415095,
      "grad_norm": 3.8243162631988525,
      "learning_rate": 3.570256436982552e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 14023048,
      "step": 21475
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 4.253672122955322,
      "learning_rate": 3.5677913361210536e-06,
      "loss": 0.2954,
      "num_input_tokens_seen": 14026408,
      "step": 21480
    },
    {
      "epoch": 12.66804245283019,
      "grad_norm": 3.681011915206909,
      "learning_rate": 3.565326614442972e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 14030632,
      "step": 21485
    },
    {
      "epoch": 12.670990566037736,
      "grad_norm": 2.8278744220733643,
      "learning_rate": 3.5628622726008523e-06,
      "loss": 0.3058,
      "num_input_tokens_seen": 14034056,
      "step": 21490
    },
    {
      "epoch": 12.673938679245284,
      "grad_norm": 3.3410496711730957,
      "learning_rate": 3.56039831124714e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 14037384,
      "step": 21495
    },
    {
      "epoch": 12.67688679245283,
      "grad_norm": 3.90824031829834,
      "learning_rate": 3.557934731034179e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 14041704,
      "step": 21500
    },
    {
      "epoch": 12.679834905660378,
      "grad_norm": 5.7740583419799805,
      "learning_rate": 3.5554715326142126e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 14044776,
      "step": 21505
    },
    {
      "epoch": 12.682783018867925,
      "grad_norm": 3.2105395793914795,
      "learning_rate": 3.553008716639384e-06,
      "loss": 0.3239,
      "num_input_tokens_seen": 14047464,
      "step": 21510
    },
    {
      "epoch": 12.685731132075471,
      "grad_norm": 3.271907091140747,
      "learning_rate": 3.5505462837617338e-06,
      "loss": 0.3782,
      "num_input_tokens_seen": 14050632,
      "step": 21515
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 2.6794464588165283,
      "learning_rate": 3.5480842346332013e-06,
      "loss": 0.2999,
      "num_input_tokens_seen": 14054376,
      "step": 21520
    },
    {
      "epoch": 12.691627358490566,
      "grad_norm": 5.829660415649414,
      "learning_rate": 3.5456225699056256e-06,
      "loss": 0.3978,
      "num_input_tokens_seen": 14056488,
      "step": 21525
    },
    {
      "epoch": 12.694575471698114,
      "grad_norm": 3.8052613735198975,
      "learning_rate": 3.5431612902307426e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 14061256,
      "step": 21530
    },
    {
      "epoch": 12.69752358490566,
      "grad_norm": 2.497452974319458,
      "learning_rate": 3.540700396260186e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 14064456,
      "step": 21535
    },
    {
      "epoch": 12.700471698113208,
      "grad_norm": 4.763362407684326,
      "learning_rate": 3.538239888645489e-06,
      "loss": 0.476,
      "num_input_tokens_seen": 14068808,
      "step": 21540
    },
    {
      "epoch": 12.703419811320755,
      "grad_norm": 3.9111428260803223,
      "learning_rate": 3.535779768038082e-06,
      "loss": 0.3754,
      "num_input_tokens_seen": 14071528,
      "step": 21545
    },
    {
      "epoch": 12.706367924528301,
      "grad_norm": 2.7621002197265625,
      "learning_rate": 3.5333200350892905e-06,
      "loss": 0.2935,
      "num_input_tokens_seen": 14074920,
      "step": 21550
    },
    {
      "epoch": 12.70931603773585,
      "grad_norm": 3.9427614212036133,
      "learning_rate": 3.530860690450342e-06,
      "loss": 0.2427,
      "num_input_tokens_seen": 14077576,
      "step": 21555
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 4.892881870269775,
      "learning_rate": 3.528401734772357e-06,
      "loss": 0.2802,
      "num_input_tokens_seen": 14080456,
      "step": 21560
    },
    {
      "epoch": 12.715212264150944,
      "grad_norm": 3.174333095550537,
      "learning_rate": 3.5259431687063538e-06,
      "loss": 0.3496,
      "num_input_tokens_seen": 14083848,
      "step": 21565
    },
    {
      "epoch": 12.71816037735849,
      "grad_norm": 2.914666175842285,
      "learning_rate": 3.523484992903249e-06,
      "loss": 0.4062,
      "num_input_tokens_seen": 14087176,
      "step": 21570
    },
    {
      "epoch": 12.721108490566039,
      "grad_norm": 2.679306745529175,
      "learning_rate": 3.5210272080138573e-06,
      "loss": 0.2946,
      "num_input_tokens_seen": 14090536,
      "step": 21575
    },
    {
      "epoch": 12.724056603773585,
      "grad_norm": 2.8195674419403076,
      "learning_rate": 3.518569814688887e-06,
      "loss": 0.3142,
      "num_input_tokens_seen": 14093736,
      "step": 21580
    },
    {
      "epoch": 12.727004716981131,
      "grad_norm": 5.794471740722656,
      "learning_rate": 3.5161128135789414e-06,
      "loss": 0.309,
      "num_input_tokens_seen": 14096328,
      "step": 21585
    },
    {
      "epoch": 12.72995283018868,
      "grad_norm": 8.51036548614502,
      "learning_rate": 3.513656205334525e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 14099048,
      "step": 21590
    },
    {
      "epoch": 12.732900943396226,
      "grad_norm": 3.3468260765075684,
      "learning_rate": 3.5111999906060336e-06,
      "loss": 0.4335,
      "num_input_tokens_seen": 14102152,
      "step": 21595
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 2.531296730041504,
      "learning_rate": 3.50874417004376e-06,
      "loss": 0.3074,
      "num_input_tokens_seen": 14105288,
      "step": 21600
    },
    {
      "epoch": 12.73879716981132,
      "grad_norm": 4.7313361167907715,
      "learning_rate": 3.5062887442978956e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 14108168,
      "step": 21605
    },
    {
      "epoch": 12.741745283018869,
      "grad_norm": 3.5229759216308594,
      "learning_rate": 3.503833714018524e-06,
      "loss": 0.3537,
      "num_input_tokens_seen": 14111400,
      "step": 21610
    },
    {
      "epoch": 12.744693396226415,
      "grad_norm": 3.2553887367248535,
      "learning_rate": 3.5013790798556228e-06,
      "loss": 0.4707,
      "num_input_tokens_seen": 14114536,
      "step": 21615
    },
    {
      "epoch": 12.747641509433961,
      "grad_norm": 3.701028823852539,
      "learning_rate": 3.4989248424590705e-06,
      "loss": 0.3075,
      "num_input_tokens_seen": 14117448,
      "step": 21620
    },
    {
      "epoch": 12.75058962264151,
      "grad_norm": 3.1283645629882812,
      "learning_rate": 3.4964710024786354e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 14121352,
      "step": 21625
    },
    {
      "epoch": 12.753537735849056,
      "grad_norm": 2.4502949714660645,
      "learning_rate": 3.4940175605639813e-06,
      "loss": 0.3996,
      "num_input_tokens_seen": 14127208,
      "step": 21630
    },
    {
      "epoch": 12.756485849056604,
      "grad_norm": 2.6479060649871826,
      "learning_rate": 3.4915645173646694e-06,
      "loss": 0.4804,
      "num_input_tokens_seen": 14130056,
      "step": 21635
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 3.0141592025756836,
      "learning_rate": 3.489111873530153e-06,
      "loss": 0.3673,
      "num_input_tokens_seen": 14133480,
      "step": 21640
    },
    {
      "epoch": 12.762382075471699,
      "grad_norm": 5.657107353210449,
      "learning_rate": 3.4866596297097776e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 14136072,
      "step": 21645
    },
    {
      "epoch": 12.765330188679245,
      "grad_norm": 1.9518108367919922,
      "learning_rate": 3.484207786552789e-06,
      "loss": 0.4179,
      "num_input_tokens_seen": 14138568,
      "step": 21650
    },
    {
      "epoch": 12.768278301886792,
      "grad_norm": 2.9988982677459717,
      "learning_rate": 3.4817563447083214e-06,
      "loss": 0.3683,
      "num_input_tokens_seen": 14141064,
      "step": 21655
    },
    {
      "epoch": 12.77122641509434,
      "grad_norm": 1.9335395097732544,
      "learning_rate": 3.4793053048254044e-06,
      "loss": 0.3458,
      "num_input_tokens_seen": 14144616,
      "step": 21660
    },
    {
      "epoch": 12.774174528301886,
      "grad_norm": 3.8155980110168457,
      "learning_rate": 3.47685466755296e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 14147784,
      "step": 21665
    },
    {
      "epoch": 12.777122641509434,
      "grad_norm": 5.824955463409424,
      "learning_rate": 3.474404433539809e-06,
      "loss": 0.2797,
      "num_input_tokens_seen": 14150696,
      "step": 21670
    },
    {
      "epoch": 12.78007075471698,
      "grad_norm": 4.176796913146973,
      "learning_rate": 3.4719546034346598e-06,
      "loss": 0.3357,
      "num_input_tokens_seen": 14153800,
      "step": 21675
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 3.1004860401153564,
      "learning_rate": 3.4695051778861125e-06,
      "loss": 0.4383,
      "num_input_tokens_seen": 14156808,
      "step": 21680
    },
    {
      "epoch": 12.785966981132075,
      "grad_norm": 2.2996182441711426,
      "learning_rate": 3.4670561575426677e-06,
      "loss": 0.2614,
      "num_input_tokens_seen": 14159656,
      "step": 21685
    },
    {
      "epoch": 12.788915094339622,
      "grad_norm": 2.7980830669403076,
      "learning_rate": 3.4646075430527115e-06,
      "loss": 0.4345,
      "num_input_tokens_seen": 14163560,
      "step": 21690
    },
    {
      "epoch": 12.79186320754717,
      "grad_norm": 2.5118000507354736,
      "learning_rate": 3.4621593350645236e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 14166984,
      "step": 21695
    },
    {
      "epoch": 12.794811320754716,
      "grad_norm": 7.4615254402160645,
      "learning_rate": 3.4597115342262817e-06,
      "loss": 0.3991,
      "num_input_tokens_seen": 14170120,
      "step": 21700
    },
    {
      "epoch": 12.797759433962264,
      "grad_norm": 1.8916300535202026,
      "learning_rate": 3.4572641411860484e-06,
      "loss": 0.2486,
      "num_input_tokens_seen": 14173864,
      "step": 21705
    },
    {
      "epoch": 12.80070754716981,
      "grad_norm": 3.637920379638672,
      "learning_rate": 3.454817156591782e-06,
      "loss": 0.5266,
      "num_input_tokens_seen": 14177224,
      "step": 21710
    },
    {
      "epoch": 12.803655660377359,
      "grad_norm": 4.433169841766357,
      "learning_rate": 3.4523705810913344e-06,
      "loss": 0.2575,
      "num_input_tokens_seen": 14180520,
      "step": 21715
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 2.1424367427825928,
      "learning_rate": 3.449924415332443e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 14188360,
      "step": 21720
    },
    {
      "epoch": 12.809551886792454,
      "grad_norm": 5.58449649810791,
      "learning_rate": 3.447478659962745e-06,
      "loss": 0.378,
      "num_input_tokens_seen": 14192136,
      "step": 21725
    },
    {
      "epoch": 12.8125,
      "grad_norm": 4.494443893432617,
      "learning_rate": 3.4450333156297625e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 14195496,
      "step": 21730
    },
    {
      "epoch": 12.815448113207546,
      "grad_norm": 2.9863317012786865,
      "learning_rate": 3.44258838298091e-06,
      "loss": 0.3435,
      "num_input_tokens_seen": 14199176,
      "step": 21735
    },
    {
      "epoch": 12.818396226415095,
      "grad_norm": 3.4902408123016357,
      "learning_rate": 3.440143862663497e-06,
      "loss": 0.4856,
      "num_input_tokens_seen": 14201800,
      "step": 21740
    },
    {
      "epoch": 12.821344339622641,
      "grad_norm": 3.3114216327667236,
      "learning_rate": 3.4376997553247183e-06,
      "loss": 0.3981,
      "num_input_tokens_seen": 14206120,
      "step": 21745
    },
    {
      "epoch": 12.82429245283019,
      "grad_norm": 1.901628851890564,
      "learning_rate": 3.4352560616116617e-06,
      "loss": 0.3052,
      "num_input_tokens_seen": 14208552,
      "step": 21750
    },
    {
      "epoch": 12.827240566037736,
      "grad_norm": 2.28065824508667,
      "learning_rate": 3.4328127821713077e-06,
      "loss": 0.308,
      "num_input_tokens_seen": 14210824,
      "step": 21755
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 3.3978164196014404,
      "learning_rate": 3.430369917650521e-06,
      "loss": 0.3258,
      "num_input_tokens_seen": 14214248,
      "step": 21760
    },
    {
      "epoch": 12.83313679245283,
      "grad_norm": 3.46160888671875,
      "learning_rate": 3.427927468696066e-06,
      "loss": 0.3158,
      "num_input_tokens_seen": 14217256,
      "step": 21765
    },
    {
      "epoch": 12.836084905660378,
      "grad_norm": 3.102599620819092,
      "learning_rate": 3.425485435954588e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 14222088,
      "step": 21770
    },
    {
      "epoch": 12.839033018867925,
      "grad_norm": 5.633053779602051,
      "learning_rate": 3.4230438200726274e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 14224360,
      "step": 21775
    },
    {
      "epoch": 12.841981132075471,
      "grad_norm": 1.9559446573257446,
      "learning_rate": 3.4206026216966113e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 14227144,
      "step": 21780
    },
    {
      "epoch": 12.84492924528302,
      "grad_norm": 4.479199409484863,
      "learning_rate": 3.418161841472858e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 14230664,
      "step": 21785
    },
    {
      "epoch": 12.847877358490566,
      "grad_norm": 3.390227794647217,
      "learning_rate": 3.4157214800475746e-06,
      "loss": 0.4452,
      "num_input_tokens_seen": 14233544,
      "step": 21790
    },
    {
      "epoch": 12.850825471698114,
      "grad_norm": 2.3024675846099854,
      "learning_rate": 3.4132815380668577e-06,
      "loss": 0.3691,
      "num_input_tokens_seen": 14236744,
      "step": 21795
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 2.436985492706299,
      "learning_rate": 3.410842016176691e-06,
      "loss": 0.2779,
      "num_input_tokens_seen": 14240456,
      "step": 21800
    },
    {
      "epoch": 12.856721698113208,
      "grad_norm": 4.819080829620361,
      "learning_rate": 3.4084029150229503e-06,
      "loss": 0.4467,
      "num_input_tokens_seen": 14243496,
      "step": 21805
    },
    {
      "epoch": 12.859669811320755,
      "grad_norm": 2.3516790866851807,
      "learning_rate": 3.4059642352513965e-06,
      "loss": 0.2988,
      "num_input_tokens_seen": 14246792,
      "step": 21810
    },
    {
      "epoch": 12.862617924528301,
      "grad_norm": 3.366542339324951,
      "learning_rate": 3.4035259775076813e-06,
      "loss": 0.3702,
      "num_input_tokens_seen": 14249160,
      "step": 21815
    },
    {
      "epoch": 12.86556603773585,
      "grad_norm": 3.2092792987823486,
      "learning_rate": 3.401088142437344e-06,
      "loss": 0.302,
      "num_input_tokens_seen": 14253608,
      "step": 21820
    },
    {
      "epoch": 12.868514150943396,
      "grad_norm": 3.760228395462036,
      "learning_rate": 3.398650730685813e-06,
      "loss": 0.3098,
      "num_input_tokens_seen": 14255976,
      "step": 21825
    },
    {
      "epoch": 12.871462264150944,
      "grad_norm": 3.425318479537964,
      "learning_rate": 3.396213742898401e-06,
      "loss": 0.2837,
      "num_input_tokens_seen": 14258248,
      "step": 21830
    },
    {
      "epoch": 12.87441037735849,
      "grad_norm": 6.024099349975586,
      "learning_rate": 3.3937771797203134e-06,
      "loss": 0.4995,
      "num_input_tokens_seen": 14261736,
      "step": 21835
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 2.1552140712738037,
      "learning_rate": 3.391341041796641e-06,
      "loss": 0.3238,
      "num_input_tokens_seen": 14264840,
      "step": 21840
    },
    {
      "epoch": 12.880306603773585,
      "grad_norm": 5.012811183929443,
      "learning_rate": 3.3889053297723585e-06,
      "loss": 0.4597,
      "num_input_tokens_seen": 14267688,
      "step": 21845
    },
    {
      "epoch": 12.883254716981131,
      "grad_norm": 2.8513922691345215,
      "learning_rate": 3.3864700442923342e-06,
      "loss": 0.3609,
      "num_input_tokens_seen": 14270728,
      "step": 21850
    },
    {
      "epoch": 12.88620283018868,
      "grad_norm": 1.6877036094665527,
      "learning_rate": 3.384035186001318e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 14274856,
      "step": 21855
    },
    {
      "epoch": 12.889150943396226,
      "grad_norm": 5.6138505935668945,
      "learning_rate": 3.381600755543953e-06,
      "loss": 0.4629,
      "num_input_tokens_seen": 14277928,
      "step": 21860
    },
    {
      "epoch": 12.892099056603774,
      "grad_norm": 3.577425479888916,
      "learning_rate": 3.3791667535647615e-06,
      "loss": 0.3219,
      "num_input_tokens_seen": 14280520,
      "step": 21865
    },
    {
      "epoch": 12.89504716981132,
      "grad_norm": 3.446071147918701,
      "learning_rate": 3.3767331807081584e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 14283912,
      "step": 21870
    },
    {
      "epoch": 12.897995283018869,
      "grad_norm": 3.4100711345672607,
      "learning_rate": 3.374300037618442e-06,
      "loss": 0.3312,
      "num_input_tokens_seen": 14287816,
      "step": 21875
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 3.2369730472564697,
      "learning_rate": 3.371867324939796e-06,
      "loss": 0.4031,
      "num_input_tokens_seen": 14290568,
      "step": 21880
    },
    {
      "epoch": 12.903891509433961,
      "grad_norm": 4.009652614593506,
      "learning_rate": 3.369435043316293e-06,
      "loss": 0.3072,
      "num_input_tokens_seen": 14293288,
      "step": 21885
    },
    {
      "epoch": 12.90683962264151,
      "grad_norm": 2.7834365367889404,
      "learning_rate": 3.36700319339189e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 14296648,
      "step": 21890
    },
    {
      "epoch": 12.909787735849056,
      "grad_norm": 3.74892520904541,
      "learning_rate": 3.3645717758104286e-06,
      "loss": 0.2193,
      "num_input_tokens_seen": 14299272,
      "step": 21895
    },
    {
      "epoch": 12.912735849056604,
      "grad_norm": 4.38859224319458,
      "learning_rate": 3.3621407912156383e-06,
      "loss": 0.3009,
      "num_input_tokens_seen": 14302280,
      "step": 21900
    },
    {
      "epoch": 12.91568396226415,
      "grad_norm": 3.277360439300537,
      "learning_rate": 3.3597102402511326e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 14306056,
      "step": 21905
    },
    {
      "epoch": 12.918632075471699,
      "grad_norm": 2.9219894409179688,
      "learning_rate": 3.3572801235604093e-06,
      "loss": 0.3363,
      "num_input_tokens_seen": 14309832,
      "step": 21910
    },
    {
      "epoch": 12.921580188679245,
      "grad_norm": 2.9536354541778564,
      "learning_rate": 3.3548504417868538e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 14312584,
      "step": 21915
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 3.0462090969085693,
      "learning_rate": 3.352421195573734e-06,
      "loss": 0.2871,
      "num_input_tokens_seen": 14315304,
      "step": 21920
    },
    {
      "epoch": 12.92747641509434,
      "grad_norm": 3.3603479862213135,
      "learning_rate": 3.3499923855642026e-06,
      "loss": 0.4661,
      "num_input_tokens_seen": 14317544,
      "step": 21925
    },
    {
      "epoch": 12.930424528301886,
      "grad_norm": 3.1318767070770264,
      "learning_rate": 3.3475640124012986e-06,
      "loss": 0.2786,
      "num_input_tokens_seen": 14320264,
      "step": 21930
    },
    {
      "epoch": 12.933372641509434,
      "grad_norm": 3.3394534587860107,
      "learning_rate": 3.345136076727945e-06,
      "loss": 0.3372,
      "num_input_tokens_seen": 14323464,
      "step": 21935
    },
    {
      "epoch": 12.93632075471698,
      "grad_norm": 2.32151198387146,
      "learning_rate": 3.3427085791869453e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 14327784,
      "step": 21940
    },
    {
      "epoch": 12.939268867924529,
      "grad_norm": 3.7095251083374023,
      "learning_rate": 3.3402815204209926e-06,
      "loss": 0.5594,
      "num_input_tokens_seen": 14330792,
      "step": 21945
    },
    {
      "epoch": 12.942216981132075,
      "grad_norm": 2.98435640335083,
      "learning_rate": 3.337854901072659e-06,
      "loss": 0.2373,
      "num_input_tokens_seen": 14333320,
      "step": 21950
    },
    {
      "epoch": 12.945165094339622,
      "grad_norm": 2.6257152557373047,
      "learning_rate": 3.3354287217844056e-06,
      "loss": 0.4029,
      "num_input_tokens_seen": 14336616,
      "step": 21955
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 4.785362720489502,
      "learning_rate": 3.3330029831985712e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 14340712,
      "step": 21960
    },
    {
      "epoch": 12.951061320754716,
      "grad_norm": 1.8789197206497192,
      "learning_rate": 3.330577685957382e-06,
      "loss": 0.4093,
      "num_input_tokens_seen": 14343816,
      "step": 21965
    },
    {
      "epoch": 12.954009433962264,
      "grad_norm": 3.110896587371826,
      "learning_rate": 3.3281528307029454e-06,
      "loss": 0.4216,
      "num_input_tokens_seen": 14347048,
      "step": 21970
    },
    {
      "epoch": 12.95695754716981,
      "grad_norm": 3.685577392578125,
      "learning_rate": 3.325728418077251e-06,
      "loss": 0.2809,
      "num_input_tokens_seen": 14351368,
      "step": 21975
    },
    {
      "epoch": 12.959905660377359,
      "grad_norm": 3.1376872062683105,
      "learning_rate": 3.3233044487221744e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 14354696,
      "step": 21980
    },
    {
      "epoch": 12.962853773584905,
      "grad_norm": 3.973897695541382,
      "learning_rate": 3.3208809232794715e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 14357640,
      "step": 21985
    },
    {
      "epoch": 12.965801886792454,
      "grad_norm": 3.2455453872680664,
      "learning_rate": 3.3184578423907797e-06,
      "loss": 0.3355,
      "num_input_tokens_seen": 14360520,
      "step": 21990
    },
    {
      "epoch": 12.96875,
      "grad_norm": 3.908141613006592,
      "learning_rate": 3.3160352066976224e-06,
      "loss": 0.4173,
      "num_input_tokens_seen": 14363112,
      "step": 21995
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 2.0486953258514404,
      "learning_rate": 3.3136130168414003e-06,
      "loss": 0.4102,
      "num_input_tokens_seen": 14366792,
      "step": 22000
    },
    {
      "epoch": 12.974646226415095,
      "grad_norm": 2.33557391166687,
      "learning_rate": 3.311191273463401e-06,
      "loss": 0.2942,
      "num_input_tokens_seen": 14372104,
      "step": 22005
    },
    {
      "epoch": 12.977594339622641,
      "grad_norm": 3.4925718307495117,
      "learning_rate": 3.3087699772047908e-06,
      "loss": 0.3461,
      "num_input_tokens_seen": 14376872,
      "step": 22010
    },
    {
      "epoch": 12.98054245283019,
      "grad_norm": 4.298727035522461,
      "learning_rate": 3.3063491287066164e-06,
      "loss": 0.3709,
      "num_input_tokens_seen": 14379336,
      "step": 22015
    },
    {
      "epoch": 12.983490566037736,
      "grad_norm": 3.2458348274230957,
      "learning_rate": 3.303928728609811e-06,
      "loss": 0.33,
      "num_input_tokens_seen": 14381960,
      "step": 22020
    },
    {
      "epoch": 12.986438679245284,
      "grad_norm": 2.6012728214263916,
      "learning_rate": 3.3015087775551835e-06,
      "loss": 0.5452,
      "num_input_tokens_seen": 14385896,
      "step": 22025
    },
    {
      "epoch": 12.98938679245283,
      "grad_norm": 3.7339859008789062,
      "learning_rate": 3.299089276183427e-06,
      "loss": 0.2802,
      "num_input_tokens_seen": 14388904,
      "step": 22030
    },
    {
      "epoch": 12.992334905660378,
      "grad_norm": 3.119081497192383,
      "learning_rate": 3.2966702251351157e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 14392200,
      "step": 22035
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 4.157709121704102,
      "learning_rate": 3.2942516250507035e-06,
      "loss": 0.3192,
      "num_input_tokens_seen": 14395080,
      "step": 22040
    },
    {
      "epoch": 12.998231132075471,
      "grad_norm": 2.216370105743408,
      "learning_rate": 3.2918334765705227e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 14398024,
      "step": 22045
    },
    {
      "epoch": 13.00117924528302,
      "grad_norm": 4.646637916564941,
      "learning_rate": 3.289415780334792e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 14401528,
      "step": 22050
    },
    {
      "epoch": 13.004127358490566,
      "grad_norm": 2.9593827724456787,
      "learning_rate": 3.2869985369836067e-06,
      "loss": 0.2627,
      "num_input_tokens_seen": 14404152,
      "step": 22055
    },
    {
      "epoch": 13.007075471698114,
      "grad_norm": 6.005983352661133,
      "learning_rate": 3.2845817471569406e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 14408152,
      "step": 22060
    },
    {
      "epoch": 13.01002358490566,
      "grad_norm": 2.255528450012207,
      "learning_rate": 3.28216541149465e-06,
      "loss": 0.3165,
      "num_input_tokens_seen": 14411192,
      "step": 22065
    },
    {
      "epoch": 13.012971698113208,
      "grad_norm": 2.5795624256134033,
      "learning_rate": 3.2797495306364707e-06,
      "loss": 0.2817,
      "num_input_tokens_seen": 14414008,
      "step": 22070
    },
    {
      "epoch": 13.015919811320755,
      "grad_norm": 2.527611494064331,
      "learning_rate": 3.2773341052220174e-06,
      "loss": 0.2757,
      "num_input_tokens_seen": 14417496,
      "step": 22075
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 2.7541284561157227,
      "learning_rate": 3.274919135890783e-06,
      "loss": 0.2228,
      "num_input_tokens_seen": 14421176,
      "step": 22080
    },
    {
      "epoch": 13.02181603773585,
      "grad_norm": 4.224510192871094,
      "learning_rate": 3.2725046232821424e-06,
      "loss": 0.4533,
      "num_input_tokens_seen": 14423960,
      "step": 22085
    },
    {
      "epoch": 13.024764150943396,
      "grad_norm": 2.616170644760132,
      "learning_rate": 3.270090568035348e-06,
      "loss": 0.3354,
      "num_input_tokens_seen": 14427480,
      "step": 22090
    },
    {
      "epoch": 13.027712264150944,
      "grad_norm": 3.0696022510528564,
      "learning_rate": 3.2676769707895306e-06,
      "loss": 0.3075,
      "num_input_tokens_seen": 14431704,
      "step": 22095
    },
    {
      "epoch": 13.03066037735849,
      "grad_norm": 2.204313039779663,
      "learning_rate": 3.2652638321837015e-06,
      "loss": 0.2925,
      "num_input_tokens_seen": 14434968,
      "step": 22100
    },
    {
      "epoch": 13.033608490566039,
      "grad_norm": 3.906071424484253,
      "learning_rate": 3.2628511528567497e-06,
      "loss": 0.2397,
      "num_input_tokens_seen": 14437816,
      "step": 22105
    },
    {
      "epoch": 13.036556603773585,
      "grad_norm": 4.082236289978027,
      "learning_rate": 3.2604389334474407e-06,
      "loss": 0.2366,
      "num_input_tokens_seen": 14440664,
      "step": 22110
    },
    {
      "epoch": 13.039504716981131,
      "grad_norm": 2.2839884757995605,
      "learning_rate": 3.2580271745944224e-06,
      "loss": 0.6123,
      "num_input_tokens_seen": 14443704,
      "step": 22115
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 3.110607624053955,
      "learning_rate": 3.255615876936217e-06,
      "loss": 0.3247,
      "num_input_tokens_seen": 14447064,
      "step": 22120
    },
    {
      "epoch": 13.045400943396226,
      "grad_norm": 2.036837339401245,
      "learning_rate": 3.2532050411112248e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 14450872,
      "step": 22125
    },
    {
      "epoch": 13.048349056603774,
      "grad_norm": 3.9381394386291504,
      "learning_rate": 3.2507946677577274e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 14454264,
      "step": 22130
    },
    {
      "epoch": 13.05129716981132,
      "grad_norm": 4.689486980438232,
      "learning_rate": 3.2483847575138807e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 14458200,
      "step": 22135
    },
    {
      "epoch": 13.054245283018869,
      "grad_norm": 3.764192581176758,
      "learning_rate": 3.245975311017716e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 14461496,
      "step": 22140
    },
    {
      "epoch": 13.057193396226415,
      "grad_norm": 5.570861339569092,
      "learning_rate": 3.2435663289071486e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 14464376,
      "step": 22145
    },
    {
      "epoch": 13.060141509433961,
      "grad_norm": 4.5474853515625,
      "learning_rate": 3.241157811819966e-06,
      "loss": 0.3907,
      "num_input_tokens_seen": 14467608,
      "step": 22150
    },
    {
      "epoch": 13.06308962264151,
      "grad_norm": 3.2539613246917725,
      "learning_rate": 3.2387497603938327e-06,
      "loss": 0.3975,
      "num_input_tokens_seen": 14470808,
      "step": 22155
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 1.9355475902557373,
      "learning_rate": 3.2363421752662903e-06,
      "loss": 0.4252,
      "num_input_tokens_seen": 14473208,
      "step": 22160
    },
    {
      "epoch": 13.068985849056604,
      "grad_norm": 7.4225873947143555,
      "learning_rate": 3.233935057074759e-06,
      "loss": 0.4194,
      "num_input_tokens_seen": 14475992,
      "step": 22165
    },
    {
      "epoch": 13.07193396226415,
      "grad_norm": 3.8156473636627197,
      "learning_rate": 3.2315284064565324e-06,
      "loss": 0.4629,
      "num_input_tokens_seen": 14479320,
      "step": 22170
    },
    {
      "epoch": 13.074882075471699,
      "grad_norm": 2.471977710723877,
      "learning_rate": 3.2291222240487813e-06,
      "loss": 0.3188,
      "num_input_tokens_seen": 14482136,
      "step": 22175
    },
    {
      "epoch": 13.077830188679245,
      "grad_norm": 2.3766062259674072,
      "learning_rate": 3.226716510488554e-06,
      "loss": 0.3814,
      "num_input_tokens_seen": 14485624,
      "step": 22180
    },
    {
      "epoch": 13.080778301886792,
      "grad_norm": 2.902137517929077,
      "learning_rate": 3.224311266412773e-06,
      "loss": 0.4103,
      "num_input_tokens_seen": 14488056,
      "step": 22185
    },
    {
      "epoch": 13.08372641509434,
      "grad_norm": 4.566019058227539,
      "learning_rate": 3.2219064924582366e-06,
      "loss": 0.308,
      "num_input_tokens_seen": 14491064,
      "step": 22190
    },
    {
      "epoch": 13.086674528301886,
      "grad_norm": 4.006875514984131,
      "learning_rate": 3.2195021892616197e-06,
      "loss": 0.3079,
      "num_input_tokens_seen": 14494392,
      "step": 22195
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 2.515698194503784,
      "learning_rate": 3.217098357459472e-06,
      "loss": 0.3929,
      "num_input_tokens_seen": 14497752,
      "step": 22200
    },
    {
      "epoch": 13.09257075471698,
      "grad_norm": 8.718137741088867,
      "learning_rate": 3.214694997688217e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 14501592,
      "step": 22205
    },
    {
      "epoch": 13.095518867924529,
      "grad_norm": 4.432575225830078,
      "learning_rate": 3.2122921105841572e-06,
      "loss": 0.2461,
      "num_input_tokens_seen": 14504120,
      "step": 22210
    },
    {
      "epoch": 13.098466981132075,
      "grad_norm": 3.068951368331909,
      "learning_rate": 3.2098896967834647e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 14507352,
      "step": 22215
    },
    {
      "epoch": 13.101415094339623,
      "grad_norm": 4.468780040740967,
      "learning_rate": 3.2074877569221896e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 14509880,
      "step": 22220
    },
    {
      "epoch": 13.10436320754717,
      "grad_norm": 3.566251754760742,
      "learning_rate": 3.205086291636257e-06,
      "loss": 0.2774,
      "num_input_tokens_seen": 14512440,
      "step": 22225
    },
    {
      "epoch": 13.107311320754716,
      "grad_norm": 2.8369638919830322,
      "learning_rate": 3.202685301561463e-06,
      "loss": 0.3832,
      "num_input_tokens_seen": 14519960,
      "step": 22230
    },
    {
      "epoch": 13.110259433962264,
      "grad_norm": 3.0071258544921875,
      "learning_rate": 3.200284787333482e-06,
      "loss": 0.3324,
      "num_input_tokens_seen": 14524344,
      "step": 22235
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 2.6962947845458984,
      "learning_rate": 3.1978847495878595e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 14527096,
      "step": 22240
    },
    {
      "epoch": 13.116155660377359,
      "grad_norm": 2.2467117309570312,
      "learning_rate": 3.1954851889600176e-06,
      "loss": 0.4646,
      "num_input_tokens_seen": 14530584,
      "step": 22245
    },
    {
      "epoch": 13.119103773584905,
      "grad_norm": 8.0064115524292,
      "learning_rate": 3.1930861060852485e-06,
      "loss": 0.306,
      "num_input_tokens_seen": 14533208,
      "step": 22250
    },
    {
      "epoch": 13.122051886792454,
      "grad_norm": 2.3811051845550537,
      "learning_rate": 3.1906875015987194e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 14536120,
      "step": 22255
    },
    {
      "epoch": 13.125,
      "grad_norm": 4.627654075622559,
      "learning_rate": 3.188289376135473e-06,
      "loss": 0.2925,
      "num_input_tokens_seen": 14539320,
      "step": 22260
    },
    {
      "epoch": 13.127948113207546,
      "grad_norm": 2.816824436187744,
      "learning_rate": 3.1858917303304213e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 14541816,
      "step": 22265
    },
    {
      "epoch": 13.130896226415095,
      "grad_norm": 2.778346300125122,
      "learning_rate": 3.1834945648183535e-06,
      "loss": 0.2855,
      "num_input_tokens_seen": 14546008,
      "step": 22270
    },
    {
      "epoch": 13.133844339622641,
      "grad_norm": 3.406909227371216,
      "learning_rate": 3.1810978802339283e-06,
      "loss": 0.2882,
      "num_input_tokens_seen": 14548760,
      "step": 22275
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 2.1889188289642334,
      "learning_rate": 3.1787016772116767e-06,
      "loss": 0.262,
      "num_input_tokens_seen": 14551352,
      "step": 22280
    },
    {
      "epoch": 13.139740566037736,
      "grad_norm": 2.7602121829986572,
      "learning_rate": 3.1763059563860073e-06,
      "loss": 0.4762,
      "num_input_tokens_seen": 14555032,
      "step": 22285
    },
    {
      "epoch": 13.142688679245284,
      "grad_norm": 4.991405010223389,
      "learning_rate": 3.1739107183911953e-06,
      "loss": 0.2672,
      "num_input_tokens_seen": 14559704,
      "step": 22290
    },
    {
      "epoch": 13.14563679245283,
      "grad_norm": 3.3546197414398193,
      "learning_rate": 3.1715159638613898e-06,
      "loss": 0.1725,
      "num_input_tokens_seen": 14563320,
      "step": 22295
    },
    {
      "epoch": 13.148584905660377,
      "grad_norm": 2.2204582691192627,
      "learning_rate": 3.1691216934306134e-06,
      "loss": 0.3341,
      "num_input_tokens_seen": 14567864,
      "step": 22300
    },
    {
      "epoch": 13.151533018867925,
      "grad_norm": 3.074469804763794,
      "learning_rate": 3.16672790773276e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 14570808,
      "step": 22305
    },
    {
      "epoch": 13.154481132075471,
      "grad_norm": 2.3233509063720703,
      "learning_rate": 3.164334607401593e-06,
      "loss": 0.3638,
      "num_input_tokens_seen": 14573784,
      "step": 22310
    },
    {
      "epoch": 13.15742924528302,
      "grad_norm": 2.663646697998047,
      "learning_rate": 3.1619417930707506e-06,
      "loss": 0.3214,
      "num_input_tokens_seen": 14577304,
      "step": 22315
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 2.198265314102173,
      "learning_rate": 3.1595494653737408e-06,
      "loss": 0.2513,
      "num_input_tokens_seen": 14579992,
      "step": 22320
    },
    {
      "epoch": 13.163325471698114,
      "grad_norm": 3.081104040145874,
      "learning_rate": 3.1571576249439408e-06,
      "loss": 0.3274,
      "num_input_tokens_seen": 14583096,
      "step": 22325
    },
    {
      "epoch": 13.16627358490566,
      "grad_norm": 3.0762689113616943,
      "learning_rate": 3.1547662724146e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 14586424,
      "step": 22330
    },
    {
      "epoch": 13.169221698113208,
      "grad_norm": 3.297476053237915,
      "learning_rate": 3.1523754084188436e-06,
      "loss": 0.303,
      "num_input_tokens_seen": 14589080,
      "step": 22335
    },
    {
      "epoch": 13.172169811320755,
      "grad_norm": 2.9362716674804688,
      "learning_rate": 3.149985033589661e-06,
      "loss": 0.3839,
      "num_input_tokens_seen": 14592600,
      "step": 22340
    },
    {
      "epoch": 13.175117924528301,
      "grad_norm": 4.702031135559082,
      "learning_rate": 3.147595148559912e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 14596216,
      "step": 22345
    },
    {
      "epoch": 13.17806603773585,
      "grad_norm": 3.12243390083313,
      "learning_rate": 3.1452057539623328e-06,
      "loss": 0.2926,
      "num_input_tokens_seen": 14599640,
      "step": 22350
    },
    {
      "epoch": 13.181014150943396,
      "grad_norm": 3.0624701976776123,
      "learning_rate": 3.142816850429523e-06,
      "loss": 0.3689,
      "num_input_tokens_seen": 14602424,
      "step": 22355
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 2.1947414875030518,
      "learning_rate": 3.1404284385939552e-06,
      "loss": 0.3285,
      "num_input_tokens_seen": 14604984,
      "step": 22360
    },
    {
      "epoch": 13.18691037735849,
      "grad_norm": 3.1189255714416504,
      "learning_rate": 3.138040519087975e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 14608056,
      "step": 22365
    },
    {
      "epoch": 13.189858490566039,
      "grad_norm": 3.3615269660949707,
      "learning_rate": 3.13565309254379e-06,
      "loss": 0.2382,
      "num_input_tokens_seen": 14610520,
      "step": 22370
    },
    {
      "epoch": 13.192806603773585,
      "grad_norm": 2.096233606338501,
      "learning_rate": 3.1332661595934845e-06,
      "loss": 0.3003,
      "num_input_tokens_seen": 14614904,
      "step": 22375
    },
    {
      "epoch": 13.195754716981131,
      "grad_norm": 2.0336666107177734,
      "learning_rate": 3.130879720869008e-06,
      "loss": 0.32,
      "num_input_tokens_seen": 14618456,
      "step": 22380
    },
    {
      "epoch": 13.19870283018868,
      "grad_norm": 5.367611408233643,
      "learning_rate": 3.1284937770021815e-06,
      "loss": 0.3777,
      "num_input_tokens_seen": 14621208,
      "step": 22385
    },
    {
      "epoch": 13.201650943396226,
      "grad_norm": 4.460502624511719,
      "learning_rate": 3.1261083286246916e-06,
      "loss": 0.3233,
      "num_input_tokens_seen": 14623704,
      "step": 22390
    },
    {
      "epoch": 13.204599056603774,
      "grad_norm": 4.48661470413208,
      "learning_rate": 3.1237233763680997e-06,
      "loss": 0.3131,
      "num_input_tokens_seen": 14626968,
      "step": 22395
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 2.9051289558410645,
      "learning_rate": 3.1213389208638303e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 14629976,
      "step": 22400
    },
    {
      "epoch": 13.210495283018869,
      "grad_norm": 3.3281054496765137,
      "learning_rate": 3.1189549627431757e-06,
      "loss": 0.3085,
      "num_input_tokens_seen": 14632728,
      "step": 22405
    },
    {
      "epoch": 13.213443396226415,
      "grad_norm": 3.5119810104370117,
      "learning_rate": 3.116571502637304e-06,
      "loss": 0.3988,
      "num_input_tokens_seen": 14635352,
      "step": 22410
    },
    {
      "epoch": 13.216391509433961,
      "grad_norm": 1.7424523830413818,
      "learning_rate": 3.1141885411772434e-06,
      "loss": 0.3662,
      "num_input_tokens_seen": 14638872,
      "step": 22415
    },
    {
      "epoch": 13.21933962264151,
      "grad_norm": 3.375314235687256,
      "learning_rate": 3.111806078993893e-06,
      "loss": 0.3245,
      "num_input_tokens_seen": 14642584,
      "step": 22420
    },
    {
      "epoch": 13.222287735849056,
      "grad_norm": 3.886932611465454,
      "learning_rate": 3.1094241167180223e-06,
      "loss": 0.3457,
      "num_input_tokens_seen": 14645144,
      "step": 22425
    },
    {
      "epoch": 13.225235849056604,
      "grad_norm": 2.6072957515716553,
      "learning_rate": 3.1070426549802623e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 14648792,
      "step": 22430
    },
    {
      "epoch": 13.22818396226415,
      "grad_norm": 2.1343703269958496,
      "learning_rate": 3.1046616944111196e-06,
      "loss": 0.2418,
      "num_input_tokens_seen": 14651864,
      "step": 22435
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 3.852203845977783,
      "learning_rate": 3.1022812356409606e-06,
      "loss": 0.3015,
      "num_input_tokens_seen": 14655160,
      "step": 22440
    },
    {
      "epoch": 13.234080188679245,
      "grad_norm": 2.5447423458099365,
      "learning_rate": 3.0999012793000244e-06,
      "loss": 0.3164,
      "num_input_tokens_seen": 14658584,
      "step": 22445
    },
    {
      "epoch": 13.237028301886792,
      "grad_norm": 2.9778213500976562,
      "learning_rate": 3.097521826018414e-06,
      "loss": 0.3364,
      "num_input_tokens_seen": 14661528,
      "step": 22450
    },
    {
      "epoch": 13.23997641509434,
      "grad_norm": 2.021454095840454,
      "learning_rate": 3.0951428764260973e-06,
      "loss": 0.2522,
      "num_input_tokens_seen": 14664440,
      "step": 22455
    },
    {
      "epoch": 13.242924528301886,
      "grad_norm": 2.299699306488037,
      "learning_rate": 3.092764431152915e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 14667512,
      "step": 22460
    },
    {
      "epoch": 13.245872641509434,
      "grad_norm": 4.1679253578186035,
      "learning_rate": 3.0903864908285693e-06,
      "loss": 0.3867,
      "num_input_tokens_seen": 14672056,
      "step": 22465
    },
    {
      "epoch": 13.24882075471698,
      "grad_norm": 3.7334322929382324,
      "learning_rate": 3.088009056082629e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 14674744,
      "step": 22470
    },
    {
      "epoch": 13.251768867924529,
      "grad_norm": 2.2841432094573975,
      "learning_rate": 3.0856321275445324e-06,
      "loss": 0.3045,
      "num_input_tokens_seen": 14679064,
      "step": 22475
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 3.067225217819214,
      "learning_rate": 3.0832557058435808e-06,
      "loss": 0.3336,
      "num_input_tokens_seen": 14681592,
      "step": 22480
    },
    {
      "epoch": 13.257665094339623,
      "grad_norm": 2.4814722537994385,
      "learning_rate": 3.0808797916089405e-06,
      "loss": 0.3359,
      "num_input_tokens_seen": 14684472,
      "step": 22485
    },
    {
      "epoch": 13.26061320754717,
      "grad_norm": 2.9223203659057617,
      "learning_rate": 3.078504385469647e-06,
      "loss": 0.275,
      "num_input_tokens_seen": 14687608,
      "step": 22490
    },
    {
      "epoch": 13.263561320754716,
      "grad_norm": 1.9499902725219727,
      "learning_rate": 3.076129488054599e-06,
      "loss": 0.2416,
      "num_input_tokens_seen": 14691512,
      "step": 22495
    },
    {
      "epoch": 13.266509433962264,
      "grad_norm": 1.3858519792556763,
      "learning_rate": 3.0737550999925604e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 14695256,
      "step": 22500
    },
    {
      "epoch": 13.26945754716981,
      "grad_norm": 6.072397708892822,
      "learning_rate": 3.0713812219121604e-06,
      "loss": 0.3847,
      "num_input_tokens_seen": 14698872,
      "step": 22505
    },
    {
      "epoch": 13.272405660377359,
      "grad_norm": 4.0680766105651855,
      "learning_rate": 3.0690078544418934e-06,
      "loss": 0.3909,
      "num_input_tokens_seen": 14702968,
      "step": 22510
    },
    {
      "epoch": 13.275353773584905,
      "grad_norm": 2.610506296157837,
      "learning_rate": 3.0666349982101198e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 14706200,
      "step": 22515
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 2.6379482746124268,
      "learning_rate": 3.0642626538450627e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 14709816,
      "step": 22520
    },
    {
      "epoch": 13.28125,
      "grad_norm": 4.217516899108887,
      "learning_rate": 3.061890821974809e-06,
      "loss": 0.3322,
      "num_input_tokens_seen": 14712440,
      "step": 22525
    },
    {
      "epoch": 13.284198113207546,
      "grad_norm": 2.4660167694091797,
      "learning_rate": 3.059519503227313e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 14715832,
      "step": 22530
    },
    {
      "epoch": 13.287146226415095,
      "grad_norm": 4.443805694580078,
      "learning_rate": 3.057148698230393e-06,
      "loss": 0.2512,
      "num_input_tokens_seen": 14718776,
      "step": 22535
    },
    {
      "epoch": 13.290094339622641,
      "grad_norm": 2.502333641052246,
      "learning_rate": 3.0547784076117294e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 14722360,
      "step": 22540
    },
    {
      "epoch": 13.29304245283019,
      "grad_norm": 3.991642475128174,
      "learning_rate": 3.0524086319988635e-06,
      "loss": 0.4854,
      "num_input_tokens_seen": 14725112,
      "step": 22545
    },
    {
      "epoch": 13.295990566037736,
      "grad_norm": 3.3328707218170166,
      "learning_rate": 3.0500393720192074e-06,
      "loss": 0.3355,
      "num_input_tokens_seen": 14728056,
      "step": 22550
    },
    {
      "epoch": 13.298938679245284,
      "grad_norm": 2.3518803119659424,
      "learning_rate": 3.047670628300031e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 14730744,
      "step": 22555
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 3.4272477626800537,
      "learning_rate": 3.0453024014684694e-06,
      "loss": 0.3381,
      "num_input_tokens_seen": 14734296,
      "step": 22560
    },
    {
      "epoch": 13.304834905660377,
      "grad_norm": 3.963034152984619,
      "learning_rate": 3.0429346921515225e-06,
      "loss": 0.2624,
      "num_input_tokens_seen": 14737208,
      "step": 22565
    },
    {
      "epoch": 13.307783018867925,
      "grad_norm": 2.552511215209961,
      "learning_rate": 3.04056750097605e-06,
      "loss": 0.3265,
      "num_input_tokens_seen": 14739896,
      "step": 22570
    },
    {
      "epoch": 13.310731132075471,
      "grad_norm": 6.1869940757751465,
      "learning_rate": 3.0382008285687754e-06,
      "loss": 0.3026,
      "num_input_tokens_seen": 14742584,
      "step": 22575
    },
    {
      "epoch": 13.31367924528302,
      "grad_norm": 4.401779651641846,
      "learning_rate": 3.035834675556287e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 14745240,
      "step": 22580
    },
    {
      "epoch": 13.316627358490566,
      "grad_norm": 3.314774513244629,
      "learning_rate": 3.0334690425650336e-06,
      "loss": 0.3252,
      "num_input_tokens_seen": 14748152,
      "step": 22585
    },
    {
      "epoch": 13.319575471698114,
      "grad_norm": 4.139735221862793,
      "learning_rate": 3.031103930221325e-06,
      "loss": 0.5225,
      "num_input_tokens_seen": 14751160,
      "step": 22590
    },
    {
      "epoch": 13.32252358490566,
      "grad_norm": 3.607975721359253,
      "learning_rate": 3.028739339151338e-06,
      "loss": 0.3466,
      "num_input_tokens_seen": 14753656,
      "step": 22595
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 2.770235776901245,
      "learning_rate": 3.0263752699811067e-06,
      "loss": 0.3488,
      "num_input_tokens_seen": 14757144,
      "step": 22600
    },
    {
      "epoch": 13.328419811320755,
      "grad_norm": 12.468465805053711,
      "learning_rate": 3.0240117233365267e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 14760536,
      "step": 22605
    },
    {
      "epoch": 13.331367924528301,
      "grad_norm": 4.6778950691223145,
      "learning_rate": 3.0216486998433604e-06,
      "loss": 0.3494,
      "num_input_tokens_seen": 14763640,
      "step": 22610
    },
    {
      "epoch": 13.33431603773585,
      "grad_norm": 2.9534060955047607,
      "learning_rate": 3.0192862001272273e-06,
      "loss": 0.3173,
      "num_input_tokens_seen": 14766360,
      "step": 22615
    },
    {
      "epoch": 13.337264150943396,
      "grad_norm": 2.578451156616211,
      "learning_rate": 3.0169242248136066e-06,
      "loss": 0.3145,
      "num_input_tokens_seen": 14769176,
      "step": 22620
    },
    {
      "epoch": 13.340212264150944,
      "grad_norm": 3.584364175796509,
      "learning_rate": 3.0145627745278457e-06,
      "loss": 0.3003,
      "num_input_tokens_seen": 14772184,
      "step": 22625
    },
    {
      "epoch": 13.34316037735849,
      "grad_norm": 3.1898036003112793,
      "learning_rate": 3.0122018498951478e-06,
      "loss": 0.3141,
      "num_input_tokens_seen": 14775096,
      "step": 22630
    },
    {
      "epoch": 13.346108490566039,
      "grad_norm": 2.5837345123291016,
      "learning_rate": 3.0098414515405765e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 14778616,
      "step": 22635
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 2.264890432357788,
      "learning_rate": 3.0074815800890576e-06,
      "loss": 0.2884,
      "num_input_tokens_seen": 14781496,
      "step": 22640
    },
    {
      "epoch": 13.352004716981131,
      "grad_norm": 2.2633955478668213,
      "learning_rate": 3.005122236165378e-06,
      "loss": 0.3021,
      "num_input_tokens_seen": 14783640,
      "step": 22645
    },
    {
      "epoch": 13.35495283018868,
      "grad_norm": 3.103914737701416,
      "learning_rate": 3.0027634203941847e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 14786392,
      "step": 22650
    },
    {
      "epoch": 13.357900943396226,
      "grad_norm": 4.498966693878174,
      "learning_rate": 3.0004051333999816e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 14788824,
      "step": 22655
    },
    {
      "epoch": 13.360849056603774,
      "grad_norm": 2.798330068588257,
      "learning_rate": 2.998047375807139e-06,
      "loss": 0.3793,
      "num_input_tokens_seen": 14792312,
      "step": 22660
    },
    {
      "epoch": 13.36379716981132,
      "grad_norm": 2.7378623485565186,
      "learning_rate": 2.995690148239881e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 14795480,
      "step": 22665
    },
    {
      "epoch": 13.366745283018869,
      "grad_norm": 2.8847906589508057,
      "learning_rate": 2.993333451322293e-06,
      "loss": 0.2916,
      "num_input_tokens_seen": 14798616,
      "step": 22670
    },
    {
      "epoch": 13.369693396226415,
      "grad_norm": 10.304115295410156,
      "learning_rate": 2.9909772856783242e-06,
      "loss": 0.4411,
      "num_input_tokens_seen": 14802520,
      "step": 22675
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 3.3466625213623047,
      "learning_rate": 2.988621651931777e-06,
      "loss": 0.368,
      "num_input_tokens_seen": 14805208,
      "step": 22680
    },
    {
      "epoch": 13.37558962264151,
      "grad_norm": 2.6595749855041504,
      "learning_rate": 2.986266550706315e-06,
      "loss": 0.4072,
      "num_input_tokens_seen": 14808152,
      "step": 22685
    },
    {
      "epoch": 13.378537735849056,
      "grad_norm": 4.3998517990112305,
      "learning_rate": 2.9839119826254627e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 14811256,
      "step": 22690
    },
    {
      "epoch": 13.381485849056604,
      "grad_norm": 5.467985153198242,
      "learning_rate": 2.981557948312602e-06,
      "loss": 0.2802,
      "num_input_tokens_seen": 14814232,
      "step": 22695
    },
    {
      "epoch": 13.38443396226415,
      "grad_norm": 5.0532307624816895,
      "learning_rate": 2.9792044483909733e-06,
      "loss": 0.3184,
      "num_input_tokens_seen": 14817816,
      "step": 22700
    },
    {
      "epoch": 13.387382075471699,
      "grad_norm": 2.8483521938323975,
      "learning_rate": 2.9768514834836767e-06,
      "loss": 0.3418,
      "num_input_tokens_seen": 14820920,
      "step": 22705
    },
    {
      "epoch": 13.390330188679245,
      "grad_norm": 4.755228519439697,
      "learning_rate": 2.9744990542136685e-06,
      "loss": 0.3451,
      "num_input_tokens_seen": 14824568,
      "step": 22710
    },
    {
      "epoch": 13.393278301886792,
      "grad_norm": 2.346679925918579,
      "learning_rate": 2.9721471612037637e-06,
      "loss": 0.2854,
      "num_input_tokens_seen": 14828888,
      "step": 22715
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 2.3905208110809326,
      "learning_rate": 2.9697958050766385e-06,
      "loss": 0.3539,
      "num_input_tokens_seen": 14832984,
      "step": 22720
    },
    {
      "epoch": 13.399174528301886,
      "grad_norm": 3.2932074069976807,
      "learning_rate": 2.967444986454825e-06,
      "loss": 0.3934,
      "num_input_tokens_seen": 14835704,
      "step": 22725
    },
    {
      "epoch": 13.402122641509434,
      "grad_norm": 1.9457067251205444,
      "learning_rate": 2.9650947059607106e-06,
      "loss": 0.3247,
      "num_input_tokens_seen": 14838808,
      "step": 22730
    },
    {
      "epoch": 13.40507075471698,
      "grad_norm": 3.4821388721466064,
      "learning_rate": 2.962744964216542e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 14841816,
      "step": 22735
    },
    {
      "epoch": 13.408018867924529,
      "grad_norm": 4.08318567276001,
      "learning_rate": 2.960395761844425e-06,
      "loss": 0.3598,
      "num_input_tokens_seen": 14845368,
      "step": 22740
    },
    {
      "epoch": 13.410966981132075,
      "grad_norm": 4.338070392608643,
      "learning_rate": 2.95804709946632e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 14848312,
      "step": 22745
    },
    {
      "epoch": 13.413915094339623,
      "grad_norm": 2.908982515335083,
      "learning_rate": 2.9556989777040457e-06,
      "loss": 0.2448,
      "num_input_tokens_seen": 14850936,
      "step": 22750
    },
    {
      "epoch": 13.41686320754717,
      "grad_norm": 4.186826705932617,
      "learning_rate": 2.9533513971792776e-06,
      "loss": 0.3298,
      "num_input_tokens_seen": 14853432,
      "step": 22755
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 5.449368476867676,
      "learning_rate": 2.9510043585135473e-06,
      "loss": 0.398,
      "num_input_tokens_seen": 14857336,
      "step": 22760
    },
    {
      "epoch": 13.422759433962264,
      "grad_norm": 5.416469097137451,
      "learning_rate": 2.948657862328244e-06,
      "loss": 0.2888,
      "num_input_tokens_seen": 14860472,
      "step": 22765
    },
    {
      "epoch": 13.42570754716981,
      "grad_norm": 4.543337345123291,
      "learning_rate": 2.946311909244613e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 14863032,
      "step": 22770
    },
    {
      "epoch": 13.428655660377359,
      "grad_norm": 3.575605869293213,
      "learning_rate": 2.9439664998837538e-06,
      "loss": 0.5328,
      "num_input_tokens_seen": 14866200,
      "step": 22775
    },
    {
      "epoch": 13.431603773584905,
      "grad_norm": 2.5062243938446045,
      "learning_rate": 2.941621634866626e-06,
      "loss": 0.2018,
      "num_input_tokens_seen": 14872536,
      "step": 22780
    },
    {
      "epoch": 13.434551886792454,
      "grad_norm": 3.2343502044677734,
      "learning_rate": 2.9392773148140406e-06,
      "loss": 0.3787,
      "num_input_tokens_seen": 14875320,
      "step": 22785
    },
    {
      "epoch": 13.4375,
      "grad_norm": 3.5390167236328125,
      "learning_rate": 2.9369335403466676e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 14878424,
      "step": 22790
    },
    {
      "epoch": 13.440448113207546,
      "grad_norm": 3.2383785247802734,
      "learning_rate": 2.9345903120850318e-06,
      "loss": 0.3775,
      "num_input_tokens_seen": 14881624,
      "step": 22795
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 6.765531539916992,
      "learning_rate": 2.932247630649512e-06,
      "loss": 0.3038,
      "num_input_tokens_seen": 14886136,
      "step": 22800
    },
    {
      "epoch": 13.446344339622641,
      "grad_norm": 5.012436866760254,
      "learning_rate": 2.9299054966603424e-06,
      "loss": 0.4231,
      "num_input_tokens_seen": 14889656,
      "step": 22805
    },
    {
      "epoch": 13.44929245283019,
      "grad_norm": 3.895576238632202,
      "learning_rate": 2.927563910737613e-06,
      "loss": 0.3251,
      "num_input_tokens_seen": 14892568,
      "step": 22810
    },
    {
      "epoch": 13.452240566037736,
      "grad_norm": 2.7525594234466553,
      "learning_rate": 2.9252228735012722e-06,
      "loss": 0.314,
      "num_input_tokens_seen": 14895704,
      "step": 22815
    },
    {
      "epoch": 13.455188679245284,
      "grad_norm": 1.8481998443603516,
      "learning_rate": 2.9228823855711174e-06,
      "loss": 0.4693,
      "num_input_tokens_seen": 14900024,
      "step": 22820
    },
    {
      "epoch": 13.45813679245283,
      "grad_norm": 3.468940019607544,
      "learning_rate": 2.920542447566802e-06,
      "loss": 0.3621,
      "num_input_tokens_seen": 14903224,
      "step": 22825
    },
    {
      "epoch": 13.461084905660377,
      "grad_norm": 2.8650736808776855,
      "learning_rate": 2.918203060107837e-06,
      "loss": 0.4482,
      "num_input_tokens_seen": 14906488,
      "step": 22830
    },
    {
      "epoch": 13.464033018867925,
      "grad_norm": 3.2954506874084473,
      "learning_rate": 2.9158642238135813e-06,
      "loss": 0.3897,
      "num_input_tokens_seen": 14910168,
      "step": 22835
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 3.461789846420288,
      "learning_rate": 2.913525939303257e-06,
      "loss": 0.3463,
      "num_input_tokens_seen": 14913592,
      "step": 22840
    },
    {
      "epoch": 13.46992924528302,
      "grad_norm": 5.079789638519287,
      "learning_rate": 2.9111882071959317e-06,
      "loss": 0.4279,
      "num_input_tokens_seen": 14916376,
      "step": 22845
    },
    {
      "epoch": 13.472877358490566,
      "grad_norm": 3.116241693496704,
      "learning_rate": 2.908851028110532e-06,
      "loss": 0.4315,
      "num_input_tokens_seen": 14920568,
      "step": 22850
    },
    {
      "epoch": 13.475825471698114,
      "grad_norm": 3.646602153778076,
      "learning_rate": 2.906514402665834e-06,
      "loss": 0.3598,
      "num_input_tokens_seen": 14923416,
      "step": 22855
    },
    {
      "epoch": 13.47877358490566,
      "grad_norm": 3.346450090408325,
      "learning_rate": 2.9041783314804705e-06,
      "loss": 0.3073,
      "num_input_tokens_seen": 14926424,
      "step": 22860
    },
    {
      "epoch": 13.481721698113208,
      "grad_norm": 3.050476551055908,
      "learning_rate": 2.9018428151729238e-06,
      "loss": 0.341,
      "num_input_tokens_seen": 14929720,
      "step": 22865
    },
    {
      "epoch": 13.484669811320755,
      "grad_norm": 2.636143922805786,
      "learning_rate": 2.899507854361537e-06,
      "loss": 0.3746,
      "num_input_tokens_seen": 14933080,
      "step": 22870
    },
    {
      "epoch": 13.487617924528301,
      "grad_norm": 4.547954082489014,
      "learning_rate": 2.8971734496644975e-06,
      "loss": 0.3791,
      "num_input_tokens_seen": 14936856,
      "step": 22875
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 4.942732810974121,
      "learning_rate": 2.894839601699851e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 14940952,
      "step": 22880
    },
    {
      "epoch": 13.493514150943396,
      "grad_norm": 4.507939338684082,
      "learning_rate": 2.8925063110854923e-06,
      "loss": 0.3133,
      "num_input_tokens_seen": 14944856,
      "step": 22885
    },
    {
      "epoch": 13.496462264150944,
      "grad_norm": 3.697765827178955,
      "learning_rate": 2.8901735784391683e-06,
      "loss": 0.3566,
      "num_input_tokens_seen": 14947864,
      "step": 22890
    },
    {
      "epoch": 13.49941037735849,
      "grad_norm": 6.375060081481934,
      "learning_rate": 2.8878414043784844e-06,
      "loss": 0.403,
      "num_input_tokens_seen": 14951224,
      "step": 22895
    },
    {
      "epoch": 13.502358490566039,
      "grad_norm": 5.10914945602417,
      "learning_rate": 2.885509789520891e-06,
      "loss": 0.3235,
      "num_input_tokens_seen": 14954040,
      "step": 22900
    },
    {
      "epoch": 13.505306603773585,
      "grad_norm": 4.0120062828063965,
      "learning_rate": 2.8831787344836926e-06,
      "loss": 0.3156,
      "num_input_tokens_seen": 14958104,
      "step": 22905
    },
    {
      "epoch": 13.508254716981131,
      "grad_norm": 3.4494800567626953,
      "learning_rate": 2.880848239884049e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 14961560,
      "step": 22910
    },
    {
      "epoch": 13.51120283018868,
      "grad_norm": 2.278385639190674,
      "learning_rate": 2.8785183063389667e-06,
      "loss": 0.3036,
      "num_input_tokens_seen": 14964440,
      "step": 22915
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 3.199136257171631,
      "learning_rate": 2.876188934465306e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 14967896,
      "step": 22920
    },
    {
      "epoch": 13.517099056603774,
      "grad_norm": 1.832895040512085,
      "learning_rate": 2.8738601248797758e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 14971352,
      "step": 22925
    },
    {
      "epoch": 13.52004716981132,
      "grad_norm": 1.9681646823883057,
      "learning_rate": 2.8715318781989432e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 14975640,
      "step": 22930
    },
    {
      "epoch": 13.522995283018869,
      "grad_norm": 3.4381494522094727,
      "learning_rate": 2.869204195039219e-06,
      "loss": 0.2423,
      "num_input_tokens_seen": 14978744,
      "step": 22935
    },
    {
      "epoch": 13.525943396226415,
      "grad_norm": 3.328511953353882,
      "learning_rate": 2.8668770760168673e-06,
      "loss": 0.3946,
      "num_input_tokens_seen": 14982328,
      "step": 22940
    },
    {
      "epoch": 13.528891509433961,
      "grad_norm": 2.195735216140747,
      "learning_rate": 2.864550521748003e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 14985048,
      "step": 22945
    },
    {
      "epoch": 13.53183962264151,
      "grad_norm": 2.909877300262451,
      "learning_rate": 2.862224532848591e-06,
      "loss": 0.3804,
      "num_input_tokens_seen": 14987896,
      "step": 22950
    },
    {
      "epoch": 13.534787735849056,
      "grad_norm": 3.266657829284668,
      "learning_rate": 2.8598991099344455e-06,
      "loss": 0.3893,
      "num_input_tokens_seen": 14991000,
      "step": 22955
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 3.2199769020080566,
      "learning_rate": 2.857574253621236e-06,
      "loss": 0.3798,
      "num_input_tokens_seen": 14994904,
      "step": 22960
    },
    {
      "epoch": 13.54068396226415,
      "grad_norm": 3.9809350967407227,
      "learning_rate": 2.855249964524476e-06,
      "loss": 0.3766,
      "num_input_tokens_seen": 14997784,
      "step": 22965
    },
    {
      "epoch": 13.543632075471699,
      "grad_norm": 4.5312323570251465,
      "learning_rate": 2.852926243259531e-06,
      "loss": 0.5431,
      "num_input_tokens_seen": 15000920,
      "step": 22970
    },
    {
      "epoch": 13.546580188679245,
      "grad_norm": 3.094437599182129,
      "learning_rate": 2.850603090441617e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 15003992,
      "step": 22975
    },
    {
      "epoch": 13.549528301886792,
      "grad_norm": 2.3040835857391357,
      "learning_rate": 2.848280506685798e-06,
      "loss": 0.3413,
      "num_input_tokens_seen": 15007448,
      "step": 22980
    },
    {
      "epoch": 13.55247641509434,
      "grad_norm": 3.2578186988830566,
      "learning_rate": 2.845958492606986e-06,
      "loss": 0.3537,
      "num_input_tokens_seen": 15009816,
      "step": 22985
    },
    {
      "epoch": 13.555424528301886,
      "grad_norm": 3.1971216201782227,
      "learning_rate": 2.843637048819949e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 15012856,
      "step": 22990
    },
    {
      "epoch": 13.558372641509434,
      "grad_norm": 1.707671880722046,
      "learning_rate": 2.8413161759392966e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 15016408,
      "step": 22995
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 5.056138038635254,
      "learning_rate": 2.8389958745794878e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 15019416,
      "step": 23000
    },
    {
      "epoch": 13.564268867924529,
      "grad_norm": 9.105999946594238,
      "learning_rate": 2.8366761453548366e-06,
      "loss": 0.3934,
      "num_input_tokens_seen": 15022872,
      "step": 23005
    },
    {
      "epoch": 13.567216981132075,
      "grad_norm": 4.4150471687316895,
      "learning_rate": 2.8343569888795e-06,
      "loss": 0.2644,
      "num_input_tokens_seen": 15025944,
      "step": 23010
    },
    {
      "epoch": 13.570165094339622,
      "grad_norm": 2.279388189315796,
      "learning_rate": 2.832038405767483e-06,
      "loss": 0.2918,
      "num_input_tokens_seen": 15029336,
      "step": 23015
    },
    {
      "epoch": 13.57311320754717,
      "grad_norm": 4.4637250900268555,
      "learning_rate": 2.8297203966326397e-06,
      "loss": 0.4835,
      "num_input_tokens_seen": 15032632,
      "step": 23020
    },
    {
      "epoch": 13.576061320754716,
      "grad_norm": 4.085013389587402,
      "learning_rate": 2.8274029620886773e-06,
      "loss": 0.4519,
      "num_input_tokens_seen": 15036056,
      "step": 23025
    },
    {
      "epoch": 13.579009433962264,
      "grad_norm": 4.077208042144775,
      "learning_rate": 2.825086102749144e-06,
      "loss": 0.3025,
      "num_input_tokens_seen": 15038328,
      "step": 23030
    },
    {
      "epoch": 13.58195754716981,
      "grad_norm": 2.421710968017578,
      "learning_rate": 2.822769819227438e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 15040952,
      "step": 23035
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 3.514065980911255,
      "learning_rate": 2.8204541121368055e-06,
      "loss": 0.3252,
      "num_input_tokens_seen": 15043480,
      "step": 23040
    },
    {
      "epoch": 13.587853773584905,
      "grad_norm": 2.571617841720581,
      "learning_rate": 2.8181389820903402e-06,
      "loss": 0.3573,
      "num_input_tokens_seen": 15047096,
      "step": 23045
    },
    {
      "epoch": 13.590801886792454,
      "grad_norm": 6.504537105560303,
      "learning_rate": 2.8158244297009814e-06,
      "loss": 0.405,
      "num_input_tokens_seen": 15050072,
      "step": 23050
    },
    {
      "epoch": 13.59375,
      "grad_norm": 4.816720962524414,
      "learning_rate": 2.8135104555815196e-06,
      "loss": 0.3705,
      "num_input_tokens_seen": 15053016,
      "step": 23055
    },
    {
      "epoch": 13.596698113207546,
      "grad_norm": 3.5920798778533936,
      "learning_rate": 2.811197060344588e-06,
      "loss": 0.3106,
      "num_input_tokens_seen": 15055960,
      "step": 23060
    },
    {
      "epoch": 13.599646226415095,
      "grad_norm": 3.0486326217651367,
      "learning_rate": 2.8088842446026677e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 15058808,
      "step": 23065
    },
    {
      "epoch": 13.602594339622641,
      "grad_norm": 5.190282821655273,
      "learning_rate": 2.806572008968087e-06,
      "loss": 0.3303,
      "num_input_tokens_seen": 15061688,
      "step": 23070
    },
    {
      "epoch": 13.60554245283019,
      "grad_norm": 4.3470282554626465,
      "learning_rate": 2.80426035405302e-06,
      "loss": 0.5184,
      "num_input_tokens_seen": 15064664,
      "step": 23075
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 4.01296854019165,
      "learning_rate": 2.8019492804694852e-06,
      "loss": 0.4465,
      "num_input_tokens_seen": 15069048,
      "step": 23080
    },
    {
      "epoch": 13.611438679245284,
      "grad_norm": 2.0902390480041504,
      "learning_rate": 2.799638788829354e-06,
      "loss": 0.2465,
      "num_input_tokens_seen": 15072152,
      "step": 23085
    },
    {
      "epoch": 13.61438679245283,
      "grad_norm": 3.1195223331451416,
      "learning_rate": 2.7973288797443367e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 15074648,
      "step": 23090
    },
    {
      "epoch": 13.617334905660378,
      "grad_norm": 2.2866008281707764,
      "learning_rate": 2.7950195538259884e-06,
      "loss": 0.4335,
      "num_input_tokens_seen": 15077336,
      "step": 23095
    },
    {
      "epoch": 13.620283018867925,
      "grad_norm": 4.121583461761475,
      "learning_rate": 2.792710811685719e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 15080152,
      "step": 23100
    },
    {
      "epoch": 13.623231132075471,
      "grad_norm": 3.588010311126709,
      "learning_rate": 2.7904026539347743e-06,
      "loss": 0.3063,
      "num_input_tokens_seen": 15083928,
      "step": 23105
    },
    {
      "epoch": 13.62617924528302,
      "grad_norm": 4.003051280975342,
      "learning_rate": 2.7880950811842507e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 15087928,
      "step": 23110
    },
    {
      "epoch": 13.629127358490566,
      "grad_norm": 6.085855960845947,
      "learning_rate": 2.785788094045085e-06,
      "loss": 0.2647,
      "num_input_tokens_seen": 15090616,
      "step": 23115
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 2.434609889984131,
      "learning_rate": 2.7834816931280655e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 15094936,
      "step": 23120
    },
    {
      "epoch": 13.63502358490566,
      "grad_norm": 3.2948315143585205,
      "learning_rate": 2.781175879043821e-06,
      "loss": 0.2703,
      "num_input_tokens_seen": 15097752,
      "step": 23125
    },
    {
      "epoch": 13.637971698113208,
      "grad_norm": 3.2358627319335938,
      "learning_rate": 2.778870652402825e-06,
      "loss": 0.2296,
      "num_input_tokens_seen": 15101496,
      "step": 23130
    },
    {
      "epoch": 13.640919811320755,
      "grad_norm": 8.36298942565918,
      "learning_rate": 2.776566013815396e-06,
      "loss": 0.382,
      "num_input_tokens_seen": 15104536,
      "step": 23135
    },
    {
      "epoch": 13.643867924528301,
      "grad_norm": 5.010889053344727,
      "learning_rate": 2.774261963891698e-06,
      "loss": 0.3795,
      "num_input_tokens_seen": 15107896,
      "step": 23140
    },
    {
      "epoch": 13.64681603773585,
      "grad_norm": 5.611725807189941,
      "learning_rate": 2.771958503241735e-06,
      "loss": 0.3594,
      "num_input_tokens_seen": 15110744,
      "step": 23145
    },
    {
      "epoch": 13.649764150943396,
      "grad_norm": 2.1291275024414062,
      "learning_rate": 2.769655632475362e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 15114712,
      "step": 23150
    },
    {
      "epoch": 13.652712264150944,
      "grad_norm": 3.012312889099121,
      "learning_rate": 2.7673533522022733e-06,
      "loss": 0.4054,
      "num_input_tokens_seen": 15117624,
      "step": 23155
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 6.592465877532959,
      "learning_rate": 2.765051663032007e-06,
      "loss": 0.4097,
      "num_input_tokens_seen": 15119928,
      "step": 23160
    },
    {
      "epoch": 13.658608490566039,
      "grad_norm": 2.6685683727264404,
      "learning_rate": 2.7627505655739446e-06,
      "loss": 0.4172,
      "num_input_tokens_seen": 15122840,
      "step": 23165
    },
    {
      "epoch": 13.661556603773585,
      "grad_norm": 3.174525499343872,
      "learning_rate": 2.7604500604373097e-06,
      "loss": 0.2484,
      "num_input_tokens_seen": 15126808,
      "step": 23170
    },
    {
      "epoch": 13.664504716981131,
      "grad_norm": 5.820985317230225,
      "learning_rate": 2.7581501482311757e-06,
      "loss": 0.3541,
      "num_input_tokens_seen": 15129400,
      "step": 23175
    },
    {
      "epoch": 13.66745283018868,
      "grad_norm": 2.2530391216278076,
      "learning_rate": 2.7558508295644513e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 15133112,
      "step": 23180
    },
    {
      "epoch": 13.670400943396226,
      "grad_norm": 2.0957775115966797,
      "learning_rate": 2.7535521050458922e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 15136952,
      "step": 23185
    },
    {
      "epoch": 13.673349056603774,
      "grad_norm": 3.055593967437744,
      "learning_rate": 2.7512539752840926e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 15139832,
      "step": 23190
    },
    {
      "epoch": 13.67629716981132,
      "grad_norm": 3.305277109146118,
      "learning_rate": 2.748956440887497e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 15142936,
      "step": 23195
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 3.238431453704834,
      "learning_rate": 2.7466595024643843e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 15146008,
      "step": 23200
    },
    {
      "epoch": 13.682193396226415,
      "grad_norm": 2.3532469272613525,
      "learning_rate": 2.744363160622878e-06,
      "loss": 0.4134,
      "num_input_tokens_seen": 15149752,
      "step": 23205
    },
    {
      "epoch": 13.685141509433961,
      "grad_norm": 4.634951114654541,
      "learning_rate": 2.742067415970948e-06,
      "loss": 0.2897,
      "num_input_tokens_seen": 15154104,
      "step": 23210
    },
    {
      "epoch": 13.68808962264151,
      "grad_norm": 3.3066370487213135,
      "learning_rate": 2.739772269116402e-06,
      "loss": 0.372,
      "num_input_tokens_seen": 15157720,
      "step": 23215
    },
    {
      "epoch": 13.691037735849056,
      "grad_norm": 3.629554510116577,
      "learning_rate": 2.7374777206668874e-06,
      "loss": 0.3354,
      "num_input_tokens_seen": 15160600,
      "step": 23220
    },
    {
      "epoch": 13.693985849056604,
      "grad_norm": 3.251127004623413,
      "learning_rate": 2.735183771229898e-06,
      "loss": 0.2815,
      "num_input_tokens_seen": 15163288,
      "step": 23225
    },
    {
      "epoch": 13.69693396226415,
      "grad_norm": 4.2906999588012695,
      "learning_rate": 2.732890421412765e-06,
      "loss": 0.2992,
      "num_input_tokens_seen": 15166072,
      "step": 23230
    },
    {
      "epoch": 13.699882075471699,
      "grad_norm": 5.926146507263184,
      "learning_rate": 2.7305976718226624e-06,
      "loss": 0.3897,
      "num_input_tokens_seen": 15169240,
      "step": 23235
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 3.0695698261260986,
      "learning_rate": 2.728305523066609e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 15172632,
      "step": 23240
    },
    {
      "epoch": 13.705778301886792,
      "grad_norm": 3.94073224067688,
      "learning_rate": 2.726013975751458e-06,
      "loss": 0.3498,
      "num_input_tokens_seen": 15175128,
      "step": 23245
    },
    {
      "epoch": 13.70872641509434,
      "grad_norm": 2.3054556846618652,
      "learning_rate": 2.723723030483908e-06,
      "loss": 0.3169,
      "num_input_tokens_seen": 15178680,
      "step": 23250
    },
    {
      "epoch": 13.711674528301886,
      "grad_norm": 3.185265064239502,
      "learning_rate": 2.7214326878704953e-06,
      "loss": 0.3645,
      "num_input_tokens_seen": 15181496,
      "step": 23255
    },
    {
      "epoch": 13.714622641509434,
      "grad_norm": 11.26134967803955,
      "learning_rate": 2.7191429485175993e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 15184184,
      "step": 23260
    },
    {
      "epoch": 13.71757075471698,
      "grad_norm": 2.9774694442749023,
      "learning_rate": 2.716853813031435e-06,
      "loss": 0.2694,
      "num_input_tokens_seen": 15187576,
      "step": 23265
    },
    {
      "epoch": 13.720518867924529,
      "grad_norm": 3.32824969291687,
      "learning_rate": 2.714565282018066e-06,
      "loss": 0.3008,
      "num_input_tokens_seen": 15190616,
      "step": 23270
    },
    {
      "epoch": 13.723466981132075,
      "grad_norm": 2.8113903999328613,
      "learning_rate": 2.7122773560833877e-06,
      "loss": 0.3346,
      "num_input_tokens_seen": 15193944,
      "step": 23275
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 4.833603858947754,
      "learning_rate": 2.709990035833139e-06,
      "loss": 0.4349,
      "num_input_tokens_seen": 15197400,
      "step": 23280
    },
    {
      "epoch": 13.72936320754717,
      "grad_norm": 3.0644757747650146,
      "learning_rate": 2.707703321872896e-06,
      "loss": 0.3574,
      "num_input_tokens_seen": 15200056,
      "step": 23285
    },
    {
      "epoch": 13.732311320754716,
      "grad_norm": 4.596850872039795,
      "learning_rate": 2.705417214808079e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 15203288,
      "step": 23290
    },
    {
      "epoch": 13.735259433962264,
      "grad_norm": 2.1580405235290527,
      "learning_rate": 2.703131715243945e-06,
      "loss": 0.2201,
      "num_input_tokens_seen": 15206712,
      "step": 23295
    },
    {
      "epoch": 13.73820754716981,
      "grad_norm": 2.6703169345855713,
      "learning_rate": 2.7008468237855855e-06,
      "loss": 0.3192,
      "num_input_tokens_seen": 15209432,
      "step": 23300
    },
    {
      "epoch": 13.741155660377359,
      "grad_norm": 1.532977819442749,
      "learning_rate": 2.69856254103794e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 15212568,
      "step": 23305
    },
    {
      "epoch": 13.744103773584905,
      "grad_norm": 5.046432018280029,
      "learning_rate": 2.6962788676057806e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 15215608,
      "step": 23310
    },
    {
      "epoch": 13.747051886792454,
      "grad_norm": 3.3253440856933594,
      "learning_rate": 2.69399580409372e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 15219640,
      "step": 23315
    },
    {
      "epoch": 13.75,
      "grad_norm": 3.6201250553131104,
      "learning_rate": 2.6917133511062076e-06,
      "loss": 0.2558,
      "num_input_tokens_seen": 15223256,
      "step": 23320
    },
    {
      "epoch": 13.752948113207546,
      "grad_norm": 5.303553581237793,
      "learning_rate": 2.6894315092475342e-06,
      "loss": 0.4444,
      "num_input_tokens_seen": 15226584,
      "step": 23325
    },
    {
      "epoch": 13.755896226415095,
      "grad_norm": 2.843135356903076,
      "learning_rate": 2.6871502791218245e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 15233688,
      "step": 23330
    },
    {
      "epoch": 13.758844339622641,
      "grad_norm": 2.060163736343384,
      "learning_rate": 2.684869661333048e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 15236792,
      "step": 23335
    },
    {
      "epoch": 13.76179245283019,
      "grad_norm": 2.2506160736083984,
      "learning_rate": 2.6825896564850074e-06,
      "loss": 0.2484,
      "num_input_tokens_seen": 15240760,
      "step": 23340
    },
    {
      "epoch": 13.764740566037736,
      "grad_norm": 2.4230294227600098,
      "learning_rate": 2.6803102651813416e-06,
      "loss": 0.321,
      "num_input_tokens_seen": 15243640,
      "step": 23345
    },
    {
      "epoch": 13.767688679245284,
      "grad_norm": 3.165419101715088,
      "learning_rate": 2.6780314880255307e-06,
      "loss": 0.3279,
      "num_input_tokens_seen": 15247320,
      "step": 23350
    },
    {
      "epoch": 13.77063679245283,
      "grad_norm": 3.231797456741333,
      "learning_rate": 2.675753325620891e-06,
      "loss": 0.4188,
      "num_input_tokens_seen": 15249816,
      "step": 23355
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 3.1095938682556152,
      "learning_rate": 2.6734757785705727e-06,
      "loss": 0.3061,
      "num_input_tokens_seen": 15253688,
      "step": 23360
    },
    {
      "epoch": 13.776533018867925,
      "grad_norm": 2.7202489376068115,
      "learning_rate": 2.6711988474775712e-06,
      "loss": 0.3512,
      "num_input_tokens_seen": 15256696,
      "step": 23365
    },
    {
      "epoch": 13.779481132075471,
      "grad_norm": 3.722193479537964,
      "learning_rate": 2.668922532944711e-06,
      "loss": 0.4754,
      "num_input_tokens_seen": 15259544,
      "step": 23370
    },
    {
      "epoch": 13.78242924528302,
      "grad_norm": 2.826719284057617,
      "learning_rate": 2.6666468355746566e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 15262936,
      "step": 23375
    },
    {
      "epoch": 13.785377358490566,
      "grad_norm": 3.6061758995056152,
      "learning_rate": 2.6643717559699073e-06,
      "loss": 0.2046,
      "num_input_tokens_seen": 15265176,
      "step": 23380
    },
    {
      "epoch": 13.788325471698114,
      "grad_norm": 3.950965642929077,
      "learning_rate": 2.662097294732803e-06,
      "loss": 0.3863,
      "num_input_tokens_seen": 15268152,
      "step": 23385
    },
    {
      "epoch": 13.79127358490566,
      "grad_norm": 6.298252582550049,
      "learning_rate": 2.6598234524655165e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 15273080,
      "step": 23390
    },
    {
      "epoch": 13.794221698113208,
      "grad_norm": 4.113889217376709,
      "learning_rate": 2.657550229770054e-06,
      "loss": 0.3184,
      "num_input_tokens_seen": 15276120,
      "step": 23395
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 3.015651226043701,
      "learning_rate": 2.655277627248265e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 15278648,
      "step": 23400
    },
    {
      "epoch": 13.800117924528301,
      "grad_norm": 4.35060453414917,
      "learning_rate": 2.65300564550183e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 15283160,
      "step": 23405
    },
    {
      "epoch": 13.80306603773585,
      "grad_norm": 2.762899875640869,
      "learning_rate": 2.6507342851322647e-06,
      "loss": 0.2786,
      "num_input_tokens_seen": 15286488,
      "step": 23410
    },
    {
      "epoch": 13.806014150943396,
      "grad_norm": 1.8176004886627197,
      "learning_rate": 2.6484635467409233e-06,
      "loss": 0.2504,
      "num_input_tokens_seen": 15290104,
      "step": 23415
    },
    {
      "epoch": 13.808962264150944,
      "grad_norm": 3.8518803119659424,
      "learning_rate": 2.64619343092899e-06,
      "loss": 0.4972,
      "num_input_tokens_seen": 15292952,
      "step": 23420
    },
    {
      "epoch": 13.81191037735849,
      "grad_norm": 3.0917935371398926,
      "learning_rate": 2.643923938297492e-06,
      "loss": 0.4743,
      "num_input_tokens_seen": 15296152,
      "step": 23425
    },
    {
      "epoch": 13.814858490566039,
      "grad_norm": 4.7573347091674805,
      "learning_rate": 2.6416550694472855e-06,
      "loss": 0.2666,
      "num_input_tokens_seen": 15301240,
      "step": 23430
    },
    {
      "epoch": 13.817806603773585,
      "grad_norm": 4.003369331359863,
      "learning_rate": 2.639386824979063e-06,
      "loss": 0.2579,
      "num_input_tokens_seen": 15304536,
      "step": 23435
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 2.3557286262512207,
      "learning_rate": 2.6371192054933525e-06,
      "loss": 0.4011,
      "num_input_tokens_seen": 15307640,
      "step": 23440
    },
    {
      "epoch": 13.82370283018868,
      "grad_norm": 1.657974362373352,
      "learning_rate": 2.634852211590516e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 15311576,
      "step": 23445
    },
    {
      "epoch": 13.826650943396226,
      "grad_norm": 3.0391407012939453,
      "learning_rate": 2.6325858438707473e-06,
      "loss": 0.2869,
      "num_input_tokens_seen": 15314936,
      "step": 23450
    },
    {
      "epoch": 13.829599056603774,
      "grad_norm": 2.658358573913574,
      "learning_rate": 2.630320102934082e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 15319320,
      "step": 23455
    },
    {
      "epoch": 13.83254716981132,
      "grad_norm": 1.9850744009017944,
      "learning_rate": 2.628054989380382e-06,
      "loss": 0.3102,
      "num_input_tokens_seen": 15322616,
      "step": 23460
    },
    {
      "epoch": 13.835495283018869,
      "grad_norm": 3.0649874210357666,
      "learning_rate": 2.625790503809346e-06,
      "loss": 0.3356,
      "num_input_tokens_seen": 15325208,
      "step": 23465
    },
    {
      "epoch": 13.838443396226415,
      "grad_norm": 4.396548748016357,
      "learning_rate": 2.6235266468205067e-06,
      "loss": 0.361,
      "num_input_tokens_seen": 15327640,
      "step": 23470
    },
    {
      "epoch": 13.841391509433961,
      "grad_norm": 4.224795818328857,
      "learning_rate": 2.621263419013227e-06,
      "loss": 0.4204,
      "num_input_tokens_seen": 15331064,
      "step": 23475
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 2.9841465950012207,
      "learning_rate": 2.619000820986711e-06,
      "loss": 0.3789,
      "num_input_tokens_seen": 15334328,
      "step": 23480
    },
    {
      "epoch": 13.847287735849056,
      "grad_norm": 2.8993804454803467,
      "learning_rate": 2.616738853339988e-06,
      "loss": 0.4255,
      "num_input_tokens_seen": 15337208,
      "step": 23485
    },
    {
      "epoch": 13.850235849056604,
      "grad_norm": 4.992593765258789,
      "learning_rate": 2.614477516671926e-06,
      "loss": 0.2846,
      "num_input_tokens_seen": 15340216,
      "step": 23490
    },
    {
      "epoch": 13.85318396226415,
      "grad_norm": 3.756143569946289,
      "learning_rate": 2.612216811581223e-06,
      "loss": 0.3311,
      "num_input_tokens_seen": 15343384,
      "step": 23495
    },
    {
      "epoch": 13.856132075471699,
      "grad_norm": 3.92010760307312,
      "learning_rate": 2.6099567386664095e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 15346008,
      "step": 23500
    },
    {
      "epoch": 13.859080188679245,
      "grad_norm": 4.758578300476074,
      "learning_rate": 2.60769729852585e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 15349336,
      "step": 23505
    },
    {
      "epoch": 13.862028301886792,
      "grad_norm": 2.572927951812744,
      "learning_rate": 2.6054384917577413e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 15352760,
      "step": 23510
    },
    {
      "epoch": 13.86497641509434,
      "grad_norm": 4.747901916503906,
      "learning_rate": 2.60318031896011e-06,
      "loss": 0.3432,
      "num_input_tokens_seen": 15355736,
      "step": 23515
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 11.3092622756958,
      "learning_rate": 2.60092278073082e-06,
      "loss": 0.4006,
      "num_input_tokens_seen": 15359512,
      "step": 23520
    },
    {
      "epoch": 13.870872641509434,
      "grad_norm": 6.63578987121582,
      "learning_rate": 2.5986658776675644e-06,
      "loss": 0.3829,
      "num_input_tokens_seen": 15361528,
      "step": 23525
    },
    {
      "epoch": 13.87382075471698,
      "grad_norm": 3.168782949447632,
      "learning_rate": 2.5964096103678666e-06,
      "loss": 0.2679,
      "num_input_tokens_seen": 15363960,
      "step": 23530
    },
    {
      "epoch": 13.876768867924529,
      "grad_norm": 3.585327386856079,
      "learning_rate": 2.5941539794290833e-06,
      "loss": 0.357,
      "num_input_tokens_seen": 15366264,
      "step": 23535
    },
    {
      "epoch": 13.879716981132075,
      "grad_norm": 3.2238824367523193,
      "learning_rate": 2.5918989854484024e-06,
      "loss": 0.2458,
      "num_input_tokens_seen": 15369048,
      "step": 23540
    },
    {
      "epoch": 13.882665094339622,
      "grad_norm": 2.7704858779907227,
      "learning_rate": 2.5896446290228417e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 15371672,
      "step": 23545
    },
    {
      "epoch": 13.88561320754717,
      "grad_norm": 2.847304344177246,
      "learning_rate": 2.5873909107492547e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 15375224,
      "step": 23550
    },
    {
      "epoch": 13.888561320754716,
      "grad_norm": 2.2743980884552,
      "learning_rate": 2.5851378312243224e-06,
      "loss": 0.2623,
      "num_input_tokens_seen": 15378360,
      "step": 23555
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 2.431802272796631,
      "learning_rate": 2.5828853910445572e-06,
      "loss": 0.1947,
      "num_input_tokens_seen": 15381208,
      "step": 23560
    },
    {
      "epoch": 13.89445754716981,
      "grad_norm": 3.6665127277374268,
      "learning_rate": 2.5806335908063012e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 15384280,
      "step": 23565
    },
    {
      "epoch": 13.897405660377359,
      "grad_norm": 6.253374099731445,
      "learning_rate": 2.5783824311057293e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 15387640,
      "step": 23570
    },
    {
      "epoch": 13.900353773584905,
      "grad_norm": 3.437129259109497,
      "learning_rate": 2.5761319125388433e-06,
      "loss": 0.4057,
      "num_input_tokens_seen": 15390712,
      "step": 23575
    },
    {
      "epoch": 13.903301886792454,
      "grad_norm": 3.6318392753601074,
      "learning_rate": 2.57388203570148e-06,
      "loss": 0.2436,
      "num_input_tokens_seen": 15393304,
      "step": 23580
    },
    {
      "epoch": 13.90625,
      "grad_norm": 4.578932762145996,
      "learning_rate": 2.5716328011893055e-06,
      "loss": 0.397,
      "num_input_tokens_seen": 15396792,
      "step": 23585
    },
    {
      "epoch": 13.909198113207546,
      "grad_norm": 3.2455573081970215,
      "learning_rate": 2.5693842095978127e-06,
      "loss": 0.2673,
      "num_input_tokens_seen": 15399832,
      "step": 23590
    },
    {
      "epoch": 13.912146226415095,
      "grad_norm": 4.046075820922852,
      "learning_rate": 2.567136261522325e-06,
      "loss": 0.268,
      "num_input_tokens_seen": 15402168,
      "step": 23595
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 2.585984706878662,
      "learning_rate": 2.5648889575579985e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 15405144,
      "step": 23600
    },
    {
      "epoch": 13.91804245283019,
      "grad_norm": 4.26426362991333,
      "learning_rate": 2.562642298299814e-06,
      "loss": 0.3701,
      "num_input_tokens_seen": 15408792,
      "step": 23605
    },
    {
      "epoch": 13.920990566037736,
      "grad_norm": 2.7056455612182617,
      "learning_rate": 2.560396284342584e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 15411992,
      "step": 23610
    },
    {
      "epoch": 13.923938679245284,
      "grad_norm": 3.3884406089782715,
      "learning_rate": 2.558150916280954e-06,
      "loss": 0.4584,
      "num_input_tokens_seen": 15415256,
      "step": 23615
    },
    {
      "epoch": 13.92688679245283,
      "grad_norm": 3.610980272293091,
      "learning_rate": 2.555906194709392e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 15418488,
      "step": 23620
    },
    {
      "epoch": 13.929834905660378,
      "grad_norm": 3.494593858718872,
      "learning_rate": 2.553662120222199e-06,
      "loss": 0.2336,
      "num_input_tokens_seen": 15421240,
      "step": 23625
    },
    {
      "epoch": 13.932783018867925,
      "grad_norm": 2.851644515991211,
      "learning_rate": 2.5514186934135026e-06,
      "loss": 0.3022,
      "num_input_tokens_seen": 15425528,
      "step": 23630
    },
    {
      "epoch": 13.935731132075471,
      "grad_norm": 3.775855779647827,
      "learning_rate": 2.54917591487726e-06,
      "loss": 0.3087,
      "num_input_tokens_seen": 15428824,
      "step": 23635
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 3.8765957355499268,
      "learning_rate": 2.5469337852072547e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 15432024,
      "step": 23640
    },
    {
      "epoch": 13.941627358490566,
      "grad_norm": 3.3204081058502197,
      "learning_rate": 2.5446923049971035e-06,
      "loss": 0.308,
      "num_input_tokens_seen": 15435864,
      "step": 23645
    },
    {
      "epoch": 13.944575471698114,
      "grad_norm": 3.9448254108428955,
      "learning_rate": 2.5424514748402463e-06,
      "loss": 0.3327,
      "num_input_tokens_seen": 15438840,
      "step": 23650
    },
    {
      "epoch": 13.94752358490566,
      "grad_norm": 2.533252716064453,
      "learning_rate": 2.540211295329953e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 15442392,
      "step": 23655
    },
    {
      "epoch": 13.950471698113208,
      "grad_norm": 5.663008213043213,
      "learning_rate": 2.5379717670593197e-06,
      "loss": 0.2476,
      "num_input_tokens_seen": 15444824,
      "step": 23660
    },
    {
      "epoch": 13.953419811320755,
      "grad_norm": 2.3578288555145264,
      "learning_rate": 2.53573289062127e-06,
      "loss": 0.4387,
      "num_input_tokens_seen": 15448024,
      "step": 23665
    },
    {
      "epoch": 13.956367924528301,
      "grad_norm": 4.130578517913818,
      "learning_rate": 2.5334946666085605e-06,
      "loss": 0.2723,
      "num_input_tokens_seen": 15451128,
      "step": 23670
    },
    {
      "epoch": 13.95931603773585,
      "grad_norm": 3.140275001525879,
      "learning_rate": 2.531257095613766e-06,
      "loss": 0.3589,
      "num_input_tokens_seen": 15454936,
      "step": 23675
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 5.4899067878723145,
      "learning_rate": 2.529020178229297e-06,
      "loss": 0.3396,
      "num_input_tokens_seen": 15458040,
      "step": 23680
    },
    {
      "epoch": 13.965212264150944,
      "grad_norm": 2.731123447418213,
      "learning_rate": 2.5267839150473846e-06,
      "loss": 0.3888,
      "num_input_tokens_seen": 15461688,
      "step": 23685
    },
    {
      "epoch": 13.96816037735849,
      "grad_norm": 2.892503023147583,
      "learning_rate": 2.5245483066600896e-06,
      "loss": 0.3678,
      "num_input_tokens_seen": 15466008,
      "step": 23690
    },
    {
      "epoch": 13.971108490566039,
      "grad_norm": 5.565773010253906,
      "learning_rate": 2.5223133536592996e-06,
      "loss": 0.4614,
      "num_input_tokens_seen": 15469016,
      "step": 23695
    },
    {
      "epoch": 13.974056603773585,
      "grad_norm": 3.147486686706543,
      "learning_rate": 2.520079056636725e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 15471640,
      "step": 23700
    },
    {
      "epoch": 13.977004716981131,
      "grad_norm": 4.436188697814941,
      "learning_rate": 2.5178454161839106e-06,
      "loss": 0.2727,
      "num_input_tokens_seen": 15475384,
      "step": 23705
    },
    {
      "epoch": 13.97995283018868,
      "grad_norm": 3.00807785987854,
      "learning_rate": 2.5156124328922195e-06,
      "loss": 0.2639,
      "num_input_tokens_seen": 15478424,
      "step": 23710
    },
    {
      "epoch": 13.982900943396226,
      "grad_norm": 3.43365478515625,
      "learning_rate": 2.513380107352844e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 15482872,
      "step": 23715
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 3.077228307723999,
      "learning_rate": 2.5111484401568014e-06,
      "loss": 0.2237,
      "num_input_tokens_seen": 15486552,
      "step": 23720
    },
    {
      "epoch": 13.98879716981132,
      "grad_norm": 2.2663214206695557,
      "learning_rate": 2.508917431894936e-06,
      "loss": 0.2797,
      "num_input_tokens_seen": 15490424,
      "step": 23725
    },
    {
      "epoch": 13.991745283018869,
      "grad_norm": 3.0091145038604736,
      "learning_rate": 2.5066870831579144e-06,
      "loss": 0.3325,
      "num_input_tokens_seen": 15493240,
      "step": 23730
    },
    {
      "epoch": 13.994693396226415,
      "grad_norm": 3.476914882659912,
      "learning_rate": 2.504457394536235e-06,
      "loss": 0.2264,
      "num_input_tokens_seen": 15495736,
      "step": 23735
    },
    {
      "epoch": 13.997641509433961,
      "grad_norm": 1.9446858167648315,
      "learning_rate": 2.502228366620216e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 15502264,
      "step": 23740
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.5563265681266785,
      "eval_runtime": 19.0579,
      "eval_samples_per_second": 88.992,
      "eval_steps_per_second": 22.248,
      "num_input_tokens_seen": 15504440,
      "step": 23744
    },
    {
      "epoch": 14.00058962264151,
      "grad_norm": 2.700174570083618,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.3551,
      "num_input_tokens_seen": 15505080,
      "step": 23745
    },
    {
      "epoch": 14.003537735849056,
      "grad_norm": 3.200284004211426,
      "learning_rate": 2.497772295265561e-06,
      "loss": 0.5061,
      "num_input_tokens_seen": 15507864,
      "step": 23750
    },
    {
      "epoch": 14.006485849056604,
      "grad_norm": 2.823657512664795,
      "learning_rate": 2.4955452530066897e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 15510616,
      "step": 23755
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 3.764455556869507,
      "learning_rate": 2.4933188738130043e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 15513464,
      "step": 23760
    },
    {
      "epoch": 14.012382075471699,
      "grad_norm": 3.342912435531616,
      "learning_rate": 2.49109315827395e-06,
      "loss": 0.3524,
      "num_input_tokens_seen": 15515992,
      "step": 23765
    },
    {
      "epoch": 14.015330188679245,
      "grad_norm": 2.971245527267456,
      "learning_rate": 2.4888681069787975e-06,
      "loss": 0.2845,
      "num_input_tokens_seen": 15519896,
      "step": 23770
    },
    {
      "epoch": 14.018278301886792,
      "grad_norm": 2.746807336807251,
      "learning_rate": 2.4866437205166353e-06,
      "loss": 0.2538,
      "num_input_tokens_seen": 15523160,
      "step": 23775
    },
    {
      "epoch": 14.02122641509434,
      "grad_norm": 3.306149482727051,
      "learning_rate": 2.4844199994763803e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 15526136,
      "step": 23780
    },
    {
      "epoch": 14.024174528301886,
      "grad_norm": 2.992251396179199,
      "learning_rate": 2.482196944446772e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 15529848,
      "step": 23785
    },
    {
      "epoch": 14.027122641509434,
      "grad_norm": 1.8172892332077026,
      "learning_rate": 2.4799745560163736e-06,
      "loss": 0.4176,
      "num_input_tokens_seen": 15533944,
      "step": 23790
    },
    {
      "epoch": 14.03007075471698,
      "grad_norm": 3.6563730239868164,
      "learning_rate": 2.4777528347735707e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 15536632,
      "step": 23795
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 2.4731602668762207,
      "learning_rate": 2.4755317813065766e-06,
      "loss": 0.2862,
      "num_input_tokens_seen": 15540376,
      "step": 23800
    },
    {
      "epoch": 14.035966981132075,
      "grad_norm": 2.8666608333587646,
      "learning_rate": 2.4733113962034234e-06,
      "loss": 0.3252,
      "num_input_tokens_seen": 15547512,
      "step": 23805
    },
    {
      "epoch": 14.038915094339623,
      "grad_norm": 2.9459025859832764,
      "learning_rate": 2.4710916800519674e-06,
      "loss": 0.2948,
      "num_input_tokens_seen": 15550200,
      "step": 23810
    },
    {
      "epoch": 14.04186320754717,
      "grad_norm": 3.9507367610931396,
      "learning_rate": 2.4688726334398883e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 15553592,
      "step": 23815
    },
    {
      "epoch": 14.044811320754716,
      "grad_norm": 3.4612033367156982,
      "learning_rate": 2.466654256954688e-06,
      "loss": 0.3442,
      "num_input_tokens_seen": 15557912,
      "step": 23820
    },
    {
      "epoch": 14.047759433962264,
      "grad_norm": 3.1247634887695312,
      "learning_rate": 2.4644365511836895e-06,
      "loss": 0.3245,
      "num_input_tokens_seen": 15561656,
      "step": 23825
    },
    {
      "epoch": 14.05070754716981,
      "grad_norm": 3.800326347351074,
      "learning_rate": 2.4622195167140432e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 15564344,
      "step": 23830
    },
    {
      "epoch": 14.053655660377359,
      "grad_norm": 3.5639641284942627,
      "learning_rate": 2.4600031541327173e-06,
      "loss": 0.2838,
      "num_input_tokens_seen": 15567928,
      "step": 23835
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 3.5447018146514893,
      "learning_rate": 2.457787464026503e-06,
      "loss": 0.1807,
      "num_input_tokens_seen": 15570680,
      "step": 23840
    },
    {
      "epoch": 14.059551886792454,
      "grad_norm": 2.460700035095215,
      "learning_rate": 2.455572446982014e-06,
      "loss": 0.3016,
      "num_input_tokens_seen": 15573464,
      "step": 23845
    },
    {
      "epoch": 14.0625,
      "grad_norm": 3.396411418914795,
      "learning_rate": 2.453358103585686e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 15576760,
      "step": 23850
    },
    {
      "epoch": 14.065448113207546,
      "grad_norm": 2.6430442333221436,
      "learning_rate": 2.4511444344237733e-06,
      "loss": 0.3232,
      "num_input_tokens_seen": 15579928,
      "step": 23855
    },
    {
      "epoch": 14.068396226415095,
      "grad_norm": 5.919976234436035,
      "learning_rate": 2.4489314400823567e-06,
      "loss": 0.2907,
      "num_input_tokens_seen": 15582552,
      "step": 23860
    },
    {
      "epoch": 14.071344339622641,
      "grad_norm": 2.4246199131011963,
      "learning_rate": 2.446719121147337e-06,
      "loss": 0.2837,
      "num_input_tokens_seen": 15585944,
      "step": 23865
    },
    {
      "epoch": 14.07429245283019,
      "grad_norm": 5.797910690307617,
      "learning_rate": 2.4445074782044347e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 15588792,
      "step": 23870
    },
    {
      "epoch": 14.077240566037736,
      "grad_norm": 2.1635212898254395,
      "learning_rate": 2.442296511839191e-06,
      "loss": 0.2721,
      "num_input_tokens_seen": 15592344,
      "step": 23875
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 4.248343467712402,
      "learning_rate": 2.4400862226369687e-06,
      "loss": 0.4079,
      "num_input_tokens_seen": 15596184,
      "step": 23880
    },
    {
      "epoch": 14.08313679245283,
      "grad_norm": 6.113920211791992,
      "learning_rate": 2.4378766111829514e-06,
      "loss": 0.3483,
      "num_input_tokens_seen": 15599064,
      "step": 23885
    },
    {
      "epoch": 14.086084905660377,
      "grad_norm": 4.220350742340088,
      "learning_rate": 2.435667678062142e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 15602168,
      "step": 23890
    },
    {
      "epoch": 14.089033018867925,
      "grad_norm": 2.0245089530944824,
      "learning_rate": 2.4334594238593682e-06,
      "loss": 0.2605,
      "num_input_tokens_seen": 15605560,
      "step": 23895
    },
    {
      "epoch": 14.091981132075471,
      "grad_norm": 7.152381420135498,
      "learning_rate": 2.4312518491592727e-06,
      "loss": 0.3345,
      "num_input_tokens_seen": 15609656,
      "step": 23900
    },
    {
      "epoch": 14.09492924528302,
      "grad_norm": 2.9948017597198486,
      "learning_rate": 2.429044954546322e-06,
      "loss": 0.2889,
      "num_input_tokens_seen": 15612792,
      "step": 23905
    },
    {
      "epoch": 14.097877358490566,
      "grad_norm": 2.1704938411712646,
      "learning_rate": 2.426838740604799e-06,
      "loss": 0.2998,
      "num_input_tokens_seen": 15616312,
      "step": 23910
    },
    {
      "epoch": 14.100825471698114,
      "grad_norm": 3.8572373390197754,
      "learning_rate": 2.4246332079188066e-06,
      "loss": 0.4825,
      "num_input_tokens_seen": 15620504,
      "step": 23915
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 3.8055877685546875,
      "learning_rate": 2.4224283570722745e-06,
      "loss": 0.3836,
      "num_input_tokens_seen": 15623672,
      "step": 23920
    },
    {
      "epoch": 14.106721698113208,
      "grad_norm": 1.9803801774978638,
      "learning_rate": 2.420224188648943e-06,
      "loss": 0.2374,
      "num_input_tokens_seen": 15627480,
      "step": 23925
    },
    {
      "epoch": 14.109669811320755,
      "grad_norm": 2.5447030067443848,
      "learning_rate": 2.418020703232376e-06,
      "loss": 0.3356,
      "num_input_tokens_seen": 15630648,
      "step": 23930
    },
    {
      "epoch": 14.112617924528301,
      "grad_norm": 3.443413257598877,
      "learning_rate": 2.4158179014059556e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 15634936,
      "step": 23935
    },
    {
      "epoch": 14.11556603773585,
      "grad_norm": 1.9529718160629272,
      "learning_rate": 2.413615783752883e-06,
      "loss": 0.2709,
      "num_input_tokens_seen": 15638488,
      "step": 23940
    },
    {
      "epoch": 14.118514150943396,
      "grad_norm": 1.821315884590149,
      "learning_rate": 2.4114143508561767e-06,
      "loss": 0.2893,
      "num_input_tokens_seen": 15642072,
      "step": 23945
    },
    {
      "epoch": 14.121462264150944,
      "grad_norm": 8.359922409057617,
      "learning_rate": 2.4092136032986783e-06,
      "loss": 0.3983,
      "num_input_tokens_seen": 15644984,
      "step": 23950
    },
    {
      "epoch": 14.12441037735849,
      "grad_norm": 3.345102310180664,
      "learning_rate": 2.407013541663043e-06,
      "loss": 0.378,
      "num_input_tokens_seen": 15649400,
      "step": 23955
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 3.142261266708374,
      "learning_rate": 2.4048141665317493e-06,
      "loss": 0.3392,
      "num_input_tokens_seen": 15652952,
      "step": 23960
    },
    {
      "epoch": 14.130306603773585,
      "grad_norm": 4.062790393829346,
      "learning_rate": 2.40261547848709e-06,
      "loss": 0.2775,
      "num_input_tokens_seen": 15655864,
      "step": 23965
    },
    {
      "epoch": 14.133254716981131,
      "grad_norm": 6.66457986831665,
      "learning_rate": 2.400417478111176e-06,
      "loss": 0.3003,
      "num_input_tokens_seen": 15658680,
      "step": 23970
    },
    {
      "epoch": 14.13620283018868,
      "grad_norm": 1.9870926141738892,
      "learning_rate": 2.3982201659859387e-06,
      "loss": 0.212,
      "num_input_tokens_seen": 15661976,
      "step": 23975
    },
    {
      "epoch": 14.139150943396226,
      "grad_norm": 3.8333899974823,
      "learning_rate": 2.3960235426931237e-06,
      "loss": 0.4252,
      "num_input_tokens_seen": 15665112,
      "step": 23980
    },
    {
      "epoch": 14.142099056603774,
      "grad_norm": 3.343698263168335,
      "learning_rate": 2.3938276088143003e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 15667576,
      "step": 23985
    },
    {
      "epoch": 14.14504716981132,
      "grad_norm": 3.2212486267089844,
      "learning_rate": 2.391632364930849e-06,
      "loss": 0.2178,
      "num_input_tokens_seen": 15670488,
      "step": 23990
    },
    {
      "epoch": 14.147995283018869,
      "grad_norm": 3.1850037574768066,
      "learning_rate": 2.3894378116239706e-06,
      "loss": 0.3178,
      "num_input_tokens_seen": 15673368,
      "step": 23995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 4.440845489501953,
      "learning_rate": 2.387243949474683e-06,
      "loss": 0.2624,
      "num_input_tokens_seen": 15676248,
      "step": 24000
    },
    {
      "epoch": 14.153891509433961,
      "grad_norm": 3.8134968280792236,
      "learning_rate": 2.38505077906382e-06,
      "loss": 0.2876,
      "num_input_tokens_seen": 15680984,
      "step": 24005
    },
    {
      "epoch": 14.15683962264151,
      "grad_norm": 3.4116532802581787,
      "learning_rate": 2.382858300972031e-06,
      "loss": 0.2589,
      "num_input_tokens_seen": 15685560,
      "step": 24010
    },
    {
      "epoch": 14.159787735849056,
      "grad_norm": 3.0390658378601074,
      "learning_rate": 2.380666515779788e-06,
      "loss": 0.4161,
      "num_input_tokens_seen": 15689432,
      "step": 24015
    },
    {
      "epoch": 14.162735849056604,
      "grad_norm": 3.1250197887420654,
      "learning_rate": 2.3784754240673734e-06,
      "loss": 0.3692,
      "num_input_tokens_seen": 15692856,
      "step": 24020
    },
    {
      "epoch": 14.16568396226415,
      "grad_norm": 5.881439208984375,
      "learning_rate": 2.3762850264148883e-06,
      "loss": 0.4904,
      "num_input_tokens_seen": 15695224,
      "step": 24025
    },
    {
      "epoch": 14.168632075471699,
      "grad_norm": 7.135479927062988,
      "learning_rate": 2.374095323402251e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 15698680,
      "step": 24030
    },
    {
      "epoch": 14.171580188679245,
      "grad_norm": 4.393278121948242,
      "learning_rate": 2.371906315609193e-06,
      "loss": 0.3185,
      "num_input_tokens_seen": 15701880,
      "step": 24035
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 4.082833290100098,
      "learning_rate": 2.369718003615263e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 15704664,
      "step": 24040
    },
    {
      "epoch": 14.17747641509434,
      "grad_norm": 3.17533278465271,
      "learning_rate": 2.3675303879998284e-06,
      "loss": 0.276,
      "num_input_tokens_seen": 15707864,
      "step": 24045
    },
    {
      "epoch": 14.180424528301886,
      "grad_norm": 3.648939847946167,
      "learning_rate": 2.365343469342068e-06,
      "loss": 0.2816,
      "num_input_tokens_seen": 15711160,
      "step": 24050
    },
    {
      "epoch": 14.183372641509434,
      "grad_norm": 3.2173426151275635,
      "learning_rate": 2.3631572482209803e-06,
      "loss": 0.2518,
      "num_input_tokens_seen": 15714008,
      "step": 24055
    },
    {
      "epoch": 14.18632075471698,
      "grad_norm": 3.749488115310669,
      "learning_rate": 2.3609717252153752e-06,
      "loss": 0.2764,
      "num_input_tokens_seen": 15717144,
      "step": 24060
    },
    {
      "epoch": 14.189268867924529,
      "grad_norm": 7.296995162963867,
      "learning_rate": 2.35878690090388e-06,
      "loss": 0.2836,
      "num_input_tokens_seen": 15720088,
      "step": 24065
    },
    {
      "epoch": 14.192216981132075,
      "grad_norm": 4.150373935699463,
      "learning_rate": 2.356602775864935e-06,
      "loss": 0.3434,
      "num_input_tokens_seen": 15723288,
      "step": 24070
    },
    {
      "epoch": 14.195165094339623,
      "grad_norm": 5.714511871337891,
      "learning_rate": 2.354419350676796e-06,
      "loss": 0.3466,
      "num_input_tokens_seen": 15726552,
      "step": 24075
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 2.7193050384521484,
      "learning_rate": 2.3522366259175377e-06,
      "loss": 0.2678,
      "num_input_tokens_seen": 15729368,
      "step": 24080
    },
    {
      "epoch": 14.201061320754716,
      "grad_norm": 3.1136112213134766,
      "learning_rate": 2.350054602165044e-06,
      "loss": 0.2771,
      "num_input_tokens_seen": 15733560,
      "step": 24085
    },
    {
      "epoch": 14.204009433962264,
      "grad_norm": 3.0337936878204346,
      "learning_rate": 2.3478732799970143e-06,
      "loss": 0.3432,
      "num_input_tokens_seen": 15737144,
      "step": 24090
    },
    {
      "epoch": 14.20695754716981,
      "grad_norm": 2.9346978664398193,
      "learning_rate": 2.3456926599909646e-06,
      "loss": 0.3761,
      "num_input_tokens_seen": 15740760,
      "step": 24095
    },
    {
      "epoch": 14.209905660377359,
      "grad_norm": 2.8554601669311523,
      "learning_rate": 2.343512742724222e-06,
      "loss": 0.2832,
      "num_input_tokens_seen": 15744248,
      "step": 24100
    },
    {
      "epoch": 14.212853773584905,
      "grad_norm": 2.938037633895874,
      "learning_rate": 2.341333528773928e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 15747896,
      "step": 24105
    },
    {
      "epoch": 14.215801886792454,
      "grad_norm": 3.329021692276001,
      "learning_rate": 2.3391550187170427e-06,
      "loss": 0.3451,
      "num_input_tokens_seen": 15750936,
      "step": 24110
    },
    {
      "epoch": 14.21875,
      "grad_norm": 3.418834686279297,
      "learning_rate": 2.336977213130333e-06,
      "loss": 0.3575,
      "num_input_tokens_seen": 15754680,
      "step": 24115
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 2.88189435005188,
      "learning_rate": 2.3348001125903837e-06,
      "loss": 0.3581,
      "num_input_tokens_seen": 15757816,
      "step": 24120
    },
    {
      "epoch": 14.224646226415095,
      "grad_norm": 2.366670846939087,
      "learning_rate": 2.3326237176735905e-06,
      "loss": 0.175,
      "num_input_tokens_seen": 15762680,
      "step": 24125
    },
    {
      "epoch": 14.227594339622641,
      "grad_norm": 2.978520631790161,
      "learning_rate": 2.330448028956164e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 15766040,
      "step": 24130
    },
    {
      "epoch": 14.23054245283019,
      "grad_norm": 2.230347156524658,
      "learning_rate": 2.3282730470141255e-06,
      "loss": 0.2452,
      "num_input_tokens_seen": 15768888,
      "step": 24135
    },
    {
      "epoch": 14.233490566037736,
      "grad_norm": 4.9864702224731445,
      "learning_rate": 2.3260987724233143e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 15771768,
      "step": 24140
    },
    {
      "epoch": 14.236438679245284,
      "grad_norm": 4.303669452667236,
      "learning_rate": 2.323925205759374e-06,
      "loss": 0.2838,
      "num_input_tokens_seen": 15774808,
      "step": 24145
    },
    {
      "epoch": 14.23938679245283,
      "grad_norm": 4.27818489074707,
      "learning_rate": 2.3217523475977715e-06,
      "loss": 0.3696,
      "num_input_tokens_seen": 15778552,
      "step": 24150
    },
    {
      "epoch": 14.242334905660377,
      "grad_norm": 2.1842057704925537,
      "learning_rate": 2.3195801985137773e-06,
      "loss": 0.2842,
      "num_input_tokens_seen": 15782840,
      "step": 24155
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 2.8917927742004395,
      "learning_rate": 2.317408759082478e-06,
      "loss": 0.3058,
      "num_input_tokens_seen": 15786232,
      "step": 24160
    },
    {
      "epoch": 14.248231132075471,
      "grad_norm": 2.476942777633667,
      "learning_rate": 2.31523802987877e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 15790040,
      "step": 24165
    },
    {
      "epoch": 14.25117924528302,
      "grad_norm": 2.9435601234436035,
      "learning_rate": 2.3130680114773637e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 15793432,
      "step": 24170
    },
    {
      "epoch": 14.254127358490566,
      "grad_norm": 2.982956647872925,
      "learning_rate": 2.310898704452782e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 15797048,
      "step": 24175
    },
    {
      "epoch": 14.257075471698114,
      "grad_norm": 3.5912795066833496,
      "learning_rate": 2.3087301093793584e-06,
      "loss": 0.3364,
      "num_input_tokens_seen": 15799928,
      "step": 24180
    },
    {
      "epoch": 14.26002358490566,
      "grad_norm": 3.1030428409576416,
      "learning_rate": 2.306562226831237e-06,
      "loss": 0.2466,
      "num_input_tokens_seen": 15803032,
      "step": 24185
    },
    {
      "epoch": 14.262971698113208,
      "grad_norm": 4.421914577484131,
      "learning_rate": 2.304395057382374e-06,
      "loss": 0.4581,
      "num_input_tokens_seen": 15808952,
      "step": 24190
    },
    {
      "epoch": 14.265919811320755,
      "grad_norm": 4.445845127105713,
      "learning_rate": 2.3022286016065354e-06,
      "loss": 0.2958,
      "num_input_tokens_seen": 15812696,
      "step": 24195
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 3.1967415809631348,
      "learning_rate": 2.300062860077303e-06,
      "loss": 0.2561,
      "num_input_tokens_seen": 15815512,
      "step": 24200
    },
    {
      "epoch": 14.27181603773585,
      "grad_norm": 2.0715417861938477,
      "learning_rate": 2.297897833368064e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 15818648,
      "step": 24205
    },
    {
      "epoch": 14.274764150943396,
      "grad_norm": 3.38955020904541,
      "learning_rate": 2.2957335220520194e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 15821976,
      "step": 24210
    },
    {
      "epoch": 14.277712264150944,
      "grad_norm": 1.5393378734588623,
      "learning_rate": 2.293569926702179e-06,
      "loss": 0.263,
      "num_input_tokens_seen": 15825496,
      "step": 24215
    },
    {
      "epoch": 14.28066037735849,
      "grad_norm": 3.302006244659424,
      "learning_rate": 2.291407047891366e-06,
      "loss": 0.2404,
      "num_input_tokens_seen": 15827960,
      "step": 24220
    },
    {
      "epoch": 14.283608490566039,
      "grad_norm": 3.4368178844451904,
      "learning_rate": 2.2892448861922075e-06,
      "loss": 0.2387,
      "num_input_tokens_seen": 15832056,
      "step": 24225
    },
    {
      "epoch": 14.286556603773585,
      "grad_norm": 4.483919620513916,
      "learning_rate": 2.2870834421771505e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 15834872,
      "step": 24230
    },
    {
      "epoch": 14.289504716981131,
      "grad_norm": 3.6195271015167236,
      "learning_rate": 2.2849227164184433e-06,
      "loss": 0.3918,
      "num_input_tokens_seen": 15838456,
      "step": 24235
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 3.367466926574707,
      "learning_rate": 2.2827627094881473e-06,
      "loss": 0.4261,
      "num_input_tokens_seen": 15841464,
      "step": 24240
    },
    {
      "epoch": 14.295400943396226,
      "grad_norm": 3.781348943710327,
      "learning_rate": 2.2806034219581364e-06,
      "loss": 0.3271,
      "num_input_tokens_seen": 15844696,
      "step": 24245
    },
    {
      "epoch": 14.298349056603774,
      "grad_norm": 4.516969680786133,
      "learning_rate": 2.278444854400089e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 15847640,
      "step": 24250
    },
    {
      "epoch": 14.30129716981132,
      "grad_norm": 2.846116065979004,
      "learning_rate": 2.276287007385496e-06,
      "loss": 0.3305,
      "num_input_tokens_seen": 15850456,
      "step": 24255
    },
    {
      "epoch": 14.304245283018869,
      "grad_norm": 3.10361909866333,
      "learning_rate": 2.2741298814856542e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 15852888,
      "step": 24260
    },
    {
      "epoch": 14.307193396226415,
      "grad_norm": 2.620551347732544,
      "learning_rate": 2.2719734772716763e-06,
      "loss": 0.3106,
      "num_input_tokens_seen": 15856280,
      "step": 24265
    },
    {
      "epoch": 14.310141509433961,
      "grad_norm": 4.810905456542969,
      "learning_rate": 2.269817795314477e-06,
      "loss": 0.526,
      "num_input_tokens_seen": 15858872,
      "step": 24270
    },
    {
      "epoch": 14.31308962264151,
      "grad_norm": 3.9686224460601807,
      "learning_rate": 2.2676628361847834e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 15862328,
      "step": 24275
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 4.161613464355469,
      "learning_rate": 2.2655086004531296e-06,
      "loss": 0.3442,
      "num_input_tokens_seen": 15865496,
      "step": 24280
    },
    {
      "epoch": 14.318985849056604,
      "grad_norm": 2.2536191940307617,
      "learning_rate": 2.2633550886898583e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 15868632,
      "step": 24285
    },
    {
      "epoch": 14.32193396226415,
      "grad_norm": 5.337737083435059,
      "learning_rate": 2.26120230146512e-06,
      "loss": 0.3865,
      "num_input_tokens_seen": 15871416,
      "step": 24290
    },
    {
      "epoch": 14.324882075471699,
      "grad_norm": 5.2575507164001465,
      "learning_rate": 2.2590502393488777e-06,
      "loss": 0.3472,
      "num_input_tokens_seen": 15873944,
      "step": 24295
    },
    {
      "epoch": 14.327830188679245,
      "grad_norm": 1.5214521884918213,
      "learning_rate": 2.256898902910898e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 15877848,
      "step": 24300
    },
    {
      "epoch": 14.330778301886792,
      "grad_norm": 2.283863067626953,
      "learning_rate": 2.2547482927207548e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 15881080,
      "step": 24305
    },
    {
      "epoch": 14.33372641509434,
      "grad_norm": 6.532295227050781,
      "learning_rate": 2.252598409347833e-06,
      "loss": 0.4324,
      "num_input_tokens_seen": 15884312,
      "step": 24310
    },
    {
      "epoch": 14.336674528301886,
      "grad_norm": 2.454620599746704,
      "learning_rate": 2.250449253361323e-06,
      "loss": 0.2873,
      "num_input_tokens_seen": 15889592,
      "step": 24315
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 4.563135147094727,
      "learning_rate": 2.2483008253302214e-06,
      "loss": 0.324,
      "num_input_tokens_seen": 15892664,
      "step": 24320
    },
    {
      "epoch": 14.34257075471698,
      "grad_norm": 2.954381227493286,
      "learning_rate": 2.246153125823337e-06,
      "loss": 0.3222,
      "num_input_tokens_seen": 15895320,
      "step": 24325
    },
    {
      "epoch": 14.345518867924529,
      "grad_norm": 3.0437660217285156,
      "learning_rate": 2.2440061554092813e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 15898360,
      "step": 24330
    },
    {
      "epoch": 14.348466981132075,
      "grad_norm": 2.0895729064941406,
      "learning_rate": 2.2418599146564714e-06,
      "loss": 0.4822,
      "num_input_tokens_seen": 15901144,
      "step": 24335
    },
    {
      "epoch": 14.351415094339623,
      "grad_norm": 3.240769863128662,
      "learning_rate": 2.239714404133138e-06,
      "loss": 0.3344,
      "num_input_tokens_seen": 15903672,
      "step": 24340
    },
    {
      "epoch": 14.35436320754717,
      "grad_norm": 5.938223838806152,
      "learning_rate": 2.2375696244073126e-06,
      "loss": 0.4396,
      "num_input_tokens_seen": 15910200,
      "step": 24345
    },
    {
      "epoch": 14.357311320754716,
      "grad_norm": 4.421813011169434,
      "learning_rate": 2.235425576046834e-06,
      "loss": 0.3306,
      "num_input_tokens_seen": 15912632,
      "step": 24350
    },
    {
      "epoch": 14.360259433962264,
      "grad_norm": 4.611784934997559,
      "learning_rate": 2.233282259619347e-06,
      "loss": 0.403,
      "num_input_tokens_seen": 15915352,
      "step": 24355
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 4.816007614135742,
      "learning_rate": 2.231139675692308e-06,
      "loss": 0.2837,
      "num_input_tokens_seen": 15918296,
      "step": 24360
    },
    {
      "epoch": 14.366155660377359,
      "grad_norm": 2.789311170578003,
      "learning_rate": 2.228997824832973e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 15922104,
      "step": 24365
    },
    {
      "epoch": 14.369103773584905,
      "grad_norm": 3.1000075340270996,
      "learning_rate": 2.226856707608406e-06,
      "loss": 0.3687,
      "num_input_tokens_seen": 15925272,
      "step": 24370
    },
    {
      "epoch": 14.372051886792454,
      "grad_norm": 3.254448652267456,
      "learning_rate": 2.2247163245854768e-06,
      "loss": 0.307,
      "num_input_tokens_seen": 15928184,
      "step": 24375
    },
    {
      "epoch": 14.375,
      "grad_norm": 3.145495891571045,
      "learning_rate": 2.222576676330862e-06,
      "loss": 0.3421,
      "num_input_tokens_seen": 15930776,
      "step": 24380
    },
    {
      "epoch": 14.377948113207546,
      "grad_norm": 3.0852370262145996,
      "learning_rate": 2.2204377634110403e-06,
      "loss": 0.4154,
      "num_input_tokens_seen": 15933752,
      "step": 24385
    },
    {
      "epoch": 14.380896226415095,
      "grad_norm": 2.2430670261383057,
      "learning_rate": 2.218299586392301e-06,
      "loss": 0.2866,
      "num_input_tokens_seen": 15937912,
      "step": 24390
    },
    {
      "epoch": 14.383844339622641,
      "grad_norm": 2.94907808303833,
      "learning_rate": 2.2161621458407355e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 15940664,
      "step": 24395
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 3.1298296451568604,
      "learning_rate": 2.2140254423222398e-06,
      "loss": 0.3773,
      "num_input_tokens_seen": 15944120,
      "step": 24400
    },
    {
      "epoch": 14.389740566037736,
      "grad_norm": 2.820992946624756,
      "learning_rate": 2.2118894764025146e-06,
      "loss": 0.3283,
      "num_input_tokens_seen": 15947192,
      "step": 24405
    },
    {
      "epoch": 14.392688679245284,
      "grad_norm": 3.1221988201141357,
      "learning_rate": 2.2097542486470667e-06,
      "loss": 0.3597,
      "num_input_tokens_seen": 15949880,
      "step": 24410
    },
    {
      "epoch": 14.39563679245283,
      "grad_norm": 3.935938596725464,
      "learning_rate": 2.207619759621205e-06,
      "loss": 0.391,
      "num_input_tokens_seen": 15954168,
      "step": 24415
    },
    {
      "epoch": 14.398584905660377,
      "grad_norm": 3.471708059310913,
      "learning_rate": 2.205486009890049e-06,
      "loss": 0.6005,
      "num_input_tokens_seen": 15957784,
      "step": 24420
    },
    {
      "epoch": 14.401533018867925,
      "grad_norm": 3.2116477489471436,
      "learning_rate": 2.2033530000185146e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 15960760,
      "step": 24425
    },
    {
      "epoch": 14.404481132075471,
      "grad_norm": 4.2083539962768555,
      "learning_rate": 2.2012207305713244e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 15964600,
      "step": 24430
    },
    {
      "epoch": 14.40742924528302,
      "grad_norm": 7.238643646240234,
      "learning_rate": 2.19908920211301e-06,
      "loss": 0.2952,
      "num_input_tokens_seen": 15967320,
      "step": 24435
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 6.290116310119629,
      "learning_rate": 2.196958415207901e-06,
      "loss": 0.275,
      "num_input_tokens_seen": 15970072,
      "step": 24440
    },
    {
      "epoch": 14.413325471698114,
      "grad_norm": 2.0073981285095215,
      "learning_rate": 2.1948283704201312e-06,
      "loss": 0.3596,
      "num_input_tokens_seen": 15972856,
      "step": 24445
    },
    {
      "epoch": 14.41627358490566,
      "grad_norm": 3.609239339828491,
      "learning_rate": 2.1926990683136383e-06,
      "loss": 0.3517,
      "num_input_tokens_seen": 15975320,
      "step": 24450
    },
    {
      "epoch": 14.419221698113208,
      "grad_norm": 3.7272489070892334,
      "learning_rate": 2.1905705094521685e-06,
      "loss": 0.2565,
      "num_input_tokens_seen": 15978328,
      "step": 24455
    },
    {
      "epoch": 14.422169811320755,
      "grad_norm": 7.36161994934082,
      "learning_rate": 2.1884426943992635e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 15982104,
      "step": 24460
    },
    {
      "epoch": 14.425117924528301,
      "grad_norm": 4.012758731842041,
      "learning_rate": 2.1863156237182727e-06,
      "loss": 0.2764,
      "num_input_tokens_seen": 15985304,
      "step": 24465
    },
    {
      "epoch": 14.42806603773585,
      "grad_norm": 2.775627613067627,
      "learning_rate": 2.1841892979723466e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 15989368,
      "step": 24470
    },
    {
      "epoch": 14.431014150943396,
      "grad_norm": 3.8198800086975098,
      "learning_rate": 2.1820637177244375e-06,
      "loss": 0.4786,
      "num_input_tokens_seen": 15992664,
      "step": 24475
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 4.719549655914307,
      "learning_rate": 2.179938883537306e-06,
      "loss": 0.2594,
      "num_input_tokens_seen": 15995384,
      "step": 24480
    },
    {
      "epoch": 14.43691037735849,
      "grad_norm": 4.316824436187744,
      "learning_rate": 2.177814795973508e-06,
      "loss": 0.3201,
      "num_input_tokens_seen": 15998712,
      "step": 24485
    },
    {
      "epoch": 14.439858490566039,
      "grad_norm": 4.16162633895874,
      "learning_rate": 2.1756914555954064e-06,
      "loss": 0.3075,
      "num_input_tokens_seen": 16001848,
      "step": 24490
    },
    {
      "epoch": 14.442806603773585,
      "grad_norm": 3.8511321544647217,
      "learning_rate": 2.173568862965164e-06,
      "loss": 0.3074,
      "num_input_tokens_seen": 16004856,
      "step": 24495
    },
    {
      "epoch": 14.445754716981131,
      "grad_norm": 3.966292142868042,
      "learning_rate": 2.171447018644746e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 16007416,
      "step": 24500
    },
    {
      "epoch": 14.44870283018868,
      "grad_norm": 3.8753838539123535,
      "learning_rate": 2.1693259231959186e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 16010776,
      "step": 24505
    },
    {
      "epoch": 14.451650943396226,
      "grad_norm": 2.3124217987060547,
      "learning_rate": 2.1672055771802545e-06,
      "loss": 0.4229,
      "num_input_tokens_seen": 16014104,
      "step": 24510
    },
    {
      "epoch": 14.454599056603774,
      "grad_norm": 4.219889163970947,
      "learning_rate": 2.1650859811591224e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 16017496,
      "step": 24515
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 3.0777931213378906,
      "learning_rate": 2.1629671356936943e-06,
      "loss": 0.2812,
      "num_input_tokens_seen": 16022008,
      "step": 24520
    },
    {
      "epoch": 14.460495283018869,
      "grad_norm": 2.0894696712493896,
      "learning_rate": 2.1608490413449428e-06,
      "loss": 0.3787,
      "num_input_tokens_seen": 16025784,
      "step": 24525
    },
    {
      "epoch": 14.463443396226415,
      "grad_norm": 2.240814685821533,
      "learning_rate": 2.158731698673645e-06,
      "loss": 0.3368,
      "num_input_tokens_seen": 16029560,
      "step": 24530
    },
    {
      "epoch": 14.466391509433961,
      "grad_norm": 3.187756299972534,
      "learning_rate": 2.1566151082403752e-06,
      "loss": 0.3304,
      "num_input_tokens_seen": 16033528,
      "step": 24535
    },
    {
      "epoch": 14.46933962264151,
      "grad_norm": 3.406580686569214,
      "learning_rate": 2.154499270605508e-06,
      "loss": 0.2678,
      "num_input_tokens_seen": 16037752,
      "step": 24540
    },
    {
      "epoch": 14.472287735849056,
      "grad_norm": 2.786196231842041,
      "learning_rate": 2.1523841863292243e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 16041464,
      "step": 24545
    },
    {
      "epoch": 14.475235849056604,
      "grad_norm": 2.4996469020843506,
      "learning_rate": 2.1502698559714998e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 16044056,
      "step": 24550
    },
    {
      "epoch": 14.47818396226415,
      "grad_norm": 8.866477966308594,
      "learning_rate": 2.1481562800921125e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 16047384,
      "step": 24555
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 2.6284923553466797,
      "learning_rate": 2.146043459250641e-06,
      "loss": 0.3227,
      "num_input_tokens_seen": 16051032,
      "step": 24560
    },
    {
      "epoch": 14.484080188679245,
      "grad_norm": 3.3459198474884033,
      "learning_rate": 2.1439313940064634e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 16054360,
      "step": 24565
    },
    {
      "epoch": 14.487028301886792,
      "grad_norm": 2.8457231521606445,
      "learning_rate": 2.141820084918756e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 16057240,
      "step": 24570
    },
    {
      "epoch": 14.48997641509434,
      "grad_norm": 3.2799975872039795,
      "learning_rate": 2.1397095325465013e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 16060056,
      "step": 24575
    },
    {
      "epoch": 14.492924528301886,
      "grad_norm": 7.33595609664917,
      "learning_rate": 2.1375997374484754e-06,
      "loss": 0.3132,
      "num_input_tokens_seen": 16062776,
      "step": 24580
    },
    {
      "epoch": 14.495872641509434,
      "grad_norm": 5.147444725036621,
      "learning_rate": 2.1354907001832546e-06,
      "loss": 0.2886,
      "num_input_tokens_seen": 16066552,
      "step": 24585
    },
    {
      "epoch": 14.49882075471698,
      "grad_norm": 3.8993031978607178,
      "learning_rate": 2.133382421309217e-06,
      "loss": 0.3969,
      "num_input_tokens_seen": 16069720,
      "step": 24590
    },
    {
      "epoch": 14.501768867924529,
      "grad_norm": 2.5256786346435547,
      "learning_rate": 2.131274901384537e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 16072888,
      "step": 24595
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 3.867098093032837,
      "learning_rate": 2.1291681409671896e-06,
      "loss": 0.4877,
      "num_input_tokens_seen": 16076184,
      "step": 24600
    },
    {
      "epoch": 14.507665094339622,
      "grad_norm": 3.8528707027435303,
      "learning_rate": 2.12706214061495e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 16079288,
      "step": 24605
    },
    {
      "epoch": 14.51061320754717,
      "grad_norm": 2.214556932449341,
      "learning_rate": 2.124956900885391e-06,
      "loss": 0.4206,
      "num_input_tokens_seen": 16082904,
      "step": 24610
    },
    {
      "epoch": 14.513561320754716,
      "grad_norm": 4.183326244354248,
      "learning_rate": 2.1228524223358833e-06,
      "loss": 0.2862,
      "num_input_tokens_seen": 16086168,
      "step": 24615
    },
    {
      "epoch": 14.516509433962264,
      "grad_norm": 2.415451765060425,
      "learning_rate": 2.120748705523595e-06,
      "loss": 0.2547,
      "num_input_tokens_seen": 16089560,
      "step": 24620
    },
    {
      "epoch": 14.51945754716981,
      "grad_norm": 2.617783308029175,
      "learning_rate": 2.1186457510054976e-06,
      "loss": 0.3272,
      "num_input_tokens_seen": 16092760,
      "step": 24625
    },
    {
      "epoch": 14.522405660377359,
      "grad_norm": 3.2810592651367188,
      "learning_rate": 2.116543559338355e-06,
      "loss": 0.3401,
      "num_input_tokens_seen": 16095352,
      "step": 24630
    },
    {
      "epoch": 14.525353773584905,
      "grad_norm": 3.7945711612701416,
      "learning_rate": 2.1144421310787305e-06,
      "loss": 0.5453,
      "num_input_tokens_seen": 16098808,
      "step": 24635
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 2.2579498291015625,
      "learning_rate": 2.11234146678299e-06,
      "loss": 0.2811,
      "num_input_tokens_seen": 16102008,
      "step": 24640
    },
    {
      "epoch": 14.53125,
      "grad_norm": 2.7898025512695312,
      "learning_rate": 2.1102415670072907e-06,
      "loss": 0.3471,
      "num_input_tokens_seen": 16104856,
      "step": 24645
    },
    {
      "epoch": 14.534198113207546,
      "grad_norm": 3.331444501876831,
      "learning_rate": 2.108142432307591e-06,
      "loss": 0.3527,
      "num_input_tokens_seen": 16108344,
      "step": 24650
    },
    {
      "epoch": 14.537146226415095,
      "grad_norm": 2.3106753826141357,
      "learning_rate": 2.1060440632396456e-06,
      "loss": 0.2835,
      "num_input_tokens_seen": 16111576,
      "step": 24655
    },
    {
      "epoch": 14.540094339622641,
      "grad_norm": 3.32875394821167,
      "learning_rate": 2.103946460359007e-06,
      "loss": 0.263,
      "num_input_tokens_seen": 16114456,
      "step": 24660
    },
    {
      "epoch": 14.54304245283019,
      "grad_norm": 3.405618906021118,
      "learning_rate": 2.101849624221022e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 16116952,
      "step": 24665
    },
    {
      "epoch": 14.545990566037736,
      "grad_norm": 2.856860876083374,
      "learning_rate": 2.0997535553808417e-06,
      "loss": 0.3645,
      "num_input_tokens_seen": 16121080,
      "step": 24670
    },
    {
      "epoch": 14.548938679245284,
      "grad_norm": 2.8556859493255615,
      "learning_rate": 2.0976582543934064e-06,
      "loss": 0.328,
      "num_input_tokens_seen": 16123416,
      "step": 24675
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 2.033029556274414,
      "learning_rate": 2.0955637218134573e-06,
      "loss": 0.3322,
      "num_input_tokens_seen": 16126680,
      "step": 24680
    },
    {
      "epoch": 14.554834905660378,
      "grad_norm": 3.4544785022735596,
      "learning_rate": 2.09346995819553e-06,
      "loss": 0.378,
      "num_input_tokens_seen": 16129784,
      "step": 24685
    },
    {
      "epoch": 14.557783018867925,
      "grad_norm": 6.0805439949035645,
      "learning_rate": 2.0913769640939553e-06,
      "loss": 0.2974,
      "num_input_tokens_seen": 16135128,
      "step": 24690
    },
    {
      "epoch": 14.560731132075471,
      "grad_norm": 2.8355491161346436,
      "learning_rate": 2.0892847400628674e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 16137880,
      "step": 24695
    },
    {
      "epoch": 14.56367924528302,
      "grad_norm": 4.272465705871582,
      "learning_rate": 2.0871932866561885e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 16140888,
      "step": 24700
    },
    {
      "epoch": 14.566627358490566,
      "grad_norm": 9.879657745361328,
      "learning_rate": 2.0851026044276405e-06,
      "loss": 0.3941,
      "num_input_tokens_seen": 16143064,
      "step": 24705
    },
    {
      "epoch": 14.569575471698114,
      "grad_norm": 3.8229897022247314,
      "learning_rate": 2.083012693930741e-06,
      "loss": 0.2112,
      "num_input_tokens_seen": 16145560,
      "step": 24710
    },
    {
      "epoch": 14.57252358490566,
      "grad_norm": 1.9774161577224731,
      "learning_rate": 2.0809235557188e-06,
      "loss": 0.401,
      "num_input_tokens_seen": 16148728,
      "step": 24715
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 4.820782661437988,
      "learning_rate": 2.0788351903449307e-06,
      "loss": 0.2823,
      "num_input_tokens_seen": 16151096,
      "step": 24720
    },
    {
      "epoch": 14.578419811320755,
      "grad_norm": 3.133119821548462,
      "learning_rate": 2.0767475983620317e-06,
      "loss": 0.2864,
      "num_input_tokens_seen": 16153784,
      "step": 24725
    },
    {
      "epoch": 14.581367924528301,
      "grad_norm": 4.325511455535889,
      "learning_rate": 2.074660780322806e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 16156472,
      "step": 24730
    },
    {
      "epoch": 14.58431603773585,
      "grad_norm": 3.1263551712036133,
      "learning_rate": 2.0725747367797473e-06,
      "loss": 0.2674,
      "num_input_tokens_seen": 16160792,
      "step": 24735
    },
    {
      "epoch": 14.587264150943396,
      "grad_norm": 2.9961209297180176,
      "learning_rate": 2.070489468285143e-06,
      "loss": 0.4325,
      "num_input_tokens_seen": 16164760,
      "step": 24740
    },
    {
      "epoch": 14.590212264150944,
      "grad_norm": 2.359610080718994,
      "learning_rate": 2.068404975391077e-06,
      "loss": 0.2332,
      "num_input_tokens_seen": 16168088,
      "step": 24745
    },
    {
      "epoch": 14.59316037735849,
      "grad_norm": 2.961120843887329,
      "learning_rate": 2.0663212586494293e-06,
      "loss": 0.2889,
      "num_input_tokens_seen": 16170712,
      "step": 24750
    },
    {
      "epoch": 14.596108490566039,
      "grad_norm": 3.9842402935028076,
      "learning_rate": 2.064238318611869e-06,
      "loss": 0.2711,
      "num_input_tokens_seen": 16176664,
      "step": 24755
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 3.6218059062957764,
      "learning_rate": 2.0621561558298693e-06,
      "loss": 0.2987,
      "num_input_tokens_seen": 16180344,
      "step": 24760
    },
    {
      "epoch": 14.602004716981131,
      "grad_norm": 2.012690305709839,
      "learning_rate": 2.0600747708546877e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 16184120,
      "step": 24765
    },
    {
      "epoch": 14.60495283018868,
      "grad_norm": 4.297055244445801,
      "learning_rate": 2.0579941642373814e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 16186392,
      "step": 24770
    },
    {
      "epoch": 14.607900943396226,
      "grad_norm": 2.8719468116760254,
      "learning_rate": 2.0559143365287993e-06,
      "loss": 0.3569,
      "num_input_tokens_seen": 16189688,
      "step": 24775
    },
    {
      "epoch": 14.610849056603774,
      "grad_norm": 4.000527858734131,
      "learning_rate": 2.0538352882795846e-06,
      "loss": 0.381,
      "num_input_tokens_seen": 16193432,
      "step": 24780
    },
    {
      "epoch": 14.61379716981132,
      "grad_norm": 9.313379287719727,
      "learning_rate": 2.051757020040173e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 16196792,
      "step": 24785
    },
    {
      "epoch": 14.616745283018869,
      "grad_norm": 3.0866851806640625,
      "learning_rate": 2.0496795323607983e-06,
      "loss": 0.234,
      "num_input_tokens_seen": 16199320,
      "step": 24790
    },
    {
      "epoch": 14.619693396226415,
      "grad_norm": 7.123655319213867,
      "learning_rate": 2.0476028257914825e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 16203576,
      "step": 24795
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 4.255812168121338,
      "learning_rate": 2.0455269008820433e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 16206680,
      "step": 24800
    },
    {
      "epoch": 14.62558962264151,
      "grad_norm": 2.1244893074035645,
      "learning_rate": 2.0434517581820893e-06,
      "loss": 0.3801,
      "num_input_tokens_seen": 16210264,
      "step": 24805
    },
    {
      "epoch": 14.628537735849056,
      "grad_norm": 1.6337798833847046,
      "learning_rate": 2.041377398241025e-06,
      "loss": 0.2808,
      "num_input_tokens_seen": 16213624,
      "step": 24810
    },
    {
      "epoch": 14.631485849056604,
      "grad_norm": 2.284867525100708,
      "learning_rate": 2.0393038216080433e-06,
      "loss": 0.3234,
      "num_input_tokens_seen": 16216632,
      "step": 24815
    },
    {
      "epoch": 14.63443396226415,
      "grad_norm": 3.6389291286468506,
      "learning_rate": 2.037231028832135e-06,
      "loss": 0.4025,
      "num_input_tokens_seen": 16219192,
      "step": 24820
    },
    {
      "epoch": 14.637382075471699,
      "grad_norm": 2.6630306243896484,
      "learning_rate": 2.0351590204620823e-06,
      "loss": 0.3707,
      "num_input_tokens_seen": 16221848,
      "step": 24825
    },
    {
      "epoch": 14.640330188679245,
      "grad_norm": 3.6553938388824463,
      "learning_rate": 2.033087797046457e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 16224856,
      "step": 24830
    },
    {
      "epoch": 14.643278301886792,
      "grad_norm": 2.101240396499634,
      "learning_rate": 2.031017359133624e-06,
      "loss": 0.2857,
      "num_input_tokens_seen": 16227704,
      "step": 24835
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 2.479818820953369,
      "learning_rate": 2.0289477072717406e-06,
      "loss": 0.3505,
      "num_input_tokens_seen": 16231512,
      "step": 24840
    },
    {
      "epoch": 14.649174528301886,
      "grad_norm": 3.084902763366699,
      "learning_rate": 2.026878842008756e-06,
      "loss": 0.2665,
      "num_input_tokens_seen": 16234680,
      "step": 24845
    },
    {
      "epoch": 14.652122641509434,
      "grad_norm": 2.891141414642334,
      "learning_rate": 2.0248107638924105e-06,
      "loss": 0.2916,
      "num_input_tokens_seen": 16237304,
      "step": 24850
    },
    {
      "epoch": 14.65507075471698,
      "grad_norm": 3.0662848949432373,
      "learning_rate": 2.0227434734702386e-06,
      "loss": 0.3065,
      "num_input_tokens_seen": 16240376,
      "step": 24855
    },
    {
      "epoch": 14.658018867924529,
      "grad_norm": 2.3439228534698486,
      "learning_rate": 2.020676971289563e-06,
      "loss": 0.2992,
      "num_input_tokens_seen": 16243928,
      "step": 24860
    },
    {
      "epoch": 14.660966981132075,
      "grad_norm": 2.0607473850250244,
      "learning_rate": 2.0186112578975005e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 16246872,
      "step": 24865
    },
    {
      "epoch": 14.663915094339622,
      "grad_norm": 2.851121664047241,
      "learning_rate": 2.016546333840956e-06,
      "loss": 0.3264,
      "num_input_tokens_seen": 16249912,
      "step": 24870
    },
    {
      "epoch": 14.66686320754717,
      "grad_norm": 4.2973527908325195,
      "learning_rate": 2.014482199666627e-06,
      "loss": 0.3707,
      "num_input_tokens_seen": 16253144,
      "step": 24875
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 3.3511409759521484,
      "learning_rate": 2.0124188559210017e-06,
      "loss": 0.4443,
      "num_input_tokens_seen": 16256376,
      "step": 24880
    },
    {
      "epoch": 14.672759433962264,
      "grad_norm": 2.756486415863037,
      "learning_rate": 2.0103563031503613e-06,
      "loss": 0.2941,
      "num_input_tokens_seen": 16260856,
      "step": 24885
    },
    {
      "epoch": 14.67570754716981,
      "grad_norm": 2.4098315238952637,
      "learning_rate": 2.0082945419007745e-06,
      "loss": 0.338,
      "num_input_tokens_seen": 16264184,
      "step": 24890
    },
    {
      "epoch": 14.678655660377359,
      "grad_norm": 4.843575477600098,
      "learning_rate": 2.0062335727181007e-06,
      "loss": 0.3525,
      "num_input_tokens_seen": 16267416,
      "step": 24895
    },
    {
      "epoch": 14.681603773584905,
      "grad_norm": 3.28536057472229,
      "learning_rate": 2.004173396147992e-06,
      "loss": 0.3239,
      "num_input_tokens_seen": 16271352,
      "step": 24900
    },
    {
      "epoch": 14.684551886792454,
      "grad_norm": 2.6036696434020996,
      "learning_rate": 2.0021140127358873e-06,
      "loss": 0.4046,
      "num_input_tokens_seen": 16274392,
      "step": 24905
    },
    {
      "epoch": 14.6875,
      "grad_norm": 3.2717294692993164,
      "learning_rate": 2.0000554230270164e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 16277464,
      "step": 24910
    },
    {
      "epoch": 14.690448113207546,
      "grad_norm": 4.876547336578369,
      "learning_rate": 1.997997627566401e-06,
      "loss": 0.3606,
      "num_input_tokens_seen": 16280952,
      "step": 24915
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 4.279379367828369,
      "learning_rate": 1.9959406268988536e-06,
      "loss": 0.3635,
      "num_input_tokens_seen": 16283768,
      "step": 24920
    },
    {
      "epoch": 14.696344339622641,
      "grad_norm": 2.870016574859619,
      "learning_rate": 1.9938844215689717e-06,
      "loss": 0.2674,
      "num_input_tokens_seen": 16286872,
      "step": 24925
    },
    {
      "epoch": 14.69929245283019,
      "grad_norm": 7.634357452392578,
      "learning_rate": 1.991829012121145e-06,
      "loss": 0.4804,
      "num_input_tokens_seen": 16289816,
      "step": 24930
    },
    {
      "epoch": 14.702240566037736,
      "grad_norm": 1.817975640296936,
      "learning_rate": 1.989774399099552e-06,
      "loss": 0.2712,
      "num_input_tokens_seen": 16293176,
      "step": 24935
    },
    {
      "epoch": 14.705188679245284,
      "grad_norm": 3.8172225952148438,
      "learning_rate": 1.98772058304816e-06,
      "loss": 0.5809,
      "num_input_tokens_seen": 16296792,
      "step": 24940
    },
    {
      "epoch": 14.70813679245283,
      "grad_norm": 2.307760715484619,
      "learning_rate": 1.9856675645107244e-06,
      "loss": 0.2782,
      "num_input_tokens_seen": 16300120,
      "step": 24945
    },
    {
      "epoch": 14.711084905660378,
      "grad_norm": 3.0773813724517822,
      "learning_rate": 1.9836153440307936e-06,
      "loss": 0.4467,
      "num_input_tokens_seen": 16303352,
      "step": 24950
    },
    {
      "epoch": 14.714033018867925,
      "grad_norm": 3.479182481765747,
      "learning_rate": 1.9815639221517002e-06,
      "loss": 0.244,
      "num_input_tokens_seen": 16306552,
      "step": 24955
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 3.0518014430999756,
      "learning_rate": 1.9795132994165673e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 16309496,
      "step": 24960
    },
    {
      "epoch": 14.71992924528302,
      "grad_norm": 2.8982226848602295,
      "learning_rate": 1.977463476368306e-06,
      "loss": 0.3394,
      "num_input_tokens_seen": 16312216,
      "step": 24965
    },
    {
      "epoch": 14.722877358490566,
      "grad_norm": 3.301406145095825,
      "learning_rate": 1.975414453549614e-06,
      "loss": 0.2319,
      "num_input_tokens_seen": 16315320,
      "step": 24970
    },
    {
      "epoch": 14.725825471698114,
      "grad_norm": 4.2256975173950195,
      "learning_rate": 1.9733662315029826e-06,
      "loss": 0.2799,
      "num_input_tokens_seen": 16318040,
      "step": 24975
    },
    {
      "epoch": 14.72877358490566,
      "grad_norm": 5.638121128082275,
      "learning_rate": 1.9713188107706856e-06,
      "loss": 0.2969,
      "num_input_tokens_seen": 16320920,
      "step": 24980
    },
    {
      "epoch": 14.731721698113208,
      "grad_norm": 4.1225266456604,
      "learning_rate": 1.969272191894786e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 16324312,
      "step": 24985
    },
    {
      "epoch": 14.734669811320755,
      "grad_norm": 3.117048978805542,
      "learning_rate": 1.967226375417135e-06,
      "loss": 0.225,
      "num_input_tokens_seen": 16327640,
      "step": 24990
    },
    {
      "epoch": 14.737617924528301,
      "grad_norm": 5.894979476928711,
      "learning_rate": 1.965181361879372e-06,
      "loss": 0.3824,
      "num_input_tokens_seen": 16330968,
      "step": 24995
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 3.177720069885254,
      "learning_rate": 1.9631371518229214e-06,
      "loss": 0.3566,
      "num_input_tokens_seen": 16333272,
      "step": 25000
    },
    {
      "epoch": 14.743514150943396,
      "grad_norm": 3.3175885677337646,
      "learning_rate": 1.9610937457889975e-06,
      "loss": 0.3711,
      "num_input_tokens_seen": 16336280,
      "step": 25005
    },
    {
      "epoch": 14.746462264150944,
      "grad_norm": 3.0084574222564697,
      "learning_rate": 1.9590511443186032e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 16338936,
      "step": 25010
    },
    {
      "epoch": 14.74941037735849,
      "grad_norm": 3.507432222366333,
      "learning_rate": 1.9570093479525243e-06,
      "loss": 0.2435,
      "num_input_tokens_seen": 16341208,
      "step": 25015
    },
    {
      "epoch": 14.752358490566039,
      "grad_norm": 3.255910634994507,
      "learning_rate": 1.954968357231335e-06,
      "loss": 0.2928,
      "num_input_tokens_seen": 16344344,
      "step": 25020
    },
    {
      "epoch": 14.755306603773585,
      "grad_norm": 2.1222736835479736,
      "learning_rate": 1.9529281726953964e-06,
      "loss": 0.321,
      "num_input_tokens_seen": 16347960,
      "step": 25025
    },
    {
      "epoch": 14.758254716981131,
      "grad_norm": 2.5876290798187256,
      "learning_rate": 1.9508887948848564e-06,
      "loss": 0.3167,
      "num_input_tokens_seen": 16351224,
      "step": 25030
    },
    {
      "epoch": 14.76120283018868,
      "grad_norm": 4.098039627075195,
      "learning_rate": 1.9488502243396475e-06,
      "loss": 0.3952,
      "num_input_tokens_seen": 16353816,
      "step": 25035
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 3.489785671234131,
      "learning_rate": 1.946812461599492e-06,
      "loss": 0.3396,
      "num_input_tokens_seen": 16356792,
      "step": 25040
    },
    {
      "epoch": 14.767099056603774,
      "grad_norm": 2.625620126724243,
      "learning_rate": 1.944775507203897e-06,
      "loss": 0.358,
      "num_input_tokens_seen": 16359480,
      "step": 25045
    },
    {
      "epoch": 14.77004716981132,
      "grad_norm": 1.6890298128128052,
      "learning_rate": 1.942739361692153e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 16363128,
      "step": 25050
    },
    {
      "epoch": 14.772995283018869,
      "grad_norm": 2.322218894958496,
      "learning_rate": 1.94070402560334e-06,
      "loss": 0.2643,
      "num_input_tokens_seen": 16365880,
      "step": 25055
    },
    {
      "epoch": 14.775943396226415,
      "grad_norm": 3.18544602394104,
      "learning_rate": 1.93866949947632e-06,
      "loss": 0.2197,
      "num_input_tokens_seen": 16368952,
      "step": 25060
    },
    {
      "epoch": 14.778891509433961,
      "grad_norm": 3.0208044052124023,
      "learning_rate": 1.9366357838497423e-06,
      "loss": 0.3563,
      "num_input_tokens_seen": 16371800,
      "step": 25065
    },
    {
      "epoch": 14.78183962264151,
      "grad_norm": 3.219998598098755,
      "learning_rate": 1.9346028792620454e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 16374392,
      "step": 25070
    },
    {
      "epoch": 14.784787735849056,
      "grad_norm": 6.357417106628418,
      "learning_rate": 1.9325707862514464e-06,
      "loss": 0.3038,
      "num_input_tokens_seen": 16377592,
      "step": 25075
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 3.9820072650909424,
      "learning_rate": 1.930539505355952e-06,
      "loss": 0.4449,
      "num_input_tokens_seen": 16380856,
      "step": 25080
    },
    {
      "epoch": 14.79068396226415,
      "grad_norm": 5.146214962005615,
      "learning_rate": 1.9285090371133524e-06,
      "loss": 0.4505,
      "num_input_tokens_seen": 16383384,
      "step": 25085
    },
    {
      "epoch": 14.793632075471699,
      "grad_norm": 4.329617977142334,
      "learning_rate": 1.9264793820612228e-06,
      "loss": 0.25,
      "num_input_tokens_seen": 16386840,
      "step": 25090
    },
    {
      "epoch": 14.796580188679245,
      "grad_norm": 1.8048756122589111,
      "learning_rate": 1.924450540736921e-06,
      "loss": 0.3032,
      "num_input_tokens_seen": 16390456,
      "step": 25095
    },
    {
      "epoch": 14.799528301886792,
      "grad_norm": 2.065546751022339,
      "learning_rate": 1.922422513677593e-06,
      "loss": 0.2741,
      "num_input_tokens_seen": 16393848,
      "step": 25100
    },
    {
      "epoch": 14.80247641509434,
      "grad_norm": 4.9970173835754395,
      "learning_rate": 1.9203953014201703e-06,
      "loss": 0.2579,
      "num_input_tokens_seen": 16396664,
      "step": 25105
    },
    {
      "epoch": 14.805424528301886,
      "grad_norm": 5.100818634033203,
      "learning_rate": 1.918368904501364e-06,
      "loss": 0.3727,
      "num_input_tokens_seen": 16400152,
      "step": 25110
    },
    {
      "epoch": 14.808372641509434,
      "grad_norm": 2.3724799156188965,
      "learning_rate": 1.9163433234576713e-06,
      "loss": 0.2809,
      "num_input_tokens_seen": 16404216,
      "step": 25115
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 3.781308650970459,
      "learning_rate": 1.9143185588253733e-06,
      "loss": 0.3424,
      "num_input_tokens_seen": 16407544,
      "step": 25120
    },
    {
      "epoch": 14.814268867924529,
      "grad_norm": 5.014944076538086,
      "learning_rate": 1.9122946111405354e-06,
      "loss": 0.4849,
      "num_input_tokens_seen": 16409880,
      "step": 25125
    },
    {
      "epoch": 14.817216981132075,
      "grad_norm": 2.6074254512786865,
      "learning_rate": 1.910271480939005e-06,
      "loss": 0.2368,
      "num_input_tokens_seen": 16413304,
      "step": 25130
    },
    {
      "epoch": 14.820165094339622,
      "grad_norm": 2.72615909576416,
      "learning_rate": 1.9082491687564176e-06,
      "loss": 0.3254,
      "num_input_tokens_seen": 16415832,
      "step": 25135
    },
    {
      "epoch": 14.82311320754717,
      "grad_norm": 2.9088611602783203,
      "learning_rate": 1.9062276751281872e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 16418552,
      "step": 25140
    },
    {
      "epoch": 14.826061320754716,
      "grad_norm": 3.593059778213501,
      "learning_rate": 1.9042070005895136e-06,
      "loss": 0.4219,
      "num_input_tokens_seen": 16422168,
      "step": 25145
    },
    {
      "epoch": 14.829009433962264,
      "grad_norm": 2.8558707237243652,
      "learning_rate": 1.9021871456753788e-06,
      "loss": 0.2835,
      "num_input_tokens_seen": 16425272,
      "step": 25150
    },
    {
      "epoch": 14.83195754716981,
      "grad_norm": 16.020322799682617,
      "learning_rate": 1.9001681109205478e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 16428312,
      "step": 25155
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 4.500168800354004,
      "learning_rate": 1.898149896859567e-06,
      "loss": 0.2802,
      "num_input_tokens_seen": 16431256,
      "step": 25160
    },
    {
      "epoch": 14.837853773584905,
      "grad_norm": 6.404747486114502,
      "learning_rate": 1.8961325040267714e-06,
      "loss": 0.3261,
      "num_input_tokens_seen": 16433880,
      "step": 25165
    },
    {
      "epoch": 14.840801886792454,
      "grad_norm": 2.5434370040893555,
      "learning_rate": 1.894115932956272e-06,
      "loss": 0.5103,
      "num_input_tokens_seen": 16436888,
      "step": 25170
    },
    {
      "epoch": 14.84375,
      "grad_norm": 3.6563103199005127,
      "learning_rate": 1.8921001841819652e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 16439832,
      "step": 25175
    },
    {
      "epoch": 14.846698113207546,
      "grad_norm": 5.400711536407471,
      "learning_rate": 1.8900852582375284e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 16442968,
      "step": 25180
    },
    {
      "epoch": 14.849646226415095,
      "grad_norm": 3.6123716831207275,
      "learning_rate": 1.8880711556564214e-06,
      "loss": 0.3145,
      "num_input_tokens_seen": 16445272,
      "step": 25185
    },
    {
      "epoch": 14.852594339622641,
      "grad_norm": 4.1285600662231445,
      "learning_rate": 1.8860578769718891e-06,
      "loss": 0.3576,
      "num_input_tokens_seen": 16448920,
      "step": 25190
    },
    {
      "epoch": 14.85554245283019,
      "grad_norm": 4.622864723205566,
      "learning_rate": 1.8840454227169525e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 16451320,
      "step": 25195
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 2.3195698261260986,
      "learning_rate": 1.882033793424421e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 16454520,
      "step": 25200
    },
    {
      "epoch": 14.861438679245284,
      "grad_norm": 5.511350154876709,
      "learning_rate": 1.88002298962688e-06,
      "loss": 0.458,
      "num_input_tokens_seen": 16457400,
      "step": 25205
    },
    {
      "epoch": 14.86438679245283,
      "grad_norm": 7.232254981994629,
      "learning_rate": 1.8780130118566996e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 16460344,
      "step": 25210
    },
    {
      "epoch": 14.867334905660378,
      "grad_norm": 5.035064220428467,
      "learning_rate": 1.876003860646029e-06,
      "loss": 0.3924,
      "num_input_tokens_seen": 16463608,
      "step": 25215
    },
    {
      "epoch": 14.870283018867925,
      "grad_norm": 4.126266956329346,
      "learning_rate": 1.8739955365267997e-06,
      "loss": 0.4715,
      "num_input_tokens_seen": 16466168,
      "step": 25220
    },
    {
      "epoch": 14.873231132075471,
      "grad_norm": 2.907674789428711,
      "learning_rate": 1.8719880400307228e-06,
      "loss": 0.2524,
      "num_input_tokens_seen": 16469336,
      "step": 25225
    },
    {
      "epoch": 14.87617924528302,
      "grad_norm": 3.071489095687866,
      "learning_rate": 1.869981371689295e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 16472248,
      "step": 25230
    },
    {
      "epoch": 14.879127358490566,
      "grad_norm": 3.5328316688537598,
      "learning_rate": 1.867975532033789e-06,
      "loss": 0.2676,
      "num_input_tokens_seen": 16474968,
      "step": 25235
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 4.197782039642334,
      "learning_rate": 1.8659705215952589e-06,
      "loss": 0.3952,
      "num_input_tokens_seen": 16479032,
      "step": 25240
    },
    {
      "epoch": 14.88502358490566,
      "grad_norm": 1.345593810081482,
      "learning_rate": 1.8639663409045405e-06,
      "loss": 0.3343,
      "num_input_tokens_seen": 16481720,
      "step": 25245
    },
    {
      "epoch": 14.887971698113208,
      "grad_norm": 2.714120864868164,
      "learning_rate": 1.8619629904922466e-06,
      "loss": 0.3605,
      "num_input_tokens_seen": 16485016,
      "step": 25250
    },
    {
      "epoch": 14.890919811320755,
      "grad_norm": 3.1914124488830566,
      "learning_rate": 1.859960470888777e-06,
      "loss": 0.3305,
      "num_input_tokens_seen": 16488632,
      "step": 25255
    },
    {
      "epoch": 14.893867924528301,
      "grad_norm": 4.577303409576416,
      "learning_rate": 1.857958782624306e-06,
      "loss": 0.3753,
      "num_input_tokens_seen": 16491896,
      "step": 25260
    },
    {
      "epoch": 14.89681603773585,
      "grad_norm": 4.490603923797607,
      "learning_rate": 1.8559579262287886e-06,
      "loss": 0.5132,
      "num_input_tokens_seen": 16495544,
      "step": 25265
    },
    {
      "epoch": 14.899764150943396,
      "grad_norm": 4.201468467712402,
      "learning_rate": 1.8539579022319599e-06,
      "loss": 0.2453,
      "num_input_tokens_seen": 16498680,
      "step": 25270
    },
    {
      "epoch": 14.902712264150944,
      "grad_norm": 1.977437973022461,
      "learning_rate": 1.8519587111633357e-06,
      "loss": 0.2693,
      "num_input_tokens_seen": 16502232,
      "step": 25275
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 2.330874443054199,
      "learning_rate": 1.8499603535522082e-06,
      "loss": 0.4664,
      "num_input_tokens_seen": 16505304,
      "step": 25280
    },
    {
      "epoch": 14.908608490566039,
      "grad_norm": 2.7671756744384766,
      "learning_rate": 1.8479628299276543e-06,
      "loss": 0.2887,
      "num_input_tokens_seen": 16508472,
      "step": 25285
    },
    {
      "epoch": 14.911556603773585,
      "grad_norm": 5.394443035125732,
      "learning_rate": 1.8459661408185241e-06,
      "loss": 0.3402,
      "num_input_tokens_seen": 16511896,
      "step": 25290
    },
    {
      "epoch": 14.914504716981131,
      "grad_norm": 2.9852945804595947,
      "learning_rate": 1.8439702867534536e-06,
      "loss": 0.4143,
      "num_input_tokens_seen": 16514968,
      "step": 25295
    },
    {
      "epoch": 14.91745283018868,
      "grad_norm": 2.138394832611084,
      "learning_rate": 1.841975268260851e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 16518232,
      "step": 25300
    },
    {
      "epoch": 14.920400943396226,
      "grad_norm": 2.5422277450561523,
      "learning_rate": 1.8399810858689066e-06,
      "loss": 0.2857,
      "num_input_tokens_seen": 16521528,
      "step": 25305
    },
    {
      "epoch": 14.923349056603774,
      "grad_norm": 5.467014312744141,
      "learning_rate": 1.8379877401055884e-06,
      "loss": 0.4246,
      "num_input_tokens_seen": 16524536,
      "step": 25310
    },
    {
      "epoch": 14.92629716981132,
      "grad_norm": 3.2085540294647217,
      "learning_rate": 1.8359952314986418e-06,
      "loss": 0.2886,
      "num_input_tokens_seen": 16527256,
      "step": 25315
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 2.9103076457977295,
      "learning_rate": 1.8340035605755957e-06,
      "loss": 0.4148,
      "num_input_tokens_seen": 16530936,
      "step": 25320
    },
    {
      "epoch": 14.932193396226415,
      "grad_norm": 3.314279556274414,
      "learning_rate": 1.8320127278637518e-06,
      "loss": 0.4706,
      "num_input_tokens_seen": 16537720,
      "step": 25325
    },
    {
      "epoch": 14.935141509433961,
      "grad_norm": 3.0978221893310547,
      "learning_rate": 1.830022733890191e-06,
      "loss": 0.2949,
      "num_input_tokens_seen": 16543160,
      "step": 25330
    },
    {
      "epoch": 14.93808962264151,
      "grad_norm": 2.3217251300811768,
      "learning_rate": 1.8280335791817733e-06,
      "loss": 0.2791,
      "num_input_tokens_seen": 16547032,
      "step": 25335
    },
    {
      "epoch": 14.941037735849056,
      "grad_norm": 3.3429245948791504,
      "learning_rate": 1.826045264265136e-06,
      "loss": 0.3202,
      "num_input_tokens_seen": 16551064,
      "step": 25340
    },
    {
      "epoch": 14.943985849056604,
      "grad_norm": 2.995203733444214,
      "learning_rate": 1.8240577896666928e-06,
      "loss": 0.3212,
      "num_input_tokens_seen": 16554552,
      "step": 25345
    },
    {
      "epoch": 14.94693396226415,
      "grad_norm": 3.3454723358154297,
      "learning_rate": 1.8220711559126382e-06,
      "loss": 0.3166,
      "num_input_tokens_seen": 16557528,
      "step": 25350
    },
    {
      "epoch": 14.949882075471699,
      "grad_norm": 3.4970862865448,
      "learning_rate": 1.8200853635289417e-06,
      "loss": 0.4881,
      "num_input_tokens_seen": 16560184,
      "step": 25355
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 3.6791491508483887,
      "learning_rate": 1.81810041304135e-06,
      "loss": 0.2556,
      "num_input_tokens_seen": 16563640,
      "step": 25360
    },
    {
      "epoch": 14.955778301886792,
      "grad_norm": 4.774307727813721,
      "learning_rate": 1.8161163049753865e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 16566104,
      "step": 25365
    },
    {
      "epoch": 14.95872641509434,
      "grad_norm": 1.428311824798584,
      "learning_rate": 1.8141330398563533e-06,
      "loss": 0.2954,
      "num_input_tokens_seen": 16569816,
      "step": 25370
    },
    {
      "epoch": 14.961674528301886,
      "grad_norm": 3.948561668395996,
      "learning_rate": 1.8121506182093268e-06,
      "loss": 0.365,
      "num_input_tokens_seen": 16573496,
      "step": 25375
    },
    {
      "epoch": 14.964622641509434,
      "grad_norm": 3.278592824935913,
      "learning_rate": 1.8101690405591643e-06,
      "loss": 0.3781,
      "num_input_tokens_seen": 16576376,
      "step": 25380
    },
    {
      "epoch": 14.96757075471698,
      "grad_norm": 3.2101192474365234,
      "learning_rate": 1.8081883074304945e-06,
      "loss": 0.2778,
      "num_input_tokens_seen": 16579832,
      "step": 25385
    },
    {
      "epoch": 14.970518867924529,
      "grad_norm": 3.706987142562866,
      "learning_rate": 1.8062084193477275e-06,
      "loss": 0.3524,
      "num_input_tokens_seen": 16583224,
      "step": 25390
    },
    {
      "epoch": 14.973466981132075,
      "grad_norm": 4.486761569976807,
      "learning_rate": 1.804229376835046e-06,
      "loss": 0.403,
      "num_input_tokens_seen": 16586104,
      "step": 25395
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 3.181718349456787,
      "learning_rate": 1.8022511804164105e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 16589016,
      "step": 25400
    },
    {
      "epoch": 14.97936320754717,
      "grad_norm": 2.817408800125122,
      "learning_rate": 1.8002738306155559e-06,
      "loss": 0.3848,
      "num_input_tokens_seen": 16592664,
      "step": 25405
    },
    {
      "epoch": 14.982311320754716,
      "grad_norm": 4.256726264953613,
      "learning_rate": 1.7982973279559935e-06,
      "loss": 0.3677,
      "num_input_tokens_seen": 16595096,
      "step": 25410
    },
    {
      "epoch": 14.985259433962264,
      "grad_norm": 4.424173355102539,
      "learning_rate": 1.7963216729610134e-06,
      "loss": 0.3672,
      "num_input_tokens_seen": 16598136,
      "step": 25415
    },
    {
      "epoch": 14.98820754716981,
      "grad_norm": 3.626483201980591,
      "learning_rate": 1.7943468661536773e-06,
      "loss": 0.2722,
      "num_input_tokens_seen": 16600888,
      "step": 25420
    },
    {
      "epoch": 14.991155660377359,
      "grad_norm": 1.9373681545257568,
      "learning_rate": 1.7923729080568242e-06,
      "loss": 0.2622,
      "num_input_tokens_seen": 16604056,
      "step": 25425
    },
    {
      "epoch": 14.994103773584905,
      "grad_norm": 4.801909923553467,
      "learning_rate": 1.7903997991930683e-06,
      "loss": 0.2848,
      "num_input_tokens_seen": 16608152,
      "step": 25430
    },
    {
      "epoch": 14.997051886792454,
      "grad_norm": 3.8285844326019287,
      "learning_rate": 1.7884275400847972e-06,
      "loss": 0.2604,
      "num_input_tokens_seen": 16610968,
      "step": 25435
    },
    {
      "epoch": 15.0,
      "grad_norm": 5.81403923034668,
      "learning_rate": 1.786456131254175e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 16614080,
      "step": 25440
    },
    {
      "epoch": 15.002948113207546,
      "grad_norm": 2.455164670944214,
      "learning_rate": 1.784485573223143e-06,
      "loss": 0.2394,
      "num_input_tokens_seen": 16617504,
      "step": 25445
    },
    {
      "epoch": 15.005896226415095,
      "grad_norm": 2.611452102661133,
      "learning_rate": 1.782515866513414e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 16620896,
      "step": 25450
    },
    {
      "epoch": 15.008844339622641,
      "grad_norm": 2.939984083175659,
      "learning_rate": 1.7805470116464758e-06,
      "loss": 0.2258,
      "num_input_tokens_seen": 16624256,
      "step": 25455
    },
    {
      "epoch": 15.01179245283019,
      "grad_norm": 4.573910236358643,
      "learning_rate": 1.7785790091435911e-06,
      "loss": 0.2677,
      "num_input_tokens_seen": 16627424,
      "step": 25460
    },
    {
      "epoch": 15.014740566037736,
      "grad_norm": 5.3294878005981445,
      "learning_rate": 1.776611859525796e-06,
      "loss": 0.3223,
      "num_input_tokens_seen": 16630912,
      "step": 25465
    },
    {
      "epoch": 15.017688679245284,
      "grad_norm": 2.611436128616333,
      "learning_rate": 1.7746455633139042e-06,
      "loss": 0.2271,
      "num_input_tokens_seen": 16633856,
      "step": 25470
    },
    {
      "epoch": 15.02063679245283,
      "grad_norm": 6.233902454376221,
      "learning_rate": 1.7726801210285005e-06,
      "loss": 0.3639,
      "num_input_tokens_seen": 16636608,
      "step": 25475
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 3.929884672164917,
      "learning_rate": 1.7707155331899418e-06,
      "loss": 0.3474,
      "num_input_tokens_seen": 16640160,
      "step": 25480
    },
    {
      "epoch": 15.026533018867925,
      "grad_norm": 3.6325838565826416,
      "learning_rate": 1.7687518003183645e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 16644384,
      "step": 25485
    },
    {
      "epoch": 15.029481132075471,
      "grad_norm": 3.584102153778076,
      "learning_rate": 1.766788922933675e-06,
      "loss": 0.2386,
      "num_input_tokens_seen": 16647488,
      "step": 25490
    },
    {
      "epoch": 15.03242924528302,
      "grad_norm": 3.053321361541748,
      "learning_rate": 1.7648269015555514e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 16650752,
      "step": 25495
    },
    {
      "epoch": 15.035377358490566,
      "grad_norm": 5.695224285125732,
      "learning_rate": 1.7628657367034474e-06,
      "loss": 0.4689,
      "num_input_tokens_seen": 16654048,
      "step": 25500
    },
    {
      "epoch": 15.038325471698114,
      "grad_norm": 5.521780014038086,
      "learning_rate": 1.7609054288965922e-06,
      "loss": 0.4417,
      "num_input_tokens_seen": 16656544,
      "step": 25505
    },
    {
      "epoch": 15.04127358490566,
      "grad_norm": 3.248173236846924,
      "learning_rate": 1.7589459786539847e-06,
      "loss": 0.3187,
      "num_input_tokens_seen": 16659776,
      "step": 25510
    },
    {
      "epoch": 15.044221698113208,
      "grad_norm": 4.402312278747559,
      "learning_rate": 1.7569873864943975e-06,
      "loss": 0.2868,
      "num_input_tokens_seen": 16662528,
      "step": 25515
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 5.557966709136963,
      "learning_rate": 1.7550296529363764e-06,
      "loss": 0.3394,
      "num_input_tokens_seen": 16665792,
      "step": 25520
    },
    {
      "epoch": 15.050117924528301,
      "grad_norm": 5.431990146636963,
      "learning_rate": 1.7530727784982393e-06,
      "loss": 0.3138,
      "num_input_tokens_seen": 16668544,
      "step": 25525
    },
    {
      "epoch": 15.05306603773585,
      "grad_norm": 3.5140373706817627,
      "learning_rate": 1.7511167636980765e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 16671296,
      "step": 25530
    },
    {
      "epoch": 15.056014150943396,
      "grad_norm": 3.083238124847412,
      "learning_rate": 1.7491616090537539e-06,
      "loss": 0.3287,
      "num_input_tokens_seen": 16674816,
      "step": 25535
    },
    {
      "epoch": 15.058962264150944,
      "grad_norm": 2.002474546432495,
      "learning_rate": 1.7472073150829056e-06,
      "loss": 0.2607,
      "num_input_tokens_seen": 16679104,
      "step": 25540
    },
    {
      "epoch": 15.06191037735849,
      "grad_norm": 2.7674267292022705,
      "learning_rate": 1.745253882302939e-06,
      "loss": 0.2836,
      "num_input_tokens_seen": 16682112,
      "step": 25545
    },
    {
      "epoch": 15.064858490566039,
      "grad_norm": 3.3668465614318848,
      "learning_rate": 1.743301311231035e-06,
      "loss": 0.2796,
      "num_input_tokens_seen": 16685312,
      "step": 25550
    },
    {
      "epoch": 15.067806603773585,
      "grad_norm": 4.063653945922852,
      "learning_rate": 1.7413496023841437e-06,
      "loss": 0.2946,
      "num_input_tokens_seen": 16687808,
      "step": 25555
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 3.1032655239105225,
      "learning_rate": 1.7393987562789876e-06,
      "loss": 0.362,
      "num_input_tokens_seen": 16690784,
      "step": 25560
    },
    {
      "epoch": 15.07370283018868,
      "grad_norm": 4.290898323059082,
      "learning_rate": 1.7374487734320655e-06,
      "loss": 0.3678,
      "num_input_tokens_seen": 16693600,
      "step": 25565
    },
    {
      "epoch": 15.076650943396226,
      "grad_norm": 4.4035162925720215,
      "learning_rate": 1.7354996543596408e-06,
      "loss": 0.3129,
      "num_input_tokens_seen": 16696768,
      "step": 25570
    },
    {
      "epoch": 15.079599056603774,
      "grad_norm": 3.3471879959106445,
      "learning_rate": 1.7335513995777504e-06,
      "loss": 0.3452,
      "num_input_tokens_seen": 16700800,
      "step": 25575
    },
    {
      "epoch": 15.08254716981132,
      "grad_norm": 2.9022021293640137,
      "learning_rate": 1.7316040096022062e-06,
      "loss": 0.451,
      "num_input_tokens_seen": 16705312,
      "step": 25580
    },
    {
      "epoch": 15.085495283018869,
      "grad_norm": 4.85904598236084,
      "learning_rate": 1.7296574849485863e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 16708064,
      "step": 25585
    },
    {
      "epoch": 15.088443396226415,
      "grad_norm": 3.711435556411743,
      "learning_rate": 1.7277118261322423e-06,
      "loss": 0.4217,
      "num_input_tokens_seen": 16711744,
      "step": 25590
    },
    {
      "epoch": 15.091391509433961,
      "grad_norm": 7.741339683532715,
      "learning_rate": 1.7257670336682925e-06,
      "loss": 0.3321,
      "num_input_tokens_seen": 16714464,
      "step": 25595
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 3.9206113815307617,
      "learning_rate": 1.7238231080716339e-06,
      "loss": 0.2412,
      "num_input_tokens_seen": 16717408,
      "step": 25600
    },
    {
      "epoch": 15.097287735849056,
      "grad_norm": 4.184138298034668,
      "learning_rate": 1.721880049856927e-06,
      "loss": 0.4318,
      "num_input_tokens_seen": 16720544,
      "step": 25605
    },
    {
      "epoch": 15.100235849056604,
      "grad_norm": 6.575077056884766,
      "learning_rate": 1.7199378595386046e-06,
      "loss": 0.3471,
      "num_input_tokens_seen": 16724000,
      "step": 25610
    },
    {
      "epoch": 15.10318396226415,
      "grad_norm": 4.336565971374512,
      "learning_rate": 1.7179965376308705e-06,
      "loss": 0.4026,
      "num_input_tokens_seen": 16728160,
      "step": 25615
    },
    {
      "epoch": 15.106132075471699,
      "grad_norm": 3.5227935314178467,
      "learning_rate": 1.7160560846476976e-06,
      "loss": 0.2568,
      "num_input_tokens_seen": 16731680,
      "step": 25620
    },
    {
      "epoch": 15.109080188679245,
      "grad_norm": 2.8550455570220947,
      "learning_rate": 1.7141165011028277e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 16734720,
      "step": 25625
    },
    {
      "epoch": 15.112028301886792,
      "grad_norm": 3.6576011180877686,
      "learning_rate": 1.7121777875097767e-06,
      "loss": 0.4049,
      "num_input_tokens_seen": 16737536,
      "step": 25630
    },
    {
      "epoch": 15.11497641509434,
      "grad_norm": 4.947840213775635,
      "learning_rate": 1.7102399443818268e-06,
      "loss": 0.3769,
      "num_input_tokens_seen": 16740704,
      "step": 25635
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 2.731698989868164,
      "learning_rate": 1.7083029722320294e-06,
      "loss": 0.2996,
      "num_input_tokens_seen": 16743936,
      "step": 25640
    },
    {
      "epoch": 15.120872641509434,
      "grad_norm": 4.019099235534668,
      "learning_rate": 1.7063668715732063e-06,
      "loss": 0.3387,
      "num_input_tokens_seen": 16746880,
      "step": 25645
    },
    {
      "epoch": 15.12382075471698,
      "grad_norm": 3.943763494491577,
      "learning_rate": 1.7044316429179492e-06,
      "loss": 0.2612,
      "num_input_tokens_seen": 16750112,
      "step": 25650
    },
    {
      "epoch": 15.126768867924529,
      "grad_norm": 2.760517120361328,
      "learning_rate": 1.7024972867786155e-06,
      "loss": 0.2893,
      "num_input_tokens_seen": 16753536,
      "step": 25655
    },
    {
      "epoch": 15.129716981132075,
      "grad_norm": 4.414573669433594,
      "learning_rate": 1.7005638036673389e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 16757024,
      "step": 25660
    },
    {
      "epoch": 15.132665094339623,
      "grad_norm": 3.0016605854034424,
      "learning_rate": 1.6986311940960148e-06,
      "loss": 0.3112,
      "num_input_tokens_seen": 16760992,
      "step": 25665
    },
    {
      "epoch": 15.13561320754717,
      "grad_norm": 3.5589380264282227,
      "learning_rate": 1.696699458576308e-06,
      "loss": 0.3347,
      "num_input_tokens_seen": 16763552,
      "step": 25670
    },
    {
      "epoch": 15.138561320754716,
      "grad_norm": 2.1354055404663086,
      "learning_rate": 1.6947685976196581e-06,
      "loss": 0.371,
      "num_input_tokens_seen": 16768256,
      "step": 25675
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 3.967299461364746,
      "learning_rate": 1.692838611737267e-06,
      "loss": 0.3522,
      "num_input_tokens_seen": 16771712,
      "step": 25680
    },
    {
      "epoch": 15.14445754716981,
      "grad_norm": 3.7380030155181885,
      "learning_rate": 1.690909501440106e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 16774336,
      "step": 25685
    },
    {
      "epoch": 15.147405660377359,
      "grad_norm": 3.0924575328826904,
      "learning_rate": 1.688981267238915e-06,
      "loss": 0.3701,
      "num_input_tokens_seen": 16777728,
      "step": 25690
    },
    {
      "epoch": 15.150353773584905,
      "grad_norm": 2.96093487739563,
      "learning_rate": 1.687053909644204e-06,
      "loss": 0.3415,
      "num_input_tokens_seen": 16780672,
      "step": 25695
    },
    {
      "epoch": 15.153301886792454,
      "grad_norm": 4.575915336608887,
      "learning_rate": 1.685127429166249e-06,
      "loss": 0.3229,
      "num_input_tokens_seen": 16783456,
      "step": 25700
    },
    {
      "epoch": 15.15625,
      "grad_norm": 4.7501983642578125,
      "learning_rate": 1.683201826315093e-06,
      "loss": 0.2404,
      "num_input_tokens_seen": 16787104,
      "step": 25705
    },
    {
      "epoch": 15.159198113207546,
      "grad_norm": 4.462807655334473,
      "learning_rate": 1.681277101600548e-06,
      "loss": 0.4042,
      "num_input_tokens_seen": 16790272,
      "step": 25710
    },
    {
      "epoch": 15.162146226415095,
      "grad_norm": 3.2985916137695312,
      "learning_rate": 1.6793532555321939e-06,
      "loss": 0.2727,
      "num_input_tokens_seen": 16793344,
      "step": 25715
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 4.255829334259033,
      "learning_rate": 1.6774302886193744e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 16796224,
      "step": 25720
    },
    {
      "epoch": 15.16804245283019,
      "grad_norm": 5.561375617980957,
      "learning_rate": 1.6755082013712076e-06,
      "loss": 0.1607,
      "num_input_tokens_seen": 16798912,
      "step": 25725
    },
    {
      "epoch": 15.170990566037736,
      "grad_norm": 18.601160049438477,
      "learning_rate": 1.6735869942965716e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 16801792,
      "step": 25730
    },
    {
      "epoch": 15.173938679245284,
      "grad_norm": 4.540792465209961,
      "learning_rate": 1.6716666679041155e-06,
      "loss": 0.3053,
      "num_input_tokens_seen": 16804448,
      "step": 25735
    },
    {
      "epoch": 15.17688679245283,
      "grad_norm": 2.5266470909118652,
      "learning_rate": 1.6697472227022533e-06,
      "loss": 0.3052,
      "num_input_tokens_seen": 16808576,
      "step": 25740
    },
    {
      "epoch": 15.179834905660377,
      "grad_norm": 4.744549751281738,
      "learning_rate": 1.6678286591991644e-06,
      "loss": 0.2524,
      "num_input_tokens_seen": 16811168,
      "step": 25745
    },
    {
      "epoch": 15.182783018867925,
      "grad_norm": 3.2243540287017822,
      "learning_rate": 1.665910977902801e-06,
      "loss": 0.3215,
      "num_input_tokens_seen": 16814304,
      "step": 25750
    },
    {
      "epoch": 15.185731132075471,
      "grad_norm": 3.3473057746887207,
      "learning_rate": 1.6639941793208747e-06,
      "loss": 0.2686,
      "num_input_tokens_seen": 16817984,
      "step": 25755
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 4.1207990646362305,
      "learning_rate": 1.6620782639608674e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 16821216,
      "step": 25760
    },
    {
      "epoch": 15.191627358490566,
      "grad_norm": 3.845247507095337,
      "learning_rate": 1.6601632323300231e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 16823936,
      "step": 25765
    },
    {
      "epoch": 15.194575471698114,
      "grad_norm": 4.704751491546631,
      "learning_rate": 1.6582490849353595e-06,
      "loss": 0.3214,
      "num_input_tokens_seen": 16826816,
      "step": 25770
    },
    {
      "epoch": 15.19752358490566,
      "grad_norm": 5.3994832038879395,
      "learning_rate": 1.6563358222836523e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 16830432,
      "step": 25775
    },
    {
      "epoch": 15.200471698113208,
      "grad_norm": 7.824821949005127,
      "learning_rate": 1.654423444881445e-06,
      "loss": 0.4466,
      "num_input_tokens_seen": 16833472,
      "step": 25780
    },
    {
      "epoch": 15.203419811320755,
      "grad_norm": 5.109165668487549,
      "learning_rate": 1.652511953235051e-06,
      "loss": 0.2383,
      "num_input_tokens_seen": 16836832,
      "step": 25785
    },
    {
      "epoch": 15.206367924528301,
      "grad_norm": 3.0378525257110596,
      "learning_rate": 1.650601347850544e-06,
      "loss": 0.293,
      "num_input_tokens_seen": 16839296,
      "step": 25790
    },
    {
      "epoch": 15.20931603773585,
      "grad_norm": 3.6430389881134033,
      "learning_rate": 1.6486916292337652e-06,
      "loss": 0.3499,
      "num_input_tokens_seen": 16842368,
      "step": 25795
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 4.2585601806640625,
      "learning_rate": 1.6467827978903212e-06,
      "loss": 0.367,
      "num_input_tokens_seen": 16848000,
      "step": 25800
    },
    {
      "epoch": 15.215212264150944,
      "grad_norm": 3.883925199508667,
      "learning_rate": 1.6448748543255827e-06,
      "loss": 0.4007,
      "num_input_tokens_seen": 16850784,
      "step": 25805
    },
    {
      "epoch": 15.21816037735849,
      "grad_norm": 3.79347562789917,
      "learning_rate": 1.6429677990446845e-06,
      "loss": 0.3585,
      "num_input_tokens_seen": 16854656,
      "step": 25810
    },
    {
      "epoch": 15.221108490566039,
      "grad_norm": 3.4180898666381836,
      "learning_rate": 1.6410616325525319e-06,
      "loss": 0.2151,
      "num_input_tokens_seen": 16857248,
      "step": 25815
    },
    {
      "epoch": 15.224056603773585,
      "grad_norm": 3.0276033878326416,
      "learning_rate": 1.6391563553537875e-06,
      "loss": 0.3931,
      "num_input_tokens_seen": 16860768,
      "step": 25820
    },
    {
      "epoch": 15.227004716981131,
      "grad_norm": 9.823476791381836,
      "learning_rate": 1.6372519679528832e-06,
      "loss": 0.3842,
      "num_input_tokens_seen": 16863104,
      "step": 25825
    },
    {
      "epoch": 15.22995283018868,
      "grad_norm": 2.7237777709960938,
      "learning_rate": 1.6353484708540124e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 16866624,
      "step": 25830
    },
    {
      "epoch": 15.232900943396226,
      "grad_norm": 1.3380271196365356,
      "learning_rate": 1.633445864561135e-06,
      "loss": 0.2767,
      "num_input_tokens_seen": 16870304,
      "step": 25835
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 39.41473388671875,
      "learning_rate": 1.6315441495779726e-06,
      "loss": 0.358,
      "num_input_tokens_seen": 16874560,
      "step": 25840
    },
    {
      "epoch": 15.23879716981132,
      "grad_norm": 5.549872875213623,
      "learning_rate": 1.6296433264080152e-06,
      "loss": 0.2847,
      "num_input_tokens_seen": 16878240,
      "step": 25845
    },
    {
      "epoch": 15.241745283018869,
      "grad_norm": 5.1191534996032715,
      "learning_rate": 1.627743395554513e-06,
      "loss": 0.4081,
      "num_input_tokens_seen": 16880320,
      "step": 25850
    },
    {
      "epoch": 15.244693396226415,
      "grad_norm": 2.6979405879974365,
      "learning_rate": 1.6258443575204802e-06,
      "loss": 0.2704,
      "num_input_tokens_seen": 16886176,
      "step": 25855
    },
    {
      "epoch": 15.247641509433961,
      "grad_norm": 4.546374320983887,
      "learning_rate": 1.6239462128086936e-06,
      "loss": 0.3669,
      "num_input_tokens_seen": 16889728,
      "step": 25860
    },
    {
      "epoch": 15.25058962264151,
      "grad_norm": 4.035529136657715,
      "learning_rate": 1.6220489619216988e-06,
      "loss": 0.3025,
      "num_input_tokens_seen": 16893472,
      "step": 25865
    },
    {
      "epoch": 15.253537735849056,
      "grad_norm": 1.8950949907302856,
      "learning_rate": 1.6201526053618e-06,
      "loss": 0.2693,
      "num_input_tokens_seen": 16896224,
      "step": 25870
    },
    {
      "epoch": 15.256485849056604,
      "grad_norm": 2.19242787361145,
      "learning_rate": 1.6182571436310634e-06,
      "loss": 0.2779,
      "num_input_tokens_seen": 16898720,
      "step": 25875
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 2.786064386367798,
      "learning_rate": 1.616362577231324e-06,
      "loss": 0.2864,
      "num_input_tokens_seen": 16902656,
      "step": 25880
    },
    {
      "epoch": 15.262382075471699,
      "grad_norm": 3.9407155513763428,
      "learning_rate": 1.614468906664175e-06,
      "loss": 0.283,
      "num_input_tokens_seen": 16905568,
      "step": 25885
    },
    {
      "epoch": 15.265330188679245,
      "grad_norm": 3.314628839492798,
      "learning_rate": 1.612576132430974e-06,
      "loss": 0.2918,
      "num_input_tokens_seen": 16908096,
      "step": 25890
    },
    {
      "epoch": 15.268278301886792,
      "grad_norm": 2.907628059387207,
      "learning_rate": 1.6106842550328406e-06,
      "loss": 0.3618,
      "num_input_tokens_seen": 16910976,
      "step": 25895
    },
    {
      "epoch": 15.27122641509434,
      "grad_norm": 3.437227249145508,
      "learning_rate": 1.6087932749706582e-06,
      "loss": 0.2758,
      "num_input_tokens_seen": 16913824,
      "step": 25900
    },
    {
      "epoch": 15.274174528301886,
      "grad_norm": 2.293837308883667,
      "learning_rate": 1.6069031927450696e-06,
      "loss": 0.3681,
      "num_input_tokens_seen": 16917056,
      "step": 25905
    },
    {
      "epoch": 15.277122641509434,
      "grad_norm": 4.190698623657227,
      "learning_rate": 1.605014008856486e-06,
      "loss": 0.2938,
      "num_input_tokens_seen": 16920160,
      "step": 25910
    },
    {
      "epoch": 15.28007075471698,
      "grad_norm": 3.9480674266815186,
      "learning_rate": 1.6031257238050745e-06,
      "loss": 0.239,
      "num_input_tokens_seen": 16923520,
      "step": 25915
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 2.7868776321411133,
      "learning_rate": 1.601238338090768e-06,
      "loss": 0.1862,
      "num_input_tokens_seen": 16926240,
      "step": 25920
    },
    {
      "epoch": 15.285966981132075,
      "grad_norm": 3.596871852874756,
      "learning_rate": 1.5993518522132595e-06,
      "loss": 0.2974,
      "num_input_tokens_seen": 16929728,
      "step": 25925
    },
    {
      "epoch": 15.288915094339623,
      "grad_norm": 3.8529109954833984,
      "learning_rate": 1.5974662666720037e-06,
      "loss": 0.3413,
      "num_input_tokens_seen": 16932448,
      "step": 25930
    },
    {
      "epoch": 15.29186320754717,
      "grad_norm": 4.3924665451049805,
      "learning_rate": 1.5955815819662162e-06,
      "loss": 0.2537,
      "num_input_tokens_seen": 16935168,
      "step": 25935
    },
    {
      "epoch": 15.294811320754716,
      "grad_norm": 3.1609251499176025,
      "learning_rate": 1.5936977985948788e-06,
      "loss": 0.3142,
      "num_input_tokens_seen": 16938592,
      "step": 25940
    },
    {
      "epoch": 15.297759433962264,
      "grad_norm": 3.526771306991577,
      "learning_rate": 1.5918149170567298e-06,
      "loss": 0.2709,
      "num_input_tokens_seen": 16942432,
      "step": 25945
    },
    {
      "epoch": 15.30070754716981,
      "grad_norm": 3.078281879425049,
      "learning_rate": 1.5899329378502698e-06,
      "loss": 0.3401,
      "num_input_tokens_seen": 16946240,
      "step": 25950
    },
    {
      "epoch": 15.303655660377359,
      "grad_norm": 4.365112781524658,
      "learning_rate": 1.588051861473761e-06,
      "loss": 0.2711,
      "num_input_tokens_seen": 16948864,
      "step": 25955
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 2.5447943210601807,
      "learning_rate": 1.5861716884252253e-06,
      "loss": 0.3009,
      "num_input_tokens_seen": 16953088,
      "step": 25960
    },
    {
      "epoch": 15.309551886792454,
      "grad_norm": 3.1284024715423584,
      "learning_rate": 1.5842924192024489e-06,
      "loss": 0.2781,
      "num_input_tokens_seen": 16956352,
      "step": 25965
    },
    {
      "epoch": 15.3125,
      "grad_norm": 5.692673683166504,
      "learning_rate": 1.5824140543029742e-06,
      "loss": 0.257,
      "num_input_tokens_seen": 16959616,
      "step": 25970
    },
    {
      "epoch": 15.315448113207546,
      "grad_norm": 2.506136178970337,
      "learning_rate": 1.5805365942241092e-06,
      "loss": 0.3622,
      "num_input_tokens_seen": 16962976,
      "step": 25975
    },
    {
      "epoch": 15.318396226415095,
      "grad_norm": 2.4727838039398193,
      "learning_rate": 1.5786600394629181e-06,
      "loss": 0.2787,
      "num_input_tokens_seen": 16967584,
      "step": 25980
    },
    {
      "epoch": 15.321344339622641,
      "grad_norm": 3.214503526687622,
      "learning_rate": 1.5767843905162261e-06,
      "loss": 0.3572,
      "num_input_tokens_seen": 16971200,
      "step": 25985
    },
    {
      "epoch": 15.32429245283019,
      "grad_norm": 4.878054141998291,
      "learning_rate": 1.5749096478806209e-06,
      "loss": 0.2985,
      "num_input_tokens_seen": 16974304,
      "step": 25990
    },
    {
      "epoch": 15.327240566037736,
      "grad_norm": 5.114634990692139,
      "learning_rate": 1.5730358120524452e-06,
      "loss": 0.2786,
      "num_input_tokens_seen": 16977056,
      "step": 25995
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 5.128930568695068,
      "learning_rate": 1.5711628835278098e-06,
      "loss": 0.3319,
      "num_input_tokens_seen": 16979840,
      "step": 26000
    },
    {
      "epoch": 15.33313679245283,
      "grad_norm": 3.2422728538513184,
      "learning_rate": 1.5692908628025782e-06,
      "loss": 0.2904,
      "num_input_tokens_seen": 16984064,
      "step": 26005
    },
    {
      "epoch": 15.336084905660377,
      "grad_norm": 3.1283209323883057,
      "learning_rate": 1.5674197503723765e-06,
      "loss": 0.3422,
      "num_input_tokens_seen": 16987296,
      "step": 26010
    },
    {
      "epoch": 15.339033018867925,
      "grad_norm": 5.691107749938965,
      "learning_rate": 1.5655495467325893e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 16989984,
      "step": 26015
    },
    {
      "epoch": 15.341981132075471,
      "grad_norm": 5.568491458892822,
      "learning_rate": 1.5636802523783613e-06,
      "loss": 0.3881,
      "num_input_tokens_seen": 16993280,
      "step": 26020
    },
    {
      "epoch": 15.34492924528302,
      "grad_norm": 3.108703374862671,
      "learning_rate": 1.5618118678045947e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 16996640,
      "step": 26025
    },
    {
      "epoch": 15.347877358490566,
      "grad_norm": 4.451146125793457,
      "learning_rate": 1.5599443935059549e-06,
      "loss": 0.2941,
      "num_input_tokens_seen": 17001056,
      "step": 26030
    },
    {
      "epoch": 15.350825471698114,
      "grad_norm": 2.723281145095825,
      "learning_rate": 1.5580778299768635e-06,
      "loss": 0.479,
      "num_input_tokens_seen": 17004800,
      "step": 26035
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 1.9855400323867798,
      "learning_rate": 1.5562121777114997e-06,
      "loss": 0.2202,
      "num_input_tokens_seen": 17008928,
      "step": 26040
    },
    {
      "epoch": 15.356721698113208,
      "grad_norm": 4.118953704833984,
      "learning_rate": 1.5543474372038043e-06,
      "loss": 0.3257,
      "num_input_tokens_seen": 17012352,
      "step": 26045
    },
    {
      "epoch": 15.359669811320755,
      "grad_norm": 1.6851309537887573,
      "learning_rate": 1.5524836089474748e-06,
      "loss": 0.3673,
      "num_input_tokens_seen": 17015520,
      "step": 26050
    },
    {
      "epoch": 15.362617924528301,
      "grad_norm": 2.0552279949188232,
      "learning_rate": 1.5506206934359664e-06,
      "loss": 0.2251,
      "num_input_tokens_seen": 17017952,
      "step": 26055
    },
    {
      "epoch": 15.36556603773585,
      "grad_norm": 3.3847506046295166,
      "learning_rate": 1.5487586911624947e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 17020992,
      "step": 26060
    },
    {
      "epoch": 15.368514150943396,
      "grad_norm": 5.199976444244385,
      "learning_rate": 1.5468976026200355e-06,
      "loss": 0.349,
      "num_input_tokens_seen": 17023840,
      "step": 26065
    },
    {
      "epoch": 15.371462264150944,
      "grad_norm": 2.453526735305786,
      "learning_rate": 1.5450374283013187e-06,
      "loss": 0.2486,
      "num_input_tokens_seen": 17027392,
      "step": 26070
    },
    {
      "epoch": 15.37441037735849,
      "grad_norm": 2.633237838745117,
      "learning_rate": 1.5431781686988317e-06,
      "loss": 0.4104,
      "num_input_tokens_seen": 17031840,
      "step": 26075
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 5.038642883300781,
      "learning_rate": 1.5413198243048233e-06,
      "loss": 0.3505,
      "num_input_tokens_seen": 17035232,
      "step": 26080
    },
    {
      "epoch": 15.380306603773585,
      "grad_norm": 5.075397968292236,
      "learning_rate": 1.5394623956112974e-06,
      "loss": 0.3336,
      "num_input_tokens_seen": 17038208,
      "step": 26085
    },
    {
      "epoch": 15.383254716981131,
      "grad_norm": 4.842166423797607,
      "learning_rate": 1.537605883110015e-06,
      "loss": 0.4314,
      "num_input_tokens_seen": 17041440,
      "step": 26090
    },
    {
      "epoch": 15.38620283018868,
      "grad_norm": 3.3399548530578613,
      "learning_rate": 1.5357502872924984e-06,
      "loss": 0.2336,
      "num_input_tokens_seen": 17044896,
      "step": 26095
    },
    {
      "epoch": 15.389150943396226,
      "grad_norm": 2.8671393394470215,
      "learning_rate": 1.5338956086500235e-06,
      "loss": 0.2991,
      "num_input_tokens_seen": 17047840,
      "step": 26100
    },
    {
      "epoch": 15.392099056603774,
      "grad_norm": 4.162959098815918,
      "learning_rate": 1.5320418476736237e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 17050784,
      "step": 26105
    },
    {
      "epoch": 15.39504716981132,
      "grad_norm": 3.62001371383667,
      "learning_rate": 1.5301890048540912e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 17054112,
      "step": 26110
    },
    {
      "epoch": 15.397995283018869,
      "grad_norm": 2.5554206371307373,
      "learning_rate": 1.5283370806819743e-06,
      "loss": 0.3459,
      "num_input_tokens_seen": 17056928,
      "step": 26115
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 4.328121662139893,
      "learning_rate": 1.5264860756475752e-06,
      "loss": 0.2974,
      "num_input_tokens_seen": 17059904,
      "step": 26120
    },
    {
      "epoch": 15.403891509433961,
      "grad_norm": 3.3417398929595947,
      "learning_rate": 1.5246359902409592e-06,
      "loss": 0.3992,
      "num_input_tokens_seen": 17063168,
      "step": 26125
    },
    {
      "epoch": 15.40683962264151,
      "grad_norm": 2.0323381423950195,
      "learning_rate": 1.5227868249519423e-06,
      "loss": 0.4392,
      "num_input_tokens_seen": 17066816,
      "step": 26130
    },
    {
      "epoch": 15.409787735849056,
      "grad_norm": 4.358390808105469,
      "learning_rate": 1.5209385802700999e-06,
      "loss": 0.4724,
      "num_input_tokens_seen": 17069600,
      "step": 26135
    },
    {
      "epoch": 15.412735849056604,
      "grad_norm": 3.6345348358154297,
      "learning_rate": 1.5190912566847626e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 17072992,
      "step": 26140
    },
    {
      "epoch": 15.41568396226415,
      "grad_norm": 8.188377380371094,
      "learning_rate": 1.5172448546850166e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 17076224,
      "step": 26145
    },
    {
      "epoch": 15.418632075471699,
      "grad_norm": 3.8304197788238525,
      "learning_rate": 1.515399374759704e-06,
      "loss": 0.2583,
      "num_input_tokens_seen": 17079392,
      "step": 26150
    },
    {
      "epoch": 15.421580188679245,
      "grad_norm": 2.05532169342041,
      "learning_rate": 1.513554817397424e-06,
      "loss": 0.2845,
      "num_input_tokens_seen": 17082976,
      "step": 26155
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 4.998197555541992,
      "learning_rate": 1.5117111830865338e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 17085440,
      "step": 26160
    },
    {
      "epoch": 15.42747641509434,
      "grad_norm": 4.153213977813721,
      "learning_rate": 1.509868472315142e-06,
      "loss": 0.294,
      "num_input_tokens_seen": 17088320,
      "step": 26165
    },
    {
      "epoch": 15.430424528301886,
      "grad_norm": 1.984317660331726,
      "learning_rate": 1.508026685571113e-06,
      "loss": 0.3319,
      "num_input_tokens_seen": 17092000,
      "step": 26170
    },
    {
      "epoch": 15.433372641509434,
      "grad_norm": 3.111135244369507,
      "learning_rate": 1.506185823342069e-06,
      "loss": 0.2952,
      "num_input_tokens_seen": 17096576,
      "step": 26175
    },
    {
      "epoch": 15.43632075471698,
      "grad_norm": 3.127795696258545,
      "learning_rate": 1.504345886115386e-06,
      "loss": 0.2807,
      "num_input_tokens_seen": 17099552,
      "step": 26180
    },
    {
      "epoch": 15.439268867924529,
      "grad_norm": 3.9387967586517334,
      "learning_rate": 1.502506874378193e-06,
      "loss": 0.363,
      "num_input_tokens_seen": 17101888,
      "step": 26185
    },
    {
      "epoch": 15.442216981132075,
      "grad_norm": 4.162726402282715,
      "learning_rate": 1.5006687886173805e-06,
      "loss": 0.2674,
      "num_input_tokens_seen": 17105152,
      "step": 26190
    },
    {
      "epoch": 15.445165094339623,
      "grad_norm": 3.2764716148376465,
      "learning_rate": 1.498831629319587e-06,
      "loss": 0.3473,
      "num_input_tokens_seen": 17108288,
      "step": 26195
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 2.5265631675720215,
      "learning_rate": 1.4969953969712087e-06,
      "loss": 0.4001,
      "num_input_tokens_seen": 17112032,
      "step": 26200
    },
    {
      "epoch": 15.451061320754716,
      "grad_norm": 4.605922698974609,
      "learning_rate": 1.4951600920583963e-06,
      "loss": 0.3067,
      "num_input_tokens_seen": 17114880,
      "step": 26205
    },
    {
      "epoch": 15.454009433962264,
      "grad_norm": 4.970421314239502,
      "learning_rate": 1.493325715067055e-06,
      "loss": 0.3155,
      "num_input_tokens_seen": 17118016,
      "step": 26210
    },
    {
      "epoch": 15.45695754716981,
      "grad_norm": 3.54055118560791,
      "learning_rate": 1.4914922664828417e-06,
      "loss": 0.3284,
      "num_input_tokens_seen": 17119968,
      "step": 26215
    },
    {
      "epoch": 15.459905660377359,
      "grad_norm": 3.9058570861816406,
      "learning_rate": 1.4896597467911732e-06,
      "loss": 0.392,
      "num_input_tokens_seen": 17122528,
      "step": 26220
    },
    {
      "epoch": 15.462853773584905,
      "grad_norm": 4.184182643890381,
      "learning_rate": 1.4878281564772156e-06,
      "loss": 0.3274,
      "num_input_tokens_seen": 17124896,
      "step": 26225
    },
    {
      "epoch": 15.465801886792454,
      "grad_norm": 3.543256998062134,
      "learning_rate": 1.4859974960258898e-06,
      "loss": 0.3126,
      "num_input_tokens_seen": 17128672,
      "step": 26230
    },
    {
      "epoch": 15.46875,
      "grad_norm": 4.931568622589111,
      "learning_rate": 1.4841677659218723e-06,
      "loss": 0.3725,
      "num_input_tokens_seen": 17131296,
      "step": 26235
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 16.31775665283203,
      "learning_rate": 1.4823389666495886e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 17134048,
      "step": 26240
    },
    {
      "epoch": 15.474646226415095,
      "grad_norm": 2.2070531845092773,
      "learning_rate": 1.4805110986932258e-06,
      "loss": 0.3689,
      "num_input_tokens_seen": 17137024,
      "step": 26245
    },
    {
      "epoch": 15.477594339622641,
      "grad_norm": 8.439335823059082,
      "learning_rate": 1.4786841625367166e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 17140320,
      "step": 26250
    },
    {
      "epoch": 15.48054245283019,
      "grad_norm": 1.933902621269226,
      "learning_rate": 1.476858158663752e-06,
      "loss": 0.3034,
      "num_input_tokens_seen": 17144448,
      "step": 26255
    },
    {
      "epoch": 15.483490566037736,
      "grad_norm": 4.177387237548828,
      "learning_rate": 1.4750330875577745e-06,
      "loss": 0.3587,
      "num_input_tokens_seen": 17147488,
      "step": 26260
    },
    {
      "epoch": 15.486438679245284,
      "grad_norm": 7.512240409851074,
      "learning_rate": 1.4732089497019787e-06,
      "loss": 0.5319,
      "num_input_tokens_seen": 17150528,
      "step": 26265
    },
    {
      "epoch": 15.48938679245283,
      "grad_norm": 3.9785561561584473,
      "learning_rate": 1.471385745579313e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 17154464,
      "step": 26270
    },
    {
      "epoch": 15.492334905660377,
      "grad_norm": 4.60462760925293,
      "learning_rate": 1.4695634756724775e-06,
      "loss": 0.2183,
      "num_input_tokens_seen": 17157088,
      "step": 26275
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 4.349890232086182,
      "learning_rate": 1.4677421404639281e-06,
      "loss": 0.2915,
      "num_input_tokens_seen": 17160192,
      "step": 26280
    },
    {
      "epoch": 15.498231132075471,
      "grad_norm": 3.298415422439575,
      "learning_rate": 1.4659217404358706e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 17163520,
      "step": 26285
    },
    {
      "epoch": 15.50117924528302,
      "grad_norm": 4.086945533752441,
      "learning_rate": 1.4641022760702627e-06,
      "loss": 0.5074,
      "num_input_tokens_seen": 17166464,
      "step": 26290
    },
    {
      "epoch": 15.504127358490566,
      "grad_norm": 4.109692096710205,
      "learning_rate": 1.4622837478488172e-06,
      "loss": 0.2623,
      "num_input_tokens_seen": 17171296,
      "step": 26295
    },
    {
      "epoch": 15.507075471698114,
      "grad_norm": 4.189747333526611,
      "learning_rate": 1.4604661562529953e-06,
      "loss": 0.527,
      "num_input_tokens_seen": 17174816,
      "step": 26300
    },
    {
      "epoch": 15.51002358490566,
      "grad_norm": 2.3047966957092285,
      "learning_rate": 1.4586495017640119e-06,
      "loss": 0.304,
      "num_input_tokens_seen": 17178016,
      "step": 26305
    },
    {
      "epoch": 15.512971698113208,
      "grad_norm": 3.1910059452056885,
      "learning_rate": 1.4568337848628366e-06,
      "loss": 0.3088,
      "num_input_tokens_seen": 17181344,
      "step": 26310
    },
    {
      "epoch": 15.515919811320755,
      "grad_norm": 2.345215082168579,
      "learning_rate": 1.4550190060301872e-06,
      "loss": 0.3227,
      "num_input_tokens_seen": 17184704,
      "step": 26315
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 2.8850290775299072,
      "learning_rate": 1.4532051657465335e-06,
      "loss": 0.294,
      "num_input_tokens_seen": 17189120,
      "step": 26320
    },
    {
      "epoch": 15.52181603773585,
      "grad_norm": 1.9276872873306274,
      "learning_rate": 1.4513922644920985e-06,
      "loss": 0.1858,
      "num_input_tokens_seen": 17191744,
      "step": 26325
    },
    {
      "epoch": 15.524764150943396,
      "grad_norm": 3.623589515686035,
      "learning_rate": 1.4495803027468552e-06,
      "loss": 0.2873,
      "num_input_tokens_seen": 17194848,
      "step": 26330
    },
    {
      "epoch": 15.527712264150944,
      "grad_norm": 3.873244285583496,
      "learning_rate": 1.4477692809905263e-06,
      "loss": 0.3458,
      "num_input_tokens_seen": 17198208,
      "step": 26335
    },
    {
      "epoch": 15.53066037735849,
      "grad_norm": 3.0659663677215576,
      "learning_rate": 1.4459591997025896e-06,
      "loss": 0.3314,
      "num_input_tokens_seen": 17201344,
      "step": 26340
    },
    {
      "epoch": 15.533608490566039,
      "grad_norm": 3.82720685005188,
      "learning_rate": 1.4441500593622737e-06,
      "loss": 0.2766,
      "num_input_tokens_seen": 17205632,
      "step": 26345
    },
    {
      "epoch": 15.536556603773585,
      "grad_norm": 3.4633045196533203,
      "learning_rate": 1.4423418604485539e-06,
      "loss": 0.2916,
      "num_input_tokens_seen": 17209024,
      "step": 26350
    },
    {
      "epoch": 15.539504716981131,
      "grad_norm": 2.3283748626708984,
      "learning_rate": 1.4405346034401597e-06,
      "loss": 0.2231,
      "num_input_tokens_seen": 17212320,
      "step": 26355
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 3.123856544494629,
      "learning_rate": 1.4387282888155695e-06,
      "loss": 0.3248,
      "num_input_tokens_seen": 17215712,
      "step": 26360
    },
    {
      "epoch": 15.545400943396226,
      "grad_norm": 3.2822468280792236,
      "learning_rate": 1.436922917053013e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 17218752,
      "step": 26365
    },
    {
      "epoch": 15.548349056603774,
      "grad_norm": 2.399505615234375,
      "learning_rate": 1.4351184886304686e-06,
      "loss": 0.2974,
      "num_input_tokens_seen": 17221696,
      "step": 26370
    },
    {
      "epoch": 15.55129716981132,
      "grad_norm": 3.4341933727264404,
      "learning_rate": 1.4333150040256699e-06,
      "loss": 0.3311,
      "num_input_tokens_seen": 17225216,
      "step": 26375
    },
    {
      "epoch": 15.554245283018869,
      "grad_norm": 9.106170654296875,
      "learning_rate": 1.4315124637160954e-06,
      "loss": 0.3031,
      "num_input_tokens_seen": 17227552,
      "step": 26380
    },
    {
      "epoch": 15.557193396226415,
      "grad_norm": 2.6528518199920654,
      "learning_rate": 1.4297108681789752e-06,
      "loss": 0.3669,
      "num_input_tokens_seen": 17230528,
      "step": 26385
    },
    {
      "epoch": 15.560141509433961,
      "grad_norm": 10.670291900634766,
      "learning_rate": 1.4279102178912902e-06,
      "loss": 0.3161,
      "num_input_tokens_seen": 17235648,
      "step": 26390
    },
    {
      "epoch": 15.56308962264151,
      "grad_norm": 3.9989030361175537,
      "learning_rate": 1.4261105133297693e-06,
      "loss": 0.336,
      "num_input_tokens_seen": 17239296,
      "step": 26395
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 3.024123430252075,
      "learning_rate": 1.4243117549708913e-06,
      "loss": 0.2422,
      "num_input_tokens_seen": 17241888,
      "step": 26400
    },
    {
      "epoch": 15.568985849056604,
      "grad_norm": 2.663079023361206,
      "learning_rate": 1.422513943290888e-06,
      "loss": 0.3532,
      "num_input_tokens_seen": 17245248,
      "step": 26405
    },
    {
      "epoch": 15.57193396226415,
      "grad_norm": 4.483865261077881,
      "learning_rate": 1.4207170787657365e-06,
      "loss": 0.3158,
      "num_input_tokens_seen": 17248736,
      "step": 26410
    },
    {
      "epoch": 15.574882075471699,
      "grad_norm": 12.577589988708496,
      "learning_rate": 1.4189211618711646e-06,
      "loss": 0.4473,
      "num_input_tokens_seen": 17252512,
      "step": 26415
    },
    {
      "epoch": 15.577830188679245,
      "grad_norm": 4.962217807769775,
      "learning_rate": 1.417126193082648e-06,
      "loss": 0.2799,
      "num_input_tokens_seen": 17255360,
      "step": 26420
    },
    {
      "epoch": 15.580778301886792,
      "grad_norm": 3.008734941482544,
      "learning_rate": 1.4153321728754133e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 17258592,
      "step": 26425
    },
    {
      "epoch": 15.58372641509434,
      "grad_norm": 2.434762954711914,
      "learning_rate": 1.4135391017244338e-06,
      "loss": 0.3216,
      "num_input_tokens_seen": 17261600,
      "step": 26430
    },
    {
      "epoch": 15.586674528301886,
      "grad_norm": 3.188650131225586,
      "learning_rate": 1.4117469801044332e-06,
      "loss": 0.2482,
      "num_input_tokens_seen": 17264224,
      "step": 26435
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 5.509434223175049,
      "learning_rate": 1.4099558084898862e-06,
      "loss": 0.3033,
      "num_input_tokens_seen": 17266944,
      "step": 26440
    },
    {
      "epoch": 15.59257075471698,
      "grad_norm": 3.4705007076263428,
      "learning_rate": 1.408165587355011e-06,
      "loss": 0.3261,
      "num_input_tokens_seen": 17271296,
      "step": 26445
    },
    {
      "epoch": 15.595518867924529,
      "grad_norm": 4.0602192878723145,
      "learning_rate": 1.4063763171737766e-06,
      "loss": 0.3923,
      "num_input_tokens_seen": 17273984,
      "step": 26450
    },
    {
      "epoch": 15.598466981132075,
      "grad_norm": 2.460423469543457,
      "learning_rate": 1.4045879984198996e-06,
      "loss": 0.3474,
      "num_input_tokens_seen": 17277408,
      "step": 26455
    },
    {
      "epoch": 15.601415094339622,
      "grad_norm": 3.1442718505859375,
      "learning_rate": 1.4028006315668457e-06,
      "loss": 0.2788,
      "num_input_tokens_seen": 17279776,
      "step": 26460
    },
    {
      "epoch": 15.60436320754717,
      "grad_norm": 3.3464889526367188,
      "learning_rate": 1.4010142170878261e-06,
      "loss": 0.3223,
      "num_input_tokens_seen": 17282720,
      "step": 26465
    },
    {
      "epoch": 15.607311320754716,
      "grad_norm": 4.998191833496094,
      "learning_rate": 1.3992287554558042e-06,
      "loss": 0.2896,
      "num_input_tokens_seen": 17286208,
      "step": 26470
    },
    {
      "epoch": 15.610259433962264,
      "grad_norm": 4.4005961418151855,
      "learning_rate": 1.3974442471434885e-06,
      "loss": 0.2891,
      "num_input_tokens_seen": 17289952,
      "step": 26475
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 4.082117557525635,
      "learning_rate": 1.395660692623334e-06,
      "loss": 0.5583,
      "num_input_tokens_seen": 17293408,
      "step": 26480
    },
    {
      "epoch": 15.616155660377359,
      "grad_norm": 3.4739232063293457,
      "learning_rate": 1.3938780923675454e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 17296384,
      "step": 26485
    },
    {
      "epoch": 15.619103773584905,
      "grad_norm": 3.6855900287628174,
      "learning_rate": 1.3920964468480718e-06,
      "loss": 0.5415,
      "num_input_tokens_seen": 17299136,
      "step": 26490
    },
    {
      "epoch": 15.622051886792454,
      "grad_norm": 2.8964996337890625,
      "learning_rate": 1.3903157565366143e-06,
      "loss": 0.2704,
      "num_input_tokens_seen": 17301184,
      "step": 26495
    },
    {
      "epoch": 15.625,
      "grad_norm": 3.637993574142456,
      "learning_rate": 1.3885360219046172e-06,
      "loss": 0.373,
      "num_input_tokens_seen": 17304224,
      "step": 26500
    },
    {
      "epoch": 15.627948113207546,
      "grad_norm": 4.168931007385254,
      "learning_rate": 1.386757243423273e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 17307776,
      "step": 26505
    },
    {
      "epoch": 15.630896226415095,
      "grad_norm": 2.1661219596862793,
      "learning_rate": 1.384979421563521e-06,
      "loss": 0.2558,
      "num_input_tokens_seen": 17311040,
      "step": 26510
    },
    {
      "epoch": 15.633844339622641,
      "grad_norm": 2.7799863815307617,
      "learning_rate": 1.3832025567960465e-06,
      "loss": 0.2623,
      "num_input_tokens_seen": 17314432,
      "step": 26515
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 4.134042263031006,
      "learning_rate": 1.3814266495912815e-06,
      "loss": 0.3782,
      "num_input_tokens_seen": 17318144,
      "step": 26520
    },
    {
      "epoch": 15.639740566037736,
      "grad_norm": 3.117633819580078,
      "learning_rate": 1.3796517004194078e-06,
      "loss": 0.3149,
      "num_input_tokens_seen": 17321568,
      "step": 26525
    },
    {
      "epoch": 15.642688679245284,
      "grad_norm": 3.2849156856536865,
      "learning_rate": 1.3778777097503476e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 17324672,
      "step": 26530
    },
    {
      "epoch": 15.64563679245283,
      "grad_norm": 3.5496699810028076,
      "learning_rate": 1.3761046780537757e-06,
      "loss": 0.2793,
      "num_input_tokens_seen": 17326944,
      "step": 26535
    },
    {
      "epoch": 15.648584905660378,
      "grad_norm": 3.6858997344970703,
      "learning_rate": 1.3743326057991086e-06,
      "loss": 0.2997,
      "num_input_tokens_seen": 17329216,
      "step": 26540
    },
    {
      "epoch": 15.651533018867925,
      "grad_norm": 4.843862056732178,
      "learning_rate": 1.3725614934555093e-06,
      "loss": 0.2749,
      "num_input_tokens_seen": 17332480,
      "step": 26545
    },
    {
      "epoch": 15.654481132075471,
      "grad_norm": 4.023950576782227,
      "learning_rate": 1.3707913414918882e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 17335712,
      "step": 26550
    },
    {
      "epoch": 15.65742924528302,
      "grad_norm": 2.7278900146484375,
      "learning_rate": 1.3690221503768996e-06,
      "loss": 0.3506,
      "num_input_tokens_seen": 17338688,
      "step": 26555
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 7.402222156524658,
      "learning_rate": 1.3672539205789465e-06,
      "loss": 0.4993,
      "num_input_tokens_seen": 17345344,
      "step": 26560
    },
    {
      "epoch": 15.663325471698114,
      "grad_norm": 2.9237332344055176,
      "learning_rate": 1.3654866525661737e-06,
      "loss": 0.2725,
      "num_input_tokens_seen": 17348096,
      "step": 26565
    },
    {
      "epoch": 15.66627358490566,
      "grad_norm": 3.727510690689087,
      "learning_rate": 1.3637203468064741e-06,
      "loss": 0.3236,
      "num_input_tokens_seen": 17351264,
      "step": 26570
    },
    {
      "epoch": 15.669221698113208,
      "grad_norm": 4.803117275238037,
      "learning_rate": 1.3619550037674838e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 17353888,
      "step": 26575
    },
    {
      "epoch": 15.672169811320755,
      "grad_norm": 4.27092170715332,
      "learning_rate": 1.3601906239165857e-06,
      "loss": 0.2807,
      "num_input_tokens_seen": 17356704,
      "step": 26580
    },
    {
      "epoch": 15.675117924528301,
      "grad_norm": 2.0001773834228516,
      "learning_rate": 1.3584272077209048e-06,
      "loss": 0.2819,
      "num_input_tokens_seen": 17360288,
      "step": 26585
    },
    {
      "epoch": 15.67806603773585,
      "grad_norm": 4.00749397277832,
      "learning_rate": 1.3566647556473168e-06,
      "loss": 0.4019,
      "num_input_tokens_seen": 17367584,
      "step": 26590
    },
    {
      "epoch": 15.681014150943396,
      "grad_norm": 2.8294129371643066,
      "learning_rate": 1.3549032681624363e-06,
      "loss": 0.2208,
      "num_input_tokens_seen": 17370624,
      "step": 26595
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 2.555790901184082,
      "learning_rate": 1.3531427457326252e-06,
      "loss": 0.2707,
      "num_input_tokens_seen": 17373376,
      "step": 26600
    },
    {
      "epoch": 15.68691037735849,
      "grad_norm": 4.443591117858887,
      "learning_rate": 1.3513831888239893e-06,
      "loss": 0.3029,
      "num_input_tokens_seen": 17376576,
      "step": 26605
    },
    {
      "epoch": 15.689858490566039,
      "grad_norm": 2.647549629211426,
      "learning_rate": 1.3496245979023786e-06,
      "loss": 0.2269,
      "num_input_tokens_seen": 17379584,
      "step": 26610
    },
    {
      "epoch": 15.692806603773585,
      "grad_norm": 3.0803987979888916,
      "learning_rate": 1.3478669734333865e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 17382080,
      "step": 26615
    },
    {
      "epoch": 15.695754716981131,
      "grad_norm": 2.685371160507202,
      "learning_rate": 1.3461103158823546e-06,
      "loss": 0.3055,
      "num_input_tokens_seen": 17384768,
      "step": 26620
    },
    {
      "epoch": 15.69870283018868,
      "grad_norm": 4.202804088592529,
      "learning_rate": 1.3443546257143624e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 17388096,
      "step": 26625
    },
    {
      "epoch": 15.701650943396226,
      "grad_norm": 4.394802570343018,
      "learning_rate": 1.3425999033942395e-06,
      "loss": 0.2678,
      "num_input_tokens_seen": 17391168,
      "step": 26630
    },
    {
      "epoch": 15.704599056603774,
      "grad_norm": 6.537935256958008,
      "learning_rate": 1.3408461493865549e-06,
      "loss": 0.3464,
      "num_input_tokens_seen": 17394080,
      "step": 26635
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 2.0991427898406982,
      "learning_rate": 1.339093364155622e-06,
      "loss": 0.2813,
      "num_input_tokens_seen": 17397728,
      "step": 26640
    },
    {
      "epoch": 15.710495283018869,
      "grad_norm": 3.0356225967407227,
      "learning_rate": 1.3373415481654988e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 17402240,
      "step": 26645
    },
    {
      "epoch": 15.713443396226415,
      "grad_norm": 4.275698661804199,
      "learning_rate": 1.335590701879984e-06,
      "loss": 0.435,
      "num_input_tokens_seen": 17405440,
      "step": 26650
    },
    {
      "epoch": 15.716391509433961,
      "grad_norm": 1.7521620988845825,
      "learning_rate": 1.3338408257626257e-06,
      "loss": 0.3485,
      "num_input_tokens_seen": 17409632,
      "step": 26655
    },
    {
      "epoch": 15.71933962264151,
      "grad_norm": 3.0775625705718994,
      "learning_rate": 1.3320919202767086e-06,
      "loss": 0.3329,
      "num_input_tokens_seen": 17412576,
      "step": 26660
    },
    {
      "epoch": 15.722287735849056,
      "grad_norm": 2.5457687377929688,
      "learning_rate": 1.3303439858852636e-06,
      "loss": 0.3273,
      "num_input_tokens_seen": 17415840,
      "step": 26665
    },
    {
      "epoch": 15.725235849056604,
      "grad_norm": 5.2274065017700195,
      "learning_rate": 1.3285970230510636e-06,
      "loss": 0.3485,
      "num_input_tokens_seen": 17420512,
      "step": 26670
    },
    {
      "epoch": 15.72818396226415,
      "grad_norm": 3.0741167068481445,
      "learning_rate": 1.3268510322366246e-06,
      "loss": 0.2402,
      "num_input_tokens_seen": 17423808,
      "step": 26675
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 6.199498653411865,
      "learning_rate": 1.3251060139042038e-06,
      "loss": 0.4863,
      "num_input_tokens_seen": 17427200,
      "step": 26680
    },
    {
      "epoch": 15.734080188679245,
      "grad_norm": 6.591602325439453,
      "learning_rate": 1.3233619685158056e-06,
      "loss": 0.2838,
      "num_input_tokens_seen": 17429600,
      "step": 26685
    },
    {
      "epoch": 15.737028301886792,
      "grad_norm": 2.0515358448028564,
      "learning_rate": 1.3216188965331712e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 17433184,
      "step": 26690
    },
    {
      "epoch": 15.73997641509434,
      "grad_norm": 4.608201026916504,
      "learning_rate": 1.3198767984177869e-06,
      "loss": 0.3399,
      "num_input_tokens_seen": 17436000,
      "step": 26695
    },
    {
      "epoch": 15.742924528301886,
      "grad_norm": 3.2978322505950928,
      "learning_rate": 1.3181356746308805e-06,
      "loss": 0.239,
      "num_input_tokens_seen": 17438752,
      "step": 26700
    },
    {
      "epoch": 15.745872641509434,
      "grad_norm": 2.7258460521698,
      "learning_rate": 1.3163955256334226e-06,
      "loss": 0.2717,
      "num_input_tokens_seen": 17441568,
      "step": 26705
    },
    {
      "epoch": 15.74882075471698,
      "grad_norm": 2.704075813293457,
      "learning_rate": 1.3146563518861227e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 17444512,
      "step": 26710
    },
    {
      "epoch": 15.751768867924529,
      "grad_norm": 6.92684268951416,
      "learning_rate": 1.3129181538494384e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 17449760,
      "step": 26715
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 3.7280688285827637,
      "learning_rate": 1.3111809319835622e-06,
      "loss": 0.486,
      "num_input_tokens_seen": 17453216,
      "step": 26720
    },
    {
      "epoch": 15.757665094339622,
      "grad_norm": 2.39424467086792,
      "learning_rate": 1.3094446867484335e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 17459520,
      "step": 26725
    },
    {
      "epoch": 15.76061320754717,
      "grad_norm": 3.24554705619812,
      "learning_rate": 1.3077094186037287e-06,
      "loss": 0.2787,
      "num_input_tokens_seen": 17463136,
      "step": 26730
    },
    {
      "epoch": 15.763561320754716,
      "grad_norm": 4.399353504180908,
      "learning_rate": 1.305975128008869e-06,
      "loss": 0.271,
      "num_input_tokens_seen": 17466112,
      "step": 26735
    },
    {
      "epoch": 15.766509433962264,
      "grad_norm": 3.9377071857452393,
      "learning_rate": 1.304241815423014e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 17470656,
      "step": 26740
    },
    {
      "epoch": 15.76945754716981,
      "grad_norm": 3.468637466430664,
      "learning_rate": 1.3025094813050655e-06,
      "loss": 0.332,
      "num_input_tokens_seen": 17473920,
      "step": 26745
    },
    {
      "epoch": 15.772405660377359,
      "grad_norm": 2.6382548809051514,
      "learning_rate": 1.3007781261136675e-06,
      "loss": 0.3299,
      "num_input_tokens_seen": 17477248,
      "step": 26750
    },
    {
      "epoch": 15.775353773584905,
      "grad_norm": 4.2992329597473145,
      "learning_rate": 1.299047750307204e-06,
      "loss": 0.3695,
      "num_input_tokens_seen": 17480032,
      "step": 26755
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 4.300413608551025,
      "learning_rate": 1.297318354343799e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 17483744,
      "step": 26760
    },
    {
      "epoch": 15.78125,
      "grad_norm": 4.389379978179932,
      "learning_rate": 1.295589938681317e-06,
      "loss": 0.3587,
      "num_input_tokens_seen": 17486912,
      "step": 26765
    },
    {
      "epoch": 15.784198113207546,
      "grad_norm": 4.923178195953369,
      "learning_rate": 1.2938625037773628e-06,
      "loss": 0.5925,
      "num_input_tokens_seen": 17489888,
      "step": 26770
    },
    {
      "epoch": 15.787146226415095,
      "grad_norm": 3.431182861328125,
      "learning_rate": 1.2921360500892843e-06,
      "loss": 0.3065,
      "num_input_tokens_seen": 17492736,
      "step": 26775
    },
    {
      "epoch": 15.790094339622641,
      "grad_norm": 3.4801125526428223,
      "learning_rate": 1.290410578074167e-06,
      "loss": 0.2743,
      "num_input_tokens_seen": 17494976,
      "step": 26780
    },
    {
      "epoch": 15.79304245283019,
      "grad_norm": 3.603374719619751,
      "learning_rate": 1.2886860881888362e-06,
      "loss": 0.3338,
      "num_input_tokens_seen": 17497728,
      "step": 26785
    },
    {
      "epoch": 15.795990566037736,
      "grad_norm": 5.771636009216309,
      "learning_rate": 1.2869625808898584e-06,
      "loss": 0.438,
      "num_input_tokens_seen": 17502432,
      "step": 26790
    },
    {
      "epoch": 15.798938679245284,
      "grad_norm": 3.9014172554016113,
      "learning_rate": 1.2852400566335398e-06,
      "loss": 0.3183,
      "num_input_tokens_seen": 17506208,
      "step": 26795
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 3.4103238582611084,
      "learning_rate": 1.2835185158759244e-06,
      "loss": 0.2646,
      "num_input_tokens_seen": 17508832,
      "step": 26800
    },
    {
      "epoch": 15.804834905660378,
      "grad_norm": 4.553548812866211,
      "learning_rate": 1.2817979590728009e-06,
      "loss": 0.3773,
      "num_input_tokens_seen": 17511584,
      "step": 26805
    },
    {
      "epoch": 15.807783018867925,
      "grad_norm": 2.966296434402466,
      "learning_rate": 1.2800783866796918e-06,
      "loss": 0.3249,
      "num_input_tokens_seen": 17514432,
      "step": 26810
    },
    {
      "epoch": 15.810731132075471,
      "grad_norm": 2.565370559692383,
      "learning_rate": 1.2783597991518604e-06,
      "loss": 0.3249,
      "num_input_tokens_seen": 17517376,
      "step": 26815
    },
    {
      "epoch": 15.81367924528302,
      "grad_norm": 3.1115801334381104,
      "learning_rate": 1.2766421969443131e-06,
      "loss": 0.3597,
      "num_input_tokens_seen": 17520256,
      "step": 26820
    },
    {
      "epoch": 15.816627358490566,
      "grad_norm": 3.2397983074188232,
      "learning_rate": 1.274925580511791e-06,
      "loss": 0.2333,
      "num_input_tokens_seen": 17523488,
      "step": 26825
    },
    {
      "epoch": 15.819575471698114,
      "grad_norm": 3.4266226291656494,
      "learning_rate": 1.2732099503087757e-06,
      "loss": 0.4474,
      "num_input_tokens_seen": 17526336,
      "step": 26830
    },
    {
      "epoch": 15.82252358490566,
      "grad_norm": 3.163853645324707,
      "learning_rate": 1.2714953067894859e-06,
      "loss": 0.286,
      "num_input_tokens_seen": 17528640,
      "step": 26835
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 6.365755081176758,
      "learning_rate": 1.2697816504078847e-06,
      "loss": 0.2249,
      "num_input_tokens_seen": 17530944,
      "step": 26840
    },
    {
      "epoch": 15.828419811320755,
      "grad_norm": 3.531479597091675,
      "learning_rate": 1.2680689816176672e-06,
      "loss": 0.4747,
      "num_input_tokens_seen": 17533952,
      "step": 26845
    },
    {
      "epoch": 15.831367924528301,
      "grad_norm": 3.8300206661224365,
      "learning_rate": 1.2663573008722707e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 17536608,
      "step": 26850
    },
    {
      "epoch": 15.83431603773585,
      "grad_norm": 2.6782639026641846,
      "learning_rate": 1.2646466086248698e-06,
      "loss": 0.2602,
      "num_input_tokens_seen": 17540384,
      "step": 26855
    },
    {
      "epoch": 15.837264150943396,
      "grad_norm": 4.051733016967773,
      "learning_rate": 1.2629369053283779e-06,
      "loss": 0.3651,
      "num_input_tokens_seen": 17543200,
      "step": 26860
    },
    {
      "epoch": 15.840212264150944,
      "grad_norm": 2.873082160949707,
      "learning_rate": 1.2612281914354452e-06,
      "loss": 0.2908,
      "num_input_tokens_seen": 17545856,
      "step": 26865
    },
    {
      "epoch": 15.84316037735849,
      "grad_norm": 3.0941123962402344,
      "learning_rate": 1.259520467398463e-06,
      "loss": 0.3333,
      "num_input_tokens_seen": 17549216,
      "step": 26870
    },
    {
      "epoch": 15.846108490566039,
      "grad_norm": 3.780475378036499,
      "learning_rate": 1.2578137336695573e-06,
      "loss": 0.3272,
      "num_input_tokens_seen": 17552640,
      "step": 26875
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 3.247061014175415,
      "learning_rate": 1.256107990700594e-06,
      "loss": 0.3749,
      "num_input_tokens_seen": 17555712,
      "step": 26880
    },
    {
      "epoch": 15.852004716981131,
      "grad_norm": 3.396000623703003,
      "learning_rate": 1.2544032389431753e-06,
      "loss": 0.3918,
      "num_input_tokens_seen": 17558880,
      "step": 26885
    },
    {
      "epoch": 15.85495283018868,
      "grad_norm": 2.539729118347168,
      "learning_rate": 1.2526994788486418e-06,
      "loss": 0.3486,
      "num_input_tokens_seen": 17561664,
      "step": 26890
    },
    {
      "epoch": 15.857900943396226,
      "grad_norm": 10.517245292663574,
      "learning_rate": 1.2509967108680697e-06,
      "loss": 0.3422,
      "num_input_tokens_seen": 17564576,
      "step": 26895
    },
    {
      "epoch": 15.860849056603774,
      "grad_norm": 6.802643299102783,
      "learning_rate": 1.249294935452277e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 17567008,
      "step": 26900
    },
    {
      "epoch": 15.86379716981132,
      "grad_norm": 2.091876983642578,
      "learning_rate": 1.247594153051815e-06,
      "loss": 0.3161,
      "num_input_tokens_seen": 17570912,
      "step": 26905
    },
    {
      "epoch": 15.866745283018869,
      "grad_norm": 3.483086109161377,
      "learning_rate": 1.2458943641169718e-06,
      "loss": 0.3991,
      "num_input_tokens_seen": 17573952,
      "step": 26910
    },
    {
      "epoch": 15.869693396226415,
      "grad_norm": 3.4162611961364746,
      "learning_rate": 1.2441955690977758e-06,
      "loss": 0.2386,
      "num_input_tokens_seen": 17577408,
      "step": 26915
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 1.688330888748169,
      "learning_rate": 1.2424977684439898e-06,
      "loss": 0.3493,
      "num_input_tokens_seen": 17581472,
      "step": 26920
    },
    {
      "epoch": 15.87558962264151,
      "grad_norm": 2.3596646785736084,
      "learning_rate": 1.2408009626051137e-06,
      "loss": 0.3085,
      "num_input_tokens_seen": 17584896,
      "step": 26925
    },
    {
      "epoch": 15.878537735849056,
      "grad_norm": 3.253265857696533,
      "learning_rate": 1.2391051520303826e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 17587488,
      "step": 26930
    },
    {
      "epoch": 15.881485849056604,
      "grad_norm": 3.5980474948883057,
      "learning_rate": 1.2374103371687723e-06,
      "loss": 0.3656,
      "num_input_tokens_seen": 17591456,
      "step": 26935
    },
    {
      "epoch": 15.88443396226415,
      "grad_norm": 3.853743314743042,
      "learning_rate": 1.2357165184689906e-06,
      "loss": 0.4097,
      "num_input_tokens_seen": 17594624,
      "step": 26940
    },
    {
      "epoch": 15.887382075471699,
      "grad_norm": 3.388942003250122,
      "learning_rate": 1.2340236963794845e-06,
      "loss": 0.2828,
      "num_input_tokens_seen": 17598304,
      "step": 26945
    },
    {
      "epoch": 15.890330188679245,
      "grad_norm": 3.1036558151245117,
      "learning_rate": 1.232331871348435e-06,
      "loss": 0.3064,
      "num_input_tokens_seen": 17601920,
      "step": 26950
    },
    {
      "epoch": 15.893278301886792,
      "grad_norm": 3.9187233448028564,
      "learning_rate": 1.2306410438237603e-06,
      "loss": 0.4653,
      "num_input_tokens_seen": 17605120,
      "step": 26955
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 2.8393239974975586,
      "learning_rate": 1.228951214253113e-06,
      "loss": 0.278,
      "num_input_tokens_seen": 17607904,
      "step": 26960
    },
    {
      "epoch": 15.899174528301886,
      "grad_norm": 2.659130096435547,
      "learning_rate": 1.2272623830838854e-06,
      "loss": 0.2905,
      "num_input_tokens_seen": 17610880,
      "step": 26965
    },
    {
      "epoch": 15.902122641509434,
      "grad_norm": 2.8249855041503906,
      "learning_rate": 1.2255745507632016e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 17613856,
      "step": 26970
    },
    {
      "epoch": 15.90507075471698,
      "grad_norm": 2.327476739883423,
      "learning_rate": 1.223887717737922e-06,
      "loss": 0.346,
      "num_input_tokens_seen": 17618368,
      "step": 26975
    },
    {
      "epoch": 15.908018867924529,
      "grad_norm": 2.6572186946868896,
      "learning_rate": 1.2222018844546434e-06,
      "loss": 0.2767,
      "num_input_tokens_seen": 17621216,
      "step": 26980
    },
    {
      "epoch": 15.910966981132075,
      "grad_norm": 4.8836894035339355,
      "learning_rate": 1.2205170513596975e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 17624352,
      "step": 26985
    },
    {
      "epoch": 15.913915094339622,
      "grad_norm": 5.77083683013916,
      "learning_rate": 1.2188332188991493e-06,
      "loss": 0.448,
      "num_input_tokens_seen": 17628224,
      "step": 26990
    },
    {
      "epoch": 15.91686320754717,
      "grad_norm": 3.308159112930298,
      "learning_rate": 1.217150387518804e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 17632256,
      "step": 26995
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 3.1044516563415527,
      "learning_rate": 1.2154685576641967e-06,
      "loss": 0.334,
      "num_input_tokens_seen": 17634720,
      "step": 27000
    },
    {
      "epoch": 15.922759433962264,
      "grad_norm": 2.1713929176330566,
      "learning_rate": 1.2137877297805972e-06,
      "loss": 0.2802,
      "num_input_tokens_seen": 17637600,
      "step": 27005
    },
    {
      "epoch": 15.92570754716981,
      "grad_norm": 2.9113693237304688,
      "learning_rate": 1.2121079043130162e-06,
      "loss": 0.3249,
      "num_input_tokens_seen": 17640128,
      "step": 27010
    },
    {
      "epoch": 15.928655660377359,
      "grad_norm": 2.517552375793457,
      "learning_rate": 1.210429081706192e-06,
      "loss": 0.2524,
      "num_input_tokens_seen": 17643136,
      "step": 27015
    },
    {
      "epoch": 15.931603773584905,
      "grad_norm": 2.7059073448181152,
      "learning_rate": 1.2087512624046005e-06,
      "loss": 0.3018,
      "num_input_tokens_seen": 17645952,
      "step": 27020
    },
    {
      "epoch": 15.934551886792454,
      "grad_norm": 5.0838446617126465,
      "learning_rate": 1.2070744468524503e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 17649248,
      "step": 27025
    },
    {
      "epoch": 15.9375,
      "grad_norm": 3.4200971126556396,
      "learning_rate": 1.2053986354936887e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 17651424,
      "step": 27030
    },
    {
      "epoch": 15.940448113207546,
      "grad_norm": 2.3485045433044434,
      "learning_rate": 1.2037238287719916e-06,
      "loss": 0.3578,
      "num_input_tokens_seen": 17655520,
      "step": 27035
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 3.382570505142212,
      "learning_rate": 1.2020500271307721e-06,
      "loss": 0.3092,
      "num_input_tokens_seen": 17659488,
      "step": 27040
    },
    {
      "epoch": 15.946344339622641,
      "grad_norm": 2.58204984664917,
      "learning_rate": 1.200377231013176e-06,
      "loss": 0.3011,
      "num_input_tokens_seen": 17662784,
      "step": 27045
    },
    {
      "epoch": 15.94929245283019,
      "grad_norm": 5.367882251739502,
      "learning_rate": 1.1987054408620825e-06,
      "loss": 0.2557,
      "num_input_tokens_seen": 17665056,
      "step": 27050
    },
    {
      "epoch": 15.952240566037736,
      "grad_norm": 2.5139312744140625,
      "learning_rate": 1.197034657120107e-06,
      "loss": 0.2352,
      "num_input_tokens_seen": 17668352,
      "step": 27055
    },
    {
      "epoch": 15.955188679245284,
      "grad_norm": 5.636582374572754,
      "learning_rate": 1.1953648802295964e-06,
      "loss": 0.3636,
      "num_input_tokens_seen": 17672384,
      "step": 27060
    },
    {
      "epoch": 15.95813679245283,
      "grad_norm": 4.263881683349609,
      "learning_rate": 1.1936961106326307e-06,
      "loss": 0.3218,
      "num_input_tokens_seen": 17675456,
      "step": 27065
    },
    {
      "epoch": 15.961084905660378,
      "grad_norm": 4.525121688842773,
      "learning_rate": 1.1920283487710237e-06,
      "loss": 0.4069,
      "num_input_tokens_seen": 17678560,
      "step": 27070
    },
    {
      "epoch": 15.964033018867925,
      "grad_norm": 3.0990347862243652,
      "learning_rate": 1.1903615950863228e-06,
      "loss": 0.3328,
      "num_input_tokens_seen": 17682816,
      "step": 27075
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 3.547428607940674,
      "learning_rate": 1.1886958500198076e-06,
      "loss": 0.4774,
      "num_input_tokens_seen": 17686304,
      "step": 27080
    },
    {
      "epoch": 15.96992924528302,
      "grad_norm": 3.6857857704162598,
      "learning_rate": 1.1870311140124923e-06,
      "loss": 0.2315,
      "num_input_tokens_seen": 17689120,
      "step": 27085
    },
    {
      "epoch": 15.972877358490566,
      "grad_norm": 5.309295177459717,
      "learning_rate": 1.185367387505123e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 17692192,
      "step": 27090
    },
    {
      "epoch": 15.975825471698114,
      "grad_norm": 4.850480079650879,
      "learning_rate": 1.1837046709381783e-06,
      "loss": 0.257,
      "num_input_tokens_seen": 17695520,
      "step": 27095
    },
    {
      "epoch": 15.97877358490566,
      "grad_norm": 3.3925485610961914,
      "learning_rate": 1.1820429647518678e-06,
      "loss": 0.2896,
      "num_input_tokens_seen": 17698816,
      "step": 27100
    },
    {
      "epoch": 15.981721698113208,
      "grad_norm": 4.513973236083984,
      "learning_rate": 1.1803822693861377e-06,
      "loss": 0.3303,
      "num_input_tokens_seen": 17701632,
      "step": 27105
    },
    {
      "epoch": 15.984669811320755,
      "grad_norm": 3.9131970405578613,
      "learning_rate": 1.1787225852806639e-06,
      "loss": 0.4597,
      "num_input_tokens_seen": 17705632,
      "step": 27110
    },
    {
      "epoch": 15.987617924528301,
      "grad_norm": 2.6709017753601074,
      "learning_rate": 1.177063912874853e-06,
      "loss": 0.3189,
      "num_input_tokens_seen": 17708544,
      "step": 27115
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 3.2247161865234375,
      "learning_rate": 1.1754062526078487e-06,
      "loss": 0.3516,
      "num_input_tokens_seen": 17711904,
      "step": 27120
    },
    {
      "epoch": 15.993514150943396,
      "grad_norm": 5.055791854858398,
      "learning_rate": 1.1737496049185215e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 17715616,
      "step": 27125
    },
    {
      "epoch": 15.996462264150944,
      "grad_norm": 2.2576732635498047,
      "learning_rate": 1.172093970245477e-06,
      "loss": 0.2559,
      "num_input_tokens_seen": 17718816,
      "step": 27130
    },
    {
      "epoch": 15.99941037735849,
      "grad_norm": 7.3913702964782715,
      "learning_rate": 1.1704393490270516e-06,
      "loss": 0.4832,
      "num_input_tokens_seen": 17721824,
      "step": 27135
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.5747630000114441,
      "eval_runtime": 19.019,
      "eval_samples_per_second": 89.174,
      "eval_steps_per_second": 22.293,
      "num_input_tokens_seen": 17721880,
      "step": 27136
    },
    {
      "epoch": 16.00235849056604,
      "grad_norm": 2.5693488121032715,
      "learning_rate": 1.1687857417013126e-06,
      "loss": 0.4311,
      "num_input_tokens_seen": 17724344,
      "step": 27140
    },
    {
      "epoch": 16.005306603773583,
      "grad_norm": 2.9722683429718018,
      "learning_rate": 1.1671331487060583e-06,
      "loss": 0.2978,
      "num_input_tokens_seen": 17727864,
      "step": 27145
    },
    {
      "epoch": 16.00825471698113,
      "grad_norm": 2.372413396835327,
      "learning_rate": 1.1654815704788237e-06,
      "loss": 0.3738,
      "num_input_tokens_seen": 17731320,
      "step": 27150
    },
    {
      "epoch": 16.01120283018868,
      "grad_norm": 4.101558208465576,
      "learning_rate": 1.1638310074568687e-06,
      "loss": 0.3443,
      "num_input_tokens_seen": 17735960,
      "step": 27155
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 2.90466570854187,
      "learning_rate": 1.162181460077188e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 17738968,
      "step": 27160
    },
    {
      "epoch": 16.017099056603772,
      "grad_norm": 2.8067777156829834,
      "learning_rate": 1.1605329287765056e-06,
      "loss": 0.4418,
      "num_input_tokens_seen": 17741752,
      "step": 27165
    },
    {
      "epoch": 16.02004716981132,
      "grad_norm": 3.8389670848846436,
      "learning_rate": 1.1588854139912775e-06,
      "loss": 0.3518,
      "num_input_tokens_seen": 17744888,
      "step": 27170
    },
    {
      "epoch": 16.02299528301887,
      "grad_norm": 3.5865423679351807,
      "learning_rate": 1.1572389161576886e-06,
      "loss": 0.2777,
      "num_input_tokens_seen": 17747832,
      "step": 27175
    },
    {
      "epoch": 16.025943396226417,
      "grad_norm": 3.151651382446289,
      "learning_rate": 1.15559343571166e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 17751160,
      "step": 27180
    },
    {
      "epoch": 16.02889150943396,
      "grad_norm": 4.623836517333984,
      "learning_rate": 1.153948973088837e-06,
      "loss": 0.2531,
      "num_input_tokens_seen": 17754456,
      "step": 27185
    },
    {
      "epoch": 16.03183962264151,
      "grad_norm": 6.9170098304748535,
      "learning_rate": 1.1523055287245993e-06,
      "loss": 0.3476,
      "num_input_tokens_seen": 17757464,
      "step": 27190
    },
    {
      "epoch": 16.034787735849058,
      "grad_norm": 3.275547504425049,
      "learning_rate": 1.150663103054056e-06,
      "loss": 0.2735,
      "num_input_tokens_seen": 17760280,
      "step": 27195
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 3.937713146209717,
      "learning_rate": 1.1490216965120438e-06,
      "loss": 0.3323,
      "num_input_tokens_seen": 17762968,
      "step": 27200
    },
    {
      "epoch": 16.04068396226415,
      "grad_norm": 3.151845693588257,
      "learning_rate": 1.147381309533136e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 17766200,
      "step": 27205
    },
    {
      "epoch": 16.0436320754717,
      "grad_norm": 2.1526002883911133,
      "learning_rate": 1.1457419425516287e-06,
      "loss": 0.2734,
      "num_input_tokens_seen": 17769144,
      "step": 27210
    },
    {
      "epoch": 16.046580188679247,
      "grad_norm": 3.762903928756714,
      "learning_rate": 1.1441035960015544e-06,
      "loss": 0.2585,
      "num_input_tokens_seen": 17772664,
      "step": 27215
    },
    {
      "epoch": 16.04952830188679,
      "grad_norm": 3.5865354537963867,
      "learning_rate": 1.1424662703166716e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 17775672,
      "step": 27220
    },
    {
      "epoch": 16.05247641509434,
      "grad_norm": 2.563685894012451,
      "learning_rate": 1.1408299659304684e-06,
      "loss": 0.2545,
      "num_input_tokens_seen": 17779928,
      "step": 27225
    },
    {
      "epoch": 16.055424528301888,
      "grad_norm": 4.033276081085205,
      "learning_rate": 1.1391946832761642e-06,
      "loss": 0.3562,
      "num_input_tokens_seen": 17783736,
      "step": 27230
    },
    {
      "epoch": 16.058372641509433,
      "grad_norm": 4.6016998291015625,
      "learning_rate": 1.137560422786706e-06,
      "loss": 0.3669,
      "num_input_tokens_seen": 17787384,
      "step": 27235
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 3.6817264556884766,
      "learning_rate": 1.1359271848947712e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 17790264,
      "step": 27240
    },
    {
      "epoch": 16.06426886792453,
      "grad_norm": 4.854693412780762,
      "learning_rate": 1.1342949700327688e-06,
      "loss": 0.2816,
      "num_input_tokens_seen": 17792568,
      "step": 27245
    },
    {
      "epoch": 16.067216981132077,
      "grad_norm": 2.10819411277771,
      "learning_rate": 1.1326637786328332e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 17796088,
      "step": 27250
    },
    {
      "epoch": 16.07016509433962,
      "grad_norm": 3.434896469116211,
      "learning_rate": 1.1310336111268293e-06,
      "loss": 0.2864,
      "num_input_tokens_seen": 17801496,
      "step": 27255
    },
    {
      "epoch": 16.07311320754717,
      "grad_norm": 3.166745901107788,
      "learning_rate": 1.1294044679463517e-06,
      "loss": 0.3723,
      "num_input_tokens_seen": 17804920,
      "step": 27260
    },
    {
      "epoch": 16.076061320754718,
      "grad_norm": 2.9236419200897217,
      "learning_rate": 1.1277763495227207e-06,
      "loss": 0.2925,
      "num_input_tokens_seen": 17808632,
      "step": 27265
    },
    {
      "epoch": 16.079009433962263,
      "grad_norm": 3.6179840564727783,
      "learning_rate": 1.1261492562869913e-06,
      "loss": 0.3324,
      "num_input_tokens_seen": 17812632,
      "step": 27270
    },
    {
      "epoch": 16.08195754716981,
      "grad_norm": 2.311969041824341,
      "learning_rate": 1.1245231886699415e-06,
      "loss": 0.2554,
      "num_input_tokens_seen": 17816056,
      "step": 27275
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 3.798926591873169,
      "learning_rate": 1.12289814710208e-06,
      "loss": 0.2981,
      "num_input_tokens_seen": 17819704,
      "step": 27280
    },
    {
      "epoch": 16.087853773584907,
      "grad_norm": 5.2353363037109375,
      "learning_rate": 1.1212741320136433e-06,
      "loss": 0.3567,
      "num_input_tokens_seen": 17822264,
      "step": 27285
    },
    {
      "epoch": 16.090801886792452,
      "grad_norm": 3.403158664703369,
      "learning_rate": 1.1196511438345963e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 17825624,
      "step": 27290
    },
    {
      "epoch": 16.09375,
      "grad_norm": 4.372208595275879,
      "learning_rate": 1.118029182994631e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 17828728,
      "step": 27295
    },
    {
      "epoch": 16.096698113207548,
      "grad_norm": 5.478572368621826,
      "learning_rate": 1.1164082499231704e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 17831736,
      "step": 27300
    },
    {
      "epoch": 16.099646226415093,
      "grad_norm": 2.467630624771118,
      "learning_rate": 1.114788345049364e-06,
      "loss": 0.3409,
      "num_input_tokens_seen": 17835320,
      "step": 27305
    },
    {
      "epoch": 16.10259433962264,
      "grad_norm": 3.5947797298431396,
      "learning_rate": 1.1131694688020872e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 17839256,
      "step": 27310
    },
    {
      "epoch": 16.10554245283019,
      "grad_norm": 3.96307635307312,
      "learning_rate": 1.1115516216099453e-06,
      "loss": 0.3446,
      "num_input_tokens_seen": 17841432,
      "step": 27315
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 4.079520225524902,
      "learning_rate": 1.1099348039012698e-06,
      "loss": 0.2692,
      "num_input_tokens_seen": 17845432,
      "step": 27320
    },
    {
      "epoch": 16.111438679245282,
      "grad_norm": 4.278092384338379,
      "learning_rate": 1.1083190161041202e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 17848216,
      "step": 27325
    },
    {
      "epoch": 16.11438679245283,
      "grad_norm": 6.621017932891846,
      "learning_rate": 1.1067042586462822e-06,
      "loss": 0.2599,
      "num_input_tokens_seen": 17851896,
      "step": 27330
    },
    {
      "epoch": 16.11733490566038,
      "grad_norm": 3.9137792587280273,
      "learning_rate": 1.1050905319552718e-06,
      "loss": 0.2912,
      "num_input_tokens_seen": 17855192,
      "step": 27335
    },
    {
      "epoch": 16.120283018867923,
      "grad_norm": 4.103687763214111,
      "learning_rate": 1.1034778364583293e-06,
      "loss": 0.4309,
      "num_input_tokens_seen": 17858552,
      "step": 27340
    },
    {
      "epoch": 16.12323113207547,
      "grad_norm": 3.3836781978607178,
      "learning_rate": 1.1018661725824231e-06,
      "loss": 0.3039,
      "num_input_tokens_seen": 17861528,
      "step": 27345
    },
    {
      "epoch": 16.12617924528302,
      "grad_norm": 3.169849395751953,
      "learning_rate": 1.100255540754247e-06,
      "loss": 0.273,
      "num_input_tokens_seen": 17864824,
      "step": 27350
    },
    {
      "epoch": 16.129127358490567,
      "grad_norm": 3.1643600463867188,
      "learning_rate": 1.0986459414002244e-06,
      "loss": 0.3952,
      "num_input_tokens_seen": 17868504,
      "step": 27355
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 2.2769832611083984,
      "learning_rate": 1.0970373749465008e-06,
      "loss": 0.3616,
      "num_input_tokens_seen": 17871640,
      "step": 27360
    },
    {
      "epoch": 16.13502358490566,
      "grad_norm": 2.861651659011841,
      "learning_rate": 1.095429841818954e-06,
      "loss": 0.2231,
      "num_input_tokens_seen": 17875416,
      "step": 27365
    },
    {
      "epoch": 16.13797169811321,
      "grad_norm": 2.6641862392425537,
      "learning_rate": 1.093823342443185e-06,
      "loss": 0.2393,
      "num_input_tokens_seen": 17879448,
      "step": 27370
    },
    {
      "epoch": 16.140919811320753,
      "grad_norm": 4.690952777862549,
      "learning_rate": 1.0922178772445203e-06,
      "loss": 0.1692,
      "num_input_tokens_seen": 17882008,
      "step": 27375
    },
    {
      "epoch": 16.1438679245283,
      "grad_norm": 3.072946548461914,
      "learning_rate": 1.0906134466480146e-06,
      "loss": 0.2041,
      "num_input_tokens_seen": 17885016,
      "step": 27380
    },
    {
      "epoch": 16.14681603773585,
      "grad_norm": 4.1507792472839355,
      "learning_rate": 1.0890100510784473e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 17887832,
      "step": 27385
    },
    {
      "epoch": 16.149764150943398,
      "grad_norm": 3.376934051513672,
      "learning_rate": 1.0874076909603227e-06,
      "loss": 0.3142,
      "num_input_tokens_seen": 17891128,
      "step": 27390
    },
    {
      "epoch": 16.152712264150942,
      "grad_norm": 3.287060260772705,
      "learning_rate": 1.0858063667178747e-06,
      "loss": 0.2749,
      "num_input_tokens_seen": 17894008,
      "step": 27395
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 2.980908155441284,
      "learning_rate": 1.0842060787750614e-06,
      "loss": 0.2356,
      "num_input_tokens_seen": 17896696,
      "step": 27400
    },
    {
      "epoch": 16.15860849056604,
      "grad_norm": 3.7345030307769775,
      "learning_rate": 1.0826068275555652e-06,
      "loss": 0.3282,
      "num_input_tokens_seen": 17900088,
      "step": 27405
    },
    {
      "epoch": 16.161556603773583,
      "grad_norm": 3.476675271987915,
      "learning_rate": 1.081008613482794e-06,
      "loss": 0.3045,
      "num_input_tokens_seen": 17902968,
      "step": 27410
    },
    {
      "epoch": 16.16450471698113,
      "grad_norm": 4.947808265686035,
      "learning_rate": 1.079411436979883e-06,
      "loss": 0.4981,
      "num_input_tokens_seen": 17905880,
      "step": 27415
    },
    {
      "epoch": 16.16745283018868,
      "grad_norm": 3.312999963760376,
      "learning_rate": 1.0778152984696905e-06,
      "loss": 0.2444,
      "num_input_tokens_seen": 17908408,
      "step": 27420
    },
    {
      "epoch": 16.170400943396228,
      "grad_norm": 2.584019184112549,
      "learning_rate": 1.0762201983747993e-06,
      "loss": 0.3381,
      "num_input_tokens_seen": 17911672,
      "step": 27425
    },
    {
      "epoch": 16.173349056603772,
      "grad_norm": 6.154178142547607,
      "learning_rate": 1.0746261371175238e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 17915320,
      "step": 27430
    },
    {
      "epoch": 16.17629716981132,
      "grad_norm": 3.839902639389038,
      "learning_rate": 1.0730331151198953e-06,
      "loss": 0.3384,
      "num_input_tokens_seen": 17918328,
      "step": 27435
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 4.084625720977783,
      "learning_rate": 1.0714411328036733e-06,
      "loss": 0.2839,
      "num_input_tokens_seen": 17921048,
      "step": 27440
    },
    {
      "epoch": 16.182193396226417,
      "grad_norm": 3.3873093128204346,
      "learning_rate": 1.0698501905903435e-06,
      "loss": 0.281,
      "num_input_tokens_seen": 17924728,
      "step": 27445
    },
    {
      "epoch": 16.18514150943396,
      "grad_norm": 2.677699565887451,
      "learning_rate": 1.0682602889011134e-06,
      "loss": 0.3369,
      "num_input_tokens_seen": 17927608,
      "step": 27450
    },
    {
      "epoch": 16.18808962264151,
      "grad_norm": 3.4303476810455322,
      "learning_rate": 1.0666714281569152e-06,
      "loss": 0.2577,
      "num_input_tokens_seen": 17931800,
      "step": 27455
    },
    {
      "epoch": 16.191037735849058,
      "grad_norm": 4.232844829559326,
      "learning_rate": 1.0650836087784095e-06,
      "loss": 0.3528,
      "num_input_tokens_seen": 17935320,
      "step": 27460
    },
    {
      "epoch": 16.193985849056602,
      "grad_norm": 4.412535190582275,
      "learning_rate": 1.0634968311859768e-06,
      "loss": 0.2861,
      "num_input_tokens_seen": 17938456,
      "step": 27465
    },
    {
      "epoch": 16.19693396226415,
      "grad_norm": 2.3588192462921143,
      "learning_rate": 1.0619110957997237e-06,
      "loss": 0.2257,
      "num_input_tokens_seen": 17942680,
      "step": 27470
    },
    {
      "epoch": 16.1998820754717,
      "grad_norm": 4.600253582000732,
      "learning_rate": 1.06032640303948e-06,
      "loss": 0.2994,
      "num_input_tokens_seen": 17946808,
      "step": 27475
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 5.6942853927612305,
      "learning_rate": 1.0587427533248002e-06,
      "loss": 0.2867,
      "num_input_tokens_seen": 17949816,
      "step": 27480
    },
    {
      "epoch": 16.20577830188679,
      "grad_norm": 3.3516685962677,
      "learning_rate": 1.057160147074961e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 17952856,
      "step": 27485
    },
    {
      "epoch": 16.20872641509434,
      "grad_norm": 4.831103324890137,
      "learning_rate": 1.0555785847089657e-06,
      "loss": 0.2555,
      "num_input_tokens_seen": 17955736,
      "step": 27490
    },
    {
      "epoch": 16.211674528301888,
      "grad_norm": 7.324881553649902,
      "learning_rate": 1.0539980666455407e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 17959480,
      "step": 27495
    },
    {
      "epoch": 16.214622641509433,
      "grad_norm": 4.795602798461914,
      "learning_rate": 1.052418593303134e-06,
      "loss": 0.3481,
      "num_input_tokens_seen": 17962456,
      "step": 27500
    },
    {
      "epoch": 16.21757075471698,
      "grad_norm": 3.3316617012023926,
      "learning_rate": 1.0508401650999178e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 17965656,
      "step": 27505
    },
    {
      "epoch": 16.22051886792453,
      "grad_norm": 4.208628177642822,
      "learning_rate": 1.0492627824537877e-06,
      "loss": 0.2849,
      "num_input_tokens_seen": 17968408,
      "step": 27510
    },
    {
      "epoch": 16.223466981132077,
      "grad_norm": 6.26914644241333,
      "learning_rate": 1.0476864457823626e-06,
      "loss": 0.229,
      "num_input_tokens_seen": 17972088,
      "step": 27515
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 5.3204145431518555,
      "learning_rate": 1.0461111555029836e-06,
      "loss": 0.3048,
      "num_input_tokens_seen": 17974648,
      "step": 27520
    },
    {
      "epoch": 16.22936320754717,
      "grad_norm": 2.5922000408172607,
      "learning_rate": 1.0445369120327175e-06,
      "loss": 0.4122,
      "num_input_tokens_seen": 17985144,
      "step": 27525
    },
    {
      "epoch": 16.232311320754718,
      "grad_norm": 3.4877166748046875,
      "learning_rate": 1.0429637157883516e-06,
      "loss": 0.3366,
      "num_input_tokens_seen": 17988056,
      "step": 27530
    },
    {
      "epoch": 16.235259433962263,
      "grad_norm": 2.8826918601989746,
      "learning_rate": 1.041391567186395e-06,
      "loss": 0.3008,
      "num_input_tokens_seen": 17992056,
      "step": 27535
    },
    {
      "epoch": 16.23820754716981,
      "grad_norm": 2.9015419483184814,
      "learning_rate": 1.0398204666430821e-06,
      "loss": 0.2671,
      "num_input_tokens_seen": 17994712,
      "step": 27540
    },
    {
      "epoch": 16.24115566037736,
      "grad_norm": 3.0022430419921875,
      "learning_rate": 1.0382504145743667e-06,
      "loss": 0.3582,
      "num_input_tokens_seen": 17997784,
      "step": 27545
    },
    {
      "epoch": 16.244103773584907,
      "grad_norm": 3.996809959411621,
      "learning_rate": 1.0366814113959294e-06,
      "loss": 0.2894,
      "num_input_tokens_seen": 18001208,
      "step": 27550
    },
    {
      "epoch": 16.247051886792452,
      "grad_norm": 6.639216899871826,
      "learning_rate": 1.0351134575231697e-06,
      "loss": 0.2657,
      "num_input_tokens_seen": 18006360,
      "step": 27555
    },
    {
      "epoch": 16.25,
      "grad_norm": 3.488778591156006,
      "learning_rate": 1.0335465533712098e-06,
      "loss": 0.2973,
      "num_input_tokens_seen": 18009368,
      "step": 27560
    },
    {
      "epoch": 16.252948113207548,
      "grad_norm": 3.4848318099975586,
      "learning_rate": 1.031980699354894e-06,
      "loss": 0.2808,
      "num_input_tokens_seen": 18011800,
      "step": 27565
    },
    {
      "epoch": 16.255896226415093,
      "grad_norm": 2.976377248764038,
      "learning_rate": 1.03041589588879e-06,
      "loss": 0.3484,
      "num_input_tokens_seen": 18014584,
      "step": 27570
    },
    {
      "epoch": 16.25884433962264,
      "grad_norm": 1.9647167921066284,
      "learning_rate": 1.0288521433871834e-06,
      "loss": 0.3088,
      "num_input_tokens_seen": 18018936,
      "step": 27575
    },
    {
      "epoch": 16.26179245283019,
      "grad_norm": 3.1421468257904053,
      "learning_rate": 1.0272894422640866e-06,
      "loss": 0.2679,
      "num_input_tokens_seen": 18022136,
      "step": 27580
    },
    {
      "epoch": 16.264740566037737,
      "grad_norm": 2.849602460861206,
      "learning_rate": 1.0257277929332332e-06,
      "loss": 0.3468,
      "num_input_tokens_seen": 18025464,
      "step": 27585
    },
    {
      "epoch": 16.267688679245282,
      "grad_norm": 2.3954241275787354,
      "learning_rate": 1.0241671958080745e-06,
      "loss": 0.4129,
      "num_input_tokens_seen": 18028920,
      "step": 27590
    },
    {
      "epoch": 16.27063679245283,
      "grad_norm": 6.475631237030029,
      "learning_rate": 1.0226076513017858e-06,
      "loss": 0.3484,
      "num_input_tokens_seen": 18032184,
      "step": 27595
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 4.097650051116943,
      "learning_rate": 1.0210491598272625e-06,
      "loss": 0.4068,
      "num_input_tokens_seen": 18035352,
      "step": 27600
    },
    {
      "epoch": 16.276533018867923,
      "grad_norm": 3.7994308471679688,
      "learning_rate": 1.0194917217971229e-06,
      "loss": 0.3069,
      "num_input_tokens_seen": 18038200,
      "step": 27605
    },
    {
      "epoch": 16.27948113207547,
      "grad_norm": 2.602895498275757,
      "learning_rate": 1.0179353376237038e-06,
      "loss": 0.3345,
      "num_input_tokens_seen": 18041048,
      "step": 27610
    },
    {
      "epoch": 16.28242924528302,
      "grad_norm": 2.778775930404663,
      "learning_rate": 1.0163800077190672e-06,
      "loss": 0.3012,
      "num_input_tokens_seen": 18043192,
      "step": 27615
    },
    {
      "epoch": 16.285377358490567,
      "grad_norm": 6.3840742111206055,
      "learning_rate": 1.0148257324949916e-06,
      "loss": 0.3919,
      "num_input_tokens_seen": 18046616,
      "step": 27620
    },
    {
      "epoch": 16.288325471698112,
      "grad_norm": 3.24342942237854,
      "learning_rate": 1.0132725123629783e-06,
      "loss": 0.288,
      "num_input_tokens_seen": 18049816,
      "step": 27625
    },
    {
      "epoch": 16.29127358490566,
      "grad_norm": 3.1366195678710938,
      "learning_rate": 1.0117203477342497e-06,
      "loss": 0.389,
      "num_input_tokens_seen": 18052824,
      "step": 27630
    },
    {
      "epoch": 16.29422169811321,
      "grad_norm": 2.1237738132476807,
      "learning_rate": 1.0101692390197477e-06,
      "loss": 0.2615,
      "num_input_tokens_seen": 18056120,
      "step": 27635
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 5.0205535888671875,
      "learning_rate": 1.0086191866301331e-06,
      "loss": 0.3125,
      "num_input_tokens_seen": 18059768,
      "step": 27640
    },
    {
      "epoch": 16.3001179245283,
      "grad_norm": 3.562006950378418,
      "learning_rate": 1.0070701909757918e-06,
      "loss": 0.3592,
      "num_input_tokens_seen": 18062648,
      "step": 27645
    },
    {
      "epoch": 16.30306603773585,
      "grad_norm": 3.806166410446167,
      "learning_rate": 1.0055222524668267e-06,
      "loss": 0.3102,
      "num_input_tokens_seen": 18067032,
      "step": 27650
    },
    {
      "epoch": 16.306014150943398,
      "grad_norm": 3.267615556716919,
      "learning_rate": 1.00397537151306e-06,
      "loss": 0.3206,
      "num_input_tokens_seen": 18070712,
      "step": 27655
    },
    {
      "epoch": 16.308962264150942,
      "grad_norm": 3.6677284240722656,
      "learning_rate": 1.002429548524036e-06,
      "loss": 0.2768,
      "num_input_tokens_seen": 18073624,
      "step": 27660
    },
    {
      "epoch": 16.31191037735849,
      "grad_norm": 2.822883367538452,
      "learning_rate": 1.0008847839090175e-06,
      "loss": 0.4041,
      "num_input_tokens_seen": 18077688,
      "step": 27665
    },
    {
      "epoch": 16.31485849056604,
      "grad_norm": 4.430085182189941,
      "learning_rate": 9.993410780769862e-07,
      "loss": 0.344,
      "num_input_tokens_seen": 18080568,
      "step": 27670
    },
    {
      "epoch": 16.317806603773583,
      "grad_norm": 2.926785469055176,
      "learning_rate": 9.977984314366463e-07,
      "loss": 0.1761,
      "num_input_tokens_seen": 18083640,
      "step": 27675
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 5.778702735900879,
      "learning_rate": 9.962568443964216e-07,
      "loss": 0.4173,
      "num_input_tokens_seen": 18087032,
      "step": 27680
    },
    {
      "epoch": 16.32370283018868,
      "grad_norm": 4.4257097244262695,
      "learning_rate": 9.947163173644524e-07,
      "loss": 0.3342,
      "num_input_tokens_seen": 18090328,
      "step": 27685
    },
    {
      "epoch": 16.326650943396228,
      "grad_norm": 3.635786294937134,
      "learning_rate": 9.931768507486007e-07,
      "loss": 0.2123,
      "num_input_tokens_seen": 18092984,
      "step": 27690
    },
    {
      "epoch": 16.329599056603772,
      "grad_norm": 3.389735460281372,
      "learning_rate": 9.916384449564453e-07,
      "loss": 0.3261,
      "num_input_tokens_seen": 18096408,
      "step": 27695
    },
    {
      "epoch": 16.33254716981132,
      "grad_norm": 2.724553108215332,
      "learning_rate": 9.90101100395287e-07,
      "loss": 0.3668,
      "num_input_tokens_seen": 18099576,
      "step": 27700
    },
    {
      "epoch": 16.33549528301887,
      "grad_norm": 4.449212551116943,
      "learning_rate": 9.885648174721428e-07,
      "loss": 0.3799,
      "num_input_tokens_seen": 18101848,
      "step": 27705
    },
    {
      "epoch": 16.338443396226417,
      "grad_norm": 4.5763421058654785,
      "learning_rate": 9.870295965937532e-07,
      "loss": 0.2762,
      "num_input_tokens_seen": 18104696,
      "step": 27710
    },
    {
      "epoch": 16.34139150943396,
      "grad_norm": 7.694447040557861,
      "learning_rate": 9.854954381665727e-07,
      "loss": 0.4656,
      "num_input_tokens_seen": 18108120,
      "step": 27715
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 3.9382705688476562,
      "learning_rate": 9.83962342596776e-07,
      "loss": 0.2788,
      "num_input_tokens_seen": 18110680,
      "step": 27720
    },
    {
      "epoch": 16.347287735849058,
      "grad_norm": 4.563595294952393,
      "learning_rate": 9.824303102902576e-07,
      "loss": 0.3585,
      "num_input_tokens_seen": 18114616,
      "step": 27725
    },
    {
      "epoch": 16.350235849056602,
      "grad_norm": 5.256905555725098,
      "learning_rate": 9.808993416526292e-07,
      "loss": 0.3388,
      "num_input_tokens_seen": 18118680,
      "step": 27730
    },
    {
      "epoch": 16.35318396226415,
      "grad_norm": 3.106235980987549,
      "learning_rate": 9.793694370892204e-07,
      "loss": 0.3413,
      "num_input_tokens_seen": 18121624,
      "step": 27735
    },
    {
      "epoch": 16.3561320754717,
      "grad_norm": 3.153129816055298,
      "learning_rate": 9.77840597005082e-07,
      "loss": 0.2584,
      "num_input_tokens_seen": 18124600,
      "step": 27740
    },
    {
      "epoch": 16.359080188679247,
      "grad_norm": 4.4321608543396,
      "learning_rate": 9.763128218049806e-07,
      "loss": 0.2657,
      "num_input_tokens_seen": 18127416,
      "step": 27745
    },
    {
      "epoch": 16.36202830188679,
      "grad_norm": 3.044034719467163,
      "learning_rate": 9.747861118934005e-07,
      "loss": 0.3186,
      "num_input_tokens_seen": 18130072,
      "step": 27750
    },
    {
      "epoch": 16.36497641509434,
      "grad_norm": 4.088731288909912,
      "learning_rate": 9.732604676745443e-07,
      "loss": 0.2439,
      "num_input_tokens_seen": 18132920,
      "step": 27755
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 2.523994207382202,
      "learning_rate": 9.717358895523333e-07,
      "loss": 0.3525,
      "num_input_tokens_seen": 18135576,
      "step": 27760
    },
    {
      "epoch": 16.370872641509433,
      "grad_norm": 4.339609146118164,
      "learning_rate": 9.702123779304074e-07,
      "loss": 0.2534,
      "num_input_tokens_seen": 18138680,
      "step": 27765
    },
    {
      "epoch": 16.37382075471698,
      "grad_norm": 4.738840579986572,
      "learning_rate": 9.686899332121203e-07,
      "loss": 0.2401,
      "num_input_tokens_seen": 18141048,
      "step": 27770
    },
    {
      "epoch": 16.37676886792453,
      "grad_norm": 3.1878724098205566,
      "learning_rate": 9.671685558005488e-07,
      "loss": 0.2888,
      "num_input_tokens_seen": 18144472,
      "step": 27775
    },
    {
      "epoch": 16.379716981132077,
      "grad_norm": 6.784033298492432,
      "learning_rate": 9.656482460984828e-07,
      "loss": 0.2644,
      "num_input_tokens_seen": 18149272,
      "step": 27780
    },
    {
      "epoch": 16.38266509433962,
      "grad_norm": 2.965301752090454,
      "learning_rate": 9.641290045084307e-07,
      "loss": 0.3597,
      "num_input_tokens_seen": 18153528,
      "step": 27785
    },
    {
      "epoch": 16.38561320754717,
      "grad_norm": 6.08561897277832,
      "learning_rate": 9.626108314326182e-07,
      "loss": 0.2621,
      "num_input_tokens_seen": 18156344,
      "step": 27790
    },
    {
      "epoch": 16.388561320754718,
      "grad_norm": 3.0875353813171387,
      "learning_rate": 9.610937272729881e-07,
      "loss": 0.3791,
      "num_input_tokens_seen": 18158872,
      "step": 27795
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 3.424264430999756,
      "learning_rate": 9.595776924311996e-07,
      "loss": 0.3502,
      "num_input_tokens_seen": 18162424,
      "step": 27800
    },
    {
      "epoch": 16.39445754716981,
      "grad_norm": 5.2200703620910645,
      "learning_rate": 9.580627273086313e-07,
      "loss": 0.2293,
      "num_input_tokens_seen": 18165496,
      "step": 27805
    },
    {
      "epoch": 16.39740566037736,
      "grad_norm": 3.856473207473755,
      "learning_rate": 9.565488323063754e-07,
      "loss": 0.3647,
      "num_input_tokens_seen": 18168888,
      "step": 27810
    },
    {
      "epoch": 16.400353773584907,
      "grad_norm": 1.4556416273117065,
      "learning_rate": 9.55036007825243e-07,
      "loss": 0.2325,
      "num_input_tokens_seen": 18172344,
      "step": 27815
    },
    {
      "epoch": 16.403301886792452,
      "grad_norm": 3.853943347930908,
      "learning_rate": 9.535242542657602e-07,
      "loss": 0.2637,
      "num_input_tokens_seen": 18176568,
      "step": 27820
    },
    {
      "epoch": 16.40625,
      "grad_norm": 4.155935287475586,
      "learning_rate": 9.520135720281692e-07,
      "loss": 0.457,
      "num_input_tokens_seen": 18180472,
      "step": 27825
    },
    {
      "epoch": 16.409198113207548,
      "grad_norm": 2.720860004425049,
      "learning_rate": 9.505039615124318e-07,
      "loss": 0.3108,
      "num_input_tokens_seen": 18183352,
      "step": 27830
    },
    {
      "epoch": 16.412146226415093,
      "grad_norm": 4.236039161682129,
      "learning_rate": 9.489954231182235e-07,
      "loss": 0.4083,
      "num_input_tokens_seen": 18186552,
      "step": 27835
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 2.60440731048584,
      "learning_rate": 9.474879572449352e-07,
      "loss": 0.3332,
      "num_input_tokens_seen": 18189752,
      "step": 27840
    },
    {
      "epoch": 16.41804245283019,
      "grad_norm": 4.244143962860107,
      "learning_rate": 9.459815642916759e-07,
      "loss": 0.3472,
      "num_input_tokens_seen": 18192792,
      "step": 27845
    },
    {
      "epoch": 16.420990566037737,
      "grad_norm": 5.233787536621094,
      "learning_rate": 9.444762446572692e-07,
      "loss": 0.3075,
      "num_input_tokens_seen": 18195096,
      "step": 27850
    },
    {
      "epoch": 16.423938679245282,
      "grad_norm": 5.0689697265625,
      "learning_rate": 9.429719987402541e-07,
      "loss": 0.3051,
      "num_input_tokens_seen": 18198904,
      "step": 27855
    },
    {
      "epoch": 16.42688679245283,
      "grad_norm": 3.291005849838257,
      "learning_rate": 9.414688269388883e-07,
      "loss": 0.2878,
      "num_input_tokens_seen": 18202488,
      "step": 27860
    },
    {
      "epoch": 16.42983490566038,
      "grad_norm": 2.2840592861175537,
      "learning_rate": 9.3996672965114e-07,
      "loss": 0.4323,
      "num_input_tokens_seen": 18206552,
      "step": 27865
    },
    {
      "epoch": 16.432783018867923,
      "grad_norm": 4.77833890914917,
      "learning_rate": 9.384657072747e-07,
      "loss": 0.4775,
      "num_input_tokens_seen": 18209432,
      "step": 27870
    },
    {
      "epoch": 16.43573113207547,
      "grad_norm": 2.792123794555664,
      "learning_rate": 9.369657602069676e-07,
      "loss": 0.2323,
      "num_input_tokens_seen": 18212344,
      "step": 27875
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 5.895608425140381,
      "learning_rate": 9.354668888450608e-07,
      "loss": 0.3323,
      "num_input_tokens_seen": 18215384,
      "step": 27880
    },
    {
      "epoch": 16.441627358490567,
      "grad_norm": 2.7167255878448486,
      "learning_rate": 9.339690935858125e-07,
      "loss": 0.2707,
      "num_input_tokens_seen": 18218680,
      "step": 27885
    },
    {
      "epoch": 16.444575471698112,
      "grad_norm": 2.7555017471313477,
      "learning_rate": 9.324723748257697e-07,
      "loss": 0.3814,
      "num_input_tokens_seen": 18221752,
      "step": 27890
    },
    {
      "epoch": 16.44752358490566,
      "grad_norm": 5.075108051300049,
      "learning_rate": 9.309767329611963e-07,
      "loss": 0.3652,
      "num_input_tokens_seen": 18224632,
      "step": 27895
    },
    {
      "epoch": 16.45047169811321,
      "grad_norm": 2.936494827270508,
      "learning_rate": 9.294821683880695e-07,
      "loss": 0.287,
      "num_input_tokens_seen": 18227576,
      "step": 27900
    },
    {
      "epoch": 16.453419811320753,
      "grad_norm": 1.8588535785675049,
      "learning_rate": 9.279886815020816e-07,
      "loss": 0.2572,
      "num_input_tokens_seen": 18231640,
      "step": 27905
    },
    {
      "epoch": 16.4563679245283,
      "grad_norm": 4.848570346832275,
      "learning_rate": 9.264962726986393e-07,
      "loss": 0.2568,
      "num_input_tokens_seen": 18235832,
      "step": 27910
    },
    {
      "epoch": 16.45931603773585,
      "grad_norm": 3.4574942588806152,
      "learning_rate": 9.250049423728652e-07,
      "loss": 0.3124,
      "num_input_tokens_seen": 18238840,
      "step": 27915
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 5.993338584899902,
      "learning_rate": 9.235146909195936e-07,
      "loss": 0.2631,
      "num_input_tokens_seen": 18242328,
      "step": 27920
    },
    {
      "epoch": 16.465212264150942,
      "grad_norm": 2.175546169281006,
      "learning_rate": 9.220255187333771e-07,
      "loss": 0.2243,
      "num_input_tokens_seen": 18246040,
      "step": 27925
    },
    {
      "epoch": 16.46816037735849,
      "grad_norm": 3.4572739601135254,
      "learning_rate": 9.205374262084798e-07,
      "loss": 0.2991,
      "num_input_tokens_seen": 18248568,
      "step": 27930
    },
    {
      "epoch": 16.47110849056604,
      "grad_norm": 2.458247661590576,
      "learning_rate": 9.190504137388806e-07,
      "loss": 0.2907,
      "num_input_tokens_seen": 18252152,
      "step": 27935
    },
    {
      "epoch": 16.474056603773583,
      "grad_norm": 3.635921001434326,
      "learning_rate": 9.175644817182722e-07,
      "loss": 0.4287,
      "num_input_tokens_seen": 18254712,
      "step": 27940
    },
    {
      "epoch": 16.47700471698113,
      "grad_norm": 5.2345967292785645,
      "learning_rate": 9.16079630540061e-07,
      "loss": 0.2941,
      "num_input_tokens_seen": 18258936,
      "step": 27945
    },
    {
      "epoch": 16.47995283018868,
      "grad_norm": 4.5294928550720215,
      "learning_rate": 9.145958605973676e-07,
      "loss": 0.4095,
      "num_input_tokens_seen": 18262360,
      "step": 27950
    },
    {
      "epoch": 16.482900943396228,
      "grad_norm": 8.738910675048828,
      "learning_rate": 9.131131722830289e-07,
      "loss": 0.434,
      "num_input_tokens_seen": 18265112,
      "step": 27955
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 2.6397457122802734,
      "learning_rate": 9.116315659895892e-07,
      "loss": 0.2738,
      "num_input_tokens_seen": 18268056,
      "step": 27960
    },
    {
      "epoch": 16.48879716981132,
      "grad_norm": 7.495605945587158,
      "learning_rate": 9.10151042109314e-07,
      "loss": 0.3828,
      "num_input_tokens_seen": 18270456,
      "step": 27965
    },
    {
      "epoch": 16.49174528301887,
      "grad_norm": 3.953989028930664,
      "learning_rate": 9.086716010341767e-07,
      "loss": 0.379,
      "num_input_tokens_seen": 18273528,
      "step": 27970
    },
    {
      "epoch": 16.494693396226417,
      "grad_norm": 3.3032877445220947,
      "learning_rate": 9.071932431558655e-07,
      "loss": 0.3895,
      "num_input_tokens_seen": 18276312,
      "step": 27975
    },
    {
      "epoch": 16.49764150943396,
      "grad_norm": 3.2237021923065186,
      "learning_rate": 9.057159688657824e-07,
      "loss": 0.3397,
      "num_input_tokens_seen": 18279352,
      "step": 27980
    },
    {
      "epoch": 16.50058962264151,
      "grad_norm": 4.781652927398682,
      "learning_rate": 9.042397785550405e-07,
      "loss": 0.2221,
      "num_input_tokens_seen": 18281784,
      "step": 27985
    },
    {
      "epoch": 16.503537735849058,
      "grad_norm": 4.088425159454346,
      "learning_rate": 9.027646726144707e-07,
      "loss": 0.3444,
      "num_input_tokens_seen": 18284344,
      "step": 27990
    },
    {
      "epoch": 16.506485849056602,
      "grad_norm": 3.808122158050537,
      "learning_rate": 9.012906514346115e-07,
      "loss": 0.3737,
      "num_input_tokens_seen": 18288152,
      "step": 27995
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 3.4689464569091797,
      "learning_rate": 8.99817715405717e-07,
      "loss": 0.4255,
      "num_input_tokens_seen": 18290584,
      "step": 28000
    },
    {
      "epoch": 16.5123820754717,
      "grad_norm": 4.25819730758667,
      "learning_rate": 8.983458649177529e-07,
      "loss": 0.3397,
      "num_input_tokens_seen": 18294584,
      "step": 28005
    },
    {
      "epoch": 16.515330188679247,
      "grad_norm": 7.33821439743042,
      "learning_rate": 8.968751003603982e-07,
      "loss": 0.4659,
      "num_input_tokens_seen": 18297656,
      "step": 28010
    },
    {
      "epoch": 16.51827830188679,
      "grad_norm": 4.921268939971924,
      "learning_rate": 8.95405422123043e-07,
      "loss": 0.3642,
      "num_input_tokens_seen": 18300344,
      "step": 28015
    },
    {
      "epoch": 16.52122641509434,
      "grad_norm": 3.8926029205322266,
      "learning_rate": 8.939368305947932e-07,
      "loss": 0.2651,
      "num_input_tokens_seen": 18304120,
      "step": 28020
    },
    {
      "epoch": 16.524174528301888,
      "grad_norm": 3.7414276599884033,
      "learning_rate": 8.92469326164464e-07,
      "loss": 0.3315,
      "num_input_tokens_seen": 18306648,
      "step": 28025
    },
    {
      "epoch": 16.527122641509433,
      "grad_norm": 3.8482439517974854,
      "learning_rate": 8.910029092205829e-07,
      "loss": 0.4088,
      "num_input_tokens_seen": 18310104,
      "step": 28030
    },
    {
      "epoch": 16.53007075471698,
      "grad_norm": 4.774228572845459,
      "learning_rate": 8.895375801513906e-07,
      "loss": 0.1654,
      "num_input_tokens_seen": 18313208,
      "step": 28035
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 3.6608729362487793,
      "learning_rate": 8.880733393448377e-07,
      "loss": 0.2908,
      "num_input_tokens_seen": 18316568,
      "step": 28040
    },
    {
      "epoch": 16.535966981132077,
      "grad_norm": 4.307713985443115,
      "learning_rate": 8.866101871885907e-07,
      "loss": 0.3565,
      "num_input_tokens_seen": 18319768,
      "step": 28045
    },
    {
      "epoch": 16.53891509433962,
      "grad_norm": 2.7867419719696045,
      "learning_rate": 8.851481240700249e-07,
      "loss": 0.2677,
      "num_input_tokens_seen": 18322872,
      "step": 28050
    },
    {
      "epoch": 16.54186320754717,
      "grad_norm": 4.711722373962402,
      "learning_rate": 8.836871503762257e-07,
      "loss": 0.2703,
      "num_input_tokens_seen": 18326136,
      "step": 28055
    },
    {
      "epoch": 16.544811320754718,
      "grad_norm": 4.086217403411865,
      "learning_rate": 8.822272664939946e-07,
      "loss": 0.3812,
      "num_input_tokens_seen": 18329176,
      "step": 28060
    },
    {
      "epoch": 16.547759433962263,
      "grad_norm": 2.0990662574768066,
      "learning_rate": 8.80768472809842e-07,
      "loss": 0.3487,
      "num_input_tokens_seen": 18332984,
      "step": 28065
    },
    {
      "epoch": 16.55070754716981,
      "grad_norm": 2.3665785789489746,
      "learning_rate": 8.793107697099884e-07,
      "loss": 0.2828,
      "num_input_tokens_seen": 18337016,
      "step": 28070
    },
    {
      "epoch": 16.55365566037736,
      "grad_norm": 3.0307488441467285,
      "learning_rate": 8.778541575803673e-07,
      "loss": 0.3572,
      "num_input_tokens_seen": 18339384,
      "step": 28075
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 1.9429633617401123,
      "learning_rate": 8.763986368066241e-07,
      "loss": 0.2446,
      "num_input_tokens_seen": 18343096,
      "step": 28080
    },
    {
      "epoch": 16.559551886792452,
      "grad_norm": 6.743112087249756,
      "learning_rate": 8.749442077741138e-07,
      "loss": 0.2879,
      "num_input_tokens_seen": 18346136,
      "step": 28085
    },
    {
      "epoch": 16.5625,
      "grad_norm": 3.393894672393799,
      "learning_rate": 8.734908708679024e-07,
      "loss": 0.4024,
      "num_input_tokens_seen": 18349112,
      "step": 28090
    },
    {
      "epoch": 16.565448113207548,
      "grad_norm": 4.97408390045166,
      "learning_rate": 8.72038626472767e-07,
      "loss": 0.3102,
      "num_input_tokens_seen": 18351640,
      "step": 28095
    },
    {
      "epoch": 16.568396226415093,
      "grad_norm": 2.7364213466644287,
      "learning_rate": 8.705874749731962e-07,
      "loss": 0.3657,
      "num_input_tokens_seen": 18355512,
      "step": 28100
    },
    {
      "epoch": 16.57134433962264,
      "grad_norm": 4.576204299926758,
      "learning_rate": 8.691374167533867e-07,
      "loss": 0.3527,
      "num_input_tokens_seen": 18358168,
      "step": 28105
    },
    {
      "epoch": 16.57429245283019,
      "grad_norm": 4.26041841506958,
      "learning_rate": 8.6768845219725e-07,
      "loss": 0.4252,
      "num_input_tokens_seen": 18360824,
      "step": 28110
    },
    {
      "epoch": 16.577240566037737,
      "grad_norm": 3.1178863048553467,
      "learning_rate": 8.662405816884056e-07,
      "loss": 0.3245,
      "num_input_tokens_seen": 18363736,
      "step": 28115
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 4.269161224365234,
      "learning_rate": 8.647938056101824e-07,
      "loss": 0.3109,
      "num_input_tokens_seen": 18367064,
      "step": 28120
    },
    {
      "epoch": 16.58313679245283,
      "grad_norm": 4.333589553833008,
      "learning_rate": 8.63348124345621e-07,
      "loss": 0.3844,
      "num_input_tokens_seen": 18370008,
      "step": 28125
    },
    {
      "epoch": 16.58608490566038,
      "grad_norm": 3.8552932739257812,
      "learning_rate": 8.619035382774716e-07,
      "loss": 0.3085,
      "num_input_tokens_seen": 18372824,
      "step": 28130
    },
    {
      "epoch": 16.589033018867923,
      "grad_norm": 4.694077491760254,
      "learning_rate": 8.60460047788193e-07,
      "loss": 0.3076,
      "num_input_tokens_seen": 18376792,
      "step": 28135
    },
    {
      "epoch": 16.59198113207547,
      "grad_norm": 2.6462182998657227,
      "learning_rate": 8.590176532599587e-07,
      "loss": 0.2699,
      "num_input_tokens_seen": 18381112,
      "step": 28140
    },
    {
      "epoch": 16.59492924528302,
      "grad_norm": 3.9045355319976807,
      "learning_rate": 8.575763550746475e-07,
      "loss": 0.2788,
      "num_input_tokens_seen": 18384376,
      "step": 28145
    },
    {
      "epoch": 16.597877358490567,
      "grad_norm": 5.151724338531494,
      "learning_rate": 8.56136153613848e-07,
      "loss": 0.2771,
      "num_input_tokens_seen": 18387096,
      "step": 28150
    },
    {
      "epoch": 16.600825471698112,
      "grad_norm": 5.624205112457275,
      "learning_rate": 8.546970492588619e-07,
      "loss": 0.3517,
      "num_input_tokens_seen": 18390040,
      "step": 28155
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 4.170825958251953,
      "learning_rate": 8.532590423906973e-07,
      "loss": 0.3615,
      "num_input_tokens_seen": 18392728,
      "step": 28160
    },
    {
      "epoch": 16.60672169811321,
      "grad_norm": 6.859929084777832,
      "learning_rate": 8.518221333900728e-07,
      "loss": 0.3312,
      "num_input_tokens_seen": 18395640,
      "step": 28165
    },
    {
      "epoch": 16.609669811320753,
      "grad_norm": 4.249855041503906,
      "learning_rate": 8.503863226374148e-07,
      "loss": 0.4159,
      "num_input_tokens_seen": 18399416,
      "step": 28170
    },
    {
      "epoch": 16.6126179245283,
      "grad_norm": 4.1701555252075195,
      "learning_rate": 8.489516105128632e-07,
      "loss": 0.3957,
      "num_input_tokens_seen": 18402360,
      "step": 28175
    },
    {
      "epoch": 16.61556603773585,
      "grad_norm": 3.0852420330047607,
      "learning_rate": 8.475179973962621e-07,
      "loss": 0.2613,
      "num_input_tokens_seen": 18406040,
      "step": 28180
    },
    {
      "epoch": 16.618514150943398,
      "grad_norm": 3.5959715843200684,
      "learning_rate": 8.460854836671678e-07,
      "loss": 0.4174,
      "num_input_tokens_seen": 18409496,
      "step": 28185
    },
    {
      "epoch": 16.621462264150942,
      "grad_norm": 8.54028606414795,
      "learning_rate": 8.446540697048445e-07,
      "loss": 0.1983,
      "num_input_tokens_seen": 18416728,
      "step": 28190
    },
    {
      "epoch": 16.62441037735849,
      "grad_norm": 3.300555944442749,
      "learning_rate": 8.432237558882639e-07,
      "loss": 0.3617,
      "num_input_tokens_seen": 18420568,
      "step": 28195
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 4.357990741729736,
      "learning_rate": 8.417945425961083e-07,
      "loss": 0.4196,
      "num_input_tokens_seen": 18423448,
      "step": 28200
    },
    {
      "epoch": 16.630306603773583,
      "grad_norm": 5.202925205230713,
      "learning_rate": 8.403664302067688e-07,
      "loss": 0.3398,
      "num_input_tokens_seen": 18427544,
      "step": 28205
    },
    {
      "epoch": 16.63325471698113,
      "grad_norm": 2.252030611038208,
      "learning_rate": 8.389394190983446e-07,
      "loss": 0.3458,
      "num_input_tokens_seen": 18430808,
      "step": 28210
    },
    {
      "epoch": 16.63620283018868,
      "grad_norm": 3.337977170944214,
      "learning_rate": 8.37513509648642e-07,
      "loss": 0.2782,
      "num_input_tokens_seen": 18433336,
      "step": 28215
    },
    {
      "epoch": 16.639150943396228,
      "grad_norm": 3.377767562866211,
      "learning_rate": 8.360887022351771e-07,
      "loss": 0.3259,
      "num_input_tokens_seen": 18437304,
      "step": 28220
    },
    {
      "epoch": 16.642099056603772,
      "grad_norm": 5.031632900238037,
      "learning_rate": 8.346649972351739e-07,
      "loss": 0.4762,
      "num_input_tokens_seen": 18440440,
      "step": 28225
    },
    {
      "epoch": 16.64504716981132,
      "grad_norm": 4.0832905769348145,
      "learning_rate": 8.33242395025563e-07,
      "loss": 0.2173,
      "num_input_tokens_seen": 18444024,
      "step": 28230
    },
    {
      "epoch": 16.64799528301887,
      "grad_norm": 4.621260643005371,
      "learning_rate": 8.318208959829871e-07,
      "loss": 0.4816,
      "num_input_tokens_seen": 18447096,
      "step": 28235
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 4.438861846923828,
      "learning_rate": 8.304005004837929e-07,
      "loss": 0.2823,
      "num_input_tokens_seen": 18450104,
      "step": 28240
    },
    {
      "epoch": 16.65389150943396,
      "grad_norm": 5.168231010437012,
      "learning_rate": 8.289812089040344e-07,
      "loss": 0.4389,
      "num_input_tokens_seen": 18453176,
      "step": 28245
    },
    {
      "epoch": 16.65683962264151,
      "grad_norm": 3.3992292881011963,
      "learning_rate": 8.275630216194785e-07,
      "loss": 0.3141,
      "num_input_tokens_seen": 18457112,
      "step": 28250
    },
    {
      "epoch": 16.659787735849058,
      "grad_norm": 6.166996955871582,
      "learning_rate": 8.261459390055948e-07,
      "loss": 0.2439,
      "num_input_tokens_seen": 18459480,
      "step": 28255
    },
    {
      "epoch": 16.662735849056602,
      "grad_norm": 7.355685710906982,
      "learning_rate": 8.24729961437562e-07,
      "loss": 0.3531,
      "num_input_tokens_seen": 18462168,
      "step": 28260
    },
    {
      "epoch": 16.66568396226415,
      "grad_norm": 3.242323637008667,
      "learning_rate": 8.233150892902653e-07,
      "loss": 0.3237,
      "num_input_tokens_seen": 18465592,
      "step": 28265
    },
    {
      "epoch": 16.6686320754717,
      "grad_norm": 3.7889318466186523,
      "learning_rate": 8.219013229383005e-07,
      "loss": 0.3815,
      "num_input_tokens_seen": 18469048,
      "step": 28270
    },
    {
      "epoch": 16.671580188679247,
      "grad_norm": 4.3089799880981445,
      "learning_rate": 8.204886627559666e-07,
      "loss": 0.3819,
      "num_input_tokens_seen": 18472408,
      "step": 28275
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 3.0126426219940186,
      "learning_rate": 8.190771091172722e-07,
      "loss": 0.2651,
      "num_input_tokens_seen": 18475512,
      "step": 28280
    },
    {
      "epoch": 16.67747641509434,
      "grad_norm": 6.028242111206055,
      "learning_rate": 8.176666623959323e-07,
      "loss": 0.4102,
      "num_input_tokens_seen": 18477976,
      "step": 28285
    },
    {
      "epoch": 16.680424528301888,
      "grad_norm": 2.5954408645629883,
      "learning_rate": 8.162573229653681e-07,
      "loss": 0.3218,
      "num_input_tokens_seen": 18482136,
      "step": 28290
    },
    {
      "epoch": 16.683372641509433,
      "grad_norm": 5.931403636932373,
      "learning_rate": 8.148490911987073e-07,
      "loss": 0.3179,
      "num_input_tokens_seen": 18485976,
      "step": 28295
    },
    {
      "epoch": 16.68632075471698,
      "grad_norm": 2.8448009490966797,
      "learning_rate": 8.134419674687876e-07,
      "loss": 0.3495,
      "num_input_tokens_seen": 18488408,
      "step": 28300
    },
    {
      "epoch": 16.68926886792453,
      "grad_norm": 3.805569887161255,
      "learning_rate": 8.120359521481502e-07,
      "loss": 0.2698,
      "num_input_tokens_seen": 18491544,
      "step": 28305
    },
    {
      "epoch": 16.692216981132077,
      "grad_norm": 3.450291156768799,
      "learning_rate": 8.106310456090438e-07,
      "loss": 0.2652,
      "num_input_tokens_seen": 18494200,
      "step": 28310
    },
    {
      "epoch": 16.69516509433962,
      "grad_norm": 3.927365303039551,
      "learning_rate": 8.092272482234231e-07,
      "loss": 0.3441,
      "num_input_tokens_seen": 18497656,
      "step": 28315
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 3.3123083114624023,
      "learning_rate": 8.078245603629486e-07,
      "loss": 0.2914,
      "num_input_tokens_seen": 18500120,
      "step": 28320
    },
    {
      "epoch": 16.701061320754718,
      "grad_norm": 3.6467196941375732,
      "learning_rate": 8.0642298239899e-07,
      "loss": 0.3345,
      "num_input_tokens_seen": 18503608,
      "step": 28325
    },
    {
      "epoch": 16.704009433962263,
      "grad_norm": 1.6442614793777466,
      "learning_rate": 8.050225147026202e-07,
      "loss": 0.2588,
      "num_input_tokens_seen": 18506296,
      "step": 28330
    },
    {
      "epoch": 16.70695754716981,
      "grad_norm": 1.5211164951324463,
      "learning_rate": 8.03623157644619e-07,
      "loss": 0.313,
      "num_input_tokens_seen": 18509848,
      "step": 28335
    },
    {
      "epoch": 16.70990566037736,
      "grad_norm": 2.742154121398926,
      "learning_rate": 8.022249115954728e-07,
      "loss": 0.3008,
      "num_input_tokens_seen": 18513144,
      "step": 28340
    },
    {
      "epoch": 16.712853773584907,
      "grad_norm": 6.130876541137695,
      "learning_rate": 8.008277769253709e-07,
      "loss": 0.3002,
      "num_input_tokens_seen": 18517016,
      "step": 28345
    },
    {
      "epoch": 16.715801886792452,
      "grad_norm": 5.696763515472412,
      "learning_rate": 7.994317540042135e-07,
      "loss": 0.4764,
      "num_input_tokens_seen": 18520312,
      "step": 28350
    },
    {
      "epoch": 16.71875,
      "grad_norm": 4.041275978088379,
      "learning_rate": 7.980368432016017e-07,
      "loss": 0.2845,
      "num_input_tokens_seen": 18523480,
      "step": 28355
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 3.8337414264678955,
      "learning_rate": 7.966430448868461e-07,
      "loss": 0.2981,
      "num_input_tokens_seen": 18526744,
      "step": 28360
    },
    {
      "epoch": 16.724646226415093,
      "grad_norm": 2.952758550643921,
      "learning_rate": 7.952503594289601e-07,
      "loss": 0.3731,
      "num_input_tokens_seen": 18529720,
      "step": 28365
    },
    {
      "epoch": 16.72759433962264,
      "grad_norm": 3.137604236602783,
      "learning_rate": 7.93858787196663e-07,
      "loss": 0.3283,
      "num_input_tokens_seen": 18532760,
      "step": 28370
    },
    {
      "epoch": 16.73054245283019,
      "grad_norm": 3.6043779850006104,
      "learning_rate": 7.92468328558379e-07,
      "loss": 0.3542,
      "num_input_tokens_seen": 18535576,
      "step": 28375
    },
    {
      "epoch": 16.733490566037737,
      "grad_norm": 2.436530590057373,
      "learning_rate": 7.910789838822386e-07,
      "loss": 0.3672,
      "num_input_tokens_seen": 18539224,
      "step": 28380
    },
    {
      "epoch": 16.736438679245282,
      "grad_norm": 2.9524600505828857,
      "learning_rate": 7.89690753536076e-07,
      "loss": 0.2771,
      "num_input_tokens_seen": 18541752,
      "step": 28385
    },
    {
      "epoch": 16.73938679245283,
      "grad_norm": 3.0726377964019775,
      "learning_rate": 7.883036378874326e-07,
      "loss": 0.3002,
      "num_input_tokens_seen": 18544632,
      "step": 28390
    },
    {
      "epoch": 16.74233490566038,
      "grad_norm": 7.170734882354736,
      "learning_rate": 7.86917637303552e-07,
      "loss": 0.3191,
      "num_input_tokens_seen": 18548312,
      "step": 28395
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 3.810809373855591,
      "learning_rate": 7.855327521513851e-07,
      "loss": 0.2722,
      "num_input_tokens_seen": 18551672,
      "step": 28400
    },
    {
      "epoch": 16.74823113207547,
      "grad_norm": 3.8175883293151855,
      "learning_rate": 7.841489827975851e-07,
      "loss": 0.3593,
      "num_input_tokens_seen": 18554776,
      "step": 28405
    },
    {
      "epoch": 16.75117924528302,
      "grad_norm": 2.1779446601867676,
      "learning_rate": 7.827663296085109e-07,
      "loss": 0.2848,
      "num_input_tokens_seen": 18557368,
      "step": 28410
    },
    {
      "epoch": 16.754127358490567,
      "grad_norm": 5.474173069000244,
      "learning_rate": 7.813847929502255e-07,
      "loss": 0.4234,
      "num_input_tokens_seen": 18560184,
      "step": 28415
    },
    {
      "epoch": 16.757075471698112,
      "grad_norm": 4.243307590484619,
      "learning_rate": 7.800043731884982e-07,
      "loss": 0.4207,
      "num_input_tokens_seen": 18563224,
      "step": 28420
    },
    {
      "epoch": 16.76002358490566,
      "grad_norm": 4.5667314529418945,
      "learning_rate": 7.786250706888005e-07,
      "loss": 0.4834,
      "num_input_tokens_seen": 18565624,
      "step": 28425
    },
    {
      "epoch": 16.76297169811321,
      "grad_norm": 2.2383103370666504,
      "learning_rate": 7.772468858163085e-07,
      "loss": 0.301,
      "num_input_tokens_seen": 18569240,
      "step": 28430
    },
    {
      "epoch": 16.765919811320753,
      "grad_norm": 3.6130480766296387,
      "learning_rate": 7.758698189359026e-07,
      "loss": 0.3608,
      "num_input_tokens_seen": 18573560,
      "step": 28435
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 3.2456161975860596,
      "learning_rate": 7.744938704121658e-07,
      "loss": 0.2755,
      "num_input_tokens_seen": 18576536,
      "step": 28440
    },
    {
      "epoch": 16.77181603773585,
      "grad_norm": 3.9009268283843994,
      "learning_rate": 7.731190406093892e-07,
      "loss": 0.2987,
      "num_input_tokens_seen": 18579672,
      "step": 28445
    },
    {
      "epoch": 16.774764150943398,
      "grad_norm": 3.6364853382110596,
      "learning_rate": 7.717453298915617e-07,
      "loss": 0.3006,
      "num_input_tokens_seen": 18582328,
      "step": 28450
    },
    {
      "epoch": 16.777712264150942,
      "grad_norm": 4.5629425048828125,
      "learning_rate": 7.703727386223825e-07,
      "loss": 0.3638,
      "num_input_tokens_seen": 18585048,
      "step": 28455
    },
    {
      "epoch": 16.78066037735849,
      "grad_norm": 3.378453493118286,
      "learning_rate": 7.690012671652491e-07,
      "loss": 0.2736,
      "num_input_tokens_seen": 18587608,
      "step": 28460
    },
    {
      "epoch": 16.78360849056604,
      "grad_norm": 3.019355535507202,
      "learning_rate": 7.676309158832651e-07,
      "loss": 0.2479,
      "num_input_tokens_seen": 18590328,
      "step": 28465
    },
    {
      "epoch": 16.786556603773583,
      "grad_norm": 4.471325874328613,
      "learning_rate": 7.662616851392362e-07,
      "loss": 0.3631,
      "num_input_tokens_seen": 18593208,
      "step": 28470
    },
    {
      "epoch": 16.78950471698113,
      "grad_norm": 2.492650032043457,
      "learning_rate": 7.648935752956732e-07,
      "loss": 0.2911,
      "num_input_tokens_seen": 18597560,
      "step": 28475
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 3.572746992111206,
      "learning_rate": 7.635265867147867e-07,
      "loss": 0.2208,
      "num_input_tokens_seen": 18600792,
      "step": 28480
    },
    {
      "epoch": 16.795400943396228,
      "grad_norm": 2.094312906265259,
      "learning_rate": 7.621607197584963e-07,
      "loss": 0.3617,
      "num_input_tokens_seen": 18603448,
      "step": 28485
    },
    {
      "epoch": 16.798349056603772,
      "grad_norm": 2.6908764839172363,
      "learning_rate": 7.607959747884186e-07,
      "loss": 0.3431,
      "num_input_tokens_seen": 18606680,
      "step": 28490
    },
    {
      "epoch": 16.80129716981132,
      "grad_norm": 5.005669593811035,
      "learning_rate": 7.594323521658769e-07,
      "loss": 0.2836,
      "num_input_tokens_seen": 18609560,
      "step": 28495
    },
    {
      "epoch": 16.80424528301887,
      "grad_norm": 2.4596924781799316,
      "learning_rate": 7.580698522518958e-07,
      "loss": 0.4035,
      "num_input_tokens_seen": 18613080,
      "step": 28500
    },
    {
      "epoch": 16.807193396226417,
      "grad_norm": 6.5550007820129395,
      "learning_rate": 7.567084754072035e-07,
      "loss": 0.2846,
      "num_input_tokens_seen": 18615512,
      "step": 28505
    },
    {
      "epoch": 16.81014150943396,
      "grad_norm": 3.409705877304077,
      "learning_rate": 7.553482219922282e-07,
      "loss": 0.4973,
      "num_input_tokens_seen": 18619192,
      "step": 28510
    },
    {
      "epoch": 16.81308962264151,
      "grad_norm": 3.9285218715667725,
      "learning_rate": 7.539890923671061e-07,
      "loss": 0.4584,
      "num_input_tokens_seen": 18626392,
      "step": 28515
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 2.5780723094940186,
      "learning_rate": 7.526310868916708e-07,
      "loss": 0.3154,
      "num_input_tokens_seen": 18629560,
      "step": 28520
    },
    {
      "epoch": 16.818985849056602,
      "grad_norm": 3.920292854309082,
      "learning_rate": 7.512742059254602e-07,
      "loss": 0.2564,
      "num_input_tokens_seen": 18632248,
      "step": 28525
    },
    {
      "epoch": 16.82193396226415,
      "grad_norm": 2.7132568359375,
      "learning_rate": 7.499184498277151e-07,
      "loss": 0.2847,
      "num_input_tokens_seen": 18635672,
      "step": 28530
    },
    {
      "epoch": 16.8248820754717,
      "grad_norm": 3.2163662910461426,
      "learning_rate": 7.485638189573758e-07,
      "loss": 0.2851,
      "num_input_tokens_seen": 18639000,
      "step": 28535
    },
    {
      "epoch": 16.827830188679247,
      "grad_norm": 6.6519012451171875,
      "learning_rate": 7.472103136730891e-07,
      "loss": 0.3736,
      "num_input_tokens_seen": 18642040,
      "step": 28540
    },
    {
      "epoch": 16.83077830188679,
      "grad_norm": 2.746511220932007,
      "learning_rate": 7.458579343331996e-07,
      "loss": 0.2292,
      "num_input_tokens_seen": 18644664,
      "step": 28545
    },
    {
      "epoch": 16.83372641509434,
      "grad_norm": 3.493406295776367,
      "learning_rate": 7.445066812957569e-07,
      "loss": 0.3303,
      "num_input_tokens_seen": 18646872,
      "step": 28550
    },
    {
      "epoch": 16.836674528301888,
      "grad_norm": 8.87328052520752,
      "learning_rate": 7.43156554918511e-07,
      "loss": 0.3171,
      "num_input_tokens_seen": 18650360,
      "step": 28555
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 3.534529685974121,
      "learning_rate": 7.418075555589132e-07,
      "loss": 0.4134,
      "num_input_tokens_seen": 18653816,
      "step": 28560
    },
    {
      "epoch": 16.84257075471698,
      "grad_norm": 5.590994358062744,
      "learning_rate": 7.404596835741168e-07,
      "loss": 0.2952,
      "num_input_tokens_seen": 18656376,
      "step": 28565
    },
    {
      "epoch": 16.84551886792453,
      "grad_norm": 3.515836000442505,
      "learning_rate": 7.391129393209751e-07,
      "loss": 0.3284,
      "num_input_tokens_seen": 18659160,
      "step": 28570
    },
    {
      "epoch": 16.848466981132077,
      "grad_norm": 2.764786958694458,
      "learning_rate": 7.377673231560478e-07,
      "loss": 0.4642,
      "num_input_tokens_seen": 18662648,
      "step": 28575
    },
    {
      "epoch": 16.85141509433962,
      "grad_norm": 4.552890777587891,
      "learning_rate": 7.364228354355907e-07,
      "loss": 0.3996,
      "num_input_tokens_seen": 18665400,
      "step": 28580
    },
    {
      "epoch": 16.85436320754717,
      "grad_norm": 4.622095108032227,
      "learning_rate": 7.350794765155627e-07,
      "loss": 0.3181,
      "num_input_tokens_seen": 18667512,
      "step": 28585
    },
    {
      "epoch": 16.857311320754718,
      "grad_norm": 2.6314871311187744,
      "learning_rate": 7.337372467516246e-07,
      "loss": 0.393,
      "num_input_tokens_seen": 18670616,
      "step": 28590
    },
    {
      "epoch": 16.860259433962263,
      "grad_norm": 4.115787029266357,
      "learning_rate": 7.323961464991369e-07,
      "loss": 0.3448,
      "num_input_tokens_seen": 18674232,
      "step": 28595
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 3.0350706577301025,
      "learning_rate": 7.310561761131601e-07,
      "loss": 0.2686,
      "num_input_tokens_seen": 18679992,
      "step": 28600
    },
    {
      "epoch": 16.86615566037736,
      "grad_norm": 4.719606399536133,
      "learning_rate": 7.297173359484605e-07,
      "loss": 0.3816,
      "num_input_tokens_seen": 18682808,
      "step": 28605
    },
    {
      "epoch": 16.869103773584907,
      "grad_norm": 3.5339155197143555,
      "learning_rate": 7.283796263595e-07,
      "loss": 0.2107,
      "num_input_tokens_seen": 18686392,
      "step": 28610
    },
    {
      "epoch": 16.872051886792452,
      "grad_norm": 3.699084997177124,
      "learning_rate": 7.270430477004431e-07,
      "loss": 0.2313,
      "num_input_tokens_seen": 18689624,
      "step": 28615
    },
    {
      "epoch": 16.875,
      "grad_norm": 4.342032432556152,
      "learning_rate": 7.257076003251545e-07,
      "loss": 0.2731,
      "num_input_tokens_seen": 18692184,
      "step": 28620
    },
    {
      "epoch": 16.877948113207548,
      "grad_norm": 3.695732355117798,
      "learning_rate": 7.243732845871998e-07,
      "loss": 0.2942,
      "num_input_tokens_seen": 18694680,
      "step": 28625
    },
    {
      "epoch": 16.880896226415093,
      "grad_norm": 2.669128179550171,
      "learning_rate": 7.230401008398441e-07,
      "loss": 0.3018,
      "num_input_tokens_seen": 18698136,
      "step": 28630
    },
    {
      "epoch": 16.88384433962264,
      "grad_norm": 3.7879445552825928,
      "learning_rate": 7.217080494360546e-07,
      "loss": 0.2808,
      "num_input_tokens_seen": 18701848,
      "step": 28635
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 3.7621188163757324,
      "learning_rate": 7.20377130728498e-07,
      "loss": 0.2053,
      "num_input_tokens_seen": 18705592,
      "step": 28640
    },
    {
      "epoch": 16.889740566037737,
      "grad_norm": 3.2621262073516846,
      "learning_rate": 7.190473450695407e-07,
      "loss": 0.2663,
      "num_input_tokens_seen": 18708696,
      "step": 28645
    },
    {
      "epoch": 16.892688679245282,
      "grad_norm": 3.285537004470825,
      "learning_rate": 7.177186928112484e-07,
      "loss": 0.3019,
      "num_input_tokens_seen": 18712248,
      "step": 28650
    },
    {
      "epoch": 16.89563679245283,
      "grad_norm": 2.9174611568450928,
      "learning_rate": 7.163911743053876e-07,
      "loss": 0.4621,
      "num_input_tokens_seen": 18715320,
      "step": 28655
    },
    {
      "epoch": 16.89858490566038,
      "grad_norm": 3.6702826023101807,
      "learning_rate": 7.150647899034252e-07,
      "loss": 0.4676,
      "num_input_tokens_seen": 18718136,
      "step": 28660
    },
    {
      "epoch": 16.901533018867923,
      "grad_norm": 6.19390869140625,
      "learning_rate": 7.13739539956525e-07,
      "loss": 0.4649,
      "num_input_tokens_seen": 18721400,
      "step": 28665
    },
    {
      "epoch": 16.90448113207547,
      "grad_norm": 3.5642800331115723,
      "learning_rate": 7.124154248155562e-07,
      "loss": 0.4226,
      "num_input_tokens_seen": 18724280,
      "step": 28670
    },
    {
      "epoch": 16.90742924528302,
      "grad_norm": 5.1280436515808105,
      "learning_rate": 7.110924448310813e-07,
      "loss": 0.3947,
      "num_input_tokens_seen": 18726840,
      "step": 28675
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 4.872152805328369,
      "learning_rate": 7.097706003533666e-07,
      "loss": 0.2615,
      "num_input_tokens_seen": 18729432,
      "step": 28680
    },
    {
      "epoch": 16.913325471698112,
      "grad_norm": 3.001509189605713,
      "learning_rate": 7.084498917323751e-07,
      "loss": 0.3134,
      "num_input_tokens_seen": 18732760,
      "step": 28685
    },
    {
      "epoch": 16.91627358490566,
      "grad_norm": 3.244638681411743,
      "learning_rate": 7.071303193177698e-07,
      "loss": 0.3121,
      "num_input_tokens_seen": 18736536,
      "step": 28690
    },
    {
      "epoch": 16.91922169811321,
      "grad_norm": 3.4661624431610107,
      "learning_rate": 7.058118834589133e-07,
      "loss": 0.2676,
      "num_input_tokens_seen": 18740248,
      "step": 28695
    },
    {
      "epoch": 16.922169811320753,
      "grad_norm": 3.4911961555480957,
      "learning_rate": 7.044945845048684e-07,
      "loss": 0.2343,
      "num_input_tokens_seen": 18743064,
      "step": 28700
    },
    {
      "epoch": 16.9251179245283,
      "grad_norm": 5.081413269042969,
      "learning_rate": 7.031784228043948e-07,
      "loss": 0.2266,
      "num_input_tokens_seen": 18746168,
      "step": 28705
    },
    {
      "epoch": 16.92806603773585,
      "grad_norm": 3.6963090896606445,
      "learning_rate": 7.01863398705952e-07,
      "loss": 0.3932,
      "num_input_tokens_seen": 18749816,
      "step": 28710
    },
    {
      "epoch": 16.931014150943398,
      "grad_norm": 3.213303565979004,
      "learning_rate": 7.005495125576983e-07,
      "loss": 0.262,
      "num_input_tokens_seen": 18754136,
      "step": 28715
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 5.0019121170043945,
      "learning_rate": 6.99236764707491e-07,
      "loss": 0.4125,
      "num_input_tokens_seen": 18757240,
      "step": 28720
    },
    {
      "epoch": 16.93691037735849,
      "grad_norm": 3.1875762939453125,
      "learning_rate": 6.979251555028843e-07,
      "loss": 0.3243,
      "num_input_tokens_seen": 18761336,
      "step": 28725
    },
    {
      "epoch": 16.93985849056604,
      "grad_norm": 4.137831687927246,
      "learning_rate": 6.966146852911332e-07,
      "loss": 0.2711,
      "num_input_tokens_seen": 18764664,
      "step": 28730
    },
    {
      "epoch": 16.942806603773583,
      "grad_norm": 3.3539657592773438,
      "learning_rate": 6.953053544191923e-07,
      "loss": 0.361,
      "num_input_tokens_seen": 18768280,
      "step": 28735
    },
    {
      "epoch": 16.94575471698113,
      "grad_norm": 2.6681501865386963,
      "learning_rate": 6.939971632337111e-07,
      "loss": 0.306,
      "num_input_tokens_seen": 18771736,
      "step": 28740
    },
    {
      "epoch": 16.94870283018868,
      "grad_norm": 3.103348970413208,
      "learning_rate": 6.926901120810387e-07,
      "loss": 0.233,
      "num_input_tokens_seen": 18774584,
      "step": 28745
    },
    {
      "epoch": 16.951650943396228,
      "grad_norm": 8.518243789672852,
      "learning_rate": 6.91384201307222e-07,
      "loss": 0.3333,
      "num_input_tokens_seen": 18777144,
      "step": 28750
    },
    {
      "epoch": 16.954599056603772,
      "grad_norm": 3.552114963531494,
      "learning_rate": 6.900794312580078e-07,
      "loss": 0.2918,
      "num_input_tokens_seen": 18780056,
      "step": 28755
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 2.4765584468841553,
      "learning_rate": 6.887758022788377e-07,
      "loss": 0.3306,
      "num_input_tokens_seen": 18783576,
      "step": 28760
    },
    {
      "epoch": 16.96049528301887,
      "grad_norm": 3.8020570278167725,
      "learning_rate": 6.874733147148549e-07,
      "loss": 0.3111,
      "num_input_tokens_seen": 18786072,
      "step": 28765
    },
    {
      "epoch": 16.963443396226417,
      "grad_norm": 3.6091468334198,
      "learning_rate": 6.861719689108987e-07,
      "loss": 0.3061,
      "num_input_tokens_seen": 18788696,
      "step": 28770
    },
    {
      "epoch": 16.96639150943396,
      "grad_norm": 4.489938259124756,
      "learning_rate": 6.84871765211505e-07,
      "loss": 0.2661,
      "num_input_tokens_seen": 18792216,
      "step": 28775
    },
    {
      "epoch": 16.96933962264151,
      "grad_norm": 7.198675632476807,
      "learning_rate": 6.835727039609086e-07,
      "loss": 0.3337,
      "num_input_tokens_seen": 18795224,
      "step": 28780
    },
    {
      "epoch": 16.972287735849058,
      "grad_norm": 3.682652711868286,
      "learning_rate": 6.822747855030415e-07,
      "loss": 0.3273,
      "num_input_tokens_seen": 18799096,
      "step": 28785
    },
    {
      "epoch": 16.975235849056602,
      "grad_norm": 2.9412262439727783,
      "learning_rate": 6.809780101815322e-07,
      "loss": 0.4221,
      "num_input_tokens_seen": 18802648,
      "step": 28790
    },
    {
      "epoch": 16.97818396226415,
      "grad_norm": 3.218263864517212,
      "learning_rate": 6.796823783397099e-07,
      "loss": 0.3151,
      "num_input_tokens_seen": 18807608,
      "step": 28795
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 4.45601224899292,
      "learning_rate": 6.783878903205976e-07,
      "loss": 0.3191,
      "num_input_tokens_seen": 18810648,
      "step": 28800
    },
    {
      "epoch": 16.984080188679247,
      "grad_norm": 4.768989086151123,
      "learning_rate": 6.77094546466916e-07,
      "loss": 0.3622,
      "num_input_tokens_seen": 18813976,
      "step": 28805
    },
    {
      "epoch": 16.98702830188679,
      "grad_norm": 7.39125394821167,
      "learning_rate": 6.758023471210845e-07,
      "loss": 0.3959,
      "num_input_tokens_seen": 18817784,
      "step": 28810
    },
    {
      "epoch": 16.98997641509434,
      "grad_norm": 3.7394485473632812,
      "learning_rate": 6.745112926252162e-07,
      "loss": 0.2355,
      "num_input_tokens_seen": 18820920,
      "step": 28815
    },
    {
      "epoch": 16.992924528301888,
      "grad_norm": 9.152069091796875,
      "learning_rate": 6.732213833211265e-07,
      "loss": 0.3699,
      "num_input_tokens_seen": 18823256,
      "step": 28820
    },
    {
      "epoch": 16.995872641509433,
      "grad_norm": 9.06840991973877,
      "learning_rate": 6.719326195503218e-07,
      "loss": 0.4811,
      "num_input_tokens_seen": 18826296,
      "step": 28825
    },
    {
      "epoch": 16.99882075471698,
      "grad_norm": 3.362377405166626,
      "learning_rate": 6.706450016540094e-07,
      "loss": 0.296,
      "num_input_tokens_seen": 18830136,
      "step": 28830
    },
    {
      "epoch": 17.00176886792453,
      "grad_norm": 2.966439962387085,
      "learning_rate": 6.69358529973092e-07,
      "loss": 0.3181,
      "num_input_tokens_seen": 18832672,
      "step": 28835
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 6.123346328735352,
      "learning_rate": 6.680732048481681e-07,
      "loss": 0.2625,
      "num_input_tokens_seen": 18835584,
      "step": 28840
    },
    {
      "epoch": 17.00766509433962,
      "grad_norm": 4.47702169418335,
      "learning_rate": 6.667890266195321e-07,
      "loss": 0.3335,
      "num_input_tokens_seen": 18839104,
      "step": 28845
    },
    {
      "epoch": 17.01061320754717,
      "grad_norm": 5.603409767150879,
      "learning_rate": 6.655059956271759e-07,
      "loss": 0.2964,
      "num_input_tokens_seen": 18842080,
      "step": 28850
    },
    {
      "epoch": 17.013561320754718,
      "grad_norm": 4.436201095581055,
      "learning_rate": 6.642241122107884e-07,
      "loss": 0.3518,
      "num_input_tokens_seen": 18845152,
      "step": 28855
    },
    {
      "epoch": 17.016509433962263,
      "grad_norm": 3.6396663188934326,
      "learning_rate": 6.629433767097537e-07,
      "loss": 0.3093,
      "num_input_tokens_seen": 18848320,
      "step": 28860
    },
    {
      "epoch": 17.01945754716981,
      "grad_norm": 4.891112804412842,
      "learning_rate": 6.616637894631517e-07,
      "loss": 0.2631,
      "num_input_tokens_seen": 18851360,
      "step": 28865
    },
    {
      "epoch": 17.02240566037736,
      "grad_norm": 4.9303789138793945,
      "learning_rate": 6.603853508097591e-07,
      "loss": 0.3472,
      "num_input_tokens_seen": 18854336,
      "step": 28870
    },
    {
      "epoch": 17.025353773584907,
      "grad_norm": 4.398445129394531,
      "learning_rate": 6.591080610880468e-07,
      "loss": 0.3227,
      "num_input_tokens_seen": 18857440,
      "step": 28875
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 2.6046156883239746,
      "learning_rate": 6.578319206361828e-07,
      "loss": 0.2593,
      "num_input_tokens_seen": 18860992,
      "step": 28880
    },
    {
      "epoch": 17.03125,
      "grad_norm": 3.3965086936950684,
      "learning_rate": 6.565569297920327e-07,
      "loss": 0.3165,
      "num_input_tokens_seen": 18864320,
      "step": 28885
    },
    {
      "epoch": 17.034198113207548,
      "grad_norm": 2.8755292892456055,
      "learning_rate": 6.552830888931544e-07,
      "loss": 0.3351,
      "num_input_tokens_seen": 18866720,
      "step": 28890
    },
    {
      "epoch": 17.037146226415093,
      "grad_norm": 3.3183062076568604,
      "learning_rate": 6.540103982768031e-07,
      "loss": 0.4076,
      "num_input_tokens_seen": 18869632,
      "step": 28895
    },
    {
      "epoch": 17.04009433962264,
      "grad_norm": 2.776266098022461,
      "learning_rate": 6.527388582799293e-07,
      "loss": 0.2841,
      "num_input_tokens_seen": 18872480,
      "step": 28900
    },
    {
      "epoch": 17.04304245283019,
      "grad_norm": 3.0167744159698486,
      "learning_rate": 6.514684692391782e-07,
      "loss": 0.378,
      "num_input_tokens_seen": 18876448,
      "step": 28905
    },
    {
      "epoch": 17.045990566037737,
      "grad_norm": 2.5297694206237793,
      "learning_rate": 6.501992314908895e-07,
      "loss": 0.3043,
      "num_input_tokens_seen": 18880192,
      "step": 28910
    },
    {
      "epoch": 17.048938679245282,
      "grad_norm": 3.871122121810913,
      "learning_rate": 6.489311453711017e-07,
      "loss": 0.3466,
      "num_input_tokens_seen": 18883616,
      "step": 28915
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 3.804379463195801,
      "learning_rate": 6.476642112155457e-07,
      "loss": 0.2478,
      "num_input_tokens_seen": 18886272,
      "step": 28920
    },
    {
      "epoch": 17.05483490566038,
      "grad_norm": 2.5956969261169434,
      "learning_rate": 6.463984293596476e-07,
      "loss": 0.319,
      "num_input_tokens_seen": 18889312,
      "step": 28925
    },
    {
      "epoch": 17.057783018867923,
      "grad_norm": 2.3112807273864746,
      "learning_rate": 6.451338001385282e-07,
      "loss": 0.266,
      "num_input_tokens_seen": 18892192,
      "step": 28930
    },
    {
      "epoch": 17.06073113207547,
      "grad_norm": 2.3293423652648926,
      "learning_rate": 6.438703238870037e-07,
      "loss": 0.2784,
      "num_input_tokens_seen": 18896960,
      "step": 28935
    },
    {
      "epoch": 17.06367924528302,
      "grad_norm": 3.423689126968384,
      "learning_rate": 6.426080009395846e-07,
      "loss": 0.3505,
      "num_input_tokens_seen": 18899776,
      "step": 28940
    },
    {
      "epoch": 17.066627358490567,
      "grad_norm": 2.389876365661621,
      "learning_rate": 6.413468316304755e-07,
      "loss": 0.2253,
      "num_input_tokens_seen": 18902432,
      "step": 28945
    },
    {
      "epoch": 17.069575471698112,
      "grad_norm": 2.8807809352874756,
      "learning_rate": 6.400868162935786e-07,
      "loss": 0.3837,
      "num_input_tokens_seen": 18906368,
      "step": 28950
    },
    {
      "epoch": 17.07252358490566,
      "grad_norm": 3.1348164081573486,
      "learning_rate": 6.388279552624877e-07,
      "loss": 0.2834,
      "num_input_tokens_seen": 18909440,
      "step": 28955
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 5.183419704437256,
      "learning_rate": 6.37570248870491e-07,
      "loss": 0.377,
      "num_input_tokens_seen": 18913920,
      "step": 28960
    },
    {
      "epoch": 17.078419811320753,
      "grad_norm": 9.07493782043457,
      "learning_rate": 6.363136974505718e-07,
      "loss": 0.4311,
      "num_input_tokens_seen": 18918080,
      "step": 28965
    },
    {
      "epoch": 17.0813679245283,
      "grad_norm": 6.650572299957275,
      "learning_rate": 6.350583013354078e-07,
      "loss": 0.2876,
      "num_input_tokens_seen": 18921024,
      "step": 28970
    },
    {
      "epoch": 17.08431603773585,
      "grad_norm": 5.48406457901001,
      "learning_rate": 6.338040608573693e-07,
      "loss": 0.3196,
      "num_input_tokens_seen": 18923648,
      "step": 28975
    },
    {
      "epoch": 17.087264150943398,
      "grad_norm": 2.744981288909912,
      "learning_rate": 6.325509763485238e-07,
      "loss": 0.3548,
      "num_input_tokens_seen": 18927264,
      "step": 28980
    },
    {
      "epoch": 17.090212264150942,
      "grad_norm": 3.4904448986053467,
      "learning_rate": 6.312990481406301e-07,
      "loss": 0.4251,
      "num_input_tokens_seen": 18930528,
      "step": 28985
    },
    {
      "epoch": 17.09316037735849,
      "grad_norm": 3.3990368843078613,
      "learning_rate": 6.300482765651411e-07,
      "loss": 0.4257,
      "num_input_tokens_seen": 18932928,
      "step": 28990
    },
    {
      "epoch": 17.09610849056604,
      "grad_norm": 4.071255207061768,
      "learning_rate": 6.28798661953205e-07,
      "loss": 0.2945,
      "num_input_tokens_seen": 18935744,
      "step": 28995
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 5.159152030944824,
      "learning_rate": 6.275502046356618e-07,
      "loss": 0.3127,
      "num_input_tokens_seen": 18939168,
      "step": 29000
    },
    {
      "epoch": 17.10200471698113,
      "grad_norm": 4.656834602355957,
      "learning_rate": 6.263029049430447e-07,
      "loss": 0.3585,
      "num_input_tokens_seen": 18941696,
      "step": 29005
    },
    {
      "epoch": 17.10495283018868,
      "grad_norm": 3.062368869781494,
      "learning_rate": 6.250567632055832e-07,
      "loss": 0.4262,
      "num_input_tokens_seen": 18946304,
      "step": 29010
    },
    {
      "epoch": 17.107900943396228,
      "grad_norm": 6.153782367706299,
      "learning_rate": 6.238117797532e-07,
      "loss": 0.3643,
      "num_input_tokens_seen": 18949120,
      "step": 29015
    },
    {
      "epoch": 17.110849056603772,
      "grad_norm": 3.6918282508850098,
      "learning_rate": 6.225679549155083e-07,
      "loss": 0.2302,
      "num_input_tokens_seen": 18952000,
      "step": 29020
    },
    {
      "epoch": 17.11379716981132,
      "grad_norm": 3.0663280487060547,
      "learning_rate": 6.213252890218163e-07,
      "loss": 0.3373,
      "num_input_tokens_seen": 18954624,
      "step": 29025
    },
    {
      "epoch": 17.11674528301887,
      "grad_norm": 8.629834175109863,
      "learning_rate": 6.200837824011247e-07,
      "loss": 0.4636,
      "num_input_tokens_seen": 18957408,
      "step": 29030
    },
    {
      "epoch": 17.119693396226417,
      "grad_norm": 3.108994960784912,
      "learning_rate": 6.188434353821282e-07,
      "loss": 0.3757,
      "num_input_tokens_seen": 18961280,
      "step": 29035
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 5.891269207000732,
      "learning_rate": 6.176042482932132e-07,
      "loss": 0.2406,
      "num_input_tokens_seen": 18964288,
      "step": 29040
    },
    {
      "epoch": 17.12558962264151,
      "grad_norm": 2.64532732963562,
      "learning_rate": 6.163662214624616e-07,
      "loss": 0.3169,
      "num_input_tokens_seen": 18966816,
      "step": 29045
    },
    {
      "epoch": 17.128537735849058,
      "grad_norm": 8.014513969421387,
      "learning_rate": 6.151293552176451e-07,
      "loss": 0.3012,
      "num_input_tokens_seen": 18969600,
      "step": 29050
    },
    {
      "epoch": 17.131485849056602,
      "grad_norm": 2.249324083328247,
      "learning_rate": 6.138936498862291e-07,
      "loss": 0.3251,
      "num_input_tokens_seen": 18972608,
      "step": 29055
    },
    {
      "epoch": 17.13443396226415,
      "grad_norm": 3.489902973175049,
      "learning_rate": 6.126591057953729e-07,
      "loss": 0.3223,
      "num_input_tokens_seen": 18976512,
      "step": 29060
    },
    {
      "epoch": 17.1373820754717,
      "grad_norm": 3.265928268432617,
      "learning_rate": 6.114257232719267e-07,
      "loss": 0.325,
      "num_input_tokens_seen": 18979648,
      "step": 29065
    },
    {
      "epoch": 17.140330188679247,
      "grad_norm": 4.462798118591309,
      "learning_rate": 6.101935026424332e-07,
      "loss": 0.2892,
      "num_input_tokens_seen": 18983168,
      "step": 29070
    },
    {
      "epoch": 17.14327830188679,
      "grad_norm": 4.18132209777832,
      "learning_rate": 6.089624442331293e-07,
      "loss": 0.3615,
      "num_input_tokens_seen": 18986112,
      "step": 29075
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 3.259573459625244,
      "learning_rate": 6.077325483699432e-07,
      "loss": 0.3049,
      "num_input_tokens_seen": 18989568,
      "step": 29080
    },
    {
      "epoch": 17.149174528301888,
      "grad_norm": 3.4189705848693848,
      "learning_rate": 6.065038153784947e-07,
      "loss": 0.3752,
      "num_input_tokens_seen": 18992576,
      "step": 29085
    },
    {
      "epoch": 17.152122641509433,
      "grad_norm": 4.124879837036133,
      "learning_rate": 6.052762455840955e-07,
      "loss": 0.2757,
      "num_input_tokens_seen": 18995584,
      "step": 29090
    },
    {
      "epoch": 17.15507075471698,
      "grad_norm": 3.8333191871643066,
      "learning_rate": 6.040498393117494e-07,
      "loss": 0.2738,
      "num_input_tokens_seen": 18998528,
      "step": 29095
    },
    {
      "epoch": 17.15801886792453,
      "grad_norm": 4.250196933746338,
      "learning_rate": 6.028245968861551e-07,
      "loss": 0.3866,
      "num_input_tokens_seen": 19005152,
      "step": 29100
    },
    {
      "epoch": 17.160966981132077,
      "grad_norm": 4.011593341827393,
      "learning_rate": 6.016005186316987e-07,
      "loss": 0.2519,
      "num_input_tokens_seen": 19007552,
      "step": 29105
    },
    {
      "epoch": 17.16391509433962,
      "grad_norm": 5.956920623779297,
      "learning_rate": 6.003776048724614e-07,
      "loss": 0.2861,
      "num_input_tokens_seen": 19010368,
      "step": 29110
    },
    {
      "epoch": 17.16686320754717,
      "grad_norm": 1.3737199306488037,
      "learning_rate": 5.991558559322152e-07,
      "loss": 0.3265,
      "num_input_tokens_seen": 19015040,
      "step": 29115
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 3.059495449066162,
      "learning_rate": 5.979352721344223e-07,
      "loss": 0.2773,
      "num_input_tokens_seen": 19017664,
      "step": 29120
    },
    {
      "epoch": 17.172759433962263,
      "grad_norm": 3.6536736488342285,
      "learning_rate": 5.967158538022383e-07,
      "loss": 0.332,
      "num_input_tokens_seen": 19021760,
      "step": 29125
    },
    {
      "epoch": 17.17570754716981,
      "grad_norm": 2.9936318397521973,
      "learning_rate": 5.954976012585078e-07,
      "loss": 0.2566,
      "num_input_tokens_seen": 19024896,
      "step": 29130
    },
    {
      "epoch": 17.17865566037736,
      "grad_norm": 2.5349936485290527,
      "learning_rate": 5.942805148257713e-07,
      "loss": 0.2686,
      "num_input_tokens_seen": 19028224,
      "step": 29135
    },
    {
      "epoch": 17.181603773584907,
      "grad_norm": 2.99298095703125,
      "learning_rate": 5.930645948262553e-07,
      "loss": 0.3121,
      "num_input_tokens_seen": 19031616,
      "step": 29140
    },
    {
      "epoch": 17.184551886792452,
      "grad_norm": 4.256497383117676,
      "learning_rate": 5.918498415818813e-07,
      "loss": 0.3119,
      "num_input_tokens_seen": 19035392,
      "step": 29145
    },
    {
      "epoch": 17.1875,
      "grad_norm": 4.123080253601074,
      "learning_rate": 5.906362554142592e-07,
      "loss": 0.3412,
      "num_input_tokens_seen": 19037856,
      "step": 29150
    },
    {
      "epoch": 17.190448113207548,
      "grad_norm": 7.723404407501221,
      "learning_rate": 5.894238366446925e-07,
      "loss": 0.3754,
      "num_input_tokens_seen": 19040320,
      "step": 29155
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 2.9545974731445312,
      "learning_rate": 5.882125855941723e-07,
      "loss": 0.2548,
      "num_input_tokens_seen": 19045760,
      "step": 29160
    },
    {
      "epoch": 17.19634433962264,
      "grad_norm": 3.5143024921417236,
      "learning_rate": 5.870025025833842e-07,
      "loss": 0.3679,
      "num_input_tokens_seen": 19049696,
      "step": 29165
    },
    {
      "epoch": 17.19929245283019,
      "grad_norm": 2.2003889083862305,
      "learning_rate": 5.857935879327031e-07,
      "loss": 0.2424,
      "num_input_tokens_seen": 19052320,
      "step": 29170
    },
    {
      "epoch": 17.202240566037737,
      "grad_norm": 4.023465156555176,
      "learning_rate": 5.845858419621936e-07,
      "loss": 0.3897,
      "num_input_tokens_seen": 19055168,
      "step": 29175
    },
    {
      "epoch": 17.205188679245282,
      "grad_norm": 5.592698097229004,
      "learning_rate": 5.83379264991612e-07,
      "loss": 0.3015,
      "num_input_tokens_seen": 19058368,
      "step": 29180
    },
    {
      "epoch": 17.20813679245283,
      "grad_norm": 3.247206211090088,
      "learning_rate": 5.821738573404046e-07,
      "loss": 0.2826,
      "num_input_tokens_seen": 19061760,
      "step": 29185
    },
    {
      "epoch": 17.21108490566038,
      "grad_norm": 4.278952121734619,
      "learning_rate": 5.80969619327707e-07,
      "loss": 0.4005,
      "num_input_tokens_seen": 19065344,
      "step": 29190
    },
    {
      "epoch": 17.214033018867923,
      "grad_norm": 2.9907169342041016,
      "learning_rate": 5.797665512723488e-07,
      "loss": 0.3684,
      "num_input_tokens_seen": 19067968,
      "step": 29195
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 2.881105422973633,
      "learning_rate": 5.785646534928452e-07,
      "loss": 0.2245,
      "num_input_tokens_seen": 19071392,
      "step": 29200
    },
    {
      "epoch": 17.21992924528302,
      "grad_norm": 4.5903000831604,
      "learning_rate": 5.77363926307406e-07,
      "loss": 0.3912,
      "num_input_tokens_seen": 19074368,
      "step": 29205
    },
    {
      "epoch": 17.222877358490567,
      "grad_norm": 1.9084266424179077,
      "learning_rate": 5.761643700339281e-07,
      "loss": 0.3145,
      "num_input_tokens_seen": 19077728,
      "step": 29210
    },
    {
      "epoch": 17.225825471698112,
      "grad_norm": 5.458241939544678,
      "learning_rate": 5.749659849899985e-07,
      "loss": 0.3239,
      "num_input_tokens_seen": 19080352,
      "step": 29215
    },
    {
      "epoch": 17.22877358490566,
      "grad_norm": 4.032101154327393,
      "learning_rate": 5.737687714928953e-07,
      "loss": 0.261,
      "num_input_tokens_seen": 19083520,
      "step": 29220
    },
    {
      "epoch": 17.23172169811321,
      "grad_norm": 8.135451316833496,
      "learning_rate": 5.725727298595846e-07,
      "loss": 0.3636,
      "num_input_tokens_seen": 19086112,
      "step": 29225
    },
    {
      "epoch": 17.234669811320753,
      "grad_norm": 2.71989369392395,
      "learning_rate": 5.71377860406726e-07,
      "loss": 0.3925,
      "num_input_tokens_seen": 19089792,
      "step": 29230
    },
    {
      "epoch": 17.2376179245283,
      "grad_norm": 2.3265304565429688,
      "learning_rate": 5.701841634506655e-07,
      "loss": 0.339,
      "num_input_tokens_seen": 19092832,
      "step": 29235
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 5.591658592224121,
      "learning_rate": 5.689916393074391e-07,
      "loss": 0.41,
      "num_input_tokens_seen": 19095840,
      "step": 29240
    },
    {
      "epoch": 17.243514150943398,
      "grad_norm": 2.7656142711639404,
      "learning_rate": 5.678002882927725e-07,
      "loss": 0.3099,
      "num_input_tokens_seen": 19098848,
      "step": 29245
    },
    {
      "epoch": 17.246462264150942,
      "grad_norm": 3.5996055603027344,
      "learning_rate": 5.666101107220811e-07,
      "loss": 0.2761,
      "num_input_tokens_seen": 19101664,
      "step": 29250
    },
    {
      "epoch": 17.24941037735849,
      "grad_norm": 6.485772132873535,
      "learning_rate": 5.654211069104693e-07,
      "loss": 0.2651,
      "num_input_tokens_seen": 19104416,
      "step": 29255
    },
    {
      "epoch": 17.25235849056604,
      "grad_norm": 3.188617467880249,
      "learning_rate": 5.642332771727321e-07,
      "loss": 0.3201,
      "num_input_tokens_seen": 19107520,
      "step": 29260
    },
    {
      "epoch": 17.255306603773583,
      "grad_norm": 3.755857229232788,
      "learning_rate": 5.630466218233521e-07,
      "loss": 0.4013,
      "num_input_tokens_seen": 19110944,
      "step": 29265
    },
    {
      "epoch": 17.25825471698113,
      "grad_norm": 3.2390003204345703,
      "learning_rate": 5.618611411765007e-07,
      "loss": 0.2883,
      "num_input_tokens_seen": 19113440,
      "step": 29270
    },
    {
      "epoch": 17.26120283018868,
      "grad_norm": 3.632431983947754,
      "learning_rate": 5.606768355460401e-07,
      "loss": 0.379,
      "num_input_tokens_seen": 19116320,
      "step": 29275
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 4.229902744293213,
      "learning_rate": 5.594937052455191e-07,
      "loss": 0.3503,
      "num_input_tokens_seen": 19119552,
      "step": 29280
    },
    {
      "epoch": 17.267099056603772,
      "grad_norm": 2.6215527057647705,
      "learning_rate": 5.583117505881764e-07,
      "loss": 0.3144,
      "num_input_tokens_seen": 19123776,
      "step": 29285
    },
    {
      "epoch": 17.27004716981132,
      "grad_norm": 3.4190385341644287,
      "learning_rate": 5.571309718869417e-07,
      "loss": 0.3459,
      "num_input_tokens_seen": 19126784,
      "step": 29290
    },
    {
      "epoch": 17.27299528301887,
      "grad_norm": 3.0819878578186035,
      "learning_rate": 5.559513694544282e-07,
      "loss": 0.256,
      "num_input_tokens_seen": 19130464,
      "step": 29295
    },
    {
      "epoch": 17.275943396226417,
      "grad_norm": 3.7838897705078125,
      "learning_rate": 5.547729436029442e-07,
      "loss": 0.2067,
      "num_input_tokens_seen": 19132448,
      "step": 29300
    },
    {
      "epoch": 17.27889150943396,
      "grad_norm": 3.096438407897949,
      "learning_rate": 5.535956946444809e-07,
      "loss": 0.2412,
      "num_input_tokens_seen": 19136096,
      "step": 29305
    },
    {
      "epoch": 17.28183962264151,
      "grad_norm": 2.7098801136016846,
      "learning_rate": 5.524196228907203e-07,
      "loss": 0.2971,
      "num_input_tokens_seen": 19139264,
      "step": 29310
    },
    {
      "epoch": 17.284787735849058,
      "grad_norm": 3.8373076915740967,
      "learning_rate": 5.512447286530326e-07,
      "loss": 0.4214,
      "num_input_tokens_seen": 19142144,
      "step": 29315
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 4.3184356689453125,
      "learning_rate": 5.500710122424746e-07,
      "loss": 0.3895,
      "num_input_tokens_seen": 19145312,
      "step": 29320
    },
    {
      "epoch": 17.29068396226415,
      "grad_norm": 3.6041417121887207,
      "learning_rate": 5.488984739697961e-07,
      "loss": 0.2608,
      "num_input_tokens_seen": 19148352,
      "step": 29325
    },
    {
      "epoch": 17.2936320754717,
      "grad_norm": 4.7440505027771,
      "learning_rate": 5.477271141454294e-07,
      "loss": 0.2587,
      "num_input_tokens_seen": 19151968,
      "step": 29330
    },
    {
      "epoch": 17.296580188679247,
      "grad_norm": 4.102355480194092,
      "learning_rate": 5.465569330794974e-07,
      "loss": 0.3032,
      "num_input_tokens_seen": 19154912,
      "step": 29335
    },
    {
      "epoch": 17.29952830188679,
      "grad_norm": 2.203185796737671,
      "learning_rate": 5.453879310818105e-07,
      "loss": 0.3092,
      "num_input_tokens_seen": 19158656,
      "step": 29340
    },
    {
      "epoch": 17.30247641509434,
      "grad_norm": 2.2666819095611572,
      "learning_rate": 5.442201084618664e-07,
      "loss": 0.3089,
      "num_input_tokens_seen": 19166496,
      "step": 29345
    },
    {
      "epoch": 17.305424528301888,
      "grad_norm": 3.3439276218414307,
      "learning_rate": 5.430534655288528e-07,
      "loss": 0.2781,
      "num_input_tokens_seen": 19169600,
      "step": 29350
    },
    {
      "epoch": 17.308372641509433,
      "grad_norm": 5.6769633293151855,
      "learning_rate": 5.418880025916428e-07,
      "loss": 0.3636,
      "num_input_tokens_seen": 19171968,
      "step": 29355
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 2.2623984813690186,
      "learning_rate": 5.407237199587973e-07,
      "loss": 0.3192,
      "num_input_tokens_seen": 19175328,
      "step": 29360
    },
    {
      "epoch": 17.31426886792453,
      "grad_norm": 8.425899505615234,
      "learning_rate": 5.395606179385654e-07,
      "loss": 0.2715,
      "num_input_tokens_seen": 19178080,
      "step": 29365
    },
    {
      "epoch": 17.317216981132077,
      "grad_norm": 9.660393714904785,
      "learning_rate": 5.383986968388833e-07,
      "loss": 0.3016,
      "num_input_tokens_seen": 19181760,
      "step": 29370
    },
    {
      "epoch": 17.32016509433962,
      "grad_norm": 4.293397903442383,
      "learning_rate": 5.372379569673736e-07,
      "loss": 0.3519,
      "num_input_tokens_seen": 19185440,
      "step": 29375
    },
    {
      "epoch": 17.32311320754717,
      "grad_norm": 6.918861389160156,
      "learning_rate": 5.360783986313495e-07,
      "loss": 0.2804,
      "num_input_tokens_seen": 19188032,
      "step": 29380
    },
    {
      "epoch": 17.326061320754718,
      "grad_norm": 5.021953105926514,
      "learning_rate": 5.349200221378076e-07,
      "loss": 0.3345,
      "num_input_tokens_seen": 19190944,
      "step": 29385
    },
    {
      "epoch": 17.329009433962263,
      "grad_norm": 2.060983896255493,
      "learning_rate": 5.33762827793432e-07,
      "loss": 0.2936,
      "num_input_tokens_seen": 19194912,
      "step": 29390
    },
    {
      "epoch": 17.33195754716981,
      "grad_norm": 3.851074457168579,
      "learning_rate": 5.326068159045978e-07,
      "loss": 0.3744,
      "num_input_tokens_seen": 19197568,
      "step": 29395
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 4.0384955406188965,
      "learning_rate": 5.314519867773621e-07,
      "loss": 0.2718,
      "num_input_tokens_seen": 19200480,
      "step": 29400
    },
    {
      "epoch": 17.337853773584907,
      "grad_norm": 3.2488880157470703,
      "learning_rate": 5.302983407174711e-07,
      "loss": 0.3391,
      "num_input_tokens_seen": 19203488,
      "step": 29405
    },
    {
      "epoch": 17.340801886792452,
      "grad_norm": 3.665304183959961,
      "learning_rate": 5.291458780303572e-07,
      "loss": 0.3253,
      "num_input_tokens_seen": 19206304,
      "step": 29410
    },
    {
      "epoch": 17.34375,
      "grad_norm": 3.413327217102051,
      "learning_rate": 5.279945990211411e-07,
      "loss": 0.2935,
      "num_input_tokens_seen": 19209824,
      "step": 29415
    },
    {
      "epoch": 17.346698113207548,
      "grad_norm": 1.9510127305984497,
      "learning_rate": 5.26844503994628e-07,
      "loss": 0.3444,
      "num_input_tokens_seen": 19212896,
      "step": 29420
    },
    {
      "epoch": 17.349646226415093,
      "grad_norm": 3.977092742919922,
      "learning_rate": 5.25695593255311e-07,
      "loss": 0.2683,
      "num_input_tokens_seen": 19215744,
      "step": 29425
    },
    {
      "epoch": 17.35259433962264,
      "grad_norm": 3.72910737991333,
      "learning_rate": 5.24547867107369e-07,
      "loss": 0.3787,
      "num_input_tokens_seen": 19218656,
      "step": 29430
    },
    {
      "epoch": 17.35554245283019,
      "grad_norm": 2.639416217803955,
      "learning_rate": 5.234013258546672e-07,
      "loss": 0.2473,
      "num_input_tokens_seen": 19221248,
      "step": 29435
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 1.9417554140090942,
      "learning_rate": 5.222559698007563e-07,
      "loss": 0.3577,
      "num_input_tokens_seen": 19224288,
      "step": 29440
    },
    {
      "epoch": 17.361438679245282,
      "grad_norm": 1.9709630012512207,
      "learning_rate": 5.211117992488763e-07,
      "loss": 0.2171,
      "num_input_tokens_seen": 19227648,
      "step": 29445
    },
    {
      "epoch": 17.36438679245283,
      "grad_norm": 3.098782777786255,
      "learning_rate": 5.199688145019505e-07,
      "loss": 0.325,
      "num_input_tokens_seen": 19230272,
      "step": 29450
    },
    {
      "epoch": 17.36733490566038,
      "grad_norm": 4.2355804443359375,
      "learning_rate": 5.188270158625891e-07,
      "loss": 0.3372,
      "num_input_tokens_seen": 19234336,
      "step": 29455
    },
    {
      "epoch": 17.370283018867923,
      "grad_norm": 3.3640921115875244,
      "learning_rate": 5.176864036330875e-07,
      "loss": 0.3386,
      "num_input_tokens_seen": 19237888,
      "step": 29460
    },
    {
      "epoch": 17.37323113207547,
      "grad_norm": 2.7001700401306152,
      "learning_rate": 5.165469781154287e-07,
      "loss": 0.2663,
      "num_input_tokens_seen": 19240960,
      "step": 29465
    },
    {
      "epoch": 17.37617924528302,
      "grad_norm": 3.044316291809082,
      "learning_rate": 5.154087396112789e-07,
      "loss": 0.2991,
      "num_input_tokens_seen": 19243488,
      "step": 29470
    },
    {
      "epoch": 17.379127358490567,
      "grad_norm": 6.185435771942139,
      "learning_rate": 5.142716884219939e-07,
      "loss": 0.3923,
      "num_input_tokens_seen": 19245920,
      "step": 29475
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 2.601227045059204,
      "learning_rate": 5.131358248486118e-07,
      "loss": 0.3989,
      "num_input_tokens_seen": 19249248,
      "step": 29480
    },
    {
      "epoch": 17.38502358490566,
      "grad_norm": 3.197658061981201,
      "learning_rate": 5.120011491918564e-07,
      "loss": 0.2457,
      "num_input_tokens_seen": 19252288,
      "step": 29485
    },
    {
      "epoch": 17.38797169811321,
      "grad_norm": 7.700962066650391,
      "learning_rate": 5.108676617521402e-07,
      "loss": 0.3438,
      "num_input_tokens_seen": 19254624,
      "step": 29490
    },
    {
      "epoch": 17.390919811320753,
      "grad_norm": 2.364845037460327,
      "learning_rate": 5.097353628295571e-07,
      "loss": 0.2978,
      "num_input_tokens_seen": 19258336,
      "step": 29495
    },
    {
      "epoch": 17.3938679245283,
      "grad_norm": 3.306013822555542,
      "learning_rate": 5.086042527238893e-07,
      "loss": 0.3489,
      "num_input_tokens_seen": 19261024,
      "step": 29500
    },
    {
      "epoch": 17.39681603773585,
      "grad_norm": 3.2454755306243896,
      "learning_rate": 5.074743317346009e-07,
      "loss": 0.3357,
      "num_input_tokens_seen": 19265920,
      "step": 29505
    },
    {
      "epoch": 17.399764150943398,
      "grad_norm": 3.1324334144592285,
      "learning_rate": 5.063456001608458e-07,
      "loss": 0.3009,
      "num_input_tokens_seen": 19270016,
      "step": 29510
    },
    {
      "epoch": 17.402712264150942,
      "grad_norm": 5.678059101104736,
      "learning_rate": 5.052180583014599e-07,
      "loss": 0.2722,
      "num_input_tokens_seen": 19273312,
      "step": 29515
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 4.291169166564941,
      "learning_rate": 5.04091706454965e-07,
      "loss": 0.2178,
      "num_input_tokens_seen": 19276544,
      "step": 29520
    },
    {
      "epoch": 17.40860849056604,
      "grad_norm": 3.952301025390625,
      "learning_rate": 5.029665449195665e-07,
      "loss": 0.2678,
      "num_input_tokens_seen": 19280128,
      "step": 29525
    },
    {
      "epoch": 17.411556603773583,
      "grad_norm": 5.559089660644531,
      "learning_rate": 5.018425739931559e-07,
      "loss": 0.2512,
      "num_input_tokens_seen": 19283552,
      "step": 29530
    },
    {
      "epoch": 17.41450471698113,
      "grad_norm": 3.4568560123443604,
      "learning_rate": 5.007197939733099e-07,
      "loss": 0.2208,
      "num_input_tokens_seen": 19286656,
      "step": 29535
    },
    {
      "epoch": 17.41745283018868,
      "grad_norm": 2.780898332595825,
      "learning_rate": 4.995982051572895e-07,
      "loss": 0.3966,
      "num_input_tokens_seen": 19290144,
      "step": 29540
    },
    {
      "epoch": 17.420400943396228,
      "grad_norm": 3.1781058311462402,
      "learning_rate": 4.984778078420405e-07,
      "loss": 0.2545,
      "num_input_tokens_seen": 19293568,
      "step": 29545
    },
    {
      "epoch": 17.423349056603772,
      "grad_norm": 2.927884817123413,
      "learning_rate": 4.973586023241917e-07,
      "loss": 0.2918,
      "num_input_tokens_seen": 19296832,
      "step": 29550
    },
    {
      "epoch": 17.42629716981132,
      "grad_norm": 2.9513590335845947,
      "learning_rate": 4.962405889000588e-07,
      "loss": 0.3109,
      "num_input_tokens_seen": 19301024,
      "step": 29555
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 2.6611475944519043,
      "learning_rate": 4.951237678656396e-07,
      "loss": 0.2089,
      "num_input_tokens_seen": 19304064,
      "step": 29560
    },
    {
      "epoch": 17.432193396226417,
      "grad_norm": 2.716643810272217,
      "learning_rate": 4.940081395166174e-07,
      "loss": 0.2803,
      "num_input_tokens_seen": 19306528,
      "step": 29565
    },
    {
      "epoch": 17.43514150943396,
      "grad_norm": 3.9046738147735596,
      "learning_rate": 4.928937041483606e-07,
      "loss": 0.3496,
      "num_input_tokens_seen": 19309696,
      "step": 29570
    },
    {
      "epoch": 17.43808962264151,
      "grad_norm": 1.8936545848846436,
      "learning_rate": 4.917804620559202e-07,
      "loss": 0.2569,
      "num_input_tokens_seen": 19313056,
      "step": 29575
    },
    {
      "epoch": 17.441037735849058,
      "grad_norm": 6.465855121612549,
      "learning_rate": 4.906684135340317e-07,
      "loss": 0.3396,
      "num_input_tokens_seen": 19316480,
      "step": 29580
    },
    {
      "epoch": 17.443985849056602,
      "grad_norm": 2.7241570949554443,
      "learning_rate": 4.89557558877114e-07,
      "loss": 0.3637,
      "num_input_tokens_seen": 19320576,
      "step": 29585
    },
    {
      "epoch": 17.44693396226415,
      "grad_norm": 3.088707208633423,
      "learning_rate": 4.884478983792728e-07,
      "loss": 0.403,
      "num_input_tokens_seen": 19322816,
      "step": 29590
    },
    {
      "epoch": 17.4498820754717,
      "grad_norm": 2.9661824703216553,
      "learning_rate": 4.873394323342939e-07,
      "loss": 0.2662,
      "num_input_tokens_seen": 19325696,
      "step": 29595
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 4.370572090148926,
      "learning_rate": 4.86232161035648e-07,
      "loss": 0.3617,
      "num_input_tokens_seen": 19329152,
      "step": 29600
    },
    {
      "epoch": 17.45577830188679,
      "grad_norm": 4.57067346572876,
      "learning_rate": 4.851260847764916e-07,
      "loss": 0.3673,
      "num_input_tokens_seen": 19332128,
      "step": 29605
    },
    {
      "epoch": 17.45872641509434,
      "grad_norm": 3.1103131771087646,
      "learning_rate": 4.840212038496622e-07,
      "loss": 0.276,
      "num_input_tokens_seen": 19334688,
      "step": 29610
    },
    {
      "epoch": 17.461674528301888,
      "grad_norm": 2.6761257648468018,
      "learning_rate": 4.82917518547682e-07,
      "loss": 0.4678,
      "num_input_tokens_seen": 19338976,
      "step": 29615
    },
    {
      "epoch": 17.464622641509433,
      "grad_norm": 6.781468868255615,
      "learning_rate": 4.81815029162756e-07,
      "loss": 0.2971,
      "num_input_tokens_seen": 19342208,
      "step": 29620
    },
    {
      "epoch": 17.46757075471698,
      "grad_norm": 2.391054153442383,
      "learning_rate": 4.807137359867725e-07,
      "loss": 0.3562,
      "num_input_tokens_seen": 19345760,
      "step": 29625
    },
    {
      "epoch": 17.47051886792453,
      "grad_norm": 3.5583009719848633,
      "learning_rate": 4.79613639311306e-07,
      "loss": 0.2786,
      "num_input_tokens_seen": 19348064,
      "step": 29630
    },
    {
      "epoch": 17.473466981132077,
      "grad_norm": 3.0280792713165283,
      "learning_rate": 4.785147394276096e-07,
      "loss": 0.2818,
      "num_input_tokens_seen": 19350784,
      "step": 29635
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 8.30457592010498,
      "learning_rate": 4.774170366266223e-07,
      "loss": 0.3574,
      "num_input_tokens_seen": 19354464,
      "step": 29640
    },
    {
      "epoch": 17.47936320754717,
      "grad_norm": 2.916675329208374,
      "learning_rate": 4.763205311989666e-07,
      "loss": 0.2985,
      "num_input_tokens_seen": 19357888,
      "step": 29645
    },
    {
      "epoch": 17.482311320754718,
      "grad_norm": 3.396804094314575,
      "learning_rate": 4.752252234349458e-07,
      "loss": 0.2347,
      "num_input_tokens_seen": 19361024,
      "step": 29650
    },
    {
      "epoch": 17.485259433962263,
      "grad_norm": 3.51948881149292,
      "learning_rate": 4.7413111362454634e-07,
      "loss": 0.2759,
      "num_input_tokens_seen": 19364000,
      "step": 29655
    },
    {
      "epoch": 17.48820754716981,
      "grad_norm": 1.9789639711380005,
      "learning_rate": 4.7303820205744143e-07,
      "loss": 0.1839,
      "num_input_tokens_seen": 19366560,
      "step": 29660
    },
    {
      "epoch": 17.49115566037736,
      "grad_norm": 2.5407049655914307,
      "learning_rate": 4.7194648902298303e-07,
      "loss": 0.3224,
      "num_input_tokens_seen": 19369760,
      "step": 29665
    },
    {
      "epoch": 17.494103773584907,
      "grad_norm": 4.91910982131958,
      "learning_rate": 4.7085597481020594e-07,
      "loss": 0.4899,
      "num_input_tokens_seen": 19372512,
      "step": 29670
    },
    {
      "epoch": 17.497051886792452,
      "grad_norm": 2.899507522583008,
      "learning_rate": 4.697666597078293e-07,
      "loss": 0.2757,
      "num_input_tokens_seen": 19375232,
      "step": 29675
    },
    {
      "epoch": 17.5,
      "grad_norm": 2.9302616119384766,
      "learning_rate": 4.6867854400425237e-07,
      "loss": 0.3095,
      "num_input_tokens_seen": 19378944,
      "step": 29680
    },
    {
      "epoch": 17.502948113207548,
      "grad_norm": 4.964863300323486,
      "learning_rate": 4.6759162798756084e-07,
      "loss": 0.3216,
      "num_input_tokens_seen": 19381504,
      "step": 29685
    },
    {
      "epoch": 17.505896226415093,
      "grad_norm": 8.532380104064941,
      "learning_rate": 4.6650591194551895e-07,
      "loss": 0.3575,
      "num_input_tokens_seen": 19384320,
      "step": 29690
    },
    {
      "epoch": 17.50884433962264,
      "grad_norm": 3.8976852893829346,
      "learning_rate": 4.654213961655757e-07,
      "loss": 0.2963,
      "num_input_tokens_seen": 19387296,
      "step": 29695
    },
    {
      "epoch": 17.51179245283019,
      "grad_norm": 3.563373327255249,
      "learning_rate": 4.6433808093486075e-07,
      "loss": 0.3031,
      "num_input_tokens_seen": 19389920,
      "step": 29700
    },
    {
      "epoch": 17.514740566037737,
      "grad_norm": 4.164343357086182,
      "learning_rate": 4.63255966540187e-07,
      "loss": 0.2419,
      "num_input_tokens_seen": 19394080,
      "step": 29705
    },
    {
      "epoch": 17.517688679245282,
      "grad_norm": 2.4304306507110596,
      "learning_rate": 4.62175053268048e-07,
      "loss": 0.2198,
      "num_input_tokens_seen": 19397440,
      "step": 29710
    },
    {
      "epoch": 17.52063679245283,
      "grad_norm": 5.741192817687988,
      "learning_rate": 4.6109534140462045e-07,
      "loss": 0.3063,
      "num_input_tokens_seen": 19399872,
      "step": 29715
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 2.8868067264556885,
      "learning_rate": 4.6001683123576226e-07,
      "loss": 0.2276,
      "num_input_tokens_seen": 19402944,
      "step": 29720
    },
    {
      "epoch": 17.526533018867923,
      "grad_norm": 3.6887872219085693,
      "learning_rate": 4.589395230470145e-07,
      "loss": 0.3125,
      "num_input_tokens_seen": 19406560,
      "step": 29725
    },
    {
      "epoch": 17.52948113207547,
      "grad_norm": 3.2998878955841064,
      "learning_rate": 4.578634171235996e-07,
      "loss": 0.3198,
      "num_input_tokens_seen": 19410400,
      "step": 29730
    },
    {
      "epoch": 17.53242924528302,
      "grad_norm": 3.6902029514312744,
      "learning_rate": 4.567885137504202e-07,
      "loss": 0.4615,
      "num_input_tokens_seen": 19413632,
      "step": 29735
    },
    {
      "epoch": 17.535377358490567,
      "grad_norm": 3.0427253246307373,
      "learning_rate": 4.55714813212062e-07,
      "loss": 0.2521,
      "num_input_tokens_seen": 19417056,
      "step": 29740
    },
    {
      "epoch": 17.538325471698112,
      "grad_norm": 3.900834798812866,
      "learning_rate": 4.5464231579279206e-07,
      "loss": 0.3495,
      "num_input_tokens_seen": 19421184,
      "step": 29745
    },
    {
      "epoch": 17.54127358490566,
      "grad_norm": 3.8062846660614014,
      "learning_rate": 4.535710217765571e-07,
      "loss": 0.289,
      "num_input_tokens_seen": 19424416,
      "step": 29750
    },
    {
      "epoch": 17.54422169811321,
      "grad_norm": 2.544437885284424,
      "learning_rate": 4.5250093144698913e-07,
      "loss": 0.2687,
      "num_input_tokens_seen": 19427776,
      "step": 29755
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 2.2115142345428467,
      "learning_rate": 4.514320450873988e-07,
      "loss": 0.2935,
      "num_input_tokens_seen": 19431136,
      "step": 29760
    },
    {
      "epoch": 17.5501179245283,
      "grad_norm": 4.99397611618042,
      "learning_rate": 4.503643629807769e-07,
      "loss": 0.2508,
      "num_input_tokens_seen": 19433984,
      "step": 29765
    },
    {
      "epoch": 17.55306603773585,
      "grad_norm": 3.9469985961914062,
      "learning_rate": 4.4929788540979844e-07,
      "loss": 0.3681,
      "num_input_tokens_seen": 19437216,
      "step": 29770
    },
    {
      "epoch": 17.556014150943398,
      "grad_norm": 6.34192419052124,
      "learning_rate": 4.4823261265681596e-07,
      "loss": 0.3012,
      "num_input_tokens_seen": 19440800,
      "step": 29775
    },
    {
      "epoch": 17.558962264150942,
      "grad_norm": 4.604106903076172,
      "learning_rate": 4.471685450038671e-07,
      "loss": 0.2798,
      "num_input_tokens_seen": 19443648,
      "step": 29780
    },
    {
      "epoch": 17.56191037735849,
      "grad_norm": 5.82126522064209,
      "learning_rate": 4.4610568273266706e-07,
      "loss": 0.3886,
      "num_input_tokens_seen": 19446048,
      "step": 29785
    },
    {
      "epoch": 17.56485849056604,
      "grad_norm": 3.5019474029541016,
      "learning_rate": 4.450440261246142e-07,
      "loss": 0.2551,
      "num_input_tokens_seen": 19448512,
      "step": 29790
    },
    {
      "epoch": 17.567806603773583,
      "grad_norm": 3.4061713218688965,
      "learning_rate": 4.439835754607863e-07,
      "loss": 0.4153,
      "num_input_tokens_seen": 19452320,
      "step": 29795
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 3.8713648319244385,
      "learning_rate": 4.429243310219422e-07,
      "loss": 0.5153,
      "num_input_tokens_seen": 19455360,
      "step": 29800
    },
    {
      "epoch": 17.57370283018868,
      "grad_norm": 3.1306111812591553,
      "learning_rate": 4.418662930885215e-07,
      "loss": 0.5099,
      "num_input_tokens_seen": 19459360,
      "step": 29805
    },
    {
      "epoch": 17.576650943396228,
      "grad_norm": 4.285674095153809,
      "learning_rate": 4.408094619406439e-07,
      "loss": 0.3265,
      "num_input_tokens_seen": 19462208,
      "step": 29810
    },
    {
      "epoch": 17.579599056603772,
      "grad_norm": 14.057676315307617,
      "learning_rate": 4.3975383785810954e-07,
      "loss": 0.5465,
      "num_input_tokens_seen": 19465088,
      "step": 29815
    },
    {
      "epoch": 17.58254716981132,
      "grad_norm": 5.473825454711914,
      "learning_rate": 4.3869942112040096e-07,
      "loss": 0.2682,
      "num_input_tokens_seen": 19467808,
      "step": 29820
    },
    {
      "epoch": 17.58549528301887,
      "grad_norm": 6.082172870635986,
      "learning_rate": 4.3764621200667936e-07,
      "loss": 0.3363,
      "num_input_tokens_seen": 19472192,
      "step": 29825
    },
    {
      "epoch": 17.588443396226417,
      "grad_norm": 4.562597274780273,
      "learning_rate": 4.365942107957849e-07,
      "loss": 0.329,
      "num_input_tokens_seen": 19475456,
      "step": 29830
    },
    {
      "epoch": 17.59139150943396,
      "grad_norm": 5.717051982879639,
      "learning_rate": 4.35543417766241e-07,
      "loss": 0.3206,
      "num_input_tokens_seen": 19478912,
      "step": 29835
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 3.213571548461914,
      "learning_rate": 4.3449383319624785e-07,
      "loss": 0.2454,
      "num_input_tokens_seen": 19482112,
      "step": 29840
    },
    {
      "epoch": 17.597287735849058,
      "grad_norm": 2.5604331493377686,
      "learning_rate": 4.3344545736368926e-07,
      "loss": 0.2672,
      "num_input_tokens_seen": 19485152,
      "step": 29845
    },
    {
      "epoch": 17.600235849056602,
      "grad_norm": 4.89127254486084,
      "learning_rate": 4.323982905461266e-07,
      "loss": 0.3648,
      "num_input_tokens_seen": 19488480,
      "step": 29850
    },
    {
      "epoch": 17.60318396226415,
      "grad_norm": 5.183581352233887,
      "learning_rate": 4.313523330208019e-07,
      "loss": 0.3012,
      "num_input_tokens_seen": 19492512,
      "step": 29855
    },
    {
      "epoch": 17.6061320754717,
      "grad_norm": 3.9073667526245117,
      "learning_rate": 4.303075850646371e-07,
      "loss": 0.3482,
      "num_input_tokens_seen": 19495712,
      "step": 29860
    },
    {
      "epoch": 17.609080188679247,
      "grad_norm": 3.079094171524048,
      "learning_rate": 4.2926404695423305e-07,
      "loss": 0.3358,
      "num_input_tokens_seen": 19498624,
      "step": 29865
    },
    {
      "epoch": 17.61202830188679,
      "grad_norm": 3.158282518386841,
      "learning_rate": 4.282217189658705e-07,
      "loss": 0.2368,
      "num_input_tokens_seen": 19501952,
      "step": 29870
    },
    {
      "epoch": 17.61497641509434,
      "grad_norm": 3.3918397426605225,
      "learning_rate": 4.27180601375512e-07,
      "loss": 0.2762,
      "num_input_tokens_seen": 19505312,
      "step": 29875
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 3.518958568572998,
      "learning_rate": 4.2614069445879646e-07,
      "loss": 0.2716,
      "num_input_tokens_seen": 19507616,
      "step": 29880
    },
    {
      "epoch": 17.620872641509433,
      "grad_norm": 3.4784576892852783,
      "learning_rate": 4.251019984910448e-07,
      "loss": 0.4382,
      "num_input_tokens_seen": 19511008,
      "step": 29885
    },
    {
      "epoch": 17.62382075471698,
      "grad_norm": 3.4126553535461426,
      "learning_rate": 4.2406451374725597e-07,
      "loss": 0.2859,
      "num_input_tokens_seen": 19513920,
      "step": 29890
    },
    {
      "epoch": 17.62676886792453,
      "grad_norm": 6.6703996658325195,
      "learning_rate": 4.2302824050210855e-07,
      "loss": 0.4332,
      "num_input_tokens_seen": 19517184,
      "step": 29895
    },
    {
      "epoch": 17.629716981132077,
      "grad_norm": 1.9942600727081299,
      "learning_rate": 4.2199317902995974e-07,
      "loss": 0.3753,
      "num_input_tokens_seen": 19521024,
      "step": 29900
    },
    {
      "epoch": 17.63266509433962,
      "grad_norm": 4.006063938140869,
      "learning_rate": 4.209593296048459e-07,
      "loss": 0.3414,
      "num_input_tokens_seen": 19524256,
      "step": 29905
    },
    {
      "epoch": 17.63561320754717,
      "grad_norm": 3.7132651805877686,
      "learning_rate": 4.1992669250048524e-07,
      "loss": 0.3699,
      "num_input_tokens_seen": 19527584,
      "step": 29910
    },
    {
      "epoch": 17.638561320754718,
      "grad_norm": 3.022113800048828,
      "learning_rate": 4.188952679902719e-07,
      "loss": 0.3248,
      "num_input_tokens_seen": 19530208,
      "step": 29915
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 3.1405909061431885,
      "learning_rate": 4.178650563472797e-07,
      "loss": 0.2684,
      "num_input_tokens_seen": 19533568,
      "step": 29920
    },
    {
      "epoch": 17.64445754716981,
      "grad_norm": 2.9855844974517822,
      "learning_rate": 4.168360578442615e-07,
      "loss": 0.3172,
      "num_input_tokens_seen": 19537568,
      "step": 29925
    },
    {
      "epoch": 17.64740566037736,
      "grad_norm": 2.727465867996216,
      "learning_rate": 4.1580827275365e-07,
      "loss": 0.3552,
      "num_input_tokens_seen": 19540896,
      "step": 29930
    },
    {
      "epoch": 17.650353773584907,
      "grad_norm": 3.1367173194885254,
      "learning_rate": 4.147817013475536e-07,
      "loss": 0.3193,
      "num_input_tokens_seen": 19544128,
      "step": 29935
    },
    {
      "epoch": 17.653301886792452,
      "grad_norm": 3.1232552528381348,
      "learning_rate": 4.1375634389776375e-07,
      "loss": 0.3569,
      "num_input_tokens_seen": 19546720,
      "step": 29940
    },
    {
      "epoch": 17.65625,
      "grad_norm": 4.470157623291016,
      "learning_rate": 4.127322006757478e-07,
      "loss": 0.3128,
      "num_input_tokens_seen": 19549952,
      "step": 29945
    },
    {
      "epoch": 17.659198113207548,
      "grad_norm": 1.889742136001587,
      "learning_rate": 4.1170927195265163e-07,
      "loss": 0.3821,
      "num_input_tokens_seen": 19553472,
      "step": 29950
    },
    {
      "epoch": 17.662146226415093,
      "grad_norm": 3.0309741497039795,
      "learning_rate": 4.1068755799930026e-07,
      "loss": 0.3398,
      "num_input_tokens_seen": 19557600,
      "step": 29955
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 3.404571533203125,
      "learning_rate": 4.096670590861962e-07,
      "loss": 0.4887,
      "num_input_tokens_seen": 19561312,
      "step": 29960
    },
    {
      "epoch": 17.66804245283019,
      "grad_norm": 4.28385591506958,
      "learning_rate": 4.086477754835211e-07,
      "loss": 0.3417,
      "num_input_tokens_seen": 19565024,
      "step": 29965
    },
    {
      "epoch": 17.670990566037737,
      "grad_norm": 5.5344109535217285,
      "learning_rate": 4.0762970746113517e-07,
      "loss": 0.2805,
      "num_input_tokens_seen": 19568352,
      "step": 29970
    },
    {
      "epoch": 17.673938679245282,
      "grad_norm": 4.14816427230835,
      "learning_rate": 4.0661285528857676e-07,
      "loss": 0.4544,
      "num_input_tokens_seen": 19571200,
      "step": 29975
    },
    {
      "epoch": 17.67688679245283,
      "grad_norm": 4.6382646560668945,
      "learning_rate": 4.0559721923506155e-07,
      "loss": 0.3481,
      "num_input_tokens_seen": 19574432,
      "step": 29980
    },
    {
      "epoch": 17.67983490566038,
      "grad_norm": 4.96541690826416,
      "learning_rate": 4.045827995694834e-07,
      "loss": 0.2953,
      "num_input_tokens_seen": 19577120,
      "step": 29985
    },
    {
      "epoch": 17.682783018867923,
      "grad_norm": 2.986609697341919,
      "learning_rate": 4.035695965604142e-07,
      "loss": 0.3214,
      "num_input_tokens_seen": 19580384,
      "step": 29990
    },
    {
      "epoch": 17.68573113207547,
      "grad_norm": 3.8126585483551025,
      "learning_rate": 4.0255761047610365e-07,
      "loss": 0.6294,
      "num_input_tokens_seen": 19583808,
      "step": 29995
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 3.144735097885132,
      "learning_rate": 4.0154684158447864e-07,
      "loss": 0.3853,
      "num_input_tokens_seen": 19591552,
      "step": 30000
    },
    {
      "epoch": 17.691627358490567,
      "grad_norm": 5.623338222503662,
      "learning_rate": 4.0053729015314623e-07,
      "loss": 0.3203,
      "num_input_tokens_seen": 19593920,
      "step": 30005
    },
    {
      "epoch": 17.694575471698112,
      "grad_norm": 1.8775466680526733,
      "learning_rate": 3.9952895644938926e-07,
      "loss": 0.2216,
      "num_input_tokens_seen": 19598016,
      "step": 30010
    },
    {
      "epoch": 17.69752358490566,
      "grad_norm": 2.5288779735565186,
      "learning_rate": 3.985218407401681e-07,
      "loss": 0.2292,
      "num_input_tokens_seen": 19600608,
      "step": 30015
    },
    {
      "epoch": 17.70047169811321,
      "grad_norm": 10.68576431274414,
      "learning_rate": 3.975159432921205e-07,
      "loss": 0.2007,
      "num_input_tokens_seen": 19603488,
      "step": 30020
    },
    {
      "epoch": 17.703419811320753,
      "grad_norm": 3.746105194091797,
      "learning_rate": 3.9651126437156294e-07,
      "loss": 0.313,
      "num_input_tokens_seen": 19606368,
      "step": 30025
    },
    {
      "epoch": 17.7063679245283,
      "grad_norm": 6.187250137329102,
      "learning_rate": 3.9550780424448653e-07,
      "loss": 0.3572,
      "num_input_tokens_seen": 19610080,
      "step": 30030
    },
    {
      "epoch": 17.70931603773585,
      "grad_norm": 2.729196071624756,
      "learning_rate": 3.9450556317656487e-07,
      "loss": 0.1795,
      "num_input_tokens_seen": 19612576,
      "step": 30035
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 5.997259140014648,
      "learning_rate": 3.935045414331434e-07,
      "loss": 0.3003,
      "num_input_tokens_seen": 19615968,
      "step": 30040
    },
    {
      "epoch": 17.715212264150942,
      "grad_norm": 4.649085521697998,
      "learning_rate": 3.925047392792475e-07,
      "loss": 0.3742,
      "num_input_tokens_seen": 19618656,
      "step": 30045
    },
    {
      "epoch": 17.71816037735849,
      "grad_norm": 2.6150035858154297,
      "learning_rate": 3.9150615697957917e-07,
      "loss": 0.2224,
      "num_input_tokens_seen": 19622912,
      "step": 30050
    },
    {
      "epoch": 17.72110849056604,
      "grad_norm": 3.4615466594696045,
      "learning_rate": 3.9050879479851753e-07,
      "loss": 0.3176,
      "num_input_tokens_seen": 19626304,
      "step": 30055
    },
    {
      "epoch": 17.724056603773583,
      "grad_norm": 3.5815584659576416,
      "learning_rate": 3.89512653000117e-07,
      "loss": 0.3306,
      "num_input_tokens_seen": 19629856,
      "step": 30060
    },
    {
      "epoch": 17.72700471698113,
      "grad_norm": 4.846798896789551,
      "learning_rate": 3.8851773184811203e-07,
      "loss": 0.2558,
      "num_input_tokens_seen": 19633312,
      "step": 30065
    },
    {
      "epoch": 17.72995283018868,
      "grad_norm": 3.0845017433166504,
      "learning_rate": 3.8752403160591255e-07,
      "loss": 0.2734,
      "num_input_tokens_seen": 19635648,
      "step": 30070
    },
    {
      "epoch": 17.732900943396228,
      "grad_norm": 2.793621063232422,
      "learning_rate": 3.8653155253660477e-07,
      "loss": 0.2945,
      "num_input_tokens_seen": 19638656,
      "step": 30075
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 4.692408084869385,
      "learning_rate": 3.8554029490295073e-07,
      "loss": 0.2603,
      "num_input_tokens_seen": 19642592,
      "step": 30080
    },
    {
      "epoch": 17.73879716981132,
      "grad_norm": 4.14473295211792,
      "learning_rate": 3.8455025896739164e-07,
      "loss": 0.2582,
      "num_input_tokens_seen": 19646080,
      "step": 30085
    },
    {
      "epoch": 17.74174528301887,
      "grad_norm": 3.0899722576141357,
      "learning_rate": 3.8356144499204215e-07,
      "loss": 0.2699,
      "num_input_tokens_seen": 19649344,
      "step": 30090
    },
    {
      "epoch": 17.744693396226417,
      "grad_norm": 3.778785467147827,
      "learning_rate": 3.8257385323869576e-07,
      "loss": 0.2482,
      "num_input_tokens_seen": 19651840,
      "step": 30095
    },
    {
      "epoch": 17.74764150943396,
      "grad_norm": 3.7867870330810547,
      "learning_rate": 3.815874839688222e-07,
      "loss": 0.3624,
      "num_input_tokens_seen": 19655232,
      "step": 30100
    },
    {
      "epoch": 17.75058962264151,
      "grad_norm": 3.42706036567688,
      "learning_rate": 3.8060233744356634e-07,
      "loss": 0.3745,
      "num_input_tokens_seen": 19658400,
      "step": 30105
    },
    {
      "epoch": 17.753537735849058,
      "grad_norm": 5.567670822143555,
      "learning_rate": 3.796184139237502e-07,
      "loss": 0.428,
      "num_input_tokens_seen": 19661024,
      "step": 30110
    },
    {
      "epoch": 17.756485849056602,
      "grad_norm": 3.7278058528900146,
      "learning_rate": 3.7863571366987206e-07,
      "loss": 0.3251,
      "num_input_tokens_seen": 19665152,
      "step": 30115
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 2.46748948097229,
      "learning_rate": 3.776542369421049e-07,
      "loss": 0.2748,
      "num_input_tokens_seen": 19669600,
      "step": 30120
    },
    {
      "epoch": 17.7623820754717,
      "grad_norm": 2.3352513313293457,
      "learning_rate": 3.766739840003003e-07,
      "loss": 0.1878,
      "num_input_tokens_seen": 19680032,
      "step": 30125
    },
    {
      "epoch": 17.765330188679247,
      "grad_norm": 4.2295308113098145,
      "learning_rate": 3.756949551039835e-07,
      "loss": 0.3057,
      "num_input_tokens_seen": 19682848,
      "step": 30130
    },
    {
      "epoch": 17.76827830188679,
      "grad_norm": 3.4606220722198486,
      "learning_rate": 3.7471715051235757e-07,
      "loss": 0.2786,
      "num_input_tokens_seen": 19686240,
      "step": 30135
    },
    {
      "epoch": 17.77122641509434,
      "grad_norm": 5.799755096435547,
      "learning_rate": 3.7374057048429947e-07,
      "loss": 0.3743,
      "num_input_tokens_seen": 19689408,
      "step": 30140
    },
    {
      "epoch": 17.774174528301888,
      "grad_norm": 4.976673603057861,
      "learning_rate": 3.7276521527836396e-07,
      "loss": 0.3289,
      "num_input_tokens_seen": 19693056,
      "step": 30145
    },
    {
      "epoch": 17.777122641509433,
      "grad_norm": 2.9047014713287354,
      "learning_rate": 3.717910851527784e-07,
      "loss": 0.3671,
      "num_input_tokens_seen": 19695776,
      "step": 30150
    },
    {
      "epoch": 17.78007075471698,
      "grad_norm": 3.222412586212158,
      "learning_rate": 3.708181803654498e-07,
      "loss": 0.3272,
      "num_input_tokens_seen": 19698656,
      "step": 30155
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 3.5740814208984375,
      "learning_rate": 3.6984650117395993e-07,
      "loss": 0.3367,
      "num_input_tokens_seen": 19701664,
      "step": 30160
    },
    {
      "epoch": 17.785966981132077,
      "grad_norm": 7.154371738433838,
      "learning_rate": 3.688760478355635e-07,
      "loss": 0.3525,
      "num_input_tokens_seen": 19705472,
      "step": 30165
    },
    {
      "epoch": 17.78891509433962,
      "grad_norm": 3.691513776779175,
      "learning_rate": 3.679068206071923e-07,
      "loss": 0.4371,
      "num_input_tokens_seen": 19708160,
      "step": 30170
    },
    {
      "epoch": 17.79186320754717,
      "grad_norm": 1.8336999416351318,
      "learning_rate": 3.669388197454532e-07,
      "loss": 0.2456,
      "num_input_tokens_seen": 19711232,
      "step": 30175
    },
    {
      "epoch": 17.794811320754718,
      "grad_norm": 3.8604328632354736,
      "learning_rate": 3.6597204550662956e-07,
      "loss": 0.3837,
      "num_input_tokens_seen": 19714752,
      "step": 30180
    },
    {
      "epoch": 17.797759433962263,
      "grad_norm": 4.275971412658691,
      "learning_rate": 3.650064981466772e-07,
      "loss": 0.3259,
      "num_input_tokens_seen": 19717536,
      "step": 30185
    },
    {
      "epoch": 17.80070754716981,
      "grad_norm": 3.8279623985290527,
      "learning_rate": 3.640421779212311e-07,
      "loss": 0.3032,
      "num_input_tokens_seen": 19721664,
      "step": 30190
    },
    {
      "epoch": 17.80365566037736,
      "grad_norm": 4.411131858825684,
      "learning_rate": 3.630790850855986e-07,
      "loss": 0.3544,
      "num_input_tokens_seen": 19725760,
      "step": 30195
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 3.0097692012786865,
      "learning_rate": 3.62117219894762e-07,
      "loss": 0.2845,
      "num_input_tokens_seen": 19728960,
      "step": 30200
    },
    {
      "epoch": 17.809551886792452,
      "grad_norm": 4.018977642059326,
      "learning_rate": 3.611565826033797e-07,
      "loss": 0.2351,
      "num_input_tokens_seen": 19732960,
      "step": 30205
    },
    {
      "epoch": 17.8125,
      "grad_norm": 5.401857376098633,
      "learning_rate": 3.6019717346578445e-07,
      "loss": 0.3109,
      "num_input_tokens_seen": 19736416,
      "step": 30210
    },
    {
      "epoch": 17.815448113207548,
      "grad_norm": 2.6173548698425293,
      "learning_rate": 3.5923899273598293e-07,
      "loss": 0.3325,
      "num_input_tokens_seen": 19739264,
      "step": 30215
    },
    {
      "epoch": 17.818396226415093,
      "grad_norm": 3.3969335556030273,
      "learning_rate": 3.582820406676596e-07,
      "loss": 0.2095,
      "num_input_tokens_seen": 19742272,
      "step": 30220
    },
    {
      "epoch": 17.82134433962264,
      "grad_norm": 3.4151036739349365,
      "learning_rate": 3.5732631751417056e-07,
      "loss": 0.2913,
      "num_input_tokens_seen": 19745664,
      "step": 30225
    },
    {
      "epoch": 17.82429245283019,
      "grad_norm": 2.986926794052124,
      "learning_rate": 3.563718235285485e-07,
      "loss": 0.4025,
      "num_input_tokens_seen": 19748736,
      "step": 30230
    },
    {
      "epoch": 17.827240566037737,
      "grad_norm": 2.944758653640747,
      "learning_rate": 3.5541855896349844e-07,
      "loss": 0.2799,
      "num_input_tokens_seen": 19751616,
      "step": 30235
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 1.9186732769012451,
      "learning_rate": 3.544665240714018e-07,
      "loss": 0.257,
      "num_input_tokens_seen": 19756736,
      "step": 30240
    },
    {
      "epoch": 17.83313679245283,
      "grad_norm": 2.595649480819702,
      "learning_rate": 3.535157191043137e-07,
      "loss": 0.3217,
      "num_input_tokens_seen": 19760928,
      "step": 30245
    },
    {
      "epoch": 17.83608490566038,
      "grad_norm": 3.1745142936706543,
      "learning_rate": 3.5256614431396385e-07,
      "loss": 0.2695,
      "num_input_tokens_seen": 19763488,
      "step": 30250
    },
    {
      "epoch": 17.839033018867923,
      "grad_norm": 2.7658450603485107,
      "learning_rate": 3.516177999517578e-07,
      "loss": 0.3027,
      "num_input_tokens_seen": 19766560,
      "step": 30255
    },
    {
      "epoch": 17.84198113207547,
      "grad_norm": 4.009328842163086,
      "learning_rate": 3.50670686268772e-07,
      "loss": 0.3724,
      "num_input_tokens_seen": 19768992,
      "step": 30260
    },
    {
      "epoch": 17.84492924528302,
      "grad_norm": 3.368565559387207,
      "learning_rate": 3.497248035157602e-07,
      "loss": 0.323,
      "num_input_tokens_seen": 19771840,
      "step": 30265
    },
    {
      "epoch": 17.847877358490567,
      "grad_norm": 4.882338047027588,
      "learning_rate": 3.4878015194314773e-07,
      "loss": 0.4384,
      "num_input_tokens_seen": 19775232,
      "step": 30270
    },
    {
      "epoch": 17.850825471698112,
      "grad_norm": 2.3811850547790527,
      "learning_rate": 3.4783673180103617e-07,
      "loss": 0.3243,
      "num_input_tokens_seen": 19778400,
      "step": 30275
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 3.2734127044677734,
      "learning_rate": 3.468945433391985e-07,
      "loss": 0.2607,
      "num_input_tokens_seen": 19781088,
      "step": 30280
    },
    {
      "epoch": 17.85672169811321,
      "grad_norm": 7.587489128112793,
      "learning_rate": 3.459535868070851e-07,
      "loss": 0.3813,
      "num_input_tokens_seen": 19784224,
      "step": 30285
    },
    {
      "epoch": 17.859669811320753,
      "grad_norm": 4.9281158447265625,
      "learning_rate": 3.450138624538174e-07,
      "loss": 0.3185,
      "num_input_tokens_seen": 19787296,
      "step": 30290
    },
    {
      "epoch": 17.8626179245283,
      "grad_norm": 5.421130180358887,
      "learning_rate": 3.440753705281913e-07,
      "loss": 0.3649,
      "num_input_tokens_seen": 19789792,
      "step": 30295
    },
    {
      "epoch": 17.86556603773585,
      "grad_norm": 3.2833364009857178,
      "learning_rate": 3.4313811127867693e-07,
      "loss": 0.2799,
      "num_input_tokens_seen": 19792928,
      "step": 30300
    },
    {
      "epoch": 17.868514150943398,
      "grad_norm": 4.870251655578613,
      "learning_rate": 3.4220208495341745e-07,
      "loss": 0.3607,
      "num_input_tokens_seen": 19797056,
      "step": 30305
    },
    {
      "epoch": 17.871462264150942,
      "grad_norm": 6.219728469848633,
      "learning_rate": 3.412672918002291e-07,
      "loss": 0.2971,
      "num_input_tokens_seen": 19800224,
      "step": 30310
    },
    {
      "epoch": 17.87441037735849,
      "grad_norm": 6.017457485198975,
      "learning_rate": 3.403337320666045e-07,
      "loss": 0.4569,
      "num_input_tokens_seen": 19803296,
      "step": 30315
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 3.2538769245147705,
      "learning_rate": 3.394014059997064e-07,
      "loss": 0.3066,
      "num_input_tokens_seen": 19806848,
      "step": 30320
    },
    {
      "epoch": 17.880306603773583,
      "grad_norm": 3.7382187843322754,
      "learning_rate": 3.3847031384637185e-07,
      "loss": 0.2985,
      "num_input_tokens_seen": 19810336,
      "step": 30325
    },
    {
      "epoch": 17.88325471698113,
      "grad_norm": 5.775665283203125,
      "learning_rate": 3.3754045585311147e-07,
      "loss": 0.2634,
      "num_input_tokens_seen": 19813056,
      "step": 30330
    },
    {
      "epoch": 17.88620283018868,
      "grad_norm": 3.7589550018310547,
      "learning_rate": 3.366118322661094e-07,
      "loss": 0.2795,
      "num_input_tokens_seen": 19815840,
      "step": 30335
    },
    {
      "epoch": 17.889150943396228,
      "grad_norm": 3.2709062099456787,
      "learning_rate": 3.3568444333122283e-07,
      "loss": 0.2272,
      "num_input_tokens_seen": 19819392,
      "step": 30340
    },
    {
      "epoch": 17.892099056603772,
      "grad_norm": 4.220242023468018,
      "learning_rate": 3.347582892939816e-07,
      "loss": 0.4272,
      "num_input_tokens_seen": 19822624,
      "step": 30345
    },
    {
      "epoch": 17.89504716981132,
      "grad_norm": 2.468141794204712,
      "learning_rate": 3.338333703995905e-07,
      "loss": 0.2899,
      "num_input_tokens_seen": 19826624,
      "step": 30350
    },
    {
      "epoch": 17.89799528301887,
      "grad_norm": 3.433145523071289,
      "learning_rate": 3.329096868929238e-07,
      "loss": 0.31,
      "num_input_tokens_seen": 19829760,
      "step": 30355
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 5.156405925750732,
      "learning_rate": 3.319872390185325e-07,
      "loss": 0.4022,
      "num_input_tokens_seen": 19833152,
      "step": 30360
    },
    {
      "epoch": 17.90389150943396,
      "grad_norm": 4.258214473724365,
      "learning_rate": 3.3106602702063727e-07,
      "loss": 0.2819,
      "num_input_tokens_seen": 19835616,
      "step": 30365
    },
    {
      "epoch": 17.90683962264151,
      "grad_norm": 4.063196182250977,
      "learning_rate": 3.3014605114313316e-07,
      "loss": 0.3691,
      "num_input_tokens_seen": 19838720,
      "step": 30370
    },
    {
      "epoch": 17.909787735849058,
      "grad_norm": 3.075004816055298,
      "learning_rate": 3.2922731162958744e-07,
      "loss": 0.4173,
      "num_input_tokens_seen": 19842208,
      "step": 30375
    },
    {
      "epoch": 17.912735849056602,
      "grad_norm": 3.941873550415039,
      "learning_rate": 3.2830980872324114e-07,
      "loss": 0.3027,
      "num_input_tokens_seen": 19844672,
      "step": 30380
    },
    {
      "epoch": 17.91568396226415,
      "grad_norm": 2.580432653427124,
      "learning_rate": 3.2739354266700775e-07,
      "loss": 0.2816,
      "num_input_tokens_seen": 19848416,
      "step": 30385
    },
    {
      "epoch": 17.9186320754717,
      "grad_norm": 5.319058895111084,
      "learning_rate": 3.264785137034709e-07,
      "loss": 0.3518,
      "num_input_tokens_seen": 19851232,
      "step": 30390
    },
    {
      "epoch": 17.921580188679247,
      "grad_norm": 2.695505380630493,
      "learning_rate": 3.2556472207488977e-07,
      "loss": 0.302,
      "num_input_tokens_seen": 19854368,
      "step": 30395
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 3.143843650817871,
      "learning_rate": 3.246521680231934e-07,
      "loss": 0.2573,
      "num_input_tokens_seen": 19856992,
      "step": 30400
    },
    {
      "epoch": 17.92747641509434,
      "grad_norm": 2.450052261352539,
      "learning_rate": 3.2374085178998594e-07,
      "loss": 0.2754,
      "num_input_tokens_seen": 19860160,
      "step": 30405
    },
    {
      "epoch": 17.930424528301888,
      "grad_norm": 2.699993848800659,
      "learning_rate": 3.2283077361654145e-07,
      "loss": 0.362,
      "num_input_tokens_seen": 19862912,
      "step": 30410
    },
    {
      "epoch": 17.933372641509433,
      "grad_norm": 3.2916910648345947,
      "learning_rate": 3.2192193374380677e-07,
      "loss": 0.3415,
      "num_input_tokens_seen": 19865952,
      "step": 30415
    },
    {
      "epoch": 17.93632075471698,
      "grad_norm": 3.782191753387451,
      "learning_rate": 3.210143324124021e-07,
      "loss": 0.3503,
      "num_input_tokens_seen": 19868928,
      "step": 30420
    },
    {
      "epoch": 17.93926886792453,
      "grad_norm": 4.9786152839660645,
      "learning_rate": 3.2010796986261805e-07,
      "loss": 0.3415,
      "num_input_tokens_seen": 19872576,
      "step": 30425
    },
    {
      "epoch": 17.942216981132077,
      "grad_norm": 2.926593780517578,
      "learning_rate": 3.1920284633441713e-07,
      "loss": 0.2926,
      "num_input_tokens_seen": 19875456,
      "step": 30430
    },
    {
      "epoch": 17.94516509433962,
      "grad_norm": 4.831915378570557,
      "learning_rate": 3.1829896206743704e-07,
      "loss": 0.2803,
      "num_input_tokens_seen": 19879136,
      "step": 30435
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 4.5263872146606445,
      "learning_rate": 3.173963173009825e-07,
      "loss": 0.3087,
      "num_input_tokens_seen": 19881888,
      "step": 30440
    },
    {
      "epoch": 17.951061320754718,
      "grad_norm": 4.172593593597412,
      "learning_rate": 3.164949122740352e-07,
      "loss": 0.2982,
      "num_input_tokens_seen": 19885312,
      "step": 30445
    },
    {
      "epoch": 17.954009433962263,
      "grad_norm": 8.467395782470703,
      "learning_rate": 3.1559474722524406e-07,
      "loss": 0.324,
      "num_input_tokens_seen": 19888448,
      "step": 30450
    },
    {
      "epoch": 17.95695754716981,
      "grad_norm": 3.013568162918091,
      "learning_rate": 3.146958223929325e-07,
      "loss": 0.2535,
      "num_input_tokens_seen": 19891776,
      "step": 30455
    },
    {
      "epoch": 17.95990566037736,
      "grad_norm": 3.7759006023406982,
      "learning_rate": 3.1379813801509454e-07,
      "loss": 0.3645,
      "num_input_tokens_seen": 19895424,
      "step": 30460
    },
    {
      "epoch": 17.962853773584907,
      "grad_norm": 2.977226495742798,
      "learning_rate": 3.1290169432939556e-07,
      "loss": 0.3797,
      "num_input_tokens_seen": 19898688,
      "step": 30465
    },
    {
      "epoch": 17.965801886792452,
      "grad_norm": 3.2936694622039795,
      "learning_rate": 3.120064915731735e-07,
      "loss": 0.3106,
      "num_input_tokens_seen": 19902272,
      "step": 30470
    },
    {
      "epoch": 17.96875,
      "grad_norm": 3.5352165699005127,
      "learning_rate": 3.1111252998343723e-07,
      "loss": 0.2009,
      "num_input_tokens_seen": 19906304,
      "step": 30475
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 3.8668055534362793,
      "learning_rate": 3.102198097968662e-07,
      "loss": 0.3205,
      "num_input_tokens_seen": 19909344,
      "step": 30480
    },
    {
      "epoch": 17.974646226415093,
      "grad_norm": 4.066013813018799,
      "learning_rate": 3.093283312498124e-07,
      "loss": 0.3442,
      "num_input_tokens_seen": 19912512,
      "step": 30485
    },
    {
      "epoch": 17.97759433962264,
      "grad_norm": 4.951056480407715,
      "learning_rate": 3.084380945782989e-07,
      "loss": 0.3631,
      "num_input_tokens_seen": 19914912,
      "step": 30490
    },
    {
      "epoch": 17.98054245283019,
      "grad_norm": 4.4774169921875,
      "learning_rate": 3.0754910001801866e-07,
      "loss": 0.3207,
      "num_input_tokens_seen": 19918016,
      "step": 30495
    },
    {
      "epoch": 17.983490566037737,
      "grad_norm": 4.929359436035156,
      "learning_rate": 3.0666134780433786e-07,
      "loss": 0.4296,
      "num_input_tokens_seen": 19921568,
      "step": 30500
    },
    {
      "epoch": 17.986438679245282,
      "grad_norm": 2.6217470169067383,
      "learning_rate": 3.0577483817229306e-07,
      "loss": 0.3006,
      "num_input_tokens_seen": 19925568,
      "step": 30505
    },
    {
      "epoch": 17.98938679245283,
      "grad_norm": 3.087860584259033,
      "learning_rate": 3.0488957135659023e-07,
      "loss": 0.411,
      "num_input_tokens_seen": 19928640,
      "step": 30510
    },
    {
      "epoch": 17.99233490566038,
      "grad_norm": 3.5151138305664062,
      "learning_rate": 3.040055475916087e-07,
      "loss": 0.1954,
      "num_input_tokens_seen": 19932032,
      "step": 30515
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 2.40657901763916,
      "learning_rate": 3.0312276711139675e-07,
      "loss": 0.4101,
      "num_input_tokens_seen": 19935584,
      "step": 30520
    },
    {
      "epoch": 17.99823113207547,
      "grad_norm": 4.014575958251953,
      "learning_rate": 3.0224123014967353e-07,
      "loss": 0.3324,
      "num_input_tokens_seen": 19938880,
      "step": 30525
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.5914515256881714,
      "eval_runtime": 19.1053,
      "eval_samples_per_second": 88.771,
      "eval_steps_per_second": 22.193,
      "num_input_tokens_seen": 19940536,
      "step": 30528
    },
    {
      "epoch": 18.00117924528302,
      "grad_norm": 2.584949254989624,
      "learning_rate": 3.013609369398324e-07,
      "loss": 0.4028,
      "num_input_tokens_seen": 19942104,
      "step": 30530
    },
    {
      "epoch": 18.004127358490567,
      "grad_norm": 4.493472576141357,
      "learning_rate": 3.004818877149318e-07,
      "loss": 0.2708,
      "num_input_tokens_seen": 19947096,
      "step": 30535
    },
    {
      "epoch": 18.007075471698112,
      "grad_norm": 3.1903536319732666,
      "learning_rate": 2.9960408270770624e-07,
      "loss": 0.2859,
      "num_input_tokens_seen": 19950424,
      "step": 30540
    },
    {
      "epoch": 18.01002358490566,
      "grad_norm": 2.805661916732788,
      "learning_rate": 2.9872752215055755e-07,
      "loss": 0.2877,
      "num_input_tokens_seen": 19953176,
      "step": 30545
    },
    {
      "epoch": 18.01297169811321,
      "grad_norm": 3.1871225833892822,
      "learning_rate": 2.9785220627555844e-07,
      "loss": 0.4204,
      "num_input_tokens_seen": 19959256,
      "step": 30550
    },
    {
      "epoch": 18.015919811320753,
      "grad_norm": 3.63694429397583,
      "learning_rate": 2.9697813531445295e-07,
      "loss": 0.3366,
      "num_input_tokens_seen": 19962744,
      "step": 30555
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 4.475766181945801,
      "learning_rate": 2.9610530949865433e-07,
      "loss": 0.3486,
      "num_input_tokens_seen": 19966520,
      "step": 30560
    },
    {
      "epoch": 18.02181603773585,
      "grad_norm": 4.3609538078308105,
      "learning_rate": 2.952337290592483e-07,
      "loss": 0.2844,
      "num_input_tokens_seen": 19968984,
      "step": 30565
    },
    {
      "epoch": 18.024764150943398,
      "grad_norm": 4.265958786010742,
      "learning_rate": 2.9436339422698913e-07,
      "loss": 0.3492,
      "num_input_tokens_seen": 19973496,
      "step": 30570
    },
    {
      "epoch": 18.027712264150942,
      "grad_norm": 7.0995378494262695,
      "learning_rate": 2.934943052323008e-07,
      "loss": 0.3088,
      "num_input_tokens_seen": 19976216,
      "step": 30575
    },
    {
      "epoch": 18.03066037735849,
      "grad_norm": 3.316744089126587,
      "learning_rate": 2.926264623052799e-07,
      "loss": 0.2102,
      "num_input_tokens_seen": 19979320,
      "step": 30580
    },
    {
      "epoch": 18.03360849056604,
      "grad_norm": 2.8612141609191895,
      "learning_rate": 2.9175986567569036e-07,
      "loss": 0.3055,
      "num_input_tokens_seen": 19982552,
      "step": 30585
    },
    {
      "epoch": 18.036556603773583,
      "grad_norm": 4.165629863739014,
      "learning_rate": 2.9089451557296755e-07,
      "loss": 0.3001,
      "num_input_tokens_seen": 19985432,
      "step": 30590
    },
    {
      "epoch": 18.03950471698113,
      "grad_norm": 2.3049025535583496,
      "learning_rate": 2.9003041222621706e-07,
      "loss": 0.2748,
      "num_input_tokens_seen": 19989368,
      "step": 30595
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 3.8789641857147217,
      "learning_rate": 2.8916755586421375e-07,
      "loss": 0.3325,
      "num_input_tokens_seen": 19992472,
      "step": 30600
    },
    {
      "epoch": 18.045400943396228,
      "grad_norm": 2.974989652633667,
      "learning_rate": 2.883059467154031e-07,
      "loss": 0.332,
      "num_input_tokens_seen": 19995608,
      "step": 30605
    },
    {
      "epoch": 18.048349056603772,
      "grad_norm": 2.6470553874969482,
      "learning_rate": 2.8744558500789887e-07,
      "loss": 0.2383,
      "num_input_tokens_seen": 19998200,
      "step": 30610
    },
    {
      "epoch": 18.05129716981132,
      "grad_norm": 4.657581806182861,
      "learning_rate": 2.8658647096948546e-07,
      "loss": 0.2554,
      "num_input_tokens_seen": 20001464,
      "step": 30615
    },
    {
      "epoch": 18.05424528301887,
      "grad_norm": 3.800567388534546,
      "learning_rate": 2.8572860482761813e-07,
      "loss": 0.259,
      "num_input_tokens_seen": 20004632,
      "step": 30620
    },
    {
      "epoch": 18.057193396226417,
      "grad_norm": 2.375772714614868,
      "learning_rate": 2.8487198680942017e-07,
      "loss": 0.2637,
      "num_input_tokens_seen": 20008632,
      "step": 30625
    },
    {
      "epoch": 18.06014150943396,
      "grad_norm": 3.124358892440796,
      "learning_rate": 2.840166171416836e-07,
      "loss": 0.1955,
      "num_input_tokens_seen": 20011736,
      "step": 30630
    },
    {
      "epoch": 18.06308962264151,
      "grad_norm": 4.307984352111816,
      "learning_rate": 2.8316249605087386e-07,
      "loss": 0.4052,
      "num_input_tokens_seen": 20015384,
      "step": 30635
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 6.949258804321289,
      "learning_rate": 2.823096237631212e-07,
      "loss": 0.3479,
      "num_input_tokens_seen": 20017432,
      "step": 30640
    },
    {
      "epoch": 18.068985849056602,
      "grad_norm": 2.438525676727295,
      "learning_rate": 2.814580005042283e-07,
      "loss": 0.2896,
      "num_input_tokens_seen": 20020632,
      "step": 30645
    },
    {
      "epoch": 18.07193396226415,
      "grad_norm": 4.24185037612915,
      "learning_rate": 2.8060762649966435e-07,
      "loss": 0.3419,
      "num_input_tokens_seen": 20023544,
      "step": 30650
    },
    {
      "epoch": 18.0748820754717,
      "grad_norm": 5.070246696472168,
      "learning_rate": 2.797585019745713e-07,
      "loss": 0.3422,
      "num_input_tokens_seen": 20026712,
      "step": 30655
    },
    {
      "epoch": 18.077830188679247,
      "grad_norm": 3.8882858753204346,
      "learning_rate": 2.789106271537584e-07,
      "loss": 0.4976,
      "num_input_tokens_seen": 20029560,
      "step": 30660
    },
    {
      "epoch": 18.08077830188679,
      "grad_norm": 6.680160999298096,
      "learning_rate": 2.780640022617037e-07,
      "loss": 0.3353,
      "num_input_tokens_seen": 20032824,
      "step": 30665
    },
    {
      "epoch": 18.08372641509434,
      "grad_norm": 4.39142370223999,
      "learning_rate": 2.772186275225547e-07,
      "loss": 0.3217,
      "num_input_tokens_seen": 20035224,
      "step": 30670
    },
    {
      "epoch": 18.086674528301888,
      "grad_norm": 4.821244239807129,
      "learning_rate": 2.7637450316012836e-07,
      "loss": 0.2834,
      "num_input_tokens_seen": 20037976,
      "step": 30675
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 4.158041000366211,
      "learning_rate": 2.755316293979088e-07,
      "loss": 0.2892,
      "num_input_tokens_seen": 20041048,
      "step": 30680
    },
    {
      "epoch": 18.09257075471698,
      "grad_norm": 4.697956085205078,
      "learning_rate": 2.7469000645905295e-07,
      "loss": 0.3225,
      "num_input_tokens_seen": 20043704,
      "step": 30685
    },
    {
      "epoch": 18.09551886792453,
      "grad_norm": 3.845637321472168,
      "learning_rate": 2.738496345663827e-07,
      "loss": 0.4273,
      "num_input_tokens_seen": 20047352,
      "step": 30690
    },
    {
      "epoch": 18.098466981132077,
      "grad_norm": 2.6598920822143555,
      "learning_rate": 2.7301051394239e-07,
      "loss": 0.3355,
      "num_input_tokens_seen": 20051064,
      "step": 30695
    },
    {
      "epoch": 18.10141509433962,
      "grad_norm": 4.378327369689941,
      "learning_rate": 2.72172644809236e-07,
      "loss": 0.2415,
      "num_input_tokens_seen": 20053880,
      "step": 30700
    },
    {
      "epoch": 18.10436320754717,
      "grad_norm": 2.6873621940612793,
      "learning_rate": 2.7133602738875e-07,
      "loss": 0.418,
      "num_input_tokens_seen": 20057688,
      "step": 30705
    },
    {
      "epoch": 18.107311320754718,
      "grad_norm": 3.1068100929260254,
      "learning_rate": 2.7050066190242976e-07,
      "loss": 0.2693,
      "num_input_tokens_seen": 20061080,
      "step": 30710
    },
    {
      "epoch": 18.110259433962263,
      "grad_norm": 2.456934928894043,
      "learning_rate": 2.696665485714428e-07,
      "loss": 0.2746,
      "num_input_tokens_seen": 20064376,
      "step": 30715
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 1.9876976013183594,
      "learning_rate": 2.6883368761662367e-07,
      "loss": 0.4441,
      "num_input_tokens_seen": 20069400,
      "step": 30720
    },
    {
      "epoch": 18.11615566037736,
      "grad_norm": 2.766303062438965,
      "learning_rate": 2.680020792584759e-07,
      "loss": 0.3035,
      "num_input_tokens_seen": 20072920,
      "step": 30725
    },
    {
      "epoch": 18.119103773584907,
      "grad_norm": 4.27656888961792,
      "learning_rate": 2.6717172371717113e-07,
      "loss": 0.2627,
      "num_input_tokens_seen": 20076440,
      "step": 30730
    },
    {
      "epoch": 18.122051886792452,
      "grad_norm": 3.7751262187957764,
      "learning_rate": 2.663426212125503e-07,
      "loss": 0.2403,
      "num_input_tokens_seen": 20079544,
      "step": 30735
    },
    {
      "epoch": 18.125,
      "grad_norm": 2.886540651321411,
      "learning_rate": 2.655147719641216e-07,
      "loss": 0.2968,
      "num_input_tokens_seen": 20082904,
      "step": 30740
    },
    {
      "epoch": 18.127948113207548,
      "grad_norm": 4.316927433013916,
      "learning_rate": 2.646881761910602e-07,
      "loss": 0.3132,
      "num_input_tokens_seen": 20086520,
      "step": 30745
    },
    {
      "epoch": 18.130896226415093,
      "grad_norm": 3.915656089782715,
      "learning_rate": 2.638628341122135e-07,
      "loss": 0.3476,
      "num_input_tokens_seen": 20089240,
      "step": 30750
    },
    {
      "epoch": 18.13384433962264,
      "grad_norm": 3.190474033355713,
      "learning_rate": 2.6303874594609314e-07,
      "loss": 0.3284,
      "num_input_tokens_seen": 20092312,
      "step": 30755
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 3.183751106262207,
      "learning_rate": 2.622159119108797e-07,
      "loss": 0.2062,
      "num_input_tokens_seen": 20096120,
      "step": 30760
    },
    {
      "epoch": 18.139740566037737,
      "grad_norm": 4.666077136993408,
      "learning_rate": 2.6139433222442226e-07,
      "loss": 0.401,
      "num_input_tokens_seen": 20099512,
      "step": 30765
    },
    {
      "epoch": 18.142688679245282,
      "grad_norm": 3.6739840507507324,
      "learning_rate": 2.6057400710423787e-07,
      "loss": 0.303,
      "num_input_tokens_seen": 20101976,
      "step": 30770
    },
    {
      "epoch": 18.14563679245283,
      "grad_norm": 1.9379053115844727,
      "learning_rate": 2.5975493676751004e-07,
      "loss": 0.2922,
      "num_input_tokens_seen": 20105592,
      "step": 30775
    },
    {
      "epoch": 18.14858490566038,
      "grad_norm": 3.6811959743499756,
      "learning_rate": 2.589371214310926e-07,
      "loss": 0.3519,
      "num_input_tokens_seen": 20109112,
      "step": 30780
    },
    {
      "epoch": 18.151533018867923,
      "grad_norm": 2.8478147983551025,
      "learning_rate": 2.581205613115051e-07,
      "loss": 0.1631,
      "num_input_tokens_seen": 20112664,
      "step": 30785
    },
    {
      "epoch": 18.15448113207547,
      "grad_norm": 2.7599480152130127,
      "learning_rate": 2.573052566249357e-07,
      "loss": 0.2934,
      "num_input_tokens_seen": 20116184,
      "step": 30790
    },
    {
      "epoch": 18.15742924528302,
      "grad_norm": 3.9310390949249268,
      "learning_rate": 2.5649120758723945e-07,
      "loss": 0.3182,
      "num_input_tokens_seen": 20119864,
      "step": 30795
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 3.406463146209717,
      "learning_rate": 2.5567841441393906e-07,
      "loss": 0.2589,
      "num_input_tokens_seen": 20123448,
      "step": 30800
    },
    {
      "epoch": 18.163325471698112,
      "grad_norm": 3.501187324523926,
      "learning_rate": 2.548668773202245e-07,
      "loss": 0.2622,
      "num_input_tokens_seen": 20126712,
      "step": 30805
    },
    {
      "epoch": 18.16627358490566,
      "grad_norm": 3.722938060760498,
      "learning_rate": 2.5405659652095573e-07,
      "loss": 0.3231,
      "num_input_tokens_seen": 20130008,
      "step": 30810
    },
    {
      "epoch": 18.16922169811321,
      "grad_norm": 3.088683605194092,
      "learning_rate": 2.5324757223065655e-07,
      "loss": 0.3449,
      "num_input_tokens_seen": 20133048,
      "step": 30815
    },
    {
      "epoch": 18.172169811320753,
      "grad_norm": 4.640005111694336,
      "learning_rate": 2.524398046635207e-07,
      "loss": 0.339,
      "num_input_tokens_seen": 20136024,
      "step": 30820
    },
    {
      "epoch": 18.1751179245283,
      "grad_norm": 5.533210754394531,
      "learning_rate": 2.51633294033406e-07,
      "loss": 0.4144,
      "num_input_tokens_seen": 20138840,
      "step": 30825
    },
    {
      "epoch": 18.17806603773585,
      "grad_norm": 3.682384967803955,
      "learning_rate": 2.5082804055384214e-07,
      "loss": 0.3159,
      "num_input_tokens_seen": 20141528,
      "step": 30830
    },
    {
      "epoch": 18.181014150943398,
      "grad_norm": 3.831970453262329,
      "learning_rate": 2.50024044438022e-07,
      "loss": 0.2789,
      "num_input_tokens_seen": 20144248,
      "step": 30835
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 3.9287912845611572,
      "learning_rate": 2.492213058988069e-07,
      "loss": 0.2228,
      "num_input_tokens_seen": 20148024,
      "step": 30840
    },
    {
      "epoch": 18.18691037735849,
      "grad_norm": 4.066948413848877,
      "learning_rate": 2.4841982514872633e-07,
      "loss": 0.422,
      "num_input_tokens_seen": 20151160,
      "step": 30845
    },
    {
      "epoch": 18.18985849056604,
      "grad_norm": 6.012871742248535,
      "learning_rate": 2.4761960239997497e-07,
      "loss": 0.3974,
      "num_input_tokens_seen": 20153624,
      "step": 30850
    },
    {
      "epoch": 18.192806603773583,
      "grad_norm": 4.717536926269531,
      "learning_rate": 2.4682063786441556e-07,
      "loss": 0.3086,
      "num_input_tokens_seen": 20156952,
      "step": 30855
    },
    {
      "epoch": 18.19575471698113,
      "grad_norm": 3.774090528488159,
      "learning_rate": 2.460229317535778e-07,
      "loss": 0.3669,
      "num_input_tokens_seen": 20160216,
      "step": 30860
    },
    {
      "epoch": 18.19870283018868,
      "grad_norm": 4.0794243812561035,
      "learning_rate": 2.4522648427865725e-07,
      "loss": 0.5037,
      "num_input_tokens_seen": 20163480,
      "step": 30865
    },
    {
      "epoch": 18.201650943396228,
      "grad_norm": 2.4848995208740234,
      "learning_rate": 2.444312956505163e-07,
      "loss": 0.3304,
      "num_input_tokens_seen": 20165944,
      "step": 30870
    },
    {
      "epoch": 18.204599056603772,
      "grad_norm": 2.6667606830596924,
      "learning_rate": 2.4363736607968537e-07,
      "loss": 0.3253,
      "num_input_tokens_seen": 20170296,
      "step": 30875
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 4.449398040771484,
      "learning_rate": 2.428446957763608e-07,
      "loss": 0.2352,
      "num_input_tokens_seen": 20172728,
      "step": 30880
    },
    {
      "epoch": 18.21049528301887,
      "grad_norm": 6.630474090576172,
      "learning_rate": 2.4205328495040535e-07,
      "loss": 0.2363,
      "num_input_tokens_seen": 20175192,
      "step": 30885
    },
    {
      "epoch": 18.213443396226417,
      "grad_norm": 4.199165344238281,
      "learning_rate": 2.412631338113486e-07,
      "loss": 0.3507,
      "num_input_tokens_seen": 20179608,
      "step": 30890
    },
    {
      "epoch": 18.21639150943396,
      "grad_norm": 4.401319980621338,
      "learning_rate": 2.404742425683848e-07,
      "loss": 0.4026,
      "num_input_tokens_seen": 20183992,
      "step": 30895
    },
    {
      "epoch": 18.21933962264151,
      "grad_norm": 3.5704336166381836,
      "learning_rate": 2.3968661143037864e-07,
      "loss": 0.3641,
      "num_input_tokens_seen": 20187704,
      "step": 30900
    },
    {
      "epoch": 18.222287735849058,
      "grad_norm": 5.98125696182251,
      "learning_rate": 2.3890024060585823e-07,
      "loss": 0.3642,
      "num_input_tokens_seen": 20190296,
      "step": 30905
    },
    {
      "epoch": 18.225235849056602,
      "grad_norm": 3.734569549560547,
      "learning_rate": 2.3811513030301826e-07,
      "loss": 0.2833,
      "num_input_tokens_seen": 20193208,
      "step": 30910
    },
    {
      "epoch": 18.22818396226415,
      "grad_norm": 7.104701995849609,
      "learning_rate": 2.373312807297201e-07,
      "loss": 0.4262,
      "num_input_tokens_seen": 20197400,
      "step": 30915
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 4.676792144775391,
      "learning_rate": 2.3654869209349007e-07,
      "loss": 0.2437,
      "num_input_tokens_seen": 20200248,
      "step": 30920
    },
    {
      "epoch": 18.234080188679247,
      "grad_norm": 2.9524242877960205,
      "learning_rate": 2.357673646015246e-07,
      "loss": 0.309,
      "num_input_tokens_seen": 20204184,
      "step": 30925
    },
    {
      "epoch": 18.23702830188679,
      "grad_norm": 3.7201201915740967,
      "learning_rate": 2.3498729846068103e-07,
      "loss": 0.3211,
      "num_input_tokens_seen": 20207704,
      "step": 30930
    },
    {
      "epoch": 18.23997641509434,
      "grad_norm": 2.7681400775909424,
      "learning_rate": 2.342084938774869e-07,
      "loss": 0.2968,
      "num_input_tokens_seen": 20211352,
      "step": 30935
    },
    {
      "epoch": 18.242924528301888,
      "grad_norm": 7.092854976654053,
      "learning_rate": 2.334309510581334e-07,
      "loss": 0.3241,
      "num_input_tokens_seen": 20214840,
      "step": 30940
    },
    {
      "epoch": 18.245872641509433,
      "grad_norm": 2.7532060146331787,
      "learning_rate": 2.3265467020847864e-07,
      "loss": 0.2669,
      "num_input_tokens_seen": 20218296,
      "step": 30945
    },
    {
      "epoch": 18.24882075471698,
      "grad_norm": 2.969210624694824,
      "learning_rate": 2.31879651534046e-07,
      "loss": 0.3662,
      "num_input_tokens_seen": 20222712,
      "step": 30950
    },
    {
      "epoch": 18.25176886792453,
      "grad_norm": 4.211237907409668,
      "learning_rate": 2.311058952400247e-07,
      "loss": 0.2911,
      "num_input_tokens_seen": 20225368,
      "step": 30955
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 2.9875974655151367,
      "learning_rate": 2.3033340153127026e-07,
      "loss": 0.2525,
      "num_input_tokens_seen": 20228440,
      "step": 30960
    },
    {
      "epoch": 18.25766509433962,
      "grad_norm": 5.268301010131836,
      "learning_rate": 2.295621706123041e-07,
      "loss": 0.4988,
      "num_input_tokens_seen": 20230680,
      "step": 30965
    },
    {
      "epoch": 18.26061320754717,
      "grad_norm": 2.122607469558716,
      "learning_rate": 2.287922026873135e-07,
      "loss": 0.3075,
      "num_input_tokens_seen": 20234744,
      "step": 30970
    },
    {
      "epoch": 18.263561320754718,
      "grad_norm": 4.583444118499756,
      "learning_rate": 2.2802349796014923e-07,
      "loss": 0.3282,
      "num_input_tokens_seen": 20238200,
      "step": 30975
    },
    {
      "epoch": 18.266509433962263,
      "grad_norm": 2.6378183364868164,
      "learning_rate": 2.2725605663433013e-07,
      "loss": 0.397,
      "num_input_tokens_seen": 20241016,
      "step": 30980
    },
    {
      "epoch": 18.26945754716981,
      "grad_norm": 3.321960687637329,
      "learning_rate": 2.264898789130393e-07,
      "loss": 0.3674,
      "num_input_tokens_seen": 20244920,
      "step": 30985
    },
    {
      "epoch": 18.27240566037736,
      "grad_norm": 4.439864635467529,
      "learning_rate": 2.2572496499912554e-07,
      "loss": 0.3382,
      "num_input_tokens_seen": 20247352,
      "step": 30990
    },
    {
      "epoch": 18.275353773584907,
      "grad_norm": 3.4746596813201904,
      "learning_rate": 2.2496131509510354e-07,
      "loss": 0.2996,
      "num_input_tokens_seen": 20249752,
      "step": 30995
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 3.7877566814422607,
      "learning_rate": 2.2419892940315268e-07,
      "loss": 0.3068,
      "num_input_tokens_seen": 20252760,
      "step": 31000
    },
    {
      "epoch": 18.28125,
      "grad_norm": 3.2860357761383057,
      "learning_rate": 2.2343780812511819e-07,
      "loss": 0.224,
      "num_input_tokens_seen": 20255480,
      "step": 31005
    },
    {
      "epoch": 18.284198113207548,
      "grad_norm": 2.5814120769500732,
      "learning_rate": 2.2267795146250936e-07,
      "loss": 0.2449,
      "num_input_tokens_seen": 20258264,
      "step": 31010
    },
    {
      "epoch": 18.287146226415093,
      "grad_norm": 2.6322600841522217,
      "learning_rate": 2.2191935961650146e-07,
      "loss": 0.3085,
      "num_input_tokens_seen": 20261624,
      "step": 31015
    },
    {
      "epoch": 18.29009433962264,
      "grad_norm": 6.0532331466674805,
      "learning_rate": 2.2116203278793603e-07,
      "loss": 0.4374,
      "num_input_tokens_seen": 20264280,
      "step": 31020
    },
    {
      "epoch": 18.29304245283019,
      "grad_norm": 2.2062087059020996,
      "learning_rate": 2.2040597117731766e-07,
      "loss": 0.3627,
      "num_input_tokens_seen": 20267032,
      "step": 31025
    },
    {
      "epoch": 18.295990566037737,
      "grad_norm": 3.3155341148376465,
      "learning_rate": 2.1965117498481793e-07,
      "loss": 0.2351,
      "num_input_tokens_seen": 20270232,
      "step": 31030
    },
    {
      "epoch": 18.298938679245282,
      "grad_norm": 2.596074342727661,
      "learning_rate": 2.188976444102714e-07,
      "loss": 0.2229,
      "num_input_tokens_seen": 20273208,
      "step": 31035
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 4.348106861114502,
      "learning_rate": 2.181453796531796e-07,
      "loss": 0.3418,
      "num_input_tokens_seen": 20276312,
      "step": 31040
    },
    {
      "epoch": 18.30483490566038,
      "grad_norm": 3.846242904663086,
      "learning_rate": 2.1739438091270658e-07,
      "loss": 0.401,
      "num_input_tokens_seen": 20279480,
      "step": 31045
    },
    {
      "epoch": 18.307783018867923,
      "grad_norm": 3.563354253768921,
      "learning_rate": 2.1664464838768329e-07,
      "loss": 0.2963,
      "num_input_tokens_seen": 20282648,
      "step": 31050
    },
    {
      "epoch": 18.31073113207547,
      "grad_norm": 5.303159713745117,
      "learning_rate": 2.1589618227660426e-07,
      "loss": 0.3484,
      "num_input_tokens_seen": 20286424,
      "step": 31055
    },
    {
      "epoch": 18.31367924528302,
      "grad_norm": 3.3126232624053955,
      "learning_rate": 2.151489827776293e-07,
      "loss": 0.2717,
      "num_input_tokens_seen": 20290104,
      "step": 31060
    },
    {
      "epoch": 18.316627358490567,
      "grad_norm": 3.7731077671051025,
      "learning_rate": 2.1440305008858298e-07,
      "loss": 0.5757,
      "num_input_tokens_seen": 20293752,
      "step": 31065
    },
    {
      "epoch": 18.319575471698112,
      "grad_norm": 8.036755561828613,
      "learning_rate": 2.1365838440695397e-07,
      "loss": 0.4163,
      "num_input_tokens_seen": 20296920,
      "step": 31070
    },
    {
      "epoch": 18.32252358490566,
      "grad_norm": 2.490412712097168,
      "learning_rate": 2.129149859298957e-07,
      "loss": 0.4767,
      "num_input_tokens_seen": 20303928,
      "step": 31075
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 5.672810077667236,
      "learning_rate": 2.1217285485422622e-07,
      "loss": 0.3048,
      "num_input_tokens_seen": 20308248,
      "step": 31080
    },
    {
      "epoch": 18.328419811320753,
      "grad_norm": 2.1927437782287598,
      "learning_rate": 2.114319913764268e-07,
      "loss": 0.3267,
      "num_input_tokens_seen": 20311992,
      "step": 31085
    },
    {
      "epoch": 18.3313679245283,
      "grad_norm": 2.8401880264282227,
      "learning_rate": 2.10692395692646e-07,
      "loss": 0.3002,
      "num_input_tokens_seen": 20315832,
      "step": 31090
    },
    {
      "epoch": 18.33431603773585,
      "grad_norm": 7.499414443969727,
      "learning_rate": 2.0995406799869444e-07,
      "loss": 0.3,
      "num_input_tokens_seen": 20318648,
      "step": 31095
    },
    {
      "epoch": 18.337264150943398,
      "grad_norm": 4.294673442840576,
      "learning_rate": 2.0921700849004743e-07,
      "loss": 0.3405,
      "num_input_tokens_seen": 20321304,
      "step": 31100
    },
    {
      "epoch": 18.340212264150942,
      "grad_norm": 5.390089988708496,
      "learning_rate": 2.084812173618439e-07,
      "loss": 0.3592,
      "num_input_tokens_seen": 20325048,
      "step": 31105
    },
    {
      "epoch": 18.34316037735849,
      "grad_norm": 6.075977802276611,
      "learning_rate": 2.0774669480888853e-07,
      "loss": 0.3845,
      "num_input_tokens_seen": 20327736,
      "step": 31110
    },
    {
      "epoch": 18.34610849056604,
      "grad_norm": 2.4842214584350586,
      "learning_rate": 2.0701344102564912e-07,
      "loss": 0.2487,
      "num_input_tokens_seen": 20331800,
      "step": 31115
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 2.557091236114502,
      "learning_rate": 2.062814562062576e-07,
      "loss": 0.3395,
      "num_input_tokens_seen": 20335192,
      "step": 31120
    },
    {
      "epoch": 18.35200471698113,
      "grad_norm": 2.5936551094055176,
      "learning_rate": 2.0555074054451063e-07,
      "loss": 0.2916,
      "num_input_tokens_seen": 20339032,
      "step": 31125
    },
    {
      "epoch": 18.35495283018868,
      "grad_norm": 10.118579864501953,
      "learning_rate": 2.0482129423386843e-07,
      "loss": 0.2213,
      "num_input_tokens_seen": 20341720,
      "step": 31130
    },
    {
      "epoch": 18.357900943396228,
      "grad_norm": 7.498178482055664,
      "learning_rate": 2.040931174674543e-07,
      "loss": 0.3359,
      "num_input_tokens_seen": 20345016,
      "step": 31135
    },
    {
      "epoch": 18.360849056603772,
      "grad_norm": 6.2783308029174805,
      "learning_rate": 2.0336621043805682e-07,
      "loss": 0.4563,
      "num_input_tokens_seen": 20347992,
      "step": 31140
    },
    {
      "epoch": 18.36379716981132,
      "grad_norm": 3.128206253051758,
      "learning_rate": 2.0264057333812704e-07,
      "loss": 0.302,
      "num_input_tokens_seen": 20350872,
      "step": 31145
    },
    {
      "epoch": 18.36674528301887,
      "grad_norm": 6.744492053985596,
      "learning_rate": 2.0191620635978127e-07,
      "loss": 0.4701,
      "num_input_tokens_seen": 20355032,
      "step": 31150
    },
    {
      "epoch": 18.369693396226417,
      "grad_norm": 2.514572858810425,
      "learning_rate": 2.0119310969479833e-07,
      "loss": 0.2923,
      "num_input_tokens_seen": 20357432,
      "step": 31155
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 3.127809762954712,
      "learning_rate": 2.004712835346212e-07,
      "loss": 0.2778,
      "num_input_tokens_seen": 20361144,
      "step": 31160
    },
    {
      "epoch": 18.37558962264151,
      "grad_norm": 3.0956196784973145,
      "learning_rate": 1.99750728070357e-07,
      "loss": 0.2369,
      "num_input_tokens_seen": 20363768,
      "step": 31165
    },
    {
      "epoch": 18.378537735849058,
      "grad_norm": 3.646576404571533,
      "learning_rate": 1.9903144349277536e-07,
      "loss": 0.242,
      "num_input_tokens_seen": 20366584,
      "step": 31170
    },
    {
      "epoch": 18.381485849056602,
      "grad_norm": 4.878662109375,
      "learning_rate": 1.983134299923095e-07,
      "loss": 0.2939,
      "num_input_tokens_seen": 20369176,
      "step": 31175
    },
    {
      "epoch": 18.38443396226415,
      "grad_norm": 4.221963405609131,
      "learning_rate": 1.9759668775905737e-07,
      "loss": 0.4134,
      "num_input_tokens_seen": 20371896,
      "step": 31180
    },
    {
      "epoch": 18.3873820754717,
      "grad_norm": 4.261650085449219,
      "learning_rate": 1.9688121698277995e-07,
      "loss": 0.4276,
      "num_input_tokens_seen": 20375352,
      "step": 31185
    },
    {
      "epoch": 18.390330188679247,
      "grad_norm": 7.510025501251221,
      "learning_rate": 1.9616701785290015e-07,
      "loss": 0.2421,
      "num_input_tokens_seen": 20378744,
      "step": 31190
    },
    {
      "epoch": 18.39327830188679,
      "grad_norm": 3.232374668121338,
      "learning_rate": 1.954540905585056e-07,
      "loss": 0.2892,
      "num_input_tokens_seen": 20381496,
      "step": 31195
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 1.9882973432540894,
      "learning_rate": 1.9474243528834757e-07,
      "loss": 0.2211,
      "num_input_tokens_seen": 20384376,
      "step": 31200
    },
    {
      "epoch": 18.399174528301888,
      "grad_norm": 2.547682762145996,
      "learning_rate": 1.9403205223083866e-07,
      "loss": 0.2876,
      "num_input_tokens_seen": 20387448,
      "step": 31205
    },
    {
      "epoch": 18.402122641509433,
      "grad_norm": 3.8012006282806396,
      "learning_rate": 1.9332294157405619e-07,
      "loss": 0.3891,
      "num_input_tokens_seen": 20391160,
      "step": 31210
    },
    {
      "epoch": 18.40507075471698,
      "grad_norm": 3.65187406539917,
      "learning_rate": 1.926151035057411e-07,
      "loss": 0.3498,
      "num_input_tokens_seen": 20393784,
      "step": 31215
    },
    {
      "epoch": 18.40801886792453,
      "grad_norm": 2.359156847000122,
      "learning_rate": 1.9190853821329626e-07,
      "loss": 0.4953,
      "num_input_tokens_seen": 20396632,
      "step": 31220
    },
    {
      "epoch": 18.410966981132077,
      "grad_norm": 2.8175954818725586,
      "learning_rate": 1.9120324588378757e-07,
      "loss": 0.3723,
      "num_input_tokens_seen": 20399448,
      "step": 31225
    },
    {
      "epoch": 18.41391509433962,
      "grad_norm": 2.333630084991455,
      "learning_rate": 1.9049922670394461e-07,
      "loss": 0.3183,
      "num_input_tokens_seen": 20402840,
      "step": 31230
    },
    {
      "epoch": 18.41686320754717,
      "grad_norm": 3.080368757247925,
      "learning_rate": 1.897964808601588e-07,
      "loss": 0.2964,
      "num_input_tokens_seen": 20405944,
      "step": 31235
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 3.7742011547088623,
      "learning_rate": 1.8909500853848517e-07,
      "loss": 0.2849,
      "num_input_tokens_seen": 20409464,
      "step": 31240
    },
    {
      "epoch": 18.422759433962263,
      "grad_norm": 2.851191759109497,
      "learning_rate": 1.8839480992464243e-07,
      "loss": 0.3593,
      "num_input_tokens_seen": 20412120,
      "step": 31245
    },
    {
      "epoch": 18.42570754716981,
      "grad_norm": 4.256245136260986,
      "learning_rate": 1.8769588520401005e-07,
      "loss": 0.3417,
      "num_input_tokens_seen": 20415672,
      "step": 31250
    },
    {
      "epoch": 18.42865566037736,
      "grad_norm": 6.028692722320557,
      "learning_rate": 1.8699823456163279e-07,
      "loss": 0.3632,
      "num_input_tokens_seen": 20419160,
      "step": 31255
    },
    {
      "epoch": 18.431603773584907,
      "grad_norm": 4.172680377960205,
      "learning_rate": 1.8630185818221514e-07,
      "loss": 0.1878,
      "num_input_tokens_seen": 20425336,
      "step": 31260
    },
    {
      "epoch": 18.434551886792452,
      "grad_norm": 4.610836505889893,
      "learning_rate": 1.856067562501268e-07,
      "loss": 0.3666,
      "num_input_tokens_seen": 20427896,
      "step": 31265
    },
    {
      "epoch": 18.4375,
      "grad_norm": 3.3099303245544434,
      "learning_rate": 1.8491292894939837e-07,
      "loss": 0.3426,
      "num_input_tokens_seen": 20432120,
      "step": 31270
    },
    {
      "epoch": 18.440448113207548,
      "grad_norm": 3.316725969314575,
      "learning_rate": 1.8422037646372405e-07,
      "loss": 0.3376,
      "num_input_tokens_seen": 20434936,
      "step": 31275
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 4.867111682891846,
      "learning_rate": 1.8352909897645989e-07,
      "loss": 0.483,
      "num_input_tokens_seen": 20438136,
      "step": 31280
    },
    {
      "epoch": 18.44634433962264,
      "grad_norm": 3.2113916873931885,
      "learning_rate": 1.8283909667062448e-07,
      "loss": 0.3996,
      "num_input_tokens_seen": 20441016,
      "step": 31285
    },
    {
      "epoch": 18.44929245283019,
      "grad_norm": 3.637784004211426,
      "learning_rate": 1.82150369728899e-07,
      "loss": 0.3315,
      "num_input_tokens_seen": 20444056,
      "step": 31290
    },
    {
      "epoch": 18.452240566037737,
      "grad_norm": 4.194357395172119,
      "learning_rate": 1.814629183336275e-07,
      "loss": 0.2993,
      "num_input_tokens_seen": 20447608,
      "step": 31295
    },
    {
      "epoch": 18.455188679245282,
      "grad_norm": 4.240993499755859,
      "learning_rate": 1.807767426668139e-07,
      "loss": 0.3312,
      "num_input_tokens_seen": 20451672,
      "step": 31300
    },
    {
      "epoch": 18.45813679245283,
      "grad_norm": 2.9933085441589355,
      "learning_rate": 1.8009184291012783e-07,
      "loss": 0.2782,
      "num_input_tokens_seen": 20454200,
      "step": 31305
    },
    {
      "epoch": 18.46108490566038,
      "grad_norm": 3.3824167251586914,
      "learning_rate": 1.7940821924489926e-07,
      "loss": 0.5167,
      "num_input_tokens_seen": 20457464,
      "step": 31310
    },
    {
      "epoch": 18.464033018867923,
      "grad_norm": 4.833495140075684,
      "learning_rate": 1.7872587185212009e-07,
      "loss": 0.4452,
      "num_input_tokens_seen": 20460472,
      "step": 31315
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 4.592500686645508,
      "learning_rate": 1.7804480091244524e-07,
      "loss": 0.3728,
      "num_input_tokens_seen": 20463384,
      "step": 31320
    },
    {
      "epoch": 18.46992924528302,
      "grad_norm": 1.4283291101455688,
      "learning_rate": 1.7736500660619104e-07,
      "loss": 0.2088,
      "num_input_tokens_seen": 20466744,
      "step": 31325
    },
    {
      "epoch": 18.472877358490567,
      "grad_norm": 4.399046421051025,
      "learning_rate": 1.766864891133352e-07,
      "loss": 0.2951,
      "num_input_tokens_seen": 20470072,
      "step": 31330
    },
    {
      "epoch": 18.475825471698112,
      "grad_norm": 3.8656833171844482,
      "learning_rate": 1.7600924861351843e-07,
      "loss": 0.3224,
      "num_input_tokens_seen": 20473208,
      "step": 31335
    },
    {
      "epoch": 18.47877358490566,
      "grad_norm": 2.5741281509399414,
      "learning_rate": 1.7533328528604398e-07,
      "loss": 0.3292,
      "num_input_tokens_seen": 20481496,
      "step": 31340
    },
    {
      "epoch": 18.48172169811321,
      "grad_norm": 3.7565059661865234,
      "learning_rate": 1.746585993098754e-07,
      "loss": 0.3021,
      "num_input_tokens_seen": 20484088,
      "step": 31345
    },
    {
      "epoch": 18.484669811320753,
      "grad_norm": 3.6433916091918945,
      "learning_rate": 1.7398519086363864e-07,
      "loss": 0.3183,
      "num_input_tokens_seen": 20487160,
      "step": 31350
    },
    {
      "epoch": 18.4876179245283,
      "grad_norm": 3.045191764831543,
      "learning_rate": 1.733130601256211e-07,
      "loss": 0.2352,
      "num_input_tokens_seen": 20489752,
      "step": 31355
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 5.4711594581604,
      "learning_rate": 1.7264220727377323e-07,
      "loss": 0.2391,
      "num_input_tokens_seen": 20492632,
      "step": 31360
    },
    {
      "epoch": 18.493514150943398,
      "grad_norm": 4.1344709396362305,
      "learning_rate": 1.7197263248570517e-07,
      "loss": 0.4679,
      "num_input_tokens_seen": 20495672,
      "step": 31365
    },
    {
      "epoch": 18.496462264150942,
      "grad_norm": 3.7273194789886475,
      "learning_rate": 1.7130433593869124e-07,
      "loss": 0.2906,
      "num_input_tokens_seen": 20498552,
      "step": 31370
    },
    {
      "epoch": 18.49941037735849,
      "grad_norm": 3.7265465259552,
      "learning_rate": 1.706373178096643e-07,
      "loss": 0.3385,
      "num_input_tokens_seen": 20501624,
      "step": 31375
    },
    {
      "epoch": 18.50235849056604,
      "grad_norm": 5.693615436553955,
      "learning_rate": 1.6997157827522092e-07,
      "loss": 0.3367,
      "num_input_tokens_seen": 20504824,
      "step": 31380
    },
    {
      "epoch": 18.505306603773583,
      "grad_norm": 5.5712785720825195,
      "learning_rate": 1.6930711751161843e-07,
      "loss": 0.272,
      "num_input_tokens_seen": 20507704,
      "step": 31385
    },
    {
      "epoch": 18.50825471698113,
      "grad_norm": 2.2745532989501953,
      "learning_rate": 1.6864393569477556e-07,
      "loss": 0.3714,
      "num_input_tokens_seen": 20511320,
      "step": 31390
    },
    {
      "epoch": 18.51120283018868,
      "grad_norm": 2.842209815979004,
      "learning_rate": 1.6798203300027295e-07,
      "loss": 0.3425,
      "num_input_tokens_seen": 20515448,
      "step": 31395
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 3.3570632934570312,
      "learning_rate": 1.6732140960335152e-07,
      "loss": 0.2556,
      "num_input_tokens_seen": 20518488,
      "step": 31400
    },
    {
      "epoch": 18.517099056603772,
      "grad_norm": 3.8766613006591797,
      "learning_rate": 1.666620656789153e-07,
      "loss": 0.3043,
      "num_input_tokens_seen": 20521208,
      "step": 31405
    },
    {
      "epoch": 18.52004716981132,
      "grad_norm": 4.373570442199707,
      "learning_rate": 1.660040014015274e-07,
      "loss": 0.256,
      "num_input_tokens_seen": 20524696,
      "step": 31410
    },
    {
      "epoch": 18.52299528301887,
      "grad_norm": 4.24659538269043,
      "learning_rate": 1.6534721694541344e-07,
      "loss": 0.3203,
      "num_input_tokens_seen": 20527864,
      "step": 31415
    },
    {
      "epoch": 18.525943396226417,
      "grad_norm": 3.441295862197876,
      "learning_rate": 1.6469171248445993e-07,
      "loss": 0.3034,
      "num_input_tokens_seen": 20531832,
      "step": 31420
    },
    {
      "epoch": 18.52889150943396,
      "grad_norm": 4.110219478607178,
      "learning_rate": 1.6403748819221464e-07,
      "loss": 0.3066,
      "num_input_tokens_seen": 20535384,
      "step": 31425
    },
    {
      "epoch": 18.53183962264151,
      "grad_norm": 3.834606409072876,
      "learning_rate": 1.6338454424188632e-07,
      "loss": 0.347,
      "num_input_tokens_seen": 20538552,
      "step": 31430
    },
    {
      "epoch": 18.534787735849058,
      "grad_norm": 5.962763786315918,
      "learning_rate": 1.6273288080634442e-07,
      "loss": 0.2833,
      "num_input_tokens_seen": 20541048,
      "step": 31435
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 2.8575570583343506,
      "learning_rate": 1.6208249805811982e-07,
      "loss": 0.3263,
      "num_input_tokens_seen": 20544120,
      "step": 31440
    },
    {
      "epoch": 18.54068396226415,
      "grad_norm": 4.003894805908203,
      "learning_rate": 1.6143339616940423e-07,
      "loss": 0.2281,
      "num_input_tokens_seen": 20547448,
      "step": 31445
    },
    {
      "epoch": 18.5436320754717,
      "grad_norm": 3.633059501647949,
      "learning_rate": 1.6078557531205018e-07,
      "loss": 0.4307,
      "num_input_tokens_seen": 20553400,
      "step": 31450
    },
    {
      "epoch": 18.546580188679247,
      "grad_norm": 2.9927961826324463,
      "learning_rate": 1.601390356575705e-07,
      "loss": 0.2844,
      "num_input_tokens_seen": 20556440,
      "step": 31455
    },
    {
      "epoch": 18.54952830188679,
      "grad_norm": 3.1863956451416016,
      "learning_rate": 1.5949377737713988e-07,
      "loss": 0.2855,
      "num_input_tokens_seen": 20559416,
      "step": 31460
    },
    {
      "epoch": 18.55247641509434,
      "grad_norm": 3.322256326675415,
      "learning_rate": 1.5884980064159338e-07,
      "loss": 0.278,
      "num_input_tokens_seen": 20562552,
      "step": 31465
    },
    {
      "epoch": 18.555424528301888,
      "grad_norm": 2.8000314235687256,
      "learning_rate": 1.5820710562142627e-07,
      "loss": 0.2478,
      "num_input_tokens_seen": 20566200,
      "step": 31470
    },
    {
      "epoch": 18.558372641509433,
      "grad_norm": 4.838726997375488,
      "learning_rate": 1.575656924867952e-07,
      "loss": 0.3064,
      "num_input_tokens_seen": 20569048,
      "step": 31475
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 3.7984418869018555,
      "learning_rate": 1.5692556140751658e-07,
      "loss": 0.473,
      "num_input_tokens_seen": 20571768,
      "step": 31480
    },
    {
      "epoch": 18.56426886792453,
      "grad_norm": 5.215952396392822,
      "learning_rate": 1.5628671255306706e-07,
      "loss": 0.3146,
      "num_input_tokens_seen": 20574648,
      "step": 31485
    },
    {
      "epoch": 18.567216981132077,
      "grad_norm": 2.4808108806610107,
      "learning_rate": 1.556491460925863e-07,
      "loss": 0.3294,
      "num_input_tokens_seen": 20579256,
      "step": 31490
    },
    {
      "epoch": 18.57016509433962,
      "grad_norm": 4.925533294677734,
      "learning_rate": 1.550128621948721e-07,
      "loss": 0.2623,
      "num_input_tokens_seen": 20582072,
      "step": 31495
    },
    {
      "epoch": 18.57311320754717,
      "grad_norm": 3.4174320697784424,
      "learning_rate": 1.5437786102838413e-07,
      "loss": 0.291,
      "num_input_tokens_seen": 20585528,
      "step": 31500
    },
    {
      "epoch": 18.576061320754718,
      "grad_norm": 2.072932481765747,
      "learning_rate": 1.5374414276124017e-07,
      "loss": 0.4489,
      "num_input_tokens_seen": 20589112,
      "step": 31505
    },
    {
      "epoch": 18.579009433962263,
      "grad_norm": 4.864414691925049,
      "learning_rate": 1.5311170756122095e-07,
      "loss": 0.2395,
      "num_input_tokens_seen": 20591608,
      "step": 31510
    },
    {
      "epoch": 18.58195754716981,
      "grad_norm": 3.8076529502868652,
      "learning_rate": 1.5248055559576647e-07,
      "loss": 0.3452,
      "num_input_tokens_seen": 20594456,
      "step": 31515
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 3.093775987625122,
      "learning_rate": 1.5185068703197526e-07,
      "loss": 0.3456,
      "num_input_tokens_seen": 20599096,
      "step": 31520
    },
    {
      "epoch": 18.587853773584907,
      "grad_norm": 4.743027210235596,
      "learning_rate": 1.5122210203661004e-07,
      "loss": 0.2909,
      "num_input_tokens_seen": 20601976,
      "step": 31525
    },
    {
      "epoch": 18.590801886792452,
      "grad_norm": 2.736938953399658,
      "learning_rate": 1.505948007760899e-07,
      "loss": 0.264,
      "num_input_tokens_seen": 20604760,
      "step": 31530
    },
    {
      "epoch": 18.59375,
      "grad_norm": 3.3961496353149414,
      "learning_rate": 1.4996878341649647e-07,
      "loss": 0.304,
      "num_input_tokens_seen": 20608440,
      "step": 31535
    },
    {
      "epoch": 18.596698113207548,
      "grad_norm": 2.9841835498809814,
      "learning_rate": 1.493440501235699e-07,
      "loss": 0.3498,
      "num_input_tokens_seen": 20611800,
      "step": 31540
    },
    {
      "epoch": 18.599646226415093,
      "grad_norm": 3.5383758544921875,
      "learning_rate": 1.487206010627118e-07,
      "loss": 0.3193,
      "num_input_tokens_seen": 20615224,
      "step": 31545
    },
    {
      "epoch": 18.60259433962264,
      "grad_norm": 2.1916186809539795,
      "learning_rate": 1.4809843639898124e-07,
      "loss": 0.2579,
      "num_input_tokens_seen": 20618232,
      "step": 31550
    },
    {
      "epoch": 18.60554245283019,
      "grad_norm": 2.416882276535034,
      "learning_rate": 1.4747755629710093e-07,
      "loss": 0.2066,
      "num_input_tokens_seen": 20621432,
      "step": 31555
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 3.8703255653381348,
      "learning_rate": 1.4685796092145045e-07,
      "loss": 0.2604,
      "num_input_tokens_seen": 20624504,
      "step": 31560
    },
    {
      "epoch": 18.611438679245282,
      "grad_norm": 5.036919116973877,
      "learning_rate": 1.4623965043607135e-07,
      "loss": 0.2514,
      "num_input_tokens_seen": 20626936,
      "step": 31565
    },
    {
      "epoch": 18.61438679245283,
      "grad_norm": 4.971261024475098,
      "learning_rate": 1.4562262500466273e-07,
      "loss": 0.3164,
      "num_input_tokens_seen": 20630232,
      "step": 31570
    },
    {
      "epoch": 18.61733490566038,
      "grad_norm": 2.2076990604400635,
      "learning_rate": 1.4500688479058556e-07,
      "loss": 0.2906,
      "num_input_tokens_seen": 20632920,
      "step": 31575
    },
    {
      "epoch": 18.620283018867923,
      "grad_norm": 3.472348213195801,
      "learning_rate": 1.4439242995685943e-07,
      "loss": 0.3934,
      "num_input_tokens_seen": 20636024,
      "step": 31580
    },
    {
      "epoch": 18.62323113207547,
      "grad_norm": 2.947403907775879,
      "learning_rate": 1.4377926066616364e-07,
      "loss": 0.3753,
      "num_input_tokens_seen": 20639128,
      "step": 31585
    },
    {
      "epoch": 18.62617924528302,
      "grad_norm": 3.1181037425994873,
      "learning_rate": 1.4316737708083783e-07,
      "loss": 0.4411,
      "num_input_tokens_seen": 20642360,
      "step": 31590
    },
    {
      "epoch": 18.629127358490567,
      "grad_norm": 3.470388412475586,
      "learning_rate": 1.4255677936288127e-07,
      "loss": 0.4119,
      "num_input_tokens_seen": 20644952,
      "step": 31595
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 2.8251824378967285,
      "learning_rate": 1.4194746767395184e-07,
      "loss": 0.3353,
      "num_input_tokens_seen": 20648280,
      "step": 31600
    },
    {
      "epoch": 18.63502358490566,
      "grad_norm": 5.370226860046387,
      "learning_rate": 1.4133944217536722e-07,
      "loss": 0.4761,
      "num_input_tokens_seen": 20651096,
      "step": 31605
    },
    {
      "epoch": 18.63797169811321,
      "grad_norm": 5.281558036804199,
      "learning_rate": 1.4073270302810471e-07,
      "loss": 0.3006,
      "num_input_tokens_seen": 20653912,
      "step": 31610
    },
    {
      "epoch": 18.640919811320753,
      "grad_norm": 2.5045106410980225,
      "learning_rate": 1.4012725039280084e-07,
      "loss": 0.4319,
      "num_input_tokens_seen": 20657368,
      "step": 31615
    },
    {
      "epoch": 18.6438679245283,
      "grad_norm": 3.949568510055542,
      "learning_rate": 1.3952308442975292e-07,
      "loss": 0.2811,
      "num_input_tokens_seen": 20659864,
      "step": 31620
    },
    {
      "epoch": 18.64681603773585,
      "grad_norm": 2.651991605758667,
      "learning_rate": 1.3892020529891637e-07,
      "loss": 0.3217,
      "num_input_tokens_seen": 20663384,
      "step": 31625
    },
    {
      "epoch": 18.649764150943398,
      "grad_norm": 3.6629958152770996,
      "learning_rate": 1.3831861315990514e-07,
      "loss": 0.3257,
      "num_input_tokens_seen": 20666520,
      "step": 31630
    },
    {
      "epoch": 18.652712264150942,
      "grad_norm": 5.021721839904785,
      "learning_rate": 1.377183081719935e-07,
      "loss": 0.4521,
      "num_input_tokens_seen": 20669656,
      "step": 31635
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 3.078813314437866,
      "learning_rate": 1.3711929049411544e-07,
      "loss": 0.2361,
      "num_input_tokens_seen": 20672344,
      "step": 31640
    },
    {
      "epoch": 18.65860849056604,
      "grad_norm": 2.3300604820251465,
      "learning_rate": 1.365215602848624e-07,
      "loss": 0.3786,
      "num_input_tokens_seen": 20675288,
      "step": 31645
    },
    {
      "epoch": 18.661556603773583,
      "grad_norm": 3.96201753616333,
      "learning_rate": 1.3592511770248727e-07,
      "loss": 0.2864,
      "num_input_tokens_seen": 20677880,
      "step": 31650
    },
    {
      "epoch": 18.66450471698113,
      "grad_norm": 3.5575389862060547,
      "learning_rate": 1.3532996290490041e-07,
      "loss": 0.365,
      "num_input_tokens_seen": 20681400,
      "step": 31655
    },
    {
      "epoch": 18.66745283018868,
      "grad_norm": 3.7428221702575684,
      "learning_rate": 1.347360960496713e-07,
      "loss": 0.2258,
      "num_input_tokens_seen": 20684792,
      "step": 31660
    },
    {
      "epoch": 18.670400943396228,
      "grad_norm": 2.868083953857422,
      "learning_rate": 1.3414351729402862e-07,
      "loss": 0.231,
      "num_input_tokens_seen": 20688312,
      "step": 31665
    },
    {
      "epoch": 18.673349056603772,
      "grad_norm": 4.7220234870910645,
      "learning_rate": 1.3355222679486025e-07,
      "loss": 0.3312,
      "num_input_tokens_seen": 20692376,
      "step": 31670
    },
    {
      "epoch": 18.67629716981132,
      "grad_norm": 4.025238037109375,
      "learning_rate": 1.3296222470871367e-07,
      "loss": 0.2706,
      "num_input_tokens_seen": 20694712,
      "step": 31675
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 6.336673259735107,
      "learning_rate": 1.3237351119179287e-07,
      "loss": 0.282,
      "num_input_tokens_seen": 20697464,
      "step": 31680
    },
    {
      "epoch": 18.682193396226417,
      "grad_norm": 4.343964099884033,
      "learning_rate": 1.3178608639996425e-07,
      "loss": 0.2948,
      "num_input_tokens_seen": 20701592,
      "step": 31685
    },
    {
      "epoch": 18.68514150943396,
      "grad_norm": 2.0135045051574707,
      "learning_rate": 1.3119995048874957e-07,
      "loss": 0.3167,
      "num_input_tokens_seen": 20705016,
      "step": 31690
    },
    {
      "epoch": 18.68808962264151,
      "grad_norm": 2.0743117332458496,
      "learning_rate": 1.3061510361333186e-07,
      "loss": 0.2435,
      "num_input_tokens_seen": 20709144,
      "step": 31695
    },
    {
      "epoch": 18.691037735849058,
      "grad_norm": 2.8856263160705566,
      "learning_rate": 1.3003154592855116e-07,
      "loss": 0.2045,
      "num_input_tokens_seen": 20712504,
      "step": 31700
    },
    {
      "epoch": 18.693985849056602,
      "grad_norm": 2.2958672046661377,
      "learning_rate": 1.2944927758890668e-07,
      "loss": 0.2479,
      "num_input_tokens_seen": 20715960,
      "step": 31705
    },
    {
      "epoch": 18.69693396226415,
      "grad_norm": 6.298005104064941,
      "learning_rate": 1.2886829874855733e-07,
      "loss": 0.2909,
      "num_input_tokens_seen": 20718424,
      "step": 31710
    },
    {
      "epoch": 18.6998820754717,
      "grad_norm": 6.039846420288086,
      "learning_rate": 1.2828860956131894e-07,
      "loss": 0.3159,
      "num_input_tokens_seen": 20721144,
      "step": 31715
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 3.4153084754943848,
      "learning_rate": 1.2771021018066765e-07,
      "loss": 0.347,
      "num_input_tokens_seen": 20724056,
      "step": 31720
    },
    {
      "epoch": 18.70577830188679,
      "grad_norm": 2.9274404048919678,
      "learning_rate": 1.271331007597365e-07,
      "loss": 0.2674,
      "num_input_tokens_seen": 20727224,
      "step": 31725
    },
    {
      "epoch": 18.70872641509434,
      "grad_norm": 2.4449400901794434,
      "learning_rate": 1.2655728145131774e-07,
      "loss": 0.2986,
      "num_input_tokens_seen": 20730264,
      "step": 31730
    },
    {
      "epoch": 18.711674528301888,
      "grad_norm": 3.304661989212036,
      "learning_rate": 1.2598275240786105e-07,
      "loss": 0.3088,
      "num_input_tokens_seen": 20734200,
      "step": 31735
    },
    {
      "epoch": 18.714622641509433,
      "grad_norm": 8.008398056030273,
      "learning_rate": 1.254095137814776e-07,
      "loss": 0.4221,
      "num_input_tokens_seen": 20738328,
      "step": 31740
    },
    {
      "epoch": 18.71757075471698,
      "grad_norm": 3.5171782970428467,
      "learning_rate": 1.2483756572393368e-07,
      "loss": 0.2737,
      "num_input_tokens_seen": 20741368,
      "step": 31745
    },
    {
      "epoch": 18.72051886792453,
      "grad_norm": 4.554544448852539,
      "learning_rate": 1.242669083866549e-07,
      "loss": 0.273,
      "num_input_tokens_seen": 20744440,
      "step": 31750
    },
    {
      "epoch": 18.723466981132077,
      "grad_norm": 3.189743995666504,
      "learning_rate": 1.2369754192072537e-07,
      "loss": 0.2532,
      "num_input_tokens_seen": 20748056,
      "step": 31755
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 2.100041627883911,
      "learning_rate": 1.231294664768873e-07,
      "loss": 0.2881,
      "num_input_tokens_seen": 20751000,
      "step": 31760
    },
    {
      "epoch": 18.72936320754717,
      "grad_norm": 3.393517017364502,
      "learning_rate": 1.225626822055409e-07,
      "loss": 0.2982,
      "num_input_tokens_seen": 20754200,
      "step": 31765
    },
    {
      "epoch": 18.732311320754718,
      "grad_norm": 4.065339088439941,
      "learning_rate": 1.2199718925674508e-07,
      "loss": 0.3627,
      "num_input_tokens_seen": 20757688,
      "step": 31770
    },
    {
      "epoch": 18.735259433962263,
      "grad_norm": 3.204542398452759,
      "learning_rate": 1.2143298778021616e-07,
      "loss": 0.3479,
      "num_input_tokens_seen": 20761656,
      "step": 31775
    },
    {
      "epoch": 18.73820754716981,
      "grad_norm": 5.862226486206055,
      "learning_rate": 1.2087007792532967e-07,
      "loss": 0.3667,
      "num_input_tokens_seen": 20764376,
      "step": 31780
    },
    {
      "epoch": 18.74115566037736,
      "grad_norm": 3.2389800548553467,
      "learning_rate": 1.203084598411175e-07,
      "loss": 0.3035,
      "num_input_tokens_seen": 20767288,
      "step": 31785
    },
    {
      "epoch": 18.744103773584907,
      "grad_norm": 4.205025672912598,
      "learning_rate": 1.1974813367627124e-07,
      "loss": 0.2982,
      "num_input_tokens_seen": 20770392,
      "step": 31790
    },
    {
      "epoch": 18.747051886792452,
      "grad_norm": 3.325157880783081,
      "learning_rate": 1.1918909957913949e-07,
      "loss": 0.3186,
      "num_input_tokens_seen": 20772984,
      "step": 31795
    },
    {
      "epoch": 18.75,
      "grad_norm": 4.6184539794921875,
      "learning_rate": 1.1863135769772827e-07,
      "loss": 0.3513,
      "num_input_tokens_seen": 20776632,
      "step": 31800
    },
    {
      "epoch": 18.752948113207548,
      "grad_norm": 3.1012983322143555,
      "learning_rate": 1.1807490817970279e-07,
      "loss": 0.3095,
      "num_input_tokens_seen": 20779864,
      "step": 31805
    },
    {
      "epoch": 18.755896226415093,
      "grad_norm": 4.732428073883057,
      "learning_rate": 1.1751975117238578e-07,
      "loss": 0.3417,
      "num_input_tokens_seen": 20782552,
      "step": 31810
    },
    {
      "epoch": 18.75884433962264,
      "grad_norm": 3.60779070854187,
      "learning_rate": 1.1696588682275633e-07,
      "loss": 0.3796,
      "num_input_tokens_seen": 20785560,
      "step": 31815
    },
    {
      "epoch": 18.76179245283019,
      "grad_norm": 3.496962785720825,
      "learning_rate": 1.1641331527745325e-07,
      "loss": 0.3129,
      "num_input_tokens_seen": 20789144,
      "step": 31820
    },
    {
      "epoch": 18.764740566037737,
      "grad_norm": 3.453718423843384,
      "learning_rate": 1.1586203668277229e-07,
      "loss": 0.282,
      "num_input_tokens_seen": 20791672,
      "step": 31825
    },
    {
      "epoch": 18.767688679245282,
      "grad_norm": 3.6641793251037598,
      "learning_rate": 1.1531205118466615e-07,
      "loss": 0.2535,
      "num_input_tokens_seen": 20794904,
      "step": 31830
    },
    {
      "epoch": 18.77063679245283,
      "grad_norm": 4.2788896560668945,
      "learning_rate": 1.1476335892874669e-07,
      "loss": 0.2409,
      "num_input_tokens_seen": 20797464,
      "step": 31835
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 5.028820991516113,
      "learning_rate": 1.1421596006028157e-07,
      "loss": 0.2655,
      "num_input_tokens_seen": 20800248,
      "step": 31840
    },
    {
      "epoch": 18.776533018867923,
      "grad_norm": 2.3371870517730713,
      "learning_rate": 1.1366985472419823e-07,
      "loss": 0.3029,
      "num_input_tokens_seen": 20804440,
      "step": 31845
    },
    {
      "epoch": 18.77948113207547,
      "grad_norm": 2.0208733081817627,
      "learning_rate": 1.1312504306507987e-07,
      "loss": 0.2166,
      "num_input_tokens_seen": 20807736,
      "step": 31850
    },
    {
      "epoch": 18.78242924528302,
      "grad_norm": 5.429619312286377,
      "learning_rate": 1.1258152522716725e-07,
      "loss": 0.4165,
      "num_input_tokens_seen": 20811192,
      "step": 31855
    },
    {
      "epoch": 18.785377358490567,
      "grad_norm": 3.13814640045166,
      "learning_rate": 1.1203930135435914e-07,
      "loss": 0.3136,
      "num_input_tokens_seen": 20814264,
      "step": 31860
    },
    {
      "epoch": 18.788325471698112,
      "grad_norm": 4.228353023529053,
      "learning_rate": 1.1149837159021238e-07,
      "loss": 0.3128,
      "num_input_tokens_seen": 20818616,
      "step": 31865
    },
    {
      "epoch": 18.79127358490566,
      "grad_norm": 3.4242944717407227,
      "learning_rate": 1.1095873607793961e-07,
      "loss": 0.2974,
      "num_input_tokens_seen": 20822328,
      "step": 31870
    },
    {
      "epoch": 18.79422169811321,
      "grad_norm": 2.8911428451538086,
      "learning_rate": 1.1042039496041212e-07,
      "loss": 0.471,
      "num_input_tokens_seen": 20824824,
      "step": 31875
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 2.876142740249634,
      "learning_rate": 1.0988334838015812e-07,
      "loss": 0.3432,
      "num_input_tokens_seen": 20828376,
      "step": 31880
    },
    {
      "epoch": 18.8001179245283,
      "grad_norm": 4.165853500366211,
      "learning_rate": 1.0934759647936333e-07,
      "loss": 0.2815,
      "num_input_tokens_seen": 20831896,
      "step": 31885
    },
    {
      "epoch": 18.80306603773585,
      "grad_norm": 2.8931679725646973,
      "learning_rate": 1.0881313939986926e-07,
      "loss": 0.2724,
      "num_input_tokens_seen": 20834872,
      "step": 31890
    },
    {
      "epoch": 18.806014150943398,
      "grad_norm": 3.6217899322509766,
      "learning_rate": 1.0827997728317662e-07,
      "loss": 0.259,
      "num_input_tokens_seen": 20837624,
      "step": 31895
    },
    {
      "epoch": 18.808962264150942,
      "grad_norm": 6.679104328155518,
      "learning_rate": 1.0774811027044196e-07,
      "loss": 0.3214,
      "num_input_tokens_seen": 20841528,
      "step": 31900
    },
    {
      "epoch": 18.81191037735849,
      "grad_norm": 3.3193917274475098,
      "learning_rate": 1.0721753850247984e-07,
      "loss": 0.42,
      "num_input_tokens_seen": 20845112,
      "step": 31905
    },
    {
      "epoch": 18.81485849056604,
      "grad_norm": 2.9678609371185303,
      "learning_rate": 1.0668826211976124e-07,
      "loss": 0.2601,
      "num_input_tokens_seen": 20847864,
      "step": 31910
    },
    {
      "epoch": 18.817806603773583,
      "grad_norm": 3.1075832843780518,
      "learning_rate": 1.0616028126241407e-07,
      "loss": 0.3048,
      "num_input_tokens_seen": 20850872,
      "step": 31915
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 4.115147590637207,
      "learning_rate": 1.0563359607022372e-07,
      "loss": 0.3119,
      "num_input_tokens_seen": 20853656,
      "step": 31920
    },
    {
      "epoch": 18.82370283018868,
      "grad_norm": 3.2456154823303223,
      "learning_rate": 1.05108206682632e-07,
      "loss": 0.3586,
      "num_input_tokens_seen": 20856920,
      "step": 31925
    },
    {
      "epoch": 18.826650943396228,
      "grad_norm": 3.075684070587158,
      "learning_rate": 1.0458411323873874e-07,
      "loss": 0.3695,
      "num_input_tokens_seen": 20861080,
      "step": 31930
    },
    {
      "epoch": 18.829599056603772,
      "grad_norm": 4.133026599884033,
      "learning_rate": 1.0406131587729962e-07,
      "loss": 0.3321,
      "num_input_tokens_seen": 20863704,
      "step": 31935
    },
    {
      "epoch": 18.83254716981132,
      "grad_norm": 3.7677576541900635,
      "learning_rate": 1.035398147367278e-07,
      "loss": 0.2297,
      "num_input_tokens_seen": 20866360,
      "step": 31940
    },
    {
      "epoch": 18.83549528301887,
      "grad_norm": 3.071049213409424,
      "learning_rate": 1.030196099550923e-07,
      "loss": 0.3628,
      "num_input_tokens_seen": 20869208,
      "step": 31945
    },
    {
      "epoch": 18.838443396226417,
      "grad_norm": 2.9188904762268066,
      "learning_rate": 1.0250070167011905e-07,
      "loss": 0.246,
      "num_input_tokens_seen": 20872088,
      "step": 31950
    },
    {
      "epoch": 18.84139150943396,
      "grad_norm": 3.588911771774292,
      "learning_rate": 1.0198309001919315e-07,
      "loss": 0.2854,
      "num_input_tokens_seen": 20875576,
      "step": 31955
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 3.3322012424468994,
      "learning_rate": 1.0146677513935277e-07,
      "loss": 0.2323,
      "num_input_tokens_seen": 20879768,
      "step": 31960
    },
    {
      "epoch": 18.847287735849058,
      "grad_norm": 3.878354787826538,
      "learning_rate": 1.0095175716729578e-07,
      "loss": 0.3095,
      "num_input_tokens_seen": 20882776,
      "step": 31965
    },
    {
      "epoch": 18.850235849056602,
      "grad_norm": 2.629302978515625,
      "learning_rate": 1.004380362393742e-07,
      "loss": 0.2321,
      "num_input_tokens_seen": 20885304,
      "step": 31970
    },
    {
      "epoch": 18.85318396226415,
      "grad_norm": 4.996763706207275,
      "learning_rate": 9.99256124915987e-08,
      "loss": 0.4014,
      "num_input_tokens_seen": 20889400,
      "step": 31975
    },
    {
      "epoch": 18.8561320754717,
      "grad_norm": 6.278010845184326,
      "learning_rate": 9.941448605963577e-08,
      "loss": 0.3145,
      "num_input_tokens_seen": 20892312,
      "step": 31980
    },
    {
      "epoch": 18.859080188679247,
      "grad_norm": 4.6514763832092285,
      "learning_rate": 9.890465707880715e-08,
      "loss": 0.2929,
      "num_input_tokens_seen": 20895480,
      "step": 31985
    },
    {
      "epoch": 18.86202830188679,
      "grad_norm": 2.291416645050049,
      "learning_rate": 9.839612568409374e-08,
      "loss": 0.2154,
      "num_input_tokens_seen": 20898904,
      "step": 31990
    },
    {
      "epoch": 18.86497641509434,
      "grad_norm": 2.6157078742980957,
      "learning_rate": 9.788889201013119e-08,
      "loss": 0.3013,
      "num_input_tokens_seen": 20902936,
      "step": 31995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 2.940873146057129,
      "learning_rate": 9.738295619121097e-08,
      "loss": 0.3525,
      "num_input_tokens_seen": 20905688,
      "step": 32000
    },
    {
      "epoch": 18.870872641509433,
      "grad_norm": 2.5423765182495117,
      "learning_rate": 9.687831836128203e-08,
      "loss": 0.2355,
      "num_input_tokens_seen": 20908408,
      "step": 32005
    },
    {
      "epoch": 18.87382075471698,
      "grad_norm": 4.924959182739258,
      "learning_rate": 9.637497865395029e-08,
      "loss": 0.3234,
      "num_input_tokens_seen": 20910616,
      "step": 32010
    },
    {
      "epoch": 18.87676886792453,
      "grad_norm": 5.004456520080566,
      "learning_rate": 9.587293720247526e-08,
      "loss": 0.2678,
      "num_input_tokens_seen": 20913912,
      "step": 32015
    },
    {
      "epoch": 18.879716981132077,
      "grad_norm": 7.090242862701416,
      "learning_rate": 9.537219413977672e-08,
      "loss": 0.3788,
      "num_input_tokens_seen": 20917400,
      "step": 32020
    },
    {
      "epoch": 18.88266509433962,
      "grad_norm": 3.1778082847595215,
      "learning_rate": 9.487274959842696e-08,
      "loss": 0.3106,
      "num_input_tokens_seen": 20920728,
      "step": 32025
    },
    {
      "epoch": 18.88561320754717,
      "grad_norm": 3.1165354251861572,
      "learning_rate": 9.437460371065687e-08,
      "loss": 0.311,
      "num_input_tokens_seen": 20925208,
      "step": 32030
    },
    {
      "epoch": 18.888561320754718,
      "grad_norm": 3.892453908920288,
      "learning_rate": 9.387775660835263e-08,
      "loss": 0.4193,
      "num_input_tokens_seen": 20928120,
      "step": 32035
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 4.9133477210998535,
      "learning_rate": 9.338220842305678e-08,
      "loss": 0.3086,
      "num_input_tokens_seen": 20930712,
      "step": 32040
    },
    {
      "epoch": 18.89445754716981,
      "grad_norm": 3.5342397689819336,
      "learning_rate": 9.288795928596661e-08,
      "loss": 0.287,
      "num_input_tokens_seen": 20934776,
      "step": 32045
    },
    {
      "epoch": 18.89740566037736,
      "grad_norm": 2.929293632507324,
      "learning_rate": 9.239500932793854e-08,
      "loss": 0.4207,
      "num_input_tokens_seen": 20938232,
      "step": 32050
    },
    {
      "epoch": 18.900353773584907,
      "grad_norm": 8.258818626403809,
      "learning_rate": 9.190335867948263e-08,
      "loss": 0.283,
      "num_input_tokens_seen": 20941752,
      "step": 32055
    },
    {
      "epoch": 18.903301886792452,
      "grad_norm": 2.9520339965820312,
      "learning_rate": 9.141300747076476e-08,
      "loss": 0.3058,
      "num_input_tokens_seen": 20944984,
      "step": 32060
    },
    {
      "epoch": 18.90625,
      "grad_norm": 3.6605069637298584,
      "learning_rate": 9.092395583160773e-08,
      "loss": 0.2909,
      "num_input_tokens_seen": 20947800,
      "step": 32065
    },
    {
      "epoch": 18.909198113207548,
      "grad_norm": 3.652766466140747,
      "learning_rate": 9.043620389149021e-08,
      "loss": 0.2178,
      "num_input_tokens_seen": 20952216,
      "step": 32070
    },
    {
      "epoch": 18.912146226415093,
      "grad_norm": 3.3913583755493164,
      "learning_rate": 8.994975177954723e-08,
      "loss": 0.364,
      "num_input_tokens_seen": 20955832,
      "step": 32075
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 1.864546537399292,
      "learning_rate": 8.946459962456855e-08,
      "loss": 0.2838,
      "num_input_tokens_seen": 20958328,
      "step": 32080
    },
    {
      "epoch": 18.91804245283019,
      "grad_norm": 2.6297452449798584,
      "learning_rate": 8.89807475550003e-08,
      "loss": 0.2903,
      "num_input_tokens_seen": 20961752,
      "step": 32085
    },
    {
      "epoch": 18.920990566037737,
      "grad_norm": 3.3460910320281982,
      "learning_rate": 8.849819569894447e-08,
      "loss": 0.3267,
      "num_input_tokens_seen": 20965240,
      "step": 32090
    },
    {
      "epoch": 18.923938679245282,
      "grad_norm": 3.364112615585327,
      "learning_rate": 8.801694418415884e-08,
      "loss": 0.2227,
      "num_input_tokens_seen": 20968088,
      "step": 32095
    },
    {
      "epoch": 18.92688679245283,
      "grad_norm": 4.892765045166016,
      "learning_rate": 8.753699313805708e-08,
      "loss": 0.2764,
      "num_input_tokens_seen": 20971416,
      "step": 32100
    },
    {
      "epoch": 18.92983490566038,
      "grad_norm": 4.3793535232543945,
      "learning_rate": 8.705834268770753e-08,
      "loss": 0.2608,
      "num_input_tokens_seen": 20973816,
      "step": 32105
    },
    {
      "epoch": 18.932783018867923,
      "grad_norm": 4.712729454040527,
      "learning_rate": 8.65809929598349e-08,
      "loss": 0.337,
      "num_input_tokens_seen": 20976664,
      "step": 32110
    },
    {
      "epoch": 18.93573113207547,
      "grad_norm": 3.306551456451416,
      "learning_rate": 8.610494408082037e-08,
      "loss": 0.4078,
      "num_input_tokens_seen": 20981272,
      "step": 32115
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 3.447371244430542,
      "learning_rate": 8.563019617669977e-08,
      "loss": 0.2276,
      "num_input_tokens_seen": 20983960,
      "step": 32120
    },
    {
      "epoch": 18.941627358490567,
      "grad_norm": 3.1482737064361572,
      "learning_rate": 8.51567493731642e-08,
      "loss": 0.344,
      "num_input_tokens_seen": 20987128,
      "step": 32125
    },
    {
      "epoch": 18.944575471698112,
      "grad_norm": 9.31950855255127,
      "learning_rate": 8.468460379556176e-08,
      "loss": 0.3803,
      "num_input_tokens_seen": 20989880,
      "step": 32130
    },
    {
      "epoch": 18.94752358490566,
      "grad_norm": 3.4820754528045654,
      "learning_rate": 8.421375956889355e-08,
      "loss": 0.216,
      "num_input_tokens_seen": 20992920,
      "step": 32135
    },
    {
      "epoch": 18.95047169811321,
      "grad_norm": 2.988152503967285,
      "learning_rate": 8.374421681781819e-08,
      "loss": 0.202,
      "num_input_tokens_seen": 20995672,
      "step": 32140
    },
    {
      "epoch": 18.953419811320753,
      "grad_norm": 3.900851249694824,
      "learning_rate": 8.327597566665013e-08,
      "loss": 0.3611,
      "num_input_tokens_seen": 20999000,
      "step": 32145
    },
    {
      "epoch": 18.9563679245283,
      "grad_norm": 3.582296848297119,
      "learning_rate": 8.280903623935688e-08,
      "loss": 0.2616,
      "num_input_tokens_seen": 21002360,
      "step": 32150
    },
    {
      "epoch": 18.95931603773585,
      "grad_norm": 4.778347492218018,
      "learning_rate": 8.234339865956342e-08,
      "loss": 0.3889,
      "num_input_tokens_seen": 21005400,
      "step": 32155
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 2.9355101585388184,
      "learning_rate": 8.187906305054838e-08,
      "loss": 0.3598,
      "num_input_tokens_seen": 21008856,
      "step": 32160
    },
    {
      "epoch": 18.965212264150942,
      "grad_norm": 3.4005956649780273,
      "learning_rate": 8.141602953524841e-08,
      "loss": 0.4335,
      "num_input_tokens_seen": 21012056,
      "step": 32165
    },
    {
      "epoch": 18.96816037735849,
      "grad_norm": 3.149977445602417,
      "learning_rate": 8.095429823625212e-08,
      "loss": 0.2893,
      "num_input_tokens_seen": 21015224,
      "step": 32170
    },
    {
      "epoch": 18.97110849056604,
      "grad_norm": 2.5119237899780273,
      "learning_rate": 8.04938692758045e-08,
      "loss": 0.3491,
      "num_input_tokens_seen": 21019288,
      "step": 32175
    },
    {
      "epoch": 18.974056603773583,
      "grad_norm": 3.1651217937469482,
      "learning_rate": 8.003474277580803e-08,
      "loss": 0.4003,
      "num_input_tokens_seen": 21022456,
      "step": 32180
    },
    {
      "epoch": 18.97700471698113,
      "grad_norm": 2.109919548034668,
      "learning_rate": 7.95769188578166e-08,
      "loss": 0.2891,
      "num_input_tokens_seen": 21026200,
      "step": 32185
    },
    {
      "epoch": 18.97995283018868,
      "grad_norm": 3.2344510555267334,
      "learning_rate": 7.912039764304213e-08,
      "loss": 0.3146,
      "num_input_tokens_seen": 21028952,
      "step": 32190
    },
    {
      "epoch": 18.982900943396228,
      "grad_norm": 3.212864875793457,
      "learning_rate": 7.866517925235017e-08,
      "loss": 0.352,
      "num_input_tokens_seen": 21031512,
      "step": 32195
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 4.646742343902588,
      "learning_rate": 7.821126380626154e-08,
      "loss": 0.3065,
      "num_input_tokens_seen": 21034136,
      "step": 32200
    },
    {
      "epoch": 18.98879716981132,
      "grad_norm": 2.9951364994049072,
      "learning_rate": 7.775865142495286e-08,
      "loss": 0.4162,
      "num_input_tokens_seen": 21037592,
      "step": 32205
    },
    {
      "epoch": 18.99174528301887,
      "grad_norm": 2.7441611289978027,
      "learning_rate": 7.730734222825442e-08,
      "loss": 0.1843,
      "num_input_tokens_seen": 21040984,
      "step": 32210
    },
    {
      "epoch": 18.994693396226417,
      "grad_norm": 3.0859601497650146,
      "learning_rate": 7.68573363356534e-08,
      "loss": 0.3279,
      "num_input_tokens_seen": 21044344,
      "step": 32215
    },
    {
      "epoch": 18.99764150943396,
      "grad_norm": 5.8634138107299805,
      "learning_rate": 7.640863386629005e-08,
      "loss": 0.2966,
      "num_input_tokens_seen": 21046808,
      "step": 32220
    },
    {
      "epoch": 19.00058962264151,
      "grad_norm": 4.466851234436035,
      "learning_rate": 7.59612349389599e-08,
      "loss": 0.2178,
      "num_input_tokens_seen": 21049520,
      "step": 32225
    },
    {
      "epoch": 19.003537735849058,
      "grad_norm": 2.3910467624664307,
      "learning_rate": 7.551513967211433e-08,
      "loss": 0.2604,
      "num_input_tokens_seen": 21052336,
      "step": 32230
    },
    {
      "epoch": 19.006485849056602,
      "grad_norm": 3.574899435043335,
      "learning_rate": 7.507034818385883e-08,
      "loss": 0.3962,
      "num_input_tokens_seen": 21055664,
      "step": 32235
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 1.482347846031189,
      "learning_rate": 7.462686059195423e-08,
      "loss": 0.2733,
      "num_input_tokens_seen": 21058928,
      "step": 32240
    },
    {
      "epoch": 19.0123820754717,
      "grad_norm": 5.55985689163208,
      "learning_rate": 7.418467701381548e-08,
      "loss": 0.331,
      "num_input_tokens_seen": 21061456,
      "step": 32245
    },
    {
      "epoch": 19.015330188679247,
      "grad_norm": 3.476004123687744,
      "learning_rate": 7.374379756651285e-08,
      "loss": 0.3092,
      "num_input_tokens_seen": 21064656,
      "step": 32250
    },
    {
      "epoch": 19.01827830188679,
      "grad_norm": 4.624046325683594,
      "learning_rate": 7.330422236677015e-08,
      "loss": 0.2696,
      "num_input_tokens_seen": 21068592,
      "step": 32255
    },
    {
      "epoch": 19.02122641509434,
      "grad_norm": 6.652890205383301,
      "learning_rate": 7.286595153096765e-08,
      "loss": 0.4931,
      "num_input_tokens_seen": 21071120,
      "step": 32260
    },
    {
      "epoch": 19.024174528301888,
      "grad_norm": 5.831721782684326,
      "learning_rate": 7.242898517513864e-08,
      "loss": 0.1815,
      "num_input_tokens_seen": 21074352,
      "step": 32265
    },
    {
      "epoch": 19.027122641509433,
      "grad_norm": 3.1537668704986572,
      "learning_rate": 7.199332341497333e-08,
      "loss": 0.3782,
      "num_input_tokens_seen": 21078576,
      "step": 32270
    },
    {
      "epoch": 19.03007075471698,
      "grad_norm": 4.329867362976074,
      "learning_rate": 7.155896636581394e-08,
      "loss": 0.2692,
      "num_input_tokens_seen": 21081040,
      "step": 32275
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 4.183619499206543,
      "learning_rate": 7.112591414265901e-08,
      "loss": 0.4164,
      "num_input_tokens_seen": 21084272,
      "step": 32280
    },
    {
      "epoch": 19.035966981132077,
      "grad_norm": 9.193342208862305,
      "learning_rate": 7.069416686016018e-08,
      "loss": 0.2212,
      "num_input_tokens_seen": 21087088,
      "step": 32285
    },
    {
      "epoch": 19.03891509433962,
      "grad_norm": 3.3148770332336426,
      "learning_rate": 7.026372463262488e-08,
      "loss": 0.2802,
      "num_input_tokens_seen": 21090448,
      "step": 32290
    },
    {
      "epoch": 19.04186320754717,
      "grad_norm": 6.1764421463012695,
      "learning_rate": 6.983458757401418e-08,
      "loss": 0.32,
      "num_input_tokens_seen": 21094352,
      "step": 32295
    },
    {
      "epoch": 19.044811320754718,
      "grad_norm": 3.9617221355438232,
      "learning_rate": 6.940675579794443e-08,
      "loss": 0.2999,
      "num_input_tokens_seen": 21097296,
      "step": 32300
    },
    {
      "epoch": 19.047759433962263,
      "grad_norm": 6.074903964996338,
      "learning_rate": 6.898022941768612e-08,
      "loss": 0.3598,
      "num_input_tokens_seen": 21100144,
      "step": 32305
    },
    {
      "epoch": 19.05070754716981,
      "grad_norm": 2.772458553314209,
      "learning_rate": 6.855500854616337e-08,
      "loss": 0.2567,
      "num_input_tokens_seen": 21104816,
      "step": 32310
    },
    {
      "epoch": 19.05365566037736,
      "grad_norm": 5.450234889984131,
      "learning_rate": 6.813109329595557e-08,
      "loss": 0.2944,
      "num_input_tokens_seen": 21108464,
      "step": 32315
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 1.8981505632400513,
      "learning_rate": 6.770848377929573e-08,
      "loss": 0.2218,
      "num_input_tokens_seen": 21111728,
      "step": 32320
    },
    {
      "epoch": 19.059551886792452,
      "grad_norm": 3.8462514877319336,
      "learning_rate": 6.728718010807156e-08,
      "loss": 0.3646,
      "num_input_tokens_seen": 21115088,
      "step": 32325
    },
    {
      "epoch": 19.0625,
      "grad_norm": 4.526978492736816,
      "learning_rate": 6.68671823938255e-08,
      "loss": 0.35,
      "num_input_tokens_seen": 21117840,
      "step": 32330
    },
    {
      "epoch": 19.065448113207548,
      "grad_norm": 5.243165016174316,
      "learning_rate": 6.644849074775361e-08,
      "loss": 0.3487,
      "num_input_tokens_seen": 21120688,
      "step": 32335
    },
    {
      "epoch": 19.068396226415093,
      "grad_norm": 3.0095505714416504,
      "learning_rate": 6.603110528070667e-08,
      "loss": 0.2828,
      "num_input_tokens_seen": 21123856,
      "step": 32340
    },
    {
      "epoch": 19.07134433962264,
      "grad_norm": 3.3154735565185547,
      "learning_rate": 6.561502610318849e-08,
      "loss": 0.3645,
      "num_input_tokens_seen": 21126544,
      "step": 32345
    },
    {
      "epoch": 19.07429245283019,
      "grad_norm": 6.6543707847595215,
      "learning_rate": 6.520025332535762e-08,
      "loss": 0.3101,
      "num_input_tokens_seen": 21129136,
      "step": 32350
    },
    {
      "epoch": 19.077240566037737,
      "grad_norm": 2.266923666000366,
      "learning_rate": 6.47867870570279e-08,
      "loss": 0.3519,
      "num_input_tokens_seen": 21132432,
      "step": 32355
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 3.8087267875671387,
      "learning_rate": 6.437462740766564e-08,
      "loss": 0.2879,
      "num_input_tokens_seen": 21135728,
      "step": 32360
    },
    {
      "epoch": 19.08313679245283,
      "grad_norm": 3.6567394733428955,
      "learning_rate": 6.396377448639246e-08,
      "loss": 0.2354,
      "num_input_tokens_seen": 21139600,
      "step": 32365
    },
    {
      "epoch": 19.08608490566038,
      "grad_norm": 2.9689459800720215,
      "learning_rate": 6.3554228401983e-08,
      "loss": 0.3486,
      "num_input_tokens_seen": 21142832,
      "step": 32370
    },
    {
      "epoch": 19.089033018867923,
      "grad_norm": 4.035787582397461,
      "learning_rate": 6.314598926286663e-08,
      "loss": 0.336,
      "num_input_tokens_seen": 21146320,
      "step": 32375
    },
    {
      "epoch": 19.09198113207547,
      "grad_norm": 3.490427017211914,
      "learning_rate": 6.273905717712637e-08,
      "loss": 0.3503,
      "num_input_tokens_seen": 21149648,
      "step": 32380
    },
    {
      "epoch": 19.09492924528302,
      "grad_norm": 6.191030502319336,
      "learning_rate": 6.233343225249933e-08,
      "loss": 0.3301,
      "num_input_tokens_seen": 21152912,
      "step": 32385
    },
    {
      "epoch": 19.097877358490567,
      "grad_norm": 4.738461494445801,
      "learning_rate": 6.192911459637519e-08,
      "loss": 0.2986,
      "num_input_tokens_seen": 21155568,
      "step": 32390
    },
    {
      "epoch": 19.100825471698112,
      "grad_norm": 2.6290602684020996,
      "learning_rate": 6.152610431580052e-08,
      "loss": 0.3465,
      "num_input_tokens_seen": 21159696,
      "step": 32395
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 3.4034714698791504,
      "learning_rate": 6.112440151747389e-08,
      "loss": 0.3309,
      "num_input_tokens_seen": 21163504,
      "step": 32400
    },
    {
      "epoch": 19.10672169811321,
      "grad_norm": 2.7487645149230957,
      "learning_rate": 6.072400630774689e-08,
      "loss": 0.2777,
      "num_input_tokens_seen": 21167408,
      "step": 32405
    },
    {
      "epoch": 19.109669811320753,
      "grad_norm": 4.527805805206299,
      "learning_rate": 6.032491879262637e-08,
      "loss": 0.3598,
      "num_input_tokens_seen": 21169968,
      "step": 32410
    },
    {
      "epoch": 19.1126179245283,
      "grad_norm": 4.2484025955200195,
      "learning_rate": 5.99271390777717e-08,
      "loss": 0.3114,
      "num_input_tokens_seen": 21172784,
      "step": 32415
    },
    {
      "epoch": 19.11556603773585,
      "grad_norm": 4.512368679046631,
      "learning_rate": 5.953066726849865e-08,
      "loss": 0.2613,
      "num_input_tokens_seen": 21175472,
      "step": 32420
    },
    {
      "epoch": 19.118514150943398,
      "grad_norm": 2.9815480709075928,
      "learning_rate": 5.913550346977326e-08,
      "loss": 0.223,
      "num_input_tokens_seen": 21178928,
      "step": 32425
    },
    {
      "epoch": 19.121462264150942,
      "grad_norm": 3.645294189453125,
      "learning_rate": 5.874164778621683e-08,
      "loss": 0.3241,
      "num_input_tokens_seen": 21182160,
      "step": 32430
    },
    {
      "epoch": 19.12441037735849,
      "grad_norm": 4.07017707824707,
      "learning_rate": 5.834910032210539e-08,
      "loss": 0.3966,
      "num_input_tokens_seen": 21184944,
      "step": 32435
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 3.8011486530303955,
      "learning_rate": 5.795786118136693e-08,
      "loss": 0.3459,
      "num_input_tokens_seen": 21187760,
      "step": 32440
    },
    {
      "epoch": 19.130306603773583,
      "grad_norm": 3.9519546031951904,
      "learning_rate": 5.756793046758302e-08,
      "loss": 0.3009,
      "num_input_tokens_seen": 21190704,
      "step": 32445
    },
    {
      "epoch": 19.13325471698113,
      "grad_norm": 3.271130084991455,
      "learning_rate": 5.7179308283990544e-08,
      "loss": 0.194,
      "num_input_tokens_seen": 21196496,
      "step": 32450
    },
    {
      "epoch": 19.13620283018868,
      "grad_norm": 7.139911651611328,
      "learning_rate": 5.679199473347885e-08,
      "loss": 0.3494,
      "num_input_tokens_seen": 21198960,
      "step": 32455
    },
    {
      "epoch": 19.139150943396228,
      "grad_norm": 2.749239683151245,
      "learning_rate": 5.6405989918590366e-08,
      "loss": 0.3046,
      "num_input_tokens_seen": 21202032,
      "step": 32460
    },
    {
      "epoch": 19.142099056603772,
      "grad_norm": 3.613892078399658,
      "learning_rate": 5.6021293941522225e-08,
      "loss": 0.327,
      "num_input_tokens_seen": 21205296,
      "step": 32465
    },
    {
      "epoch": 19.14504716981132,
      "grad_norm": 6.142009735107422,
      "learning_rate": 5.563790690412352e-08,
      "loss": 0.2584,
      "num_input_tokens_seen": 21208496,
      "step": 32470
    },
    {
      "epoch": 19.14799528301887,
      "grad_norm": 2.7789769172668457,
      "learning_rate": 5.525582890789805e-08,
      "loss": 0.3574,
      "num_input_tokens_seen": 21212304,
      "step": 32475
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 4.277173042297363,
      "learning_rate": 5.4875060054002115e-08,
      "loss": 0.3209,
      "num_input_tokens_seen": 21216208,
      "step": 32480
    },
    {
      "epoch": 19.15389150943396,
      "grad_norm": 8.294990539550781,
      "learning_rate": 5.4495600443246755e-08,
      "loss": 0.3078,
      "num_input_tokens_seen": 21218896,
      "step": 32485
    },
    {
      "epoch": 19.15683962264151,
      "grad_norm": 3.7499265670776367,
      "learning_rate": 5.411745017609493e-08,
      "loss": 0.3876,
      "num_input_tokens_seen": 21222224,
      "step": 32490
    },
    {
      "epoch": 19.159787735849058,
      "grad_norm": 3.7036070823669434,
      "learning_rate": 5.374060935266434e-08,
      "loss": 0.295,
      "num_input_tokens_seen": 21224784,
      "step": 32495
    },
    {
      "epoch": 19.162735849056602,
      "grad_norm": 2.631542444229126,
      "learning_rate": 5.3365078072724065e-08,
      "loss": 0.3354,
      "num_input_tokens_seen": 21229008,
      "step": 32500
    },
    {
      "epoch": 19.16568396226415,
      "grad_norm": 4.694311141967773,
      "learning_rate": 5.299085643569846e-08,
      "loss": 0.1914,
      "num_input_tokens_seen": 21232176,
      "step": 32505
    },
    {
      "epoch": 19.1686320754717,
      "grad_norm": 3.6424858570098877,
      "learning_rate": 5.261794454066327e-08,
      "loss": 0.344,
      "num_input_tokens_seen": 21236112,
      "step": 32510
    },
    {
      "epoch": 19.171580188679247,
      "grad_norm": 2.4972641468048096,
      "learning_rate": 5.224634248635008e-08,
      "loss": 0.281,
      "num_input_tokens_seen": 21239952,
      "step": 32515
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 3.0749526023864746,
      "learning_rate": 5.187605037114129e-08,
      "loss": 0.2848,
      "num_input_tokens_seen": 21243216,
      "step": 32520
    },
    {
      "epoch": 19.17747641509434,
      "grad_norm": 3.4262595176696777,
      "learning_rate": 5.15070682930735e-08,
      "loss": 0.3238,
      "num_input_tokens_seen": 21246224,
      "step": 32525
    },
    {
      "epoch": 19.180424528301888,
      "grad_norm": 2.2313661575317383,
      "learning_rate": 5.113939634983578e-08,
      "loss": 0.4632,
      "num_input_tokens_seen": 21249424,
      "step": 32530
    },
    {
      "epoch": 19.183372641509433,
      "grad_norm": 3.0493459701538086,
      "learning_rate": 5.077303463877192e-08,
      "loss": 0.381,
      "num_input_tokens_seen": 21253360,
      "step": 32535
    },
    {
      "epoch": 19.18632075471698,
      "grad_norm": 2.591914176940918,
      "learning_rate": 5.040798325687601e-08,
      "loss": 0.2371,
      "num_input_tokens_seen": 21255952,
      "step": 32540
    },
    {
      "epoch": 19.18926886792453,
      "grad_norm": 7.067666053771973,
      "learning_rate": 5.004424230079852e-08,
      "loss": 0.3107,
      "num_input_tokens_seen": 21258576,
      "step": 32545
    },
    {
      "epoch": 19.192216981132077,
      "grad_norm": 6.682590961456299,
      "learning_rate": 4.968181186684129e-08,
      "loss": 0.2601,
      "num_input_tokens_seen": 21261296,
      "step": 32550
    },
    {
      "epoch": 19.19516509433962,
      "grad_norm": 6.076308250427246,
      "learning_rate": 4.932069205095924e-08,
      "loss": 0.4623,
      "num_input_tokens_seen": 21265392,
      "step": 32555
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 2.7938437461853027,
      "learning_rate": 4.896088294875978e-08,
      "loss": 0.427,
      "num_input_tokens_seen": 21268528,
      "step": 32560
    },
    {
      "epoch": 19.201061320754718,
      "grad_norm": 5.552639961242676,
      "learning_rate": 4.8602384655505044e-08,
      "loss": 0.354,
      "num_input_tokens_seen": 21271280,
      "step": 32565
    },
    {
      "epoch": 19.204009433962263,
      "grad_norm": 5.168237209320068,
      "learning_rate": 4.824519726610744e-08,
      "loss": 0.4567,
      "num_input_tokens_seen": 21274416,
      "step": 32570
    },
    {
      "epoch": 19.20695754716981,
      "grad_norm": 2.8187341690063477,
      "learning_rate": 4.7889320875135206e-08,
      "loss": 0.2678,
      "num_input_tokens_seen": 21278160,
      "step": 32575
    },
    {
      "epoch": 19.20990566037736,
      "grad_norm": 3.0463366508483887,
      "learning_rate": 4.753475557680742e-08,
      "loss": 0.3645,
      "num_input_tokens_seen": 21281392,
      "step": 32580
    },
    {
      "epoch": 19.212853773584907,
      "grad_norm": 2.9570560455322266,
      "learning_rate": 4.718150146499734e-08,
      "loss": 0.3641,
      "num_input_tokens_seen": 21283952,
      "step": 32585
    },
    {
      "epoch": 19.215801886792452,
      "grad_norm": 5.469176292419434,
      "learning_rate": 4.682955863323013e-08,
      "loss": 0.3011,
      "num_input_tokens_seen": 21288176,
      "step": 32590
    },
    {
      "epoch": 19.21875,
      "grad_norm": 3.495906114578247,
      "learning_rate": 4.6478927174684606e-08,
      "loss": 0.2196,
      "num_input_tokens_seen": 21291408,
      "step": 32595
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 3.7453126907348633,
      "learning_rate": 4.612960718219095e-08,
      "loss": 0.4095,
      "num_input_tokens_seen": 21294800,
      "step": 32600
    },
    {
      "epoch": 19.224646226415093,
      "grad_norm": 1.9583903551101685,
      "learning_rate": 4.578159874823407e-08,
      "loss": 0.3557,
      "num_input_tokens_seen": 21297296,
      "step": 32605
    },
    {
      "epoch": 19.22759433962264,
      "grad_norm": 2.8977105617523193,
      "learning_rate": 4.5434901964950264e-08,
      "loss": 0.2957,
      "num_input_tokens_seen": 21300112,
      "step": 32610
    },
    {
      "epoch": 19.23054245283019,
      "grad_norm": 4.314634799957275,
      "learning_rate": 4.508951692412944e-08,
      "loss": 0.2647,
      "num_input_tokens_seen": 21302992,
      "step": 32615
    },
    {
      "epoch": 19.233490566037737,
      "grad_norm": 4.861223220825195,
      "learning_rate": 4.4745443717213455e-08,
      "loss": 0.3093,
      "num_input_tokens_seen": 21305488,
      "step": 32620
    },
    {
      "epoch": 19.236438679245282,
      "grad_norm": 2.268911361694336,
      "learning_rate": 4.4402682435296666e-08,
      "loss": 0.3166,
      "num_input_tokens_seen": 21308880,
      "step": 32625
    },
    {
      "epoch": 19.23938679245283,
      "grad_norm": 5.375190258026123,
      "learning_rate": 4.406123316912758e-08,
      "loss": 0.2386,
      "num_input_tokens_seen": 21311920,
      "step": 32630
    },
    {
      "epoch": 19.24233490566038,
      "grad_norm": 5.622133255004883,
      "learning_rate": 4.372109600910612e-08,
      "loss": 0.4153,
      "num_input_tokens_seen": 21315472,
      "step": 32635
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 2.8614888191223145,
      "learning_rate": 4.338227104528414e-08,
      "loss": 0.418,
      "num_input_tokens_seen": 21319664,
      "step": 32640
    },
    {
      "epoch": 19.24823113207547,
      "grad_norm": 4.159849166870117,
      "learning_rate": 4.304475836736821e-08,
      "loss": 0.327,
      "num_input_tokens_seen": 21323056,
      "step": 32645
    },
    {
      "epoch": 19.25117924528302,
      "grad_norm": 3.2727503776550293,
      "learning_rate": 4.27085580647163e-08,
      "loss": 0.3548,
      "num_input_tokens_seen": 21325744,
      "step": 32650
    },
    {
      "epoch": 19.254127358490567,
      "grad_norm": 5.498486518859863,
      "learning_rate": 4.237367022633776e-08,
      "loss": 0.2836,
      "num_input_tokens_seen": 21328144,
      "step": 32655
    },
    {
      "epoch": 19.257075471698112,
      "grad_norm": 3.7676990032196045,
      "learning_rate": 4.204009494089612e-08,
      "loss": 0.3768,
      "num_input_tokens_seen": 21330992,
      "step": 32660
    },
    {
      "epoch": 19.26002358490566,
      "grad_norm": 5.448235988616943,
      "learning_rate": 4.170783229670739e-08,
      "loss": 0.2522,
      "num_input_tokens_seen": 21333872,
      "step": 32665
    },
    {
      "epoch": 19.26297169811321,
      "grad_norm": 4.003766059875488,
      "learning_rate": 4.137688238173898e-08,
      "loss": 0.3348,
      "num_input_tokens_seen": 21336144,
      "step": 32670
    },
    {
      "epoch": 19.265919811320753,
      "grad_norm": 3.1381678581237793,
      "learning_rate": 4.104724528361137e-08,
      "loss": 0.2491,
      "num_input_tokens_seen": 21339376,
      "step": 32675
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 2.4695334434509277,
      "learning_rate": 4.071892108959752e-08,
      "loss": 0.3864,
      "num_input_tokens_seen": 21342256,
      "step": 32680
    },
    {
      "epoch": 19.27181603773585,
      "grad_norm": 4.532027244567871,
      "learning_rate": 4.039190988662234e-08,
      "loss": 0.3516,
      "num_input_tokens_seen": 21345136,
      "step": 32685
    },
    {
      "epoch": 19.274764150943398,
      "grad_norm": 2.592947244644165,
      "learning_rate": 4.006621176126435e-08,
      "loss": 0.3829,
      "num_input_tokens_seen": 21348048,
      "step": 32690
    },
    {
      "epoch": 19.277712264150942,
      "grad_norm": 5.0856099128723145,
      "learning_rate": 3.974182679975236e-08,
      "loss": 0.3449,
      "num_input_tokens_seen": 21350832,
      "step": 32695
    },
    {
      "epoch": 19.28066037735849,
      "grad_norm": 3.242884874343872,
      "learning_rate": 3.941875508796933e-08,
      "loss": 0.2078,
      "num_input_tokens_seen": 21353200,
      "step": 32700
    },
    {
      "epoch": 19.28360849056604,
      "grad_norm": 2.823246717453003,
      "learning_rate": 3.909699671145017e-08,
      "loss": 0.2841,
      "num_input_tokens_seen": 21355888,
      "step": 32705
    },
    {
      "epoch": 19.286556603773583,
      "grad_norm": 2.5467326641082764,
      "learning_rate": 3.87765517553812e-08,
      "loss": 0.4132,
      "num_input_tokens_seen": 21359184,
      "step": 32710
    },
    {
      "epoch": 19.28950471698113,
      "grad_norm": 2.990246057510376,
      "learning_rate": 3.8457420304601756e-08,
      "loss": 0.3205,
      "num_input_tokens_seen": 21362288,
      "step": 32715
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 4.274648666381836,
      "learning_rate": 3.813960244360371e-08,
      "loss": 0.3239,
      "num_input_tokens_seen": 21365872,
      "step": 32720
    },
    {
      "epoch": 19.295400943396228,
      "grad_norm": 1.9199990034103394,
      "learning_rate": 3.7823098256529744e-08,
      "loss": 0.3985,
      "num_input_tokens_seen": 21369872,
      "step": 32725
    },
    {
      "epoch": 19.298349056603772,
      "grad_norm": 6.853724002838135,
      "learning_rate": 3.750790782717673e-08,
      "loss": 0.2373,
      "num_input_tokens_seen": 21373520,
      "step": 32730
    },
    {
      "epoch": 19.30129716981132,
      "grad_norm": 2.774139404296875,
      "learning_rate": 3.719403123899179e-08,
      "loss": 0.1239,
      "num_input_tokens_seen": 21376560,
      "step": 32735
    },
    {
      "epoch": 19.30424528301887,
      "grad_norm": 4.077716827392578,
      "learning_rate": 3.688146857507624e-08,
      "loss": 0.3119,
      "num_input_tokens_seen": 21380624,
      "step": 32740
    },
    {
      "epoch": 19.307193396226417,
      "grad_norm": 5.948211193084717,
      "learning_rate": 3.657021991818166e-08,
      "loss": 0.2767,
      "num_input_tokens_seen": 21382512,
      "step": 32745
    },
    {
      "epoch": 19.31014150943396,
      "grad_norm": 5.262592315673828,
      "learning_rate": 3.626028535071213e-08,
      "loss": 0.3168,
      "num_input_tokens_seen": 21386288,
      "step": 32750
    },
    {
      "epoch": 19.31308962264151,
      "grad_norm": 4.623315334320068,
      "learning_rate": 3.59516649547248e-08,
      "loss": 0.2417,
      "num_input_tokens_seen": 21389040,
      "step": 32755
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 5.991945266723633,
      "learning_rate": 3.564435881192818e-08,
      "loss": 0.3181,
      "num_input_tokens_seen": 21392432,
      "step": 32760
    },
    {
      "epoch": 19.318985849056602,
      "grad_norm": 4.564460754394531,
      "learning_rate": 3.5338367003682763e-08,
      "loss": 0.2324,
      "num_input_tokens_seen": 21395184,
      "step": 32765
    },
    {
      "epoch": 19.32193396226415,
      "grad_norm": 3.984793186187744,
      "learning_rate": 3.5033689611000954e-08,
      "loss": 0.3681,
      "num_input_tokens_seen": 21398768,
      "step": 32770
    },
    {
      "epoch": 19.3248820754717,
      "grad_norm": 4.808549880981445,
      "learning_rate": 3.473032671454768e-08,
      "loss": 0.3503,
      "num_input_tokens_seen": 21401808,
      "step": 32775
    },
    {
      "epoch": 19.327830188679247,
      "grad_norm": 5.868702411651611,
      "learning_rate": 3.44282783946398e-08,
      "loss": 0.2492,
      "num_input_tokens_seen": 21404784,
      "step": 32780
    },
    {
      "epoch": 19.33077830188679,
      "grad_norm": 3.042813539505005,
      "learning_rate": 3.4127544731245575e-08,
      "loss": 0.2282,
      "num_input_tokens_seen": 21410128,
      "step": 32785
    },
    {
      "epoch": 19.33372641509434,
      "grad_norm": 4.479361534118652,
      "learning_rate": 3.38281258039852e-08,
      "loss": 0.2772,
      "num_input_tokens_seen": 21414864,
      "step": 32790
    },
    {
      "epoch": 19.336674528301888,
      "grad_norm": 4.377635955810547,
      "learning_rate": 3.353002169213193e-08,
      "loss": 0.3507,
      "num_input_tokens_seen": 21417680,
      "step": 32795
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 5.001479625701904,
      "learning_rate": 3.32332324746093e-08,
      "loss": 0.5152,
      "num_input_tokens_seen": 21422064,
      "step": 32800
    },
    {
      "epoch": 19.34257075471698,
      "grad_norm": 2.761552095413208,
      "learning_rate": 3.2937758229994455e-08,
      "loss": 0.2905,
      "num_input_tokens_seen": 21424752,
      "step": 32805
    },
    {
      "epoch": 19.34551886792453,
      "grad_norm": 4.0775556564331055,
      "learning_rate": 3.2643599036514815e-08,
      "loss": 0.3559,
      "num_input_tokens_seen": 21427792,
      "step": 32810
    },
    {
      "epoch": 19.348466981132077,
      "grad_norm": 2.401578664779663,
      "learning_rate": 3.2350754972050316e-08,
      "loss": 0.348,
      "num_input_tokens_seen": 21431280,
      "step": 32815
    },
    {
      "epoch": 19.35141509433962,
      "grad_norm": 4.9675703048706055,
      "learning_rate": 3.2059226114132815e-08,
      "loss": 0.1909,
      "num_input_tokens_seen": 21434576,
      "step": 32820
    },
    {
      "epoch": 19.35436320754717,
      "grad_norm": 5.871214389801025,
      "learning_rate": 3.1769012539945575e-08,
      "loss": 0.2693,
      "num_input_tokens_seen": 21437456,
      "step": 32825
    },
    {
      "epoch": 19.357311320754718,
      "grad_norm": 2.864609479904175,
      "learning_rate": 3.1480114326324364e-08,
      "loss": 0.2955,
      "num_input_tokens_seen": 21441392,
      "step": 32830
    },
    {
      "epoch": 19.360259433962263,
      "grad_norm": 3.205639600753784,
      "learning_rate": 3.1192531549756325e-08,
      "loss": 0.3853,
      "num_input_tokens_seen": 21444176,
      "step": 32835
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 4.123002052307129,
      "learning_rate": 3.090626428638e-08,
      "loss": 0.2611,
      "num_input_tokens_seen": 21446480,
      "step": 32840
    },
    {
      "epoch": 19.36615566037736,
      "grad_norm": 6.693368434906006,
      "learning_rate": 3.062131261198531e-08,
      "loss": 0.2329,
      "num_input_tokens_seen": 21449744,
      "step": 32845
    },
    {
      "epoch": 19.369103773584907,
      "grad_norm": 5.033766746520996,
      "learning_rate": 3.033767660201525e-08,
      "loss": 0.2649,
      "num_input_tokens_seen": 21452656,
      "step": 32850
    },
    {
      "epoch": 19.372051886792452,
      "grad_norm": 3.1192452907562256,
      "learning_rate": 3.005535633156309e-08,
      "loss": 0.2433,
      "num_input_tokens_seen": 21455664,
      "step": 32855
    },
    {
      "epoch": 19.375,
      "grad_norm": 4.529199123382568,
      "learning_rate": 2.977435187537514e-08,
      "loss": 0.2658,
      "num_input_tokens_seen": 21458224,
      "step": 32860
    },
    {
      "epoch": 19.377948113207548,
      "grad_norm": 2.260039806365967,
      "learning_rate": 2.949466330784745e-08,
      "loss": 0.3515,
      "num_input_tokens_seen": 21462096,
      "step": 32865
    },
    {
      "epoch": 19.380896226415093,
      "grad_norm": 3.023681402206421,
      "learning_rate": 2.921629070302967e-08,
      "loss": 0.2896,
      "num_input_tokens_seen": 21465392,
      "step": 32870
    },
    {
      "epoch": 19.38384433962264,
      "grad_norm": 4.487349033355713,
      "learning_rate": 2.893923413462174e-08,
      "loss": 0.315,
      "num_input_tokens_seen": 21468304,
      "step": 32875
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 2.9140946865081787,
      "learning_rate": 2.866349367597554e-08,
      "loss": 0.2817,
      "num_input_tokens_seen": 21471152,
      "step": 32880
    },
    {
      "epoch": 19.389740566037737,
      "grad_norm": 2.9632973670959473,
      "learning_rate": 2.8389069400094893e-08,
      "loss": 0.3559,
      "num_input_tokens_seen": 21475280,
      "step": 32885
    },
    {
      "epoch": 19.392688679245282,
      "grad_norm": 6.871697902679443,
      "learning_rate": 2.811596137963446e-08,
      "loss": 0.2512,
      "num_input_tokens_seen": 21478256,
      "step": 32890
    },
    {
      "epoch": 19.39563679245283,
      "grad_norm": 4.2522687911987305,
      "learning_rate": 2.7844169686900844e-08,
      "loss": 0.3513,
      "num_input_tokens_seen": 21481744,
      "step": 32895
    },
    {
      "epoch": 19.39858490566038,
      "grad_norm": 4.720954895019531,
      "learning_rate": 2.7573694393852047e-08,
      "loss": 0.4536,
      "num_input_tokens_seen": 21484528,
      "step": 32900
    },
    {
      "epoch": 19.401533018867923,
      "grad_norm": 3.292646884918213,
      "learning_rate": 2.7304535572098e-08,
      "loss": 0.2931,
      "num_input_tokens_seen": 21487888,
      "step": 32905
    },
    {
      "epoch": 19.40448113207547,
      "grad_norm": 4.726192474365234,
      "learning_rate": 2.703669329289893e-08,
      "loss": 0.3097,
      "num_input_tokens_seen": 21491120,
      "step": 32910
    },
    {
      "epoch": 19.40742924528302,
      "grad_norm": 1.7103344202041626,
      "learning_rate": 2.6770167627167554e-08,
      "loss": 0.285,
      "num_input_tokens_seen": 21494832,
      "step": 32915
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 5.138693809509277,
      "learning_rate": 2.6504958645467426e-08,
      "loss": 0.3486,
      "num_input_tokens_seen": 21499024,
      "step": 32920
    },
    {
      "epoch": 19.413325471698112,
      "grad_norm": 3.486959934234619,
      "learning_rate": 2.6241066418014605e-08,
      "loss": 0.3806,
      "num_input_tokens_seen": 21502480,
      "step": 32925
    },
    {
      "epoch": 19.41627358490566,
      "grad_norm": 4.1281514167785645,
      "learning_rate": 2.5978491014674866e-08,
      "loss": 0.3129,
      "num_input_tokens_seen": 21506224,
      "step": 32930
    },
    {
      "epoch": 19.41922169811321,
      "grad_norm": 5.694957733154297,
      "learning_rate": 2.571723250496705e-08,
      "loss": 0.2961,
      "num_input_tokens_seen": 21510736,
      "step": 32935
    },
    {
      "epoch": 19.422169811320753,
      "grad_norm": 5.592071533203125,
      "learning_rate": 2.5457290958059155e-08,
      "loss": 0.275,
      "num_input_tokens_seen": 21513680,
      "step": 32940
    },
    {
      "epoch": 19.4251179245283,
      "grad_norm": 3.776582717895508,
      "learning_rate": 2.51986664427728e-08,
      "loss": 0.2848,
      "num_input_tokens_seen": 21516496,
      "step": 32945
    },
    {
      "epoch": 19.42806603773585,
      "grad_norm": 5.030200004577637,
      "learning_rate": 2.4941359027579883e-08,
      "loss": 0.3406,
      "num_input_tokens_seen": 21519696,
      "step": 32950
    },
    {
      "epoch": 19.431014150943398,
      "grad_norm": 4.194863796234131,
      "learning_rate": 2.468536878060368e-08,
      "loss": 0.2814,
      "num_input_tokens_seen": 21522320,
      "step": 32955
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 2.960634469985962,
      "learning_rate": 2.443069576961832e-08,
      "loss": 0.3458,
      "num_input_tokens_seen": 21525584,
      "step": 32960
    },
    {
      "epoch": 19.43691037735849,
      "grad_norm": 2.4196009635925293,
      "learning_rate": 2.4177340062049304e-08,
      "loss": 0.2497,
      "num_input_tokens_seen": 21529200,
      "step": 32965
    },
    {
      "epoch": 19.43985849056604,
      "grad_norm": 4.09235954284668,
      "learning_rate": 2.3925301724974647e-08,
      "loss": 0.471,
      "num_input_tokens_seen": 21532432,
      "step": 32970
    },
    {
      "epoch": 19.442806603773583,
      "grad_norm": 3.946441650390625,
      "learning_rate": 2.367458082512153e-08,
      "loss": 0.2933,
      "num_input_tokens_seen": 21535984,
      "step": 32975
    },
    {
      "epoch": 19.44575471698113,
      "grad_norm": 6.068199157714844,
      "learning_rate": 2.3425177428870737e-08,
      "loss": 0.4729,
      "num_input_tokens_seen": 21538864,
      "step": 32980
    },
    {
      "epoch": 19.44870283018868,
      "grad_norm": 3.9441142082214355,
      "learning_rate": 2.3177091602251677e-08,
      "loss": 0.2577,
      "num_input_tokens_seen": 21541616,
      "step": 32985
    },
    {
      "epoch": 19.451650943396228,
      "grad_norm": 3.844703197479248,
      "learning_rate": 2.2930323410946254e-08,
      "loss": 0.2776,
      "num_input_tokens_seen": 21544816,
      "step": 32990
    },
    {
      "epoch": 19.454599056603772,
      "grad_norm": 4.612931728363037,
      "learning_rate": 2.2684872920287758e-08,
      "loss": 0.3766,
      "num_input_tokens_seen": 21548272,
      "step": 32995
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 3.1272497177124023,
      "learning_rate": 2.2440740195260323e-08,
      "loss": 0.2904,
      "num_input_tokens_seen": 21551472,
      "step": 33000
    },
    {
      "epoch": 19.46049528301887,
      "grad_norm": 3.382800340652466,
      "learning_rate": 2.219792530049891e-08,
      "loss": 0.3312,
      "num_input_tokens_seen": 21554544,
      "step": 33005
    },
    {
      "epoch": 19.463443396226417,
      "grad_norm": 4.892094135284424,
      "learning_rate": 2.1956428300290434e-08,
      "loss": 0.2693,
      "num_input_tokens_seen": 21558000,
      "step": 33010
    },
    {
      "epoch": 19.46639150943396,
      "grad_norm": 3.404143810272217,
      "learning_rate": 2.1716249258570966e-08,
      "loss": 0.4285,
      "num_input_tokens_seen": 21561552,
      "step": 33015
    },
    {
      "epoch": 19.46933962264151,
      "grad_norm": 2.863386869430542,
      "learning_rate": 2.1477388238930196e-08,
      "loss": 0.2148,
      "num_input_tokens_seen": 21564304,
      "step": 33020
    },
    {
      "epoch": 19.472287735849058,
      "grad_norm": 2.6937942504882812,
      "learning_rate": 2.1239845304606988e-08,
      "loss": 0.2558,
      "num_input_tokens_seen": 21566992,
      "step": 33025
    },
    {
      "epoch": 19.475235849056602,
      "grad_norm": 3.171477794647217,
      "learning_rate": 2.100362051849214e-08,
      "loss": 0.2813,
      "num_input_tokens_seen": 21571152,
      "step": 33030
    },
    {
      "epoch": 19.47818396226415,
      "grad_norm": 4.330771446228027,
      "learning_rate": 2.076871394312674e-08,
      "loss": 0.4427,
      "num_input_tokens_seen": 21574576,
      "step": 33035
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 2.988088607788086,
      "learning_rate": 2.0535125640703813e-08,
      "loss": 0.3155,
      "num_input_tokens_seen": 21577680,
      "step": 33040
    },
    {
      "epoch": 19.484080188679247,
      "grad_norm": 3.346752405166626,
      "learning_rate": 2.0302855673066667e-08,
      "loss": 0.3415,
      "num_input_tokens_seen": 21580016,
      "step": 33045
    },
    {
      "epoch": 19.48702830188679,
      "grad_norm": 3.8094685077667236,
      "learning_rate": 2.0071904101710004e-08,
      "loss": 0.3748,
      "num_input_tokens_seen": 21584176,
      "step": 33050
    },
    {
      "epoch": 19.48997641509434,
      "grad_norm": 4.600164413452148,
      "learning_rate": 1.98422709877788e-08,
      "loss": 0.3711,
      "num_input_tokens_seen": 21587248,
      "step": 33055
    },
    {
      "epoch": 19.492924528301888,
      "grad_norm": 3.3654775619506836,
      "learning_rate": 1.961395639206942e-08,
      "loss": 0.2628,
      "num_input_tokens_seen": 21591056,
      "step": 33060
    },
    {
      "epoch": 19.495872641509433,
      "grad_norm": 2.9769468307495117,
      "learning_rate": 1.9386960375029628e-08,
      "loss": 0.2332,
      "num_input_tokens_seen": 21594704,
      "step": 33065
    },
    {
      "epoch": 19.49882075471698,
      "grad_norm": 4.1523027420043945,
      "learning_rate": 1.9161282996757458e-08,
      "loss": 0.384,
      "num_input_tokens_seen": 21598384,
      "step": 33070
    },
    {
      "epoch": 19.50176886792453,
      "grad_norm": 3.483713388442993,
      "learning_rate": 1.8936924317001225e-08,
      "loss": 0.2727,
      "num_input_tokens_seen": 21601424,
      "step": 33075
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 4.386260986328125,
      "learning_rate": 1.8713884395162308e-08,
      "loss": 0.2775,
      "num_input_tokens_seen": 21604336,
      "step": 33080
    },
    {
      "epoch": 19.50766509433962,
      "grad_norm": 2.9245684146881104,
      "learning_rate": 1.8492163290290132e-08,
      "loss": 0.3022,
      "num_input_tokens_seen": 21607408,
      "step": 33085
    },
    {
      "epoch": 19.51061320754717,
      "grad_norm": 2.988342523574829,
      "learning_rate": 1.827176106108719e-08,
      "loss": 0.1814,
      "num_input_tokens_seen": 21611056,
      "step": 33090
    },
    {
      "epoch": 19.513561320754718,
      "grad_norm": 6.390865325927734,
      "learning_rate": 1.8052677765905137e-08,
      "loss": 0.3412,
      "num_input_tokens_seen": 21614128,
      "step": 33095
    },
    {
      "epoch": 19.516509433962263,
      "grad_norm": 3.0881354808807373,
      "learning_rate": 1.783491346274757e-08,
      "loss": 0.3241,
      "num_input_tokens_seen": 21617104,
      "step": 33100
    },
    {
      "epoch": 19.51945754716981,
      "grad_norm": 5.010420799255371,
      "learning_rate": 1.7618468209268936e-08,
      "loss": 0.2723,
      "num_input_tokens_seen": 21619664,
      "step": 33105
    },
    {
      "epoch": 19.52240566037736,
      "grad_norm": 1.7453926801681519,
      "learning_rate": 1.7403342062773943e-08,
      "loss": 0.2366,
      "num_input_tokens_seen": 21623024,
      "step": 33110
    },
    {
      "epoch": 19.525353773584907,
      "grad_norm": 7.0845746994018555,
      "learning_rate": 1.718953508021759e-08,
      "loss": 0.3381,
      "num_input_tokens_seen": 21627312,
      "step": 33115
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 4.5036702156066895,
      "learning_rate": 1.6977047318206262e-08,
      "loss": 0.419,
      "num_input_tokens_seen": 21630256,
      "step": 33120
    },
    {
      "epoch": 19.53125,
      "grad_norm": 6.16861629486084,
      "learning_rate": 1.676587883299774e-08,
      "loss": 0.3695,
      "num_input_tokens_seen": 21633936,
      "step": 33125
    },
    {
      "epoch": 19.534198113207548,
      "grad_norm": 3.411015272140503,
      "learning_rate": 1.655602968049952e-08,
      "loss": 0.2814,
      "num_input_tokens_seen": 21638128,
      "step": 33130
    },
    {
      "epoch": 19.537146226415093,
      "grad_norm": 6.68475866317749,
      "learning_rate": 1.634749991626938e-08,
      "loss": 0.3129,
      "num_input_tokens_seen": 21640528,
      "step": 33135
    },
    {
      "epoch": 19.54009433962264,
      "grad_norm": 3.8217201232910156,
      "learning_rate": 1.6140289595517056e-08,
      "loss": 0.396,
      "num_input_tokens_seen": 21643856,
      "step": 33140
    },
    {
      "epoch": 19.54304245283019,
      "grad_norm": 8.121113777160645,
      "learning_rate": 1.5934398773102545e-08,
      "loss": 0.4211,
      "num_input_tokens_seen": 21650000,
      "step": 33145
    },
    {
      "epoch": 19.545990566037737,
      "grad_norm": 3.946362018585205,
      "learning_rate": 1.5729827503536133e-08,
      "loss": 0.3598,
      "num_input_tokens_seen": 21652976,
      "step": 33150
    },
    {
      "epoch": 19.548938679245282,
      "grad_norm": 3.1030235290527344,
      "learning_rate": 1.5526575840978942e-08,
      "loss": 0.2932,
      "num_input_tokens_seen": 21656400,
      "step": 33155
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 3.855794668197632,
      "learning_rate": 1.532464383924237e-08,
      "loss": 0.3291,
      "num_input_tokens_seen": 21659184,
      "step": 33160
    },
    {
      "epoch": 19.55483490566038,
      "grad_norm": 3.967013359069824,
      "learning_rate": 1.5124031551789208e-08,
      "loss": 0.274,
      "num_input_tokens_seen": 21663536,
      "step": 33165
    },
    {
      "epoch": 19.557783018867923,
      "grad_norm": 5.488925457000732,
      "learning_rate": 1.4924739031732527e-08,
      "loss": 0.2836,
      "num_input_tokens_seen": 21666256,
      "step": 33170
    },
    {
      "epoch": 19.56073113207547,
      "grad_norm": 2.9553914070129395,
      "learning_rate": 1.4726766331835118e-08,
      "loss": 0.3008,
      "num_input_tokens_seen": 21668752,
      "step": 33175
    },
    {
      "epoch": 19.56367924528302,
      "grad_norm": 4.54148530960083,
      "learning_rate": 1.4530113504512278e-08,
      "loss": 0.2792,
      "num_input_tokens_seen": 21671696,
      "step": 33180
    },
    {
      "epoch": 19.566627358490567,
      "grad_norm": 6.434611797332764,
      "learning_rate": 1.4334780601827914e-08,
      "loss": 0.4145,
      "num_input_tokens_seen": 21674704,
      "step": 33185
    },
    {
      "epoch": 19.569575471698112,
      "grad_norm": 6.086824417114258,
      "learning_rate": 1.4140767675497325e-08,
      "loss": 0.3354,
      "num_input_tokens_seen": 21677680,
      "step": 33190
    },
    {
      "epoch": 19.57252358490566,
      "grad_norm": 4.874594211578369,
      "learning_rate": 1.394807477688609e-08,
      "loss": 0.3275,
      "num_input_tokens_seen": 21681168,
      "step": 33195
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 4.842461585998535,
      "learning_rate": 1.3756701957011177e-08,
      "loss": 0.3443,
      "num_input_tokens_seen": 21683632,
      "step": 33200
    },
    {
      "epoch": 19.578419811320753,
      "grad_norm": 4.006941318511963,
      "learning_rate": 1.3566649266538723e-08,
      "loss": 0.4078,
      "num_input_tokens_seen": 21688592,
      "step": 33205
    },
    {
      "epoch": 19.5813679245283,
      "grad_norm": 4.261688709259033,
      "learning_rate": 1.3377916755786257e-08,
      "loss": 0.4663,
      "num_input_tokens_seen": 21691824,
      "step": 33210
    },
    {
      "epoch": 19.58431603773585,
      "grad_norm": 3.665513038635254,
      "learning_rate": 1.3190504474721588e-08,
      "loss": 0.3035,
      "num_input_tokens_seen": 21694736,
      "step": 33215
    },
    {
      "epoch": 19.587264150943398,
      "grad_norm": 5.099005699157715,
      "learning_rate": 1.3004412472962802e-08,
      "loss": 0.3504,
      "num_input_tokens_seen": 21697680,
      "step": 33220
    },
    {
      "epoch": 19.590212264150942,
      "grad_norm": 5.531406402587891,
      "learning_rate": 1.2819640799778266e-08,
      "loss": 0.4397,
      "num_input_tokens_seen": 21700304,
      "step": 33225
    },
    {
      "epoch": 19.59316037735849,
      "grad_norm": 4.5653395652771,
      "learning_rate": 1.2636189504087737e-08,
      "loss": 0.3504,
      "num_input_tokens_seen": 21703536,
      "step": 33230
    },
    {
      "epoch": 19.59610849056604,
      "grad_norm": 2.6702589988708496,
      "learning_rate": 1.2454058634460142e-08,
      "loss": 0.2799,
      "num_input_tokens_seen": 21706320,
      "step": 33235
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 3.70550537109375,
      "learning_rate": 1.2273248239115798e-08,
      "loss": 0.3164,
      "num_input_tokens_seen": 21709232,
      "step": 33240
    },
    {
      "epoch": 19.60200471698113,
      "grad_norm": 3.771127223968506,
      "learning_rate": 1.2093758365924746e-08,
      "loss": 0.3039,
      "num_input_tokens_seen": 21711952,
      "step": 33245
    },
    {
      "epoch": 19.60495283018868,
      "grad_norm": 2.102527379989624,
      "learning_rate": 1.1915589062408417e-08,
      "loss": 0.2698,
      "num_input_tokens_seen": 21715728,
      "step": 33250
    },
    {
      "epoch": 19.607900943396228,
      "grad_norm": 3.0036380290985107,
      "learning_rate": 1.1738740375736301e-08,
      "loss": 0.3159,
      "num_input_tokens_seen": 21718736,
      "step": 33255
    },
    {
      "epoch": 19.610849056603772,
      "grad_norm": 7.657852649688721,
      "learning_rate": 1.15632123527315e-08,
      "loss": 0.3193,
      "num_input_tokens_seen": 21721840,
      "step": 33260
    },
    {
      "epoch": 19.61379716981132,
      "grad_norm": 5.628442287445068,
      "learning_rate": 1.1389005039865176e-08,
      "loss": 0.4016,
      "num_input_tokens_seen": 21725104,
      "step": 33265
    },
    {
      "epoch": 19.61674528301887,
      "grad_norm": 2.6410462856292725,
      "learning_rate": 1.1216118483259875e-08,
      "loss": 0.2452,
      "num_input_tokens_seen": 21728528,
      "step": 33270
    },
    {
      "epoch": 19.619693396226417,
      "grad_norm": 9.812280654907227,
      "learning_rate": 1.1044552728687319e-08,
      "loss": 0.3336,
      "num_input_tokens_seen": 21732112,
      "step": 33275
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 2.9351677894592285,
      "learning_rate": 1.0874307821570618e-08,
      "loss": 0.2703,
      "num_input_tokens_seen": 21734832,
      "step": 33280
    },
    {
      "epoch": 19.62558962264151,
      "grad_norm": 5.605329513549805,
      "learning_rate": 1.0705383806982606e-08,
      "loss": 0.2881,
      "num_input_tokens_seen": 21739152,
      "step": 33285
    },
    {
      "epoch": 19.628537735849058,
      "grad_norm": 4.322150707244873,
      "learning_rate": 1.0537780729646952e-08,
      "loss": 0.3471,
      "num_input_tokens_seen": 21741872,
      "step": 33290
    },
    {
      "epoch": 19.631485849056602,
      "grad_norm": 3.3154237270355225,
      "learning_rate": 1.0371498633937605e-08,
      "loss": 0.2948,
      "num_input_tokens_seen": 21744816,
      "step": 33295
    },
    {
      "epoch": 19.63443396226415,
      "grad_norm": 3.0250189304351807,
      "learning_rate": 1.0206537563877683e-08,
      "loss": 0.3135,
      "num_input_tokens_seen": 21748272,
      "step": 33300
    },
    {
      "epoch": 19.6373820754717,
      "grad_norm": 3.271798849105835,
      "learning_rate": 1.0042897563141695e-08,
      "loss": 0.3092,
      "num_input_tokens_seen": 21750768,
      "step": 33305
    },
    {
      "epoch": 19.640330188679247,
      "grad_norm": 5.878098011016846,
      "learning_rate": 9.88057867505443e-09,
      "loss": 0.2617,
      "num_input_tokens_seen": 21753520,
      "step": 33310
    },
    {
      "epoch": 19.64327830188679,
      "grad_norm": 4.108860969543457,
      "learning_rate": 9.7195809425904e-09,
      "loss": 0.3225,
      "num_input_tokens_seen": 21757648,
      "step": 33315
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 3.138734817504883,
      "learning_rate": 9.559904408373844e-09,
      "loss": 0.2052,
      "num_input_tokens_seen": 21760112,
      "step": 33320
    },
    {
      "epoch": 19.649174528301888,
      "grad_norm": 2.1404061317443848,
      "learning_rate": 9.401549114680387e-09,
      "loss": 0.2451,
      "num_input_tokens_seen": 21763600,
      "step": 33325
    },
    {
      "epoch": 19.652122641509433,
      "grad_norm": 3.1537375450134277,
      "learning_rate": 9.244515103434826e-09,
      "loss": 0.4661,
      "num_input_tokens_seen": 21766288,
      "step": 33330
    },
    {
      "epoch": 19.65507075471698,
      "grad_norm": 3.3181490898132324,
      "learning_rate": 9.08880241621335e-09,
      "loss": 0.2839,
      "num_input_tokens_seen": 21769776,
      "step": 33335
    },
    {
      "epoch": 19.65801886792453,
      "grad_norm": 3.383791923522949,
      "learning_rate": 8.934411094240758e-09,
      "loss": 0.3219,
      "num_input_tokens_seen": 21772464,
      "step": 33340
    },
    {
      "epoch": 19.660966981132077,
      "grad_norm": 3.730647325515747,
      "learning_rate": 8.781341178393244e-09,
      "loss": 0.3291,
      "num_input_tokens_seen": 21779440,
      "step": 33345
    },
    {
      "epoch": 19.66391509433962,
      "grad_norm": 4.039711952209473,
      "learning_rate": 8.629592709196167e-09,
      "loss": 0.3529,
      "num_input_tokens_seen": 21783056,
      "step": 33350
    },
    {
      "epoch": 19.66686320754717,
      "grad_norm": 4.297609806060791,
      "learning_rate": 8.479165726826277e-09,
      "loss": 0.43,
      "num_input_tokens_seen": 21785968,
      "step": 33355
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 3.200453758239746,
      "learning_rate": 8.330060271109496e-09,
      "loss": 0.5033,
      "num_input_tokens_seen": 21789520,
      "step": 33360
    },
    {
      "epoch": 19.672759433962263,
      "grad_norm": 4.692076206207275,
      "learning_rate": 8.18227638152258e-09,
      "loss": 0.2909,
      "num_input_tokens_seen": 21792880,
      "step": 33365
    },
    {
      "epoch": 19.67570754716981,
      "grad_norm": 4.041689872741699,
      "learning_rate": 8.035814097191452e-09,
      "loss": 0.3138,
      "num_input_tokens_seen": 21795536,
      "step": 33370
    },
    {
      "epoch": 19.67865566037736,
      "grad_norm": 4.324225902557373,
      "learning_rate": 7.890673456892317e-09,
      "loss": 0.3557,
      "num_input_tokens_seen": 21798672,
      "step": 33375
    },
    {
      "epoch": 19.681603773584907,
      "grad_norm": 3.2564759254455566,
      "learning_rate": 7.746854499052215e-09,
      "loss": 0.2685,
      "num_input_tokens_seen": 21801456,
      "step": 33380
    },
    {
      "epoch": 19.684551886792452,
      "grad_norm": 3.7479262351989746,
      "learning_rate": 7.604357261747907e-09,
      "loss": 0.2949,
      "num_input_tokens_seen": 21804112,
      "step": 33385
    },
    {
      "epoch": 19.6875,
      "grad_norm": 8.608222007751465,
      "learning_rate": 7.463181782705886e-09,
      "loss": 0.4902,
      "num_input_tokens_seen": 21812944,
      "step": 33390
    },
    {
      "epoch": 19.690448113207548,
      "grad_norm": 4.027400493621826,
      "learning_rate": 7.3233280993034726e-09,
      "loss": 0.2879,
      "num_input_tokens_seen": 21815664,
      "step": 33395
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 3.279092788696289,
      "learning_rate": 7.184796248567161e-09,
      "loss": 0.3172,
      "num_input_tokens_seen": 21819952,
      "step": 33400
    },
    {
      "epoch": 19.69634433962264,
      "grad_norm": 3.3430347442626953,
      "learning_rate": 7.047586267173723e-09,
      "loss": 0.342,
      "num_input_tokens_seen": 21822928,
      "step": 33405
    },
    {
      "epoch": 19.69929245283019,
      "grad_norm": 3.120882034301758,
      "learning_rate": 6.9116981914502114e-09,
      "loss": 0.2987,
      "num_input_tokens_seen": 21826192,
      "step": 33410
    },
    {
      "epoch": 19.702240566037737,
      "grad_norm": 3.1174347400665283,
      "learning_rate": 6.7771320573734036e-09,
      "loss": 0.2986,
      "num_input_tokens_seen": 21829776,
      "step": 33415
    },
    {
      "epoch": 19.705188679245282,
      "grad_norm": 2.655299663543701,
      "learning_rate": 6.6438879005709114e-09,
      "loss": 0.2884,
      "num_input_tokens_seen": 21833136,
      "step": 33420
    },
    {
      "epoch": 19.70813679245283,
      "grad_norm": 3.939650774002075,
      "learning_rate": 6.511965756318961e-09,
      "loss": 0.4068,
      "num_input_tokens_seen": 21836016,
      "step": 33425
    },
    {
      "epoch": 19.71108490566038,
      "grad_norm": 4.84300422668457,
      "learning_rate": 6.381365659545169e-09,
      "loss": 0.2966,
      "num_input_tokens_seen": 21839568,
      "step": 33430
    },
    {
      "epoch": 19.714033018867923,
      "grad_norm": 5.034727096557617,
      "learning_rate": 6.252087644825766e-09,
      "loss": 0.2924,
      "num_input_tokens_seen": 21842640,
      "step": 33435
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 2.500784397125244,
      "learning_rate": 6.124131746388373e-09,
      "loss": 0.2282,
      "num_input_tokens_seen": 21845168,
      "step": 33440
    },
    {
      "epoch": 19.71992924528302,
      "grad_norm": 6.887560844421387,
      "learning_rate": 5.997497998109225e-09,
      "loss": 0.3089,
      "num_input_tokens_seen": 21848304,
      "step": 33445
    },
    {
      "epoch": 19.722877358490567,
      "grad_norm": 2.4969804286956787,
      "learning_rate": 5.8721864335153925e-09,
      "loss": 0.299,
      "num_input_tokens_seen": 21850800,
      "step": 33450
    },
    {
      "epoch": 19.725825471698112,
      "grad_norm": 2.532170534133911,
      "learning_rate": 5.748197085784224e-09,
      "loss": 0.3173,
      "num_input_tokens_seen": 21853552,
      "step": 33455
    },
    {
      "epoch": 19.72877358490566,
      "grad_norm": 3.3602864742279053,
      "learning_rate": 5.62552998774113e-09,
      "loss": 0.2613,
      "num_input_tokens_seen": 21856912,
      "step": 33460
    },
    {
      "epoch": 19.73172169811321,
      "grad_norm": 3.848642349243164,
      "learning_rate": 5.504185171864018e-09,
      "loss": 0.2687,
      "num_input_tokens_seen": 21860016,
      "step": 33465
    },
    {
      "epoch": 19.734669811320753,
      "grad_norm": 6.042895317077637,
      "learning_rate": 5.384162670278858e-09,
      "loss": 0.2753,
      "num_input_tokens_seen": 21863184,
      "step": 33470
    },
    {
      "epoch": 19.7376179245283,
      "grad_norm": 3.1063830852508545,
      "learning_rate": 5.265462514762454e-09,
      "loss": 0.2364,
      "num_input_tokens_seen": 21870384,
      "step": 33475
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 2.2517356872558594,
      "learning_rate": 5.148084736740777e-09,
      "loss": 0.297,
      "num_input_tokens_seen": 21873008,
      "step": 33480
    },
    {
      "epoch": 19.743514150943398,
      "grad_norm": 3.7433228492736816,
      "learning_rate": 5.032029367290081e-09,
      "loss": 0.2945,
      "num_input_tokens_seen": 21876144,
      "step": 33485
    },
    {
      "epoch": 19.746462264150942,
      "grad_norm": 2.637693405151367,
      "learning_rate": 4.917296437136898e-09,
      "loss": 0.2967,
      "num_input_tokens_seen": 21879440,
      "step": 33490
    },
    {
      "epoch": 19.74941037735849,
      "grad_norm": 3.584977626800537,
      "learning_rate": 4.8038859766569305e-09,
      "loss": 0.3636,
      "num_input_tokens_seen": 21882000,
      "step": 33495
    },
    {
      "epoch": 19.75235849056604,
      "grad_norm": 4.364902496337891,
      "learning_rate": 4.691798015876714e-09,
      "loss": 0.4487,
      "num_input_tokens_seen": 21884688,
      "step": 33500
    },
    {
      "epoch": 19.755306603773583,
      "grad_norm": 4.373682975769043,
      "learning_rate": 4.581032584470846e-09,
      "loss": 0.2354,
      "num_input_tokens_seen": 21887600,
      "step": 33505
    },
    {
      "epoch": 19.75825471698113,
      "grad_norm": 4.521154880523682,
      "learning_rate": 4.471589711766422e-09,
      "loss": 0.22,
      "num_input_tokens_seen": 21892016,
      "step": 33510
    },
    {
      "epoch": 19.76120283018868,
      "grad_norm": 2.131868600845337,
      "learning_rate": 4.363469426737487e-09,
      "loss": 0.4173,
      "num_input_tokens_seen": 21896112,
      "step": 33515
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 3.133582830429077,
      "learning_rate": 4.256671758010588e-09,
      "loss": 0.3978,
      "num_input_tokens_seen": 21899056,
      "step": 33520
    },
    {
      "epoch": 19.767099056603772,
      "grad_norm": 2.8089849948883057,
      "learning_rate": 4.151196733859775e-09,
      "loss": 0.2053,
      "num_input_tokens_seen": 21902416,
      "step": 33525
    },
    {
      "epoch": 19.77004716981132,
      "grad_norm": 4.769979000091553,
      "learning_rate": 4.047044382211041e-09,
      "loss": 0.3015,
      "num_input_tokens_seen": 21907440,
      "step": 33530
    },
    {
      "epoch": 19.77299528301887,
      "grad_norm": 2.682943344116211,
      "learning_rate": 3.94421473063844e-09,
      "loss": 0.3534,
      "num_input_tokens_seen": 21911408,
      "step": 33535
    },
    {
      "epoch": 19.775943396226417,
      "grad_norm": 6.695996284484863,
      "learning_rate": 3.842707806366863e-09,
      "loss": 0.2797,
      "num_input_tokens_seen": 21914000,
      "step": 33540
    },
    {
      "epoch": 19.77889150943396,
      "grad_norm": 2.43522310256958,
      "learning_rate": 3.742523636270368e-09,
      "loss": 0.2351,
      "num_input_tokens_seen": 21916368,
      "step": 33545
    },
    {
      "epoch": 19.78183962264151,
      "grad_norm": 3.321373701095581,
      "learning_rate": 3.6436622468738468e-09,
      "loss": 0.2295,
      "num_input_tokens_seen": 21920016,
      "step": 33550
    },
    {
      "epoch": 19.784787735849058,
      "grad_norm": 2.8692212104797363,
      "learning_rate": 3.546123664350254e-09,
      "loss": 0.2496,
      "num_input_tokens_seen": 21923664,
      "step": 33555
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 2.7099435329437256,
      "learning_rate": 3.449907914524486e-09,
      "loss": 0.3234,
      "num_input_tokens_seen": 21926544,
      "step": 33560
    },
    {
      "epoch": 19.79068396226415,
      "grad_norm": 4.282285213470459,
      "learning_rate": 3.355015022869501e-09,
      "loss": 0.2226,
      "num_input_tokens_seen": 21929392,
      "step": 33565
    },
    {
      "epoch": 19.7936320754717,
      "grad_norm": 3.635810613632202,
      "learning_rate": 3.261445014508535e-09,
      "loss": 0.219,
      "num_input_tokens_seen": 21932304,
      "step": 33570
    },
    {
      "epoch": 19.796580188679247,
      "grad_norm": 2.7104759216308594,
      "learning_rate": 3.1691979142145503e-09,
      "loss": 0.3956,
      "num_input_tokens_seen": 21936144,
      "step": 33575
    },
    {
      "epoch": 19.79952830188679,
      "grad_norm": 5.584103107452393,
      "learning_rate": 3.078273746410787e-09,
      "loss": 0.2324,
      "num_input_tokens_seen": 21938672,
      "step": 33580
    },
    {
      "epoch": 19.80247641509434,
      "grad_norm": 3.6690213680267334,
      "learning_rate": 2.988672535169657e-09,
      "loss": 0.2985,
      "num_input_tokens_seen": 21941072,
      "step": 33585
    },
    {
      "epoch": 19.805424528301888,
      "grad_norm": 3.7093045711517334,
      "learning_rate": 2.9003943042127393e-09,
      "loss": 0.3508,
      "num_input_tokens_seen": 21943952,
      "step": 33590
    },
    {
      "epoch": 19.808372641509433,
      "grad_norm": 3.2316548824310303,
      "learning_rate": 2.8134390769135598e-09,
      "loss": 0.2791,
      "num_input_tokens_seen": 21947376,
      "step": 33595
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 2.7603721618652344,
      "learning_rate": 2.7278068762925935e-09,
      "loss": 0.3798,
      "num_input_tokens_seen": 21951152,
      "step": 33600
    },
    {
      "epoch": 19.81426886792453,
      "grad_norm": 4.050398349761963,
      "learning_rate": 2.6434977250217043e-09,
      "loss": 0.3173,
      "num_input_tokens_seen": 21954000,
      "step": 33605
    },
    {
      "epoch": 19.817216981132077,
      "grad_norm": 3.0910770893096924,
      "learning_rate": 2.560511645422481e-09,
      "loss": 0.3954,
      "num_input_tokens_seen": 21957424,
      "step": 33610
    },
    {
      "epoch": 19.82016509433962,
      "grad_norm": 2.7562873363494873,
      "learning_rate": 2.4788486594656825e-09,
      "loss": 0.2811,
      "num_input_tokens_seen": 21961456,
      "step": 33615
    },
    {
      "epoch": 19.82311320754717,
      "grad_norm": 4.667176246643066,
      "learning_rate": 2.398508788771792e-09,
      "loss": 0.2805,
      "num_input_tokens_seen": 21964720,
      "step": 33620
    },
    {
      "epoch": 19.826061320754718,
      "grad_norm": 4.786773681640625,
      "learning_rate": 2.3194920546110166e-09,
      "loss": 0.247,
      "num_input_tokens_seen": 21967696,
      "step": 33625
    },
    {
      "epoch": 19.829009433962263,
      "grad_norm": 4.9144287109375,
      "learning_rate": 2.2417984779032896e-09,
      "loss": 0.4033,
      "num_input_tokens_seen": 21970224,
      "step": 33630
    },
    {
      "epoch": 19.83195754716981,
      "grad_norm": 2.1433475017547607,
      "learning_rate": 2.1654280792193782e-09,
      "loss": 0.3057,
      "num_input_tokens_seen": 21973744,
      "step": 33635
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 3.763214588165283,
      "learning_rate": 2.0903808787769987e-09,
      "loss": 0.197,
      "num_input_tokens_seen": 21977360,
      "step": 33640
    },
    {
      "epoch": 19.837853773584907,
      "grad_norm": 2.512707233428955,
      "learning_rate": 2.0166568964463673e-09,
      "loss": 0.2688,
      "num_input_tokens_seen": 21981104,
      "step": 33645
    },
    {
      "epoch": 19.840801886792452,
      "grad_norm": 2.961911678314209,
      "learning_rate": 1.9442561517463153e-09,
      "loss": 0.3643,
      "num_input_tokens_seen": 21984112,
      "step": 33650
    },
    {
      "epoch": 19.84375,
      "grad_norm": 4.377889156341553,
      "learning_rate": 1.8731786638442886e-09,
      "loss": 0.2481,
      "num_input_tokens_seen": 21987536,
      "step": 33655
    },
    {
      "epoch": 19.846698113207548,
      "grad_norm": 3.933340549468994,
      "learning_rate": 1.8034244515591214e-09,
      "loss": 0.25,
      "num_input_tokens_seen": 21990512,
      "step": 33660
    },
    {
      "epoch": 19.849646226415093,
      "grad_norm": 2.746096611022949,
      "learning_rate": 1.7349935333582646e-09,
      "loss": 0.2888,
      "num_input_tokens_seen": 21993520,
      "step": 33665
    },
    {
      "epoch": 19.85259433962264,
      "grad_norm": 2.0601389408111572,
      "learning_rate": 1.6678859273594471e-09,
      "loss": 0.2569,
      "num_input_tokens_seen": 21997008,
      "step": 33670
    },
    {
      "epoch": 19.85554245283019,
      "grad_norm": 3.76108980178833,
      "learning_rate": 1.6021016513295683e-09,
      "loss": 0.2559,
      "num_input_tokens_seen": 22000080,
      "step": 33675
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 3.9463653564453125,
      "learning_rate": 1.5376407226846968e-09,
      "loss": 0.3227,
      "num_input_tokens_seen": 22002896,
      "step": 33680
    },
    {
      "epoch": 19.861438679245282,
      "grad_norm": 3.4322128295898438,
      "learning_rate": 1.4745031584917357e-09,
      "loss": 0.2767,
      "num_input_tokens_seen": 22006352,
      "step": 33685
    },
    {
      "epoch": 19.86438679245283,
      "grad_norm": 3.933011293411255,
      "learning_rate": 1.4126889754667583e-09,
      "loss": 0.5691,
      "num_input_tokens_seen": 22008496,
      "step": 33690
    },
    {
      "epoch": 19.86733490566038,
      "grad_norm": 2.2166748046875,
      "learning_rate": 1.3521981899750069e-09,
      "loss": 0.3826,
      "num_input_tokens_seen": 22011536,
      "step": 33695
    },
    {
      "epoch": 19.870283018867923,
      "grad_norm": 3.210789918899536,
      "learning_rate": 1.293030818032004e-09,
      "loss": 0.3847,
      "num_input_tokens_seen": 22017008,
      "step": 33700
    },
    {
      "epoch": 19.87323113207547,
      "grad_norm": 9.213064193725586,
      "learning_rate": 1.2351868753018858e-09,
      "loss": 0.2762,
      "num_input_tokens_seen": 22020080,
      "step": 33705
    },
    {
      "epoch": 19.87617924528302,
      "grad_norm": 3.442802906036377,
      "learning_rate": 1.1786663770996242e-09,
      "loss": 0.3526,
      "num_input_tokens_seen": 22022928,
      "step": 33710
    },
    {
      "epoch": 19.879127358490567,
      "grad_norm": 3.8756141662597656,
      "learning_rate": 1.1234693383893602e-09,
      "loss": 0.3275,
      "num_input_tokens_seen": 22026160,
      "step": 33715
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 5.167281627655029,
      "learning_rate": 1.0695957737844043e-09,
      "loss": 0.3054,
      "num_input_tokens_seen": 22029104,
      "step": 33720
    },
    {
      "epoch": 19.88502358490566,
      "grad_norm": 3.897064447402954,
      "learning_rate": 1.0170456975483467e-09,
      "loss": 0.2617,
      "num_input_tokens_seen": 22032208,
      "step": 33725
    },
    {
      "epoch": 19.88797169811321,
      "grad_norm": 3.348083019256592,
      "learning_rate": 9.658191235933922e-10,
      "loss": 0.2721,
      "num_input_tokens_seen": 22035408,
      "step": 33730
    },
    {
      "epoch": 19.890919811320753,
      "grad_norm": 3.937305450439453,
      "learning_rate": 9.159160654825805e-10,
      "loss": 0.3262,
      "num_input_tokens_seen": 22038544,
      "step": 33735
    },
    {
      "epoch": 19.8938679245283,
      "grad_norm": 3.6857526302337646,
      "learning_rate": 8.673365364281205e-10,
      "loss": 0.4444,
      "num_input_tokens_seen": 22042576,
      "step": 33740
    },
    {
      "epoch": 19.89681603773585,
      "grad_norm": 4.741818904876709,
      "learning_rate": 8.200805492913911e-10,
      "loss": 0.3298,
      "num_input_tokens_seen": 22045552,
      "step": 33745
    },
    {
      "epoch": 19.899764150943398,
      "grad_norm": 5.357819557189941,
      "learning_rate": 7.741481165834952e-10,
      "loss": 0.2693,
      "num_input_tokens_seen": 22048560,
      "step": 33750
    },
    {
      "epoch": 19.902712264150942,
      "grad_norm": 4.999390602111816,
      "learning_rate": 7.29539250465261e-10,
      "loss": 0.2872,
      "num_input_tokens_seen": 22051472,
      "step": 33755
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 2.470064878463745,
      "learning_rate": 6.862539627472409e-10,
      "loss": 0.2556,
      "num_input_tokens_seen": 22054704,
      "step": 33760
    },
    {
      "epoch": 19.90860849056604,
      "grad_norm": 2.7944180965423584,
      "learning_rate": 6.442922648897121e-10,
      "loss": 0.3408,
      "num_input_tokens_seen": 22058224,
      "step": 33765
    },
    {
      "epoch": 19.911556603773583,
      "grad_norm": 6.984306812286377,
      "learning_rate": 6.036541680015662e-10,
      "loss": 0.4193,
      "num_input_tokens_seen": 22062736,
      "step": 33770
    },
    {
      "epoch": 19.91450471698113,
      "grad_norm": 2.162064790725708,
      "learning_rate": 5.643396828419745e-10,
      "loss": 0.2805,
      "num_input_tokens_seen": 22067152,
      "step": 33775
    },
    {
      "epoch": 19.91745283018868,
      "grad_norm": 3.736823797225952,
      "learning_rate": 5.26348819819833e-10,
      "loss": 0.3321,
      "num_input_tokens_seen": 22070128,
      "step": 33780
    },
    {
      "epoch": 19.920400943396228,
      "grad_norm": 3.10927152633667,
      "learning_rate": 4.896815889937622e-10,
      "loss": 0.3171,
      "num_input_tokens_seen": 22073104,
      "step": 33785
    },
    {
      "epoch": 19.923349056603772,
      "grad_norm": 4.134750843048096,
      "learning_rate": 4.543380000704423e-10,
      "loss": 0.4025,
      "num_input_tokens_seen": 22076880,
      "step": 33790
    },
    {
      "epoch": 19.92629716981132,
      "grad_norm": 4.513524532318115,
      "learning_rate": 4.203180624084979e-10,
      "loss": 0.4202,
      "num_input_tokens_seen": 22080272,
      "step": 33795
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 5.99235725402832,
      "learning_rate": 3.876217850146136e-10,
      "loss": 0.4419,
      "num_input_tokens_seen": 22082672,
      "step": 33800
    },
    {
      "epoch": 19.932193396226417,
      "grad_norm": 4.555630683898926,
      "learning_rate": 3.56249176544643e-10,
      "loss": 0.2757,
      "num_input_tokens_seen": 22085456,
      "step": 33805
    },
    {
      "epoch": 19.93514150943396,
      "grad_norm": 3.389197587966919,
      "learning_rate": 3.262002453047197e-10,
      "loss": 0.3152,
      "num_input_tokens_seen": 22089072,
      "step": 33810
    },
    {
      "epoch": 19.93808962264151,
      "grad_norm": 4.302583694458008,
      "learning_rate": 2.974749992512571e-10,
      "loss": 0.3917,
      "num_input_tokens_seen": 22091632,
      "step": 33815
    },
    {
      "epoch": 19.941037735849058,
      "grad_norm": 3.1678085327148438,
      "learning_rate": 2.700734459881726e-10,
      "loss": 0.4005,
      "num_input_tokens_seen": 22094608,
      "step": 33820
    },
    {
      "epoch": 19.943985849056602,
      "grad_norm": 4.013486385345459,
      "learning_rate": 2.439955927713289e-10,
      "loss": 0.3744,
      "num_input_tokens_seen": 22098768,
      "step": 33825
    },
    {
      "epoch": 19.94693396226415,
      "grad_norm": 4.479465484619141,
      "learning_rate": 2.1924144650409263e-10,
      "loss": 0.3322,
      "num_input_tokens_seen": 22102512,
      "step": 33830
    },
    {
      "epoch": 19.9498820754717,
      "grad_norm": 4.7454142570495605,
      "learning_rate": 1.9581101374066546e-10,
      "loss": 0.3303,
      "num_input_tokens_seen": 22105040,
      "step": 33835
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 5.639388084411621,
      "learning_rate": 1.7370430068441858e-10,
      "loss": 0.3125,
      "num_input_tokens_seen": 22108400,
      "step": 33840
    },
    {
      "epoch": 19.95577830188679,
      "grad_norm": 4.2347283363342285,
      "learning_rate": 1.529213131878926e-10,
      "loss": 0.2879,
      "num_input_tokens_seen": 22111376,
      "step": 33845
    },
    {
      "epoch": 19.95872641509434,
      "grad_norm": 3.0609421730041504,
      "learning_rate": 1.3346205675335288e-10,
      "loss": 0.3411,
      "num_input_tokens_seen": 22114416,
      "step": 33850
    },
    {
      "epoch": 19.961674528301888,
      "grad_norm": 4.61405086517334,
      "learning_rate": 1.1532653653334447e-10,
      "loss": 0.3788,
      "num_input_tokens_seen": 22117328,
      "step": 33855
    },
    {
      "epoch": 19.964622641509433,
      "grad_norm": 2.970242738723755,
      "learning_rate": 9.85147573284717e-11,
      "loss": 0.2779,
      "num_input_tokens_seen": 22120880,
      "step": 33860
    },
    {
      "epoch": 19.96757075471698,
      "grad_norm": 3.2500529289245605,
      "learning_rate": 8.302672359072894e-11,
      "loss": 0.2331,
      "num_input_tokens_seen": 22124144,
      "step": 33865
    },
    {
      "epoch": 19.97051886792453,
      "grad_norm": 4.053802013397217,
      "learning_rate": 6.886243941961468e-11,
      "loss": 0.3581,
      "num_input_tokens_seen": 22127440,
      "step": 33870
    },
    {
      "epoch": 19.973466981132077,
      "grad_norm": 3.0404775142669678,
      "learning_rate": 5.602190856601741e-11,
      "loss": 0.224,
      "num_input_tokens_seen": 22130544,
      "step": 33875
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 6.755335807800293,
      "learning_rate": 4.450513442888493e-11,
      "loss": 0.2515,
      "num_input_tokens_seen": 22134096,
      "step": 33880
    },
    {
      "epoch": 19.97936320754717,
      "grad_norm": 1.9490594863891602,
      "learning_rate": 3.4312120057999886e-11,
      "loss": 0.3336,
      "num_input_tokens_seen": 22137360,
      "step": 33885
    },
    {
      "epoch": 19.982311320754718,
      "grad_norm": 3.896151542663574,
      "learning_rate": 2.5442868151204224e-11,
      "loss": 0.3678,
      "num_input_tokens_seen": 22140752,
      "step": 33890
    },
    {
      "epoch": 19.985259433962263,
      "grad_norm": 5.548608303070068,
      "learning_rate": 1.7897381057729867e-11,
      "loss": 0.3882,
      "num_input_tokens_seen": 22143120,
      "step": 33895
    },
    {
      "epoch": 19.98820754716981,
      "grad_norm": 2.5209734439849854,
      "learning_rate": 1.1675660773757813e-11,
      "loss": 0.2653,
      "num_input_tokens_seen": 22146864,
      "step": 33900
    },
    {
      "epoch": 19.99115566037736,
      "grad_norm": 6.747762680053711,
      "learning_rate": 6.777708947969253e-12,
      "loss": 0.2641,
      "num_input_tokens_seen": 22149808,
      "step": 33905
    },
    {
      "epoch": 19.994103773584907,
      "grad_norm": 3.535848379135132,
      "learning_rate": 3.2035268765495674e-12,
      "loss": 0.199,
      "num_input_tokens_seen": 22152592,
      "step": 33910
    },
    {
      "epoch": 19.997051886792452,
      "grad_norm": 4.722427845001221,
      "learning_rate": 9.531155059638863e-13,
      "loss": 0.227,
      "num_input_tokens_seen": 22155184,
      "step": 33915
    },
    {
      "epoch": 20.0,
      "grad_norm": 17.234344482421875,
      "learning_rate": 2.6475431291750342e-14,
      "loss": 0.3648,
      "num_input_tokens_seen": 22157104,
      "step": 33920
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.5974128246307373,
      "eval_runtime": 19.0628,
      "eval_samples_per_second": 88.969,
      "eval_steps_per_second": 22.242,
      "num_input_tokens_seen": 22157104,
      "step": 33920
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 22157104,
      "step": 33920,
      "total_flos": 9.977243788180193e+17,
      "train_loss": 0.46073363143072094,
      "train_runtime": 4068.8559,
      "train_samples_per_second": 33.331,
      "train_steps_per_second": 8.336
    }
  ],
  "logging_steps": 5,
  "max_steps": 33920,
  "num_input_tokens_seen": 22157104,
  "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": 9.977243788180193e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}