{
  "best_global_step": 19080,
  "best_metric": 0.49317190051078796,
  "best_model_checkpoint": "saves/prompt-tuning/llama-3-8b-instruct/train_codealpacapy_1754891836/checkpoint-19080",
  "epoch": 10.0,
  "eval_steps": 954,
  "global_step": 19080,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002620545073375262,
      "grad_norm": 1.203125,
      "learning_rate": 1.048218029350105e-07,
      "loss": 1.2628,
      "num_input_tokens_seen": 2944,
      "step": 5
    },
    {
      "epoch": 0.005241090146750524,
      "grad_norm": 0.71484375,
      "learning_rate": 2.3584905660377358e-07,
      "loss": 1.1578,
      "num_input_tokens_seen": 6816,
      "step": 10
    },
    {
      "epoch": 0.007861635220125786,
      "grad_norm": 0.86328125,
      "learning_rate": 3.6687631027253674e-07,
      "loss": 1.2492,
      "num_input_tokens_seen": 9760,
      "step": 15
    },
    {
      "epoch": 0.010482180293501049,
      "grad_norm": 0.46875,
      "learning_rate": 4.979035639412998e-07,
      "loss": 1.1866,
      "num_input_tokens_seen": 13472,
      "step": 20
    },
    {
      "epoch": 0.01310272536687631,
      "grad_norm": 1.6171875,
      "learning_rate": 6.28930817610063e-07,
      "loss": 1.2291,
      "num_input_tokens_seen": 16864,
      "step": 25
    },
    {
      "epoch": 0.015723270440251572,
      "grad_norm": 1.109375,
      "learning_rate": 7.59958071278826e-07,
      "loss": 1.3849,
      "num_input_tokens_seen": 20352,
      "step": 30
    },
    {
      "epoch": 0.018343815513626835,
      "grad_norm": 1.09375,
      "learning_rate": 8.90985324947589e-07,
      "loss": 1.3365,
      "num_input_tokens_seen": 23008,
      "step": 35
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 0.8671875,
      "learning_rate": 1.0220125786163522e-06,
      "loss": 1.0781,
      "num_input_tokens_seen": 25920,
      "step": 40
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 1.1875,
      "learning_rate": 1.1530398322851154e-06,
      "loss": 1.2643,
      "num_input_tokens_seen": 28672,
      "step": 45
    },
    {
      "epoch": 0.02620545073375262,
      "grad_norm": 0.55078125,
      "learning_rate": 1.2840670859538784e-06,
      "loss": 1.0797,
      "num_input_tokens_seen": 31488,
      "step": 50
    },
    {
      "epoch": 0.028825995807127882,
      "grad_norm": 1.078125,
      "learning_rate": 1.4150943396226415e-06,
      "loss": 1.2637,
      "num_input_tokens_seen": 34496,
      "step": 55
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 0.53515625,
      "learning_rate": 1.5461215932914047e-06,
      "loss": 1.3338,
      "num_input_tokens_seen": 38112,
      "step": 60
    },
    {
      "epoch": 0.034067085953878404,
      "grad_norm": 0.65625,
      "learning_rate": 1.677148846960168e-06,
      "loss": 1.407,
      "num_input_tokens_seen": 40864,
      "step": 65
    },
    {
      "epoch": 0.03668763102725367,
      "grad_norm": 0.3984375,
      "learning_rate": 1.8081761006289309e-06,
      "loss": 1.3176,
      "num_input_tokens_seen": 44192,
      "step": 70
    },
    {
      "epoch": 0.03930817610062893,
      "grad_norm": 1.8359375,
      "learning_rate": 1.939203354297694e-06,
      "loss": 1.4341,
      "num_input_tokens_seen": 46784,
      "step": 75
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 2.703125,
      "learning_rate": 2.0702306079664572e-06,
      "loss": 1.3998,
      "num_input_tokens_seen": 50176,
      "step": 80
    },
    {
      "epoch": 0.044549266247379454,
      "grad_norm": 0.63671875,
      "learning_rate": 2.20125786163522e-06,
      "loss": 1.0991,
      "num_input_tokens_seen": 54400,
      "step": 85
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 1.40625,
      "learning_rate": 2.3322851153039836e-06,
      "loss": 1.4359,
      "num_input_tokens_seen": 56992,
      "step": 90
    },
    {
      "epoch": 0.04979035639412998,
      "grad_norm": 1.671875,
      "learning_rate": 2.4633123689727464e-06,
      "loss": 1.0197,
      "num_input_tokens_seen": 60896,
      "step": 95
    },
    {
      "epoch": 0.05241090146750524,
      "grad_norm": 0.490234375,
      "learning_rate": 2.5943396226415095e-06,
      "loss": 1.1701,
      "num_input_tokens_seen": 64576,
      "step": 100
    },
    {
      "epoch": 0.055031446540880505,
      "grad_norm": 1.7578125,
      "learning_rate": 2.7253668763102727e-06,
      "loss": 1.3655,
      "num_input_tokens_seen": 67616,
      "step": 105
    },
    {
      "epoch": 0.057651991614255764,
      "grad_norm": 0.56640625,
      "learning_rate": 2.8563941299790355e-06,
      "loss": 1.056,
      "num_input_tokens_seen": 72544,
      "step": 110
    },
    {
      "epoch": 0.06027253668763103,
      "grad_norm": 0.65234375,
      "learning_rate": 2.987421383647799e-06,
      "loss": 1.3432,
      "num_input_tokens_seen": 75648,
      "step": 115
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 1.4921875,
      "learning_rate": 3.118448637316562e-06,
      "loss": 1.3879,
      "num_input_tokens_seen": 78368,
      "step": 120
    },
    {
      "epoch": 0.06551362683438156,
      "grad_norm": 0.921875,
      "learning_rate": 3.249475890985325e-06,
      "loss": 1.2679,
      "num_input_tokens_seen": 81952,
      "step": 125
    },
    {
      "epoch": 0.06813417190775681,
      "grad_norm": 1.390625,
      "learning_rate": 3.380503144654088e-06,
      "loss": 1.2207,
      "num_input_tokens_seen": 84960,
      "step": 130
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 1.1875,
      "learning_rate": 3.5115303983228514e-06,
      "loss": 1.3446,
      "num_input_tokens_seen": 88576,
      "step": 135
    },
    {
      "epoch": 0.07337526205450734,
      "grad_norm": 1.203125,
      "learning_rate": 3.642557651991614e-06,
      "loss": 1.0626,
      "num_input_tokens_seen": 92160,
      "step": 140
    },
    {
      "epoch": 0.0759958071278826,
      "grad_norm": 0.73046875,
      "learning_rate": 3.7735849056603773e-06,
      "loss": 1.1339,
      "num_input_tokens_seen": 95840,
      "step": 145
    },
    {
      "epoch": 0.07861635220125786,
      "grad_norm": 1.1875,
      "learning_rate": 3.9046121593291405e-06,
      "loss": 1.2053,
      "num_input_tokens_seen": 98816,
      "step": 150
    },
    {
      "epoch": 0.08123689727463312,
      "grad_norm": 0.6875,
      "learning_rate": 4.035639412997904e-06,
      "loss": 1.0429,
      "num_input_tokens_seen": 102880,
      "step": 155
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 1.1015625,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.1913,
      "num_input_tokens_seen": 105920,
      "step": 160
    },
    {
      "epoch": 0.08647798742138364,
      "grad_norm": 0.5703125,
      "learning_rate": 4.29769392033543e-06,
      "loss": 1.3801,
      "num_input_tokens_seen": 108800,
      "step": 165
    },
    {
      "epoch": 0.08909853249475891,
      "grad_norm": 1.078125,
      "learning_rate": 4.428721174004193e-06,
      "loss": 1.1379,
      "num_input_tokens_seen": 111968,
      "step": 170
    },
    {
      "epoch": 0.09171907756813417,
      "grad_norm": 1.4375,
      "learning_rate": 4.559748427672956e-06,
      "loss": 1.1921,
      "num_input_tokens_seen": 114976,
      "step": 175
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 4.03125,
      "learning_rate": 4.69077568134172e-06,
      "loss": 1.4869,
      "num_input_tokens_seen": 117568,
      "step": 180
    },
    {
      "epoch": 0.09696016771488469,
      "grad_norm": 1.3203125,
      "learning_rate": 4.821802935010482e-06,
      "loss": 1.0642,
      "num_input_tokens_seen": 120384,
      "step": 185
    },
    {
      "epoch": 0.09958071278825996,
      "grad_norm": 0.6484375,
      "learning_rate": 4.952830188679246e-06,
      "loss": 1.0254,
      "num_input_tokens_seen": 123680,
      "step": 190
    },
    {
      "epoch": 0.10220125786163523,
      "grad_norm": 1.3203125,
      "learning_rate": 5.083857442348009e-06,
      "loss": 1.2598,
      "num_input_tokens_seen": 127168,
      "step": 195
    },
    {
      "epoch": 0.10482180293501048,
      "grad_norm": 1.2890625,
      "learning_rate": 5.2148846960167715e-06,
      "loss": 1.0045,
      "num_input_tokens_seen": 130176,
      "step": 200
    },
    {
      "epoch": 0.10744234800838574,
      "grad_norm": 1.125,
      "learning_rate": 5.345911949685535e-06,
      "loss": 1.1704,
      "num_input_tokens_seen": 133056,
      "step": 205
    },
    {
      "epoch": 0.11006289308176101,
      "grad_norm": 1.09375,
      "learning_rate": 5.476939203354298e-06,
      "loss": 1.0809,
      "num_input_tokens_seen": 136864,
      "step": 210
    },
    {
      "epoch": 0.11268343815513626,
      "grad_norm": 1.2421875,
      "learning_rate": 5.607966457023061e-06,
      "loss": 1.4802,
      "num_input_tokens_seen": 140064,
      "step": 215
    },
    {
      "epoch": 0.11530398322851153,
      "grad_norm": 1.328125,
      "learning_rate": 5.738993710691824e-06,
      "loss": 1.0521,
      "num_input_tokens_seen": 143712,
      "step": 220
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 1.078125,
      "learning_rate": 5.870020964360588e-06,
      "loss": 1.1935,
      "num_input_tokens_seen": 146880,
      "step": 225
    },
    {
      "epoch": 0.12054507337526206,
      "grad_norm": 1.640625,
      "learning_rate": 6.0010482180293506e-06,
      "loss": 1.0454,
      "num_input_tokens_seen": 150560,
      "step": 230
    },
    {
      "epoch": 0.12316561844863731,
      "grad_norm": 0.765625,
      "learning_rate": 6.132075471698113e-06,
      "loss": 1.3067,
      "num_input_tokens_seen": 153664,
      "step": 235
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 2.046875,
      "learning_rate": 6.263102725366876e-06,
      "loss": 1.028,
      "num_input_tokens_seen": 157568,
      "step": 240
    },
    {
      "epoch": 0.12840670859538783,
      "grad_norm": 1.4453125,
      "learning_rate": 6.3941299790356405e-06,
      "loss": 1.119,
      "num_input_tokens_seen": 160512,
      "step": 245
    },
    {
      "epoch": 0.1310272536687631,
      "grad_norm": 0.56640625,
      "learning_rate": 6.5251572327044024e-06,
      "loss": 1.0176,
      "num_input_tokens_seen": 164512,
      "step": 250
    },
    {
      "epoch": 0.13364779874213836,
      "grad_norm": 2.328125,
      "learning_rate": 6.656184486373165e-06,
      "loss": 1.254,
      "num_input_tokens_seen": 167200,
      "step": 255
    },
    {
      "epoch": 0.13626834381551362,
      "grad_norm": 2.109375,
      "learning_rate": 6.78721174004193e-06,
      "loss": 1.287,
      "num_input_tokens_seen": 169984,
      "step": 260
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 1.2109375,
      "learning_rate": 6.918238993710692e-06,
      "loss": 1.1272,
      "num_input_tokens_seen": 173856,
      "step": 265
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 1.125,
      "learning_rate": 7.049266247379454e-06,
      "loss": 1.4513,
      "num_input_tokens_seen": 176192,
      "step": 270
    },
    {
      "epoch": 0.1441299790356394,
      "grad_norm": 0.5859375,
      "learning_rate": 7.180293501048219e-06,
      "loss": 0.9098,
      "num_input_tokens_seen": 179840,
      "step": 275
    },
    {
      "epoch": 0.14675052410901468,
      "grad_norm": 0.88671875,
      "learning_rate": 7.3113207547169815e-06,
      "loss": 1.0298,
      "num_input_tokens_seen": 182976,
      "step": 280
    },
    {
      "epoch": 0.14937106918238993,
      "grad_norm": 0.93359375,
      "learning_rate": 7.442348008385745e-06,
      "loss": 1.0592,
      "num_input_tokens_seen": 186400,
      "step": 285
    },
    {
      "epoch": 0.1519916142557652,
      "grad_norm": 1.609375,
      "learning_rate": 7.573375262054508e-06,
      "loss": 1.1071,
      "num_input_tokens_seen": 190144,
      "step": 290
    },
    {
      "epoch": 0.15461215932914046,
      "grad_norm": 1.7109375,
      "learning_rate": 7.70440251572327e-06,
      "loss": 1.3272,
      "num_input_tokens_seen": 193344,
      "step": 295
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 1.9765625,
      "learning_rate": 7.835429769392034e-06,
      "loss": 1.0238,
      "num_input_tokens_seen": 195904,
      "step": 300
    },
    {
      "epoch": 0.159853249475891,
      "grad_norm": 1.5859375,
      "learning_rate": 7.966457023060797e-06,
      "loss": 1.158,
      "num_input_tokens_seen": 198944,
      "step": 305
    },
    {
      "epoch": 0.16247379454926625,
      "grad_norm": 3.390625,
      "learning_rate": 8.09748427672956e-06,
      "loss": 1.3459,
      "num_input_tokens_seen": 201792,
      "step": 310
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 0.84765625,
      "learning_rate": 8.228511530398324e-06,
      "loss": 1.0175,
      "num_input_tokens_seen": 206080,
      "step": 315
    },
    {
      "epoch": 0.16771488469601678,
      "grad_norm": 2.046875,
      "learning_rate": 8.359538784067087e-06,
      "loss": 1.1506,
      "num_input_tokens_seen": 209376,
      "step": 320
    },
    {
      "epoch": 0.17033542976939203,
      "grad_norm": 1.9296875,
      "learning_rate": 8.49056603773585e-06,
      "loss": 1.2901,
      "num_input_tokens_seen": 212160,
      "step": 325
    },
    {
      "epoch": 0.17295597484276728,
      "grad_norm": 0.90625,
      "learning_rate": 8.621593291404612e-06,
      "loss": 0.9216,
      "num_input_tokens_seen": 216128,
      "step": 330
    },
    {
      "epoch": 0.17557651991614256,
      "grad_norm": 1.5,
      "learning_rate": 8.752620545073375e-06,
      "loss": 1.9761,
      "num_input_tokens_seen": 219232,
      "step": 335
    },
    {
      "epoch": 0.17819706498951782,
      "grad_norm": 2.21875,
      "learning_rate": 8.883647798742138e-06,
      "loss": 0.9717,
      "num_input_tokens_seen": 222464,
      "step": 340
    },
    {
      "epoch": 0.18081761006289307,
      "grad_norm": 3.421875,
      "learning_rate": 9.014675052410902e-06,
      "loss": 1.3254,
      "num_input_tokens_seen": 225184,
      "step": 345
    },
    {
      "epoch": 0.18343815513626835,
      "grad_norm": 6.375,
      "learning_rate": 9.145702306079665e-06,
      "loss": 1.0733,
      "num_input_tokens_seen": 228000,
      "step": 350
    },
    {
      "epoch": 0.1860587002096436,
      "grad_norm": 4.21875,
      "learning_rate": 9.276729559748428e-06,
      "loss": 0.9186,
      "num_input_tokens_seen": 230944,
      "step": 355
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 20.375,
      "learning_rate": 9.40775681341719e-06,
      "loss": 0.8539,
      "num_input_tokens_seen": 233728,
      "step": 360
    },
    {
      "epoch": 0.19129979035639413,
      "grad_norm": 9.0625,
      "learning_rate": 9.538784067085953e-06,
      "loss": 0.7485,
      "num_input_tokens_seen": 236480,
      "step": 365
    },
    {
      "epoch": 0.19392033542976939,
      "grad_norm": 14.1875,
      "learning_rate": 9.669811320754718e-06,
      "loss": 0.7958,
      "num_input_tokens_seen": 239488,
      "step": 370
    },
    {
      "epoch": 0.19654088050314467,
      "grad_norm": 9.875,
      "learning_rate": 9.80083857442348e-06,
      "loss": 0.8409,
      "num_input_tokens_seen": 242144,
      "step": 375
    },
    {
      "epoch": 0.19916142557651992,
      "grad_norm": 5.46875,
      "learning_rate": 9.931865828092243e-06,
      "loss": 0.7258,
      "num_input_tokens_seen": 245216,
      "step": 380
    },
    {
      "epoch": 0.20178197064989517,
      "grad_norm": 4.71875,
      "learning_rate": 1.0062893081761008e-05,
      "loss": 0.7808,
      "num_input_tokens_seen": 249632,
      "step": 385
    },
    {
      "epoch": 0.20440251572327045,
      "grad_norm": 2.484375,
      "learning_rate": 1.019392033542977e-05,
      "loss": 0.7488,
      "num_input_tokens_seen": 253312,
      "step": 390
    },
    {
      "epoch": 0.2070230607966457,
      "grad_norm": 7.375,
      "learning_rate": 1.0324947589098532e-05,
      "loss": 0.673,
      "num_input_tokens_seen": 256384,
      "step": 395
    },
    {
      "epoch": 0.20964360587002095,
      "grad_norm": 35.5,
      "learning_rate": 1.0455974842767296e-05,
      "loss": 0.7235,
      "num_input_tokens_seen": 259744,
      "step": 400
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 11.125,
      "learning_rate": 1.0587002096436059e-05,
      "loss": 0.8728,
      "num_input_tokens_seen": 263456,
      "step": 405
    },
    {
      "epoch": 0.2148846960167715,
      "grad_norm": 6.34375,
      "learning_rate": 1.0718029350104822e-05,
      "loss": 0.6691,
      "num_input_tokens_seen": 266496,
      "step": 410
    },
    {
      "epoch": 0.21750524109014674,
      "grad_norm": 5.34375,
      "learning_rate": 1.0849056603773586e-05,
      "loss": 0.7354,
      "num_input_tokens_seen": 269600,
      "step": 415
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 5.71875,
      "learning_rate": 1.0980083857442349e-05,
      "loss": 0.7052,
      "num_input_tokens_seen": 273152,
      "step": 420
    },
    {
      "epoch": 0.22274633123689727,
      "grad_norm": 6.3125,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.8262,
      "num_input_tokens_seen": 276992,
      "step": 425
    },
    {
      "epoch": 0.22536687631027252,
      "grad_norm": 11.5625,
      "learning_rate": 1.1242138364779874e-05,
      "loss": 0.7826,
      "num_input_tokens_seen": 279296,
      "step": 430
    },
    {
      "epoch": 0.2279874213836478,
      "grad_norm": 2.625,
      "learning_rate": 1.1373165618448637e-05,
      "loss": 0.4965,
      "num_input_tokens_seen": 283296,
      "step": 435
    },
    {
      "epoch": 0.23060796645702306,
      "grad_norm": 9.5625,
      "learning_rate": 1.1504192872117402e-05,
      "loss": 0.8445,
      "num_input_tokens_seen": 286016,
      "step": 440
    },
    {
      "epoch": 0.23322851153039834,
      "grad_norm": 7.90625,
      "learning_rate": 1.1635220125786164e-05,
      "loss": 0.6474,
      "num_input_tokens_seen": 289248,
      "step": 445
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 4.65625,
      "learning_rate": 1.1766247379454927e-05,
      "loss": 0.6856,
      "num_input_tokens_seen": 292160,
      "step": 450
    },
    {
      "epoch": 0.23846960167714884,
      "grad_norm": 4.8125,
      "learning_rate": 1.1897274633123692e-05,
      "loss": 0.6325,
      "num_input_tokens_seen": 295296,
      "step": 455
    },
    {
      "epoch": 0.24109014675052412,
      "grad_norm": 3.53125,
      "learning_rate": 1.2028301886792454e-05,
      "loss": 0.6974,
      "num_input_tokens_seen": 299360,
      "step": 460
    },
    {
      "epoch": 0.24371069182389937,
      "grad_norm": 4.03125,
      "learning_rate": 1.2159329140461215e-05,
      "loss": 0.5518,
      "num_input_tokens_seen": 303072,
      "step": 465
    },
    {
      "epoch": 0.24633123689727462,
      "grad_norm": 7.84375,
      "learning_rate": 1.229035639412998e-05,
      "loss": 0.5498,
      "num_input_tokens_seen": 307616,
      "step": 470
    },
    {
      "epoch": 0.2489517819706499,
      "grad_norm": 4.4375,
      "learning_rate": 1.2421383647798743e-05,
      "loss": 0.6387,
      "num_input_tokens_seen": 310112,
      "step": 475
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 5.6875,
      "learning_rate": 1.2552410901467507e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 313472,
      "step": 480
    },
    {
      "epoch": 0.25419287211740044,
      "grad_norm": 4.34375,
      "learning_rate": 1.2683438155136268e-05,
      "loss": 0.5813,
      "num_input_tokens_seen": 316128,
      "step": 485
    },
    {
      "epoch": 0.25681341719077566,
      "grad_norm": 5.875,
      "learning_rate": 1.2814465408805033e-05,
      "loss": 0.5419,
      "num_input_tokens_seen": 319584,
      "step": 490
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 4.46875,
      "learning_rate": 1.2945492662473795e-05,
      "loss": 0.6292,
      "num_input_tokens_seen": 322496,
      "step": 495
    },
    {
      "epoch": 0.2620545073375262,
      "grad_norm": 6.625,
      "learning_rate": 1.3076519916142556e-05,
      "loss": 0.9389,
      "num_input_tokens_seen": 325120,
      "step": 500
    },
    {
      "epoch": 0.26467505241090145,
      "grad_norm": 16.375,
      "learning_rate": 1.320754716981132e-05,
      "loss": 0.6915,
      "num_input_tokens_seen": 327840,
      "step": 505
    },
    {
      "epoch": 0.2672955974842767,
      "grad_norm": 4.375,
      "learning_rate": 1.3338574423480085e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 331776,
      "step": 510
    },
    {
      "epoch": 0.269916142557652,
      "grad_norm": 22.125,
      "learning_rate": 1.3469601677148846e-05,
      "loss": 0.5575,
      "num_input_tokens_seen": 336544,
      "step": 515
    },
    {
      "epoch": 0.27253668763102723,
      "grad_norm": 5.59375,
      "learning_rate": 1.360062893081761e-05,
      "loss": 0.7599,
      "num_input_tokens_seen": 340768,
      "step": 520
    },
    {
      "epoch": 0.2751572327044025,
      "grad_norm": 9.1875,
      "learning_rate": 1.3731656184486375e-05,
      "loss": 0.7095,
      "num_input_tokens_seen": 343296,
      "step": 525
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 5.375,
      "learning_rate": 1.3862683438155136e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 346848,
      "step": 530
    },
    {
      "epoch": 0.280398322851153,
      "grad_norm": 4.25,
      "learning_rate": 1.3993710691823899e-05,
      "loss": 0.6471,
      "num_input_tokens_seen": 349632,
      "step": 535
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 6.65625,
      "learning_rate": 1.4124737945492664e-05,
      "loss": 0.6085,
      "num_input_tokens_seen": 353024,
      "step": 540
    },
    {
      "epoch": 0.2856394129979036,
      "grad_norm": 6.125,
      "learning_rate": 1.4255765199161425e-05,
      "loss": 0.6005,
      "num_input_tokens_seen": 355552,
      "step": 545
    },
    {
      "epoch": 0.2882599580712788,
      "grad_norm": 5.3125,
      "learning_rate": 1.4386792452830189e-05,
      "loss": 0.5453,
      "num_input_tokens_seen": 360128,
      "step": 550
    },
    {
      "epoch": 0.2908805031446541,
      "grad_norm": 3.71875,
      "learning_rate": 1.4517819706498954e-05,
      "loss": 0.601,
      "num_input_tokens_seen": 363360,
      "step": 555
    },
    {
      "epoch": 0.29350104821802936,
      "grad_norm": 7.53125,
      "learning_rate": 1.4648846960167716e-05,
      "loss": 0.6307,
      "num_input_tokens_seen": 366144,
      "step": 560
    },
    {
      "epoch": 0.29612159329140464,
      "grad_norm": 6.125,
      "learning_rate": 1.4779874213836479e-05,
      "loss": 0.6256,
      "num_input_tokens_seen": 369216,
      "step": 565
    },
    {
      "epoch": 0.29874213836477986,
      "grad_norm": 21.375,
      "learning_rate": 1.4910901467505242e-05,
      "loss": 0.6783,
      "num_input_tokens_seen": 372512,
      "step": 570
    },
    {
      "epoch": 0.30136268343815514,
      "grad_norm": 16.25,
      "learning_rate": 1.5041928721174006e-05,
      "loss": 0.6356,
      "num_input_tokens_seen": 374976,
      "step": 575
    },
    {
      "epoch": 0.3039832285115304,
      "grad_norm": 11.3125,
      "learning_rate": 1.5172955974842767e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 377632,
      "step": 580
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 8.1875,
      "learning_rate": 1.530398322851153e-05,
      "loss": 0.7531,
      "num_input_tokens_seen": 380768,
      "step": 585
    },
    {
      "epoch": 0.30922431865828093,
      "grad_norm": 5.46875,
      "learning_rate": 1.5435010482180296e-05,
      "loss": 0.5701,
      "num_input_tokens_seen": 383520,
      "step": 590
    },
    {
      "epoch": 0.3118448637316562,
      "grad_norm": 11.6875,
      "learning_rate": 1.5566037735849056e-05,
      "loss": 0.8314,
      "num_input_tokens_seen": 386496,
      "step": 595
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 9.0,
      "learning_rate": 1.5697064989517822e-05,
      "loss": 0.6421,
      "num_input_tokens_seen": 390464,
      "step": 600
    },
    {
      "epoch": 0.3170859538784067,
      "grad_norm": 9.3125,
      "learning_rate": 1.5828092243186584e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 393440,
      "step": 605
    },
    {
      "epoch": 0.319706498951782,
      "grad_norm": 6.0625,
      "learning_rate": 1.5959119496855347e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 398336,
      "step": 610
    },
    {
      "epoch": 0.3223270440251572,
      "grad_norm": 13.3125,
      "learning_rate": 1.609014675052411e-05,
      "loss": 0.5294,
      "num_input_tokens_seen": 401504,
      "step": 615
    },
    {
      "epoch": 0.3249475890985325,
      "grad_norm": 10.875,
      "learning_rate": 1.6221174004192873e-05,
      "loss": 0.5336,
      "num_input_tokens_seen": 404864,
      "step": 620
    },
    {
      "epoch": 0.3275681341719078,
      "grad_norm": 5.40625,
      "learning_rate": 1.6352201257861635e-05,
      "loss": 0.65,
      "num_input_tokens_seen": 409056,
      "step": 625
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 4.8125,
      "learning_rate": 1.6483228511530398e-05,
      "loss": 0.6216,
      "num_input_tokens_seen": 412544,
      "step": 630
    },
    {
      "epoch": 0.3328092243186583,
      "grad_norm": 24.0,
      "learning_rate": 1.6614255765199164e-05,
      "loss": 0.6184,
      "num_input_tokens_seen": 415520,
      "step": 635
    },
    {
      "epoch": 0.33542976939203356,
      "grad_norm": 8.25,
      "learning_rate": 1.6745283018867924e-05,
      "loss": 1.0227,
      "num_input_tokens_seen": 417760,
      "step": 640
    },
    {
      "epoch": 0.3380503144654088,
      "grad_norm": 4.1875,
      "learning_rate": 1.687631027253669e-05,
      "loss": 0.6247,
      "num_input_tokens_seen": 421728,
      "step": 645
    },
    {
      "epoch": 0.34067085953878407,
      "grad_norm": 37.25,
      "learning_rate": 1.7007337526205453e-05,
      "loss": 0.6085,
      "num_input_tokens_seen": 425664,
      "step": 650
    },
    {
      "epoch": 0.34329140461215935,
      "grad_norm": 8.125,
      "learning_rate": 1.7138364779874212e-05,
      "loss": 0.531,
      "num_input_tokens_seen": 428896,
      "step": 655
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 13.8125,
      "learning_rate": 1.7269392033542978e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 431744,
      "step": 660
    },
    {
      "epoch": 0.34853249475890985,
      "grad_norm": 8.125,
      "learning_rate": 1.740041928721174e-05,
      "loss": 0.5823,
      "num_input_tokens_seen": 435520,
      "step": 665
    },
    {
      "epoch": 0.35115303983228513,
      "grad_norm": 4.84375,
      "learning_rate": 1.7531446540880504e-05,
      "loss": 0.5513,
      "num_input_tokens_seen": 438592,
      "step": 670
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 5.0625,
      "learning_rate": 1.7662473794549266e-05,
      "loss": 0.5737,
      "num_input_tokens_seen": 441664,
      "step": 675
    },
    {
      "epoch": 0.35639412997903563,
      "grad_norm": 11.25,
      "learning_rate": 1.779350104821803e-05,
      "loss": 0.6602,
      "num_input_tokens_seen": 444832,
      "step": 680
    },
    {
      "epoch": 0.3590146750524109,
      "grad_norm": 6.4375,
      "learning_rate": 1.7924528301886792e-05,
      "loss": 0.7295,
      "num_input_tokens_seen": 447392,
      "step": 685
    },
    {
      "epoch": 0.36163522012578614,
      "grad_norm": 11.25,
      "learning_rate": 1.8055555555555555e-05,
      "loss": 0.6172,
      "num_input_tokens_seen": 450528,
      "step": 690
    },
    {
      "epoch": 0.3642557651991614,
      "grad_norm": 19.25,
      "learning_rate": 1.818658280922432e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 453504,
      "step": 695
    },
    {
      "epoch": 0.3668763102725367,
      "grad_norm": 18.5,
      "learning_rate": 1.831761006289308e-05,
      "loss": 0.8998,
      "num_input_tokens_seen": 456288,
      "step": 700
    },
    {
      "epoch": 0.3694968553459119,
      "grad_norm": 23.0,
      "learning_rate": 1.8448637316561846e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 459840,
      "step": 705
    },
    {
      "epoch": 0.3721174004192872,
      "grad_norm": 7.5625,
      "learning_rate": 1.857966457023061e-05,
      "loss": 0.4875,
      "num_input_tokens_seen": 463168,
      "step": 710
    },
    {
      "epoch": 0.3747379454926625,
      "grad_norm": 13.8125,
      "learning_rate": 1.8710691823899372e-05,
      "loss": 0.5495,
      "num_input_tokens_seen": 466112,
      "step": 715
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 14.625,
      "learning_rate": 1.8841719077568135e-05,
      "loss": 0.6965,
      "num_input_tokens_seen": 469568,
      "step": 720
    },
    {
      "epoch": 0.379979035639413,
      "grad_norm": 5.5,
      "learning_rate": 1.8972746331236897e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 472896,
      "step": 725
    },
    {
      "epoch": 0.38259958071278827,
      "grad_norm": 15.75,
      "learning_rate": 1.9103773584905664e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 475744,
      "step": 730
    },
    {
      "epoch": 0.38522012578616355,
      "grad_norm": 3.03125,
      "learning_rate": 1.9234800838574423e-05,
      "loss": 0.5742,
      "num_input_tokens_seen": 481216,
      "step": 735
    },
    {
      "epoch": 0.38784067085953877,
      "grad_norm": 14.75,
      "learning_rate": 1.936582809224319e-05,
      "loss": 0.5894,
      "num_input_tokens_seen": 483488,
      "step": 740
    },
    {
      "epoch": 0.39046121593291405,
      "grad_norm": 10.375,
      "learning_rate": 1.9496855345911952e-05,
      "loss": 0.6533,
      "num_input_tokens_seen": 487072,
      "step": 745
    },
    {
      "epoch": 0.39308176100628933,
      "grad_norm": 7.375,
      "learning_rate": 1.9627882599580715e-05,
      "loss": 0.5764,
      "num_input_tokens_seen": 490496,
      "step": 750
    },
    {
      "epoch": 0.39570230607966456,
      "grad_norm": 10.25,
      "learning_rate": 1.9758909853249477e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 493824,
      "step": 755
    },
    {
      "epoch": 0.39832285115303984,
      "grad_norm": 14.5625,
      "learning_rate": 1.988993710691824e-05,
      "loss": 0.8022,
      "num_input_tokens_seen": 496192,
      "step": 760
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 5.46875,
      "learning_rate": 2.0020964360587003e-05,
      "loss": 0.6403,
      "num_input_tokens_seen": 498880,
      "step": 765
    },
    {
      "epoch": 0.40356394129979034,
      "grad_norm": 6.75,
      "learning_rate": 2.0151991614255766e-05,
      "loss": 0.5626,
      "num_input_tokens_seen": 502016,
      "step": 770
    },
    {
      "epoch": 0.4061844863731656,
      "grad_norm": 3.1875,
      "learning_rate": 2.0283018867924532e-05,
      "loss": 0.5998,
      "num_input_tokens_seen": 505120,
      "step": 775
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 13.8125,
      "learning_rate": 2.041404612159329e-05,
      "loss": 0.5897,
      "num_input_tokens_seen": 508256,
      "step": 780
    },
    {
      "epoch": 0.4114255765199161,
      "grad_norm": 5.34375,
      "learning_rate": 2.0545073375262054e-05,
      "loss": 0.5779,
      "num_input_tokens_seen": 511776,
      "step": 785
    },
    {
      "epoch": 0.4140461215932914,
      "grad_norm": 19.625,
      "learning_rate": 2.067610062893082e-05,
      "loss": 0.5681,
      "num_input_tokens_seen": 514464,
      "step": 790
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 23.75,
      "learning_rate": 2.080712788259958e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 518144,
      "step": 795
    },
    {
      "epoch": 0.4192872117400419,
      "grad_norm": 12.125,
      "learning_rate": 2.0938155136268346e-05,
      "loss": 0.5975,
      "num_input_tokens_seen": 520768,
      "step": 800
    },
    {
      "epoch": 0.4219077568134172,
      "grad_norm": 12.75,
      "learning_rate": 2.106918238993711e-05,
      "loss": 0.7107,
      "num_input_tokens_seen": 523808,
      "step": 805
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 8.0625,
      "learning_rate": 2.120020964360587e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 527136,
      "step": 810
    },
    {
      "epoch": 0.4271488469601677,
      "grad_norm": 28.375,
      "learning_rate": 2.1331236897274634e-05,
      "loss": 0.6855,
      "num_input_tokens_seen": 530240,
      "step": 815
    },
    {
      "epoch": 0.429769392033543,
      "grad_norm": 16.5,
      "learning_rate": 2.1462264150943397e-05,
      "loss": 0.5787,
      "num_input_tokens_seen": 534496,
      "step": 820
    },
    {
      "epoch": 0.43238993710691825,
      "grad_norm": 17.75,
      "learning_rate": 2.159329140461216e-05,
      "loss": 0.5982,
      "num_input_tokens_seen": 538624,
      "step": 825
    },
    {
      "epoch": 0.4350104821802935,
      "grad_norm": 18.5,
      "learning_rate": 2.1724318658280922e-05,
      "loss": 0.5925,
      "num_input_tokens_seen": 542112,
      "step": 830
    },
    {
      "epoch": 0.43763102725366876,
      "grad_norm": 7.5,
      "learning_rate": 2.1855345911949688e-05,
      "loss": 0.6853,
      "num_input_tokens_seen": 545408,
      "step": 835
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 5.71875,
      "learning_rate": 2.1986373165618448e-05,
      "loss": 0.5473,
      "num_input_tokens_seen": 548416,
      "step": 840
    },
    {
      "epoch": 0.44287211740041926,
      "grad_norm": 15.875,
      "learning_rate": 2.2117400419287214e-05,
      "loss": 0.59,
      "num_input_tokens_seen": 551264,
      "step": 845
    },
    {
      "epoch": 0.44549266247379454,
      "grad_norm": 5.4375,
      "learning_rate": 2.2248427672955977e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 554592,
      "step": 850
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 11.75,
      "learning_rate": 2.237945492662474e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 558304,
      "step": 855
    },
    {
      "epoch": 0.45073375262054505,
      "grad_norm": 5.3125,
      "learning_rate": 2.2510482180293502e-05,
      "loss": 0.5963,
      "num_input_tokens_seen": 561248,
      "step": 860
    },
    {
      "epoch": 0.4533542976939203,
      "grad_norm": 24.625,
      "learning_rate": 2.2641509433962265e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 564352,
      "step": 865
    },
    {
      "epoch": 0.4559748427672956,
      "grad_norm": 27.5,
      "learning_rate": 2.2772536687631028e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 567712,
      "step": 870
    },
    {
      "epoch": 0.4585953878406709,
      "grad_norm": 11.0625,
      "learning_rate": 2.290356394129979e-05,
      "loss": 0.6098,
      "num_input_tokens_seen": 570240,
      "step": 875
    },
    {
      "epoch": 0.4612159329140461,
      "grad_norm": 24.25,
      "learning_rate": 2.3034591194968556e-05,
      "loss": 0.6404,
      "num_input_tokens_seen": 572768,
      "step": 880
    },
    {
      "epoch": 0.4638364779874214,
      "grad_norm": 12.0,
      "learning_rate": 2.316561844863732e-05,
      "loss": 0.5991,
      "num_input_tokens_seen": 576512,
      "step": 885
    },
    {
      "epoch": 0.46645702306079667,
      "grad_norm": 18.375,
      "learning_rate": 2.329664570230608e-05,
      "loss": 0.6079,
      "num_input_tokens_seen": 579264,
      "step": 890
    },
    {
      "epoch": 0.4690775681341719,
      "grad_norm": 7.875,
      "learning_rate": 2.3427672955974845e-05,
      "loss": 0.5918,
      "num_input_tokens_seen": 581440,
      "step": 895
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 10.75,
      "learning_rate": 2.3558700209643607e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 584544,
      "step": 900
    },
    {
      "epoch": 0.47431865828092246,
      "grad_norm": 24.875,
      "learning_rate": 2.368972746331237e-05,
      "loss": 0.6111,
      "num_input_tokens_seen": 588160,
      "step": 905
    },
    {
      "epoch": 0.4769392033542977,
      "grad_norm": 16.0,
      "learning_rate": 2.3820754716981133e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 590400,
      "step": 910
    },
    {
      "epoch": 0.47955974842767296,
      "grad_norm": 7.1875,
      "learning_rate": 2.39517819706499e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 593440,
      "step": 915
    },
    {
      "epoch": 0.48218029350104824,
      "grad_norm": 12.5,
      "learning_rate": 2.408280922431866e-05,
      "loss": 0.5453,
      "num_input_tokens_seen": 596480,
      "step": 920
    },
    {
      "epoch": 0.48480083857442346,
      "grad_norm": 8.6875,
      "learning_rate": 2.421383647798742e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 599392,
      "step": 925
    },
    {
      "epoch": 0.48742138364779874,
      "grad_norm": 14.9375,
      "learning_rate": 2.4344863731656187e-05,
      "loss": 0.448,
      "num_input_tokens_seen": 602432,
      "step": 930
    },
    {
      "epoch": 0.490041928721174,
      "grad_norm": 19.875,
      "learning_rate": 2.4475890985324947e-05,
      "loss": 0.5309,
      "num_input_tokens_seen": 604512,
      "step": 935
    },
    {
      "epoch": 0.49266247379454925,
      "grad_norm": 3.734375,
      "learning_rate": 2.4606918238993713e-05,
      "loss": 0.4966,
      "num_input_tokens_seen": 608864,
      "step": 940
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 15.3125,
      "learning_rate": 2.4737945492662476e-05,
      "loss": 0.5553,
      "num_input_tokens_seen": 612448,
      "step": 945
    },
    {
      "epoch": 0.4979035639412998,
      "grad_norm": 76.5,
      "learning_rate": 2.486897274633124e-05,
      "loss": 0.5525,
      "num_input_tokens_seen": 614944,
      "step": 950
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.5660056471824646,
      "eval_runtime": 24.1742,
      "eval_samples_per_second": 35.079,
      "eval_steps_per_second": 8.77,
      "num_input_tokens_seen": 616992,
      "step": 954
    },
    {
      "epoch": 0.500524109014675,
      "grad_norm": 18.25,
      "learning_rate": 2.5e-05,
      "loss": 0.5437,
      "num_input_tokens_seen": 617472,
      "step": 955
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 19.25,
      "learning_rate": 2.5131027253668764e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 620192,
      "step": 960
    },
    {
      "epoch": 0.5057651991614256,
      "grad_norm": 11.5,
      "learning_rate": 2.526205450733753e-05,
      "loss": 0.5962,
      "num_input_tokens_seen": 623232,
      "step": 965
    },
    {
      "epoch": 0.5083857442348009,
      "grad_norm": 10.0,
      "learning_rate": 2.5393081761006293e-05,
      "loss": 0.7233,
      "num_input_tokens_seen": 626016,
      "step": 970
    },
    {
      "epoch": 0.5110062893081762,
      "grad_norm": 14.8125,
      "learning_rate": 2.5524109014675052e-05,
      "loss": 0.5393,
      "num_input_tokens_seen": 629472,
      "step": 975
    },
    {
      "epoch": 0.5136268343815513,
      "grad_norm": 11.8125,
      "learning_rate": 2.5655136268343815e-05,
      "loss": 0.6123,
      "num_input_tokens_seen": 632608,
      "step": 980
    },
    {
      "epoch": 0.5162473794549266,
      "grad_norm": 20.75,
      "learning_rate": 2.578616352201258e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 636192,
      "step": 985
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 4.25,
      "learning_rate": 2.5917190775681344e-05,
      "loss": 0.5224,
      "num_input_tokens_seen": 640544,
      "step": 990
    },
    {
      "epoch": 0.5214884696016772,
      "grad_norm": 12.1875,
      "learning_rate": 2.6048218029350107e-05,
      "loss": 0.5264,
      "num_input_tokens_seen": 644448,
      "step": 995
    },
    {
      "epoch": 0.5241090146750524,
      "grad_norm": 11.9375,
      "learning_rate": 2.6179245283018873e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 648288,
      "step": 1000
    },
    {
      "epoch": 0.5267295597484277,
      "grad_norm": 14.1875,
      "learning_rate": 2.631027253668763e-05,
      "loss": 0.7173,
      "num_input_tokens_seen": 651296,
      "step": 1005
    },
    {
      "epoch": 0.5293501048218029,
      "grad_norm": 6.46875,
      "learning_rate": 2.6441299790356395e-05,
      "loss": 0.4911,
      "num_input_tokens_seen": 654176,
      "step": 1010
    },
    {
      "epoch": 0.5319706498951782,
      "grad_norm": 7.15625,
      "learning_rate": 2.6572327044025158e-05,
      "loss": 0.5728,
      "num_input_tokens_seen": 657664,
      "step": 1015
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 22.75,
      "learning_rate": 2.6703354297693924e-05,
      "loss": 0.637,
      "num_input_tokens_seen": 660064,
      "step": 1020
    },
    {
      "epoch": 0.5372117400419287,
      "grad_norm": 15.75,
      "learning_rate": 2.6834381551362687e-05,
      "loss": 0.6535,
      "num_input_tokens_seen": 662880,
      "step": 1025
    },
    {
      "epoch": 0.539832285115304,
      "grad_norm": 6.125,
      "learning_rate": 2.696540880503145e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 665920,
      "step": 1030
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 15.6875,
      "learning_rate": 2.709643605870021e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 668864,
      "step": 1035
    },
    {
      "epoch": 0.5450733752620545,
      "grad_norm": 15.9375,
      "learning_rate": 2.722746331236897e-05,
      "loss": 0.4906,
      "num_input_tokens_seen": 671776,
      "step": 1040
    },
    {
      "epoch": 0.5476939203354297,
      "grad_norm": 7.3125,
      "learning_rate": 2.7358490566037738e-05,
      "loss": 0.5935,
      "num_input_tokens_seen": 674912,
      "step": 1045
    },
    {
      "epoch": 0.550314465408805,
      "grad_norm": 11.8125,
      "learning_rate": 2.74895178197065e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 678368,
      "step": 1050
    },
    {
      "epoch": 0.5529350104821803,
      "grad_norm": 3.21875,
      "learning_rate": 2.7620545073375263e-05,
      "loss": 0.6151,
      "num_input_tokens_seen": 682336,
      "step": 1055
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 25.875,
      "learning_rate": 2.775157232704403e-05,
      "loss": 0.5426,
      "num_input_tokens_seen": 685760,
      "step": 1060
    },
    {
      "epoch": 0.5581761006289309,
      "grad_norm": 16.25,
      "learning_rate": 2.788259958071279e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 688480,
      "step": 1065
    },
    {
      "epoch": 0.560796645702306,
      "grad_norm": 5.84375,
      "learning_rate": 2.801362683438155e-05,
      "loss": 0.6357,
      "num_input_tokens_seen": 692064,
      "step": 1070
    },
    {
      "epoch": 0.5634171907756813,
      "grad_norm": 12.4375,
      "learning_rate": 2.8144654088050314e-05,
      "loss": 0.6344,
      "num_input_tokens_seen": 694464,
      "step": 1075
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 5.96875,
      "learning_rate": 2.827568134171908e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 697344,
      "step": 1080
    },
    {
      "epoch": 0.5686582809224319,
      "grad_norm": 29.125,
      "learning_rate": 2.8406708595387843e-05,
      "loss": 0.5473,
      "num_input_tokens_seen": 700256,
      "step": 1085
    },
    {
      "epoch": 0.5712788259958071,
      "grad_norm": 5.28125,
      "learning_rate": 2.8537735849056606e-05,
      "loss": 0.5463,
      "num_input_tokens_seen": 703968,
      "step": 1090
    },
    {
      "epoch": 0.5738993710691824,
      "grad_norm": 4.90625,
      "learning_rate": 2.8668763102725365e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 708160,
      "step": 1095
    },
    {
      "epoch": 0.5765199161425576,
      "grad_norm": 8.6875,
      "learning_rate": 2.8799790356394128e-05,
      "loss": 0.7388,
      "num_input_tokens_seen": 710976,
      "step": 1100
    },
    {
      "epoch": 0.5791404612159329,
      "grad_norm": 77.5,
      "learning_rate": 2.8930817610062894e-05,
      "loss": 0.6838,
      "num_input_tokens_seen": 713696,
      "step": 1105
    },
    {
      "epoch": 0.5817610062893082,
      "grad_norm": 12.8125,
      "learning_rate": 2.9061844863731657e-05,
      "loss": 0.5617,
      "num_input_tokens_seen": 716608,
      "step": 1110
    },
    {
      "epoch": 0.5843815513626834,
      "grad_norm": 12.0625,
      "learning_rate": 2.9192872117400423e-05,
      "loss": 0.5193,
      "num_input_tokens_seen": 719488,
      "step": 1115
    },
    {
      "epoch": 0.5870020964360587,
      "grad_norm": 9.4375,
      "learning_rate": 2.9323899371069186e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 722304,
      "step": 1120
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 15.75,
      "learning_rate": 2.945492662473795e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 725472,
      "step": 1125
    },
    {
      "epoch": 0.5922431865828093,
      "grad_norm": 10.0625,
      "learning_rate": 2.9585953878406708e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 728896,
      "step": 1130
    },
    {
      "epoch": 0.5948637316561844,
      "grad_norm": 8.0625,
      "learning_rate": 2.971698113207547e-05,
      "loss": 0.4737,
      "num_input_tokens_seen": 732000,
      "step": 1135
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 11.5,
      "learning_rate": 2.9848008385744237e-05,
      "loss": 0.5903,
      "num_input_tokens_seen": 735328,
      "step": 1140
    },
    {
      "epoch": 0.600104821802935,
      "grad_norm": 12.0,
      "learning_rate": 2.9979035639413e-05,
      "loss": 0.4405,
      "num_input_tokens_seen": 738528,
      "step": 1145
    },
    {
      "epoch": 0.6027253668763103,
      "grad_norm": 10.125,
      "learning_rate": 3.0110062893081766e-05,
      "loss": 0.6279,
      "num_input_tokens_seen": 741376,
      "step": 1150
    },
    {
      "epoch": 0.6053459119496856,
      "grad_norm": 15.125,
      "learning_rate": 3.024109014675053e-05,
      "loss": 0.5907,
      "num_input_tokens_seen": 743872,
      "step": 1155
    },
    {
      "epoch": 0.6079664570230608,
      "grad_norm": 6.0,
      "learning_rate": 3.0372117400419288e-05,
      "loss": 0.5634,
      "num_input_tokens_seen": 746720,
      "step": 1160
    },
    {
      "epoch": 0.610587002096436,
      "grad_norm": 11.125,
      "learning_rate": 3.050314465408805e-05,
      "loss": 0.6031,
      "num_input_tokens_seen": 750240,
      "step": 1165
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 18.375,
      "learning_rate": 3.063417190775681e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 753344,
      "step": 1170
    },
    {
      "epoch": 0.6158280922431866,
      "grad_norm": 26.625,
      "learning_rate": 3.076519916142558e-05,
      "loss": 0.5501,
      "num_input_tokens_seen": 755968,
      "step": 1175
    },
    {
      "epoch": 0.6184486373165619,
      "grad_norm": 3.65625,
      "learning_rate": 3.0896226415094346e-05,
      "loss": 0.6121,
      "num_input_tokens_seen": 759520,
      "step": 1180
    },
    {
      "epoch": 0.6210691823899371,
      "grad_norm": 7.75,
      "learning_rate": 3.1027253668763105e-05,
      "loss": 0.5043,
      "num_input_tokens_seen": 762048,
      "step": 1185
    },
    {
      "epoch": 0.6236897274633124,
      "grad_norm": 13.1875,
      "learning_rate": 3.1158280922431864e-05,
      "loss": 0.644,
      "num_input_tokens_seen": 765504,
      "step": 1190
    },
    {
      "epoch": 0.6263102725366876,
      "grad_norm": 11.625,
      "learning_rate": 3.128930817610063e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 768192,
      "step": 1195
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 10.5,
      "learning_rate": 3.142033542976939e-05,
      "loss": 0.5812,
      "num_input_tokens_seen": 771168,
      "step": 1200
    },
    {
      "epoch": 0.6315513626834381,
      "grad_norm": 4.59375,
      "learning_rate": 3.1551362683438156e-05,
      "loss": 0.3301,
      "num_input_tokens_seen": 777504,
      "step": 1205
    },
    {
      "epoch": 0.6341719077568134,
      "grad_norm": 11.9375,
      "learning_rate": 3.168238993710692e-05,
      "loss": 0.5176,
      "num_input_tokens_seen": 780512,
      "step": 1210
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 15.1875,
      "learning_rate": 3.181341719077569e-05,
      "loss": 0.5819,
      "num_input_tokens_seen": 784192,
      "step": 1215
    },
    {
      "epoch": 0.639412997903564,
      "grad_norm": 11.5,
      "learning_rate": 3.194444444444444e-05,
      "loss": 0.6246,
      "num_input_tokens_seen": 787552,
      "step": 1220
    },
    {
      "epoch": 0.6420335429769392,
      "grad_norm": 5.59375,
      "learning_rate": 3.207547169811321e-05,
      "loss": 0.703,
      "num_input_tokens_seen": 790720,
      "step": 1225
    },
    {
      "epoch": 0.6446540880503144,
      "grad_norm": 10.875,
      "learning_rate": 3.220649895178197e-05,
      "loss": 0.6178,
      "num_input_tokens_seen": 793696,
      "step": 1230
    },
    {
      "epoch": 0.6472746331236897,
      "grad_norm": 30.0,
      "learning_rate": 3.233752620545073e-05,
      "loss": 0.6247,
      "num_input_tokens_seen": 796640,
      "step": 1235
    },
    {
      "epoch": 0.649895178197065,
      "grad_norm": 12.6875,
      "learning_rate": 3.24685534591195e-05,
      "loss": 0.5656,
      "num_input_tokens_seen": 799648,
      "step": 1240
    },
    {
      "epoch": 0.6525157232704403,
      "grad_norm": 10.375,
      "learning_rate": 3.2599580712788265e-05,
      "loss": 0.3974,
      "num_input_tokens_seen": 802496,
      "step": 1245
    },
    {
      "epoch": 0.6551362683438156,
      "grad_norm": 11.3125,
      "learning_rate": 3.2730607966457024e-05,
      "loss": 0.5988,
      "num_input_tokens_seen": 805760,
      "step": 1250
    },
    {
      "epoch": 0.6577568134171907,
      "grad_norm": 13.25,
      "learning_rate": 3.2861635220125784e-05,
      "loss": 0.6101,
      "num_input_tokens_seen": 808832,
      "step": 1255
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 15.375,
      "learning_rate": 3.299266247379455e-05,
      "loss": 0.5093,
      "num_input_tokens_seen": 812256,
      "step": 1260
    },
    {
      "epoch": 0.6629979035639413,
      "grad_norm": 9.9375,
      "learning_rate": 3.3123689727463316e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 815680,
      "step": 1265
    },
    {
      "epoch": 0.6656184486373166,
      "grad_norm": 6.28125,
      "learning_rate": 3.3254716981132075e-05,
      "loss": 0.4114,
      "num_input_tokens_seen": 818400,
      "step": 1270
    },
    {
      "epoch": 0.6682389937106918,
      "grad_norm": 9.9375,
      "learning_rate": 3.338574423480084e-05,
      "loss": 0.5805,
      "num_input_tokens_seen": 821536,
      "step": 1275
    },
    {
      "epoch": 0.6708595387840671,
      "grad_norm": 19.875,
      "learning_rate": 3.351677148846961e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 825024,
      "step": 1280
    },
    {
      "epoch": 0.6734800838574424,
      "grad_norm": 5.5625,
      "learning_rate": 3.364779874213837e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 829248,
      "step": 1285
    },
    {
      "epoch": 0.6761006289308176,
      "grad_norm": 5.78125,
      "learning_rate": 3.3778825995807126e-05,
      "loss": 0.4924,
      "num_input_tokens_seen": 831872,
      "step": 1290
    },
    {
      "epoch": 0.6787211740041929,
      "grad_norm": 11.0625,
      "learning_rate": 3.390985324947589e-05,
      "loss": 0.6688,
      "num_input_tokens_seen": 834880,
      "step": 1295
    },
    {
      "epoch": 0.6813417190775681,
      "grad_norm": 6.9375,
      "learning_rate": 3.404088050314466e-05,
      "loss": 0.6255,
      "num_input_tokens_seen": 837824,
      "step": 1300
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 11.375,
      "learning_rate": 3.417190775681342e-05,
      "loss": 0.5663,
      "num_input_tokens_seen": 840896,
      "step": 1305
    },
    {
      "epoch": 0.6865828092243187,
      "grad_norm": 11.4375,
      "learning_rate": 3.4302935010482184e-05,
      "loss": 0.7189,
      "num_input_tokens_seen": 844000,
      "step": 1310
    },
    {
      "epoch": 0.689203354297694,
      "grad_norm": 4.75,
      "learning_rate": 3.4433962264150943e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 847840,
      "step": 1315
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 5.71875,
      "learning_rate": 3.456498951781971e-05,
      "loss": 0.6558,
      "num_input_tokens_seen": 850976,
      "step": 1320
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 6.21875,
      "learning_rate": 3.469601677148847e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 855616,
      "step": 1325
    },
    {
      "epoch": 0.6970649895178197,
      "grad_norm": 3.78125,
      "learning_rate": 3.4827044025157235e-05,
      "loss": 0.3575,
      "num_input_tokens_seen": 858528,
      "step": 1330
    },
    {
      "epoch": 0.699685534591195,
      "grad_norm": 5.34375,
      "learning_rate": 3.4958071278826e-05,
      "loss": 0.5461,
      "num_input_tokens_seen": 862144,
      "step": 1335
    },
    {
      "epoch": 0.7023060796645703,
      "grad_norm": 9.9375,
      "learning_rate": 3.508909853249476e-05,
      "loss": 0.6042,
      "num_input_tokens_seen": 865440,
      "step": 1340
    },
    {
      "epoch": 0.7049266247379455,
      "grad_norm": 8.25,
      "learning_rate": 3.522012578616352e-05,
      "loss": 0.5962,
      "num_input_tokens_seen": 868992,
      "step": 1345
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 14.125,
      "learning_rate": 3.5351153039832286e-05,
      "loss": 0.5537,
      "num_input_tokens_seen": 871968,
      "step": 1350
    },
    {
      "epoch": 0.710167714884696,
      "grad_norm": 10.9375,
      "learning_rate": 3.548218029350105e-05,
      "loss": 0.532,
      "num_input_tokens_seen": 874912,
      "step": 1355
    },
    {
      "epoch": 0.7127882599580713,
      "grad_norm": 10.9375,
      "learning_rate": 3.561320754716981e-05,
      "loss": 0.5763,
      "num_input_tokens_seen": 878112,
      "step": 1360
    },
    {
      "epoch": 0.7154088050314465,
      "grad_norm": 29.75,
      "learning_rate": 3.574423480083858e-05,
      "loss": 0.5618,
      "num_input_tokens_seen": 880928,
      "step": 1365
    },
    {
      "epoch": 0.7180293501048218,
      "grad_norm": 8.5,
      "learning_rate": 3.5875262054507344e-05,
      "loss": 0.5496,
      "num_input_tokens_seen": 884832,
      "step": 1370
    },
    {
      "epoch": 0.7206498951781971,
      "grad_norm": 4.53125,
      "learning_rate": 3.6006289308176097e-05,
      "loss": 0.5223,
      "num_input_tokens_seen": 888800,
      "step": 1375
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 6.3125,
      "learning_rate": 3.613731656184486e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 892416,
      "step": 1380
    },
    {
      "epoch": 0.7258909853249476,
      "grad_norm": 13.6875,
      "learning_rate": 3.626834381551363e-05,
      "loss": 0.637,
      "num_input_tokens_seen": 895840,
      "step": 1385
    },
    {
      "epoch": 0.7285115303983228,
      "grad_norm": 10.0,
      "learning_rate": 3.6399371069182395e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 899392,
      "step": 1390
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 5.71875,
      "learning_rate": 3.6530398322851154e-05,
      "loss": 0.4963,
      "num_input_tokens_seen": 903264,
      "step": 1395
    },
    {
      "epoch": 0.7337526205450734,
      "grad_norm": 8.5,
      "learning_rate": 3.666142557651992e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 906016,
      "step": 1400
    },
    {
      "epoch": 0.7363731656184487,
      "grad_norm": 9.875,
      "learning_rate": 3.679245283018868e-05,
      "loss": 0.6107,
      "num_input_tokens_seen": 909152,
      "step": 1405
    },
    {
      "epoch": 0.7389937106918238,
      "grad_norm": 13.625,
      "learning_rate": 3.692348008385744e-05,
      "loss": 0.6228,
      "num_input_tokens_seen": 912416,
      "step": 1410
    },
    {
      "epoch": 0.7416142557651991,
      "grad_norm": 16.25,
      "learning_rate": 3.7054507337526205e-05,
      "loss": 0.6966,
      "num_input_tokens_seen": 915616,
      "step": 1415
    },
    {
      "epoch": 0.7442348008385744,
      "grad_norm": 23.125,
      "learning_rate": 3.718553459119497e-05,
      "loss": 0.571,
      "num_input_tokens_seen": 921568,
      "step": 1420
    },
    {
      "epoch": 0.7468553459119497,
      "grad_norm": 9.875,
      "learning_rate": 3.731656184486374e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 924864,
      "step": 1425
    },
    {
      "epoch": 0.749475890985325,
      "grad_norm": 12.5625,
      "learning_rate": 3.74475890985325e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 927968,
      "step": 1430
    },
    {
      "epoch": 0.7520964360587002,
      "grad_norm": 6.65625,
      "learning_rate": 3.757861635220126e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 931584,
      "step": 1435
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 13.25,
      "learning_rate": 3.770964360587002e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 934816,
      "step": 1440
    },
    {
      "epoch": 0.7573375262054507,
      "grad_norm": 14.0,
      "learning_rate": 3.784067085953878e-05,
      "loss": 0.6036,
      "num_input_tokens_seen": 937856,
      "step": 1445
    },
    {
      "epoch": 0.759958071278826,
      "grad_norm": 10.0,
      "learning_rate": 3.797169811320755e-05,
      "loss": 0.5795,
      "num_input_tokens_seen": 945632,
      "step": 1450
    },
    {
      "epoch": 0.7625786163522013,
      "grad_norm": 9.5,
      "learning_rate": 3.8102725366876314e-05,
      "loss": 0.6385,
      "num_input_tokens_seen": 948608,
      "step": 1455
    },
    {
      "epoch": 0.7651991614255765,
      "grad_norm": 13.5625,
      "learning_rate": 3.8233752620545074e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 951488,
      "step": 1460
    },
    {
      "epoch": 0.7678197064989518,
      "grad_norm": 7.53125,
      "learning_rate": 3.836477987421384e-05,
      "loss": 0.5928,
      "num_input_tokens_seen": 955072,
      "step": 1465
    },
    {
      "epoch": 0.7704402515723271,
      "grad_norm": 9.0625,
      "learning_rate": 3.84958071278826e-05,
      "loss": 0.6225,
      "num_input_tokens_seen": 958272,
      "step": 1470
    },
    {
      "epoch": 0.7730607966457023,
      "grad_norm": 3.640625,
      "learning_rate": 3.8626834381551365e-05,
      "loss": 0.7074,
      "num_input_tokens_seen": 961888,
      "step": 1475
    },
    {
      "epoch": 0.7756813417190775,
      "grad_norm": 27.75,
      "learning_rate": 3.8757861635220125e-05,
      "loss": 0.7886,
      "num_input_tokens_seen": 964256,
      "step": 1480
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 8.25,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.566,
      "num_input_tokens_seen": 968224,
      "step": 1485
    },
    {
      "epoch": 0.7809224318658281,
      "grad_norm": 7.625,
      "learning_rate": 3.901991614255766e-05,
      "loss": 0.5215,
      "num_input_tokens_seen": 971136,
      "step": 1490
    },
    {
      "epoch": 0.7835429769392034,
      "grad_norm": 9.3125,
      "learning_rate": 3.9150943396226416e-05,
      "loss": 0.5539,
      "num_input_tokens_seen": 975936,
      "step": 1495
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 9.625,
      "learning_rate": 3.9281970649895176e-05,
      "loss": 0.6043,
      "num_input_tokens_seen": 979360,
      "step": 1500
    },
    {
      "epoch": 0.7887840670859538,
      "grad_norm": 4.34375,
      "learning_rate": 3.941299790356394e-05,
      "loss": 0.4961,
      "num_input_tokens_seen": 983616,
      "step": 1505
    },
    {
      "epoch": 0.7914046121593291,
      "grad_norm": 10.0625,
      "learning_rate": 3.954402515723271e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 986944,
      "step": 1510
    },
    {
      "epoch": 0.7940251572327044,
      "grad_norm": 7.71875,
      "learning_rate": 3.967505241090147e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 989888,
      "step": 1515
    },
    {
      "epoch": 0.7966457023060797,
      "grad_norm": 12.5,
      "learning_rate": 3.9806079664570233e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 992768,
      "step": 1520
    },
    {
      "epoch": 0.799266247379455,
      "grad_norm": 8.125,
      "learning_rate": 3.9937106918239e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 996896,
      "step": 1525
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 9.0625,
      "learning_rate": 4.006813417190776e-05,
      "loss": 0.5918,
      "num_input_tokens_seen": 999840,
      "step": 1530
    },
    {
      "epoch": 0.8045073375262054,
      "grad_norm": 6.28125,
      "learning_rate": 4.019916142557652e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 1003680,
      "step": 1535
    },
    {
      "epoch": 0.8071278825995807,
      "grad_norm": 10.5625,
      "learning_rate": 4.0330188679245284e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 1006656,
      "step": 1540
    },
    {
      "epoch": 0.809748427672956,
      "grad_norm": 18.375,
      "learning_rate": 4.046121593291405e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 1008768,
      "step": 1545
    },
    {
      "epoch": 0.8123689727463312,
      "grad_norm": 6.4375,
      "learning_rate": 4.059224318658281e-05,
      "loss": 0.5384,
      "num_input_tokens_seen": 1011968,
      "step": 1550
    },
    {
      "epoch": 0.8149895178197065,
      "grad_norm": 5.625,
      "learning_rate": 4.0723270440251576e-05,
      "loss": 0.5764,
      "num_input_tokens_seen": 1018784,
      "step": 1555
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 6.125,
      "learning_rate": 4.0854297693920336e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 1023520,
      "step": 1560
    },
    {
      "epoch": 0.820230607966457,
      "grad_norm": 5.3125,
      "learning_rate": 4.09853249475891e-05,
      "loss": 0.5191,
      "num_input_tokens_seen": 1027200,
      "step": 1565
    },
    {
      "epoch": 0.8228511530398323,
      "grad_norm": 3.8125,
      "learning_rate": 4.111635220125786e-05,
      "loss": 0.8592,
      "num_input_tokens_seen": 1030848,
      "step": 1570
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 8.75,
      "learning_rate": 4.124737945492663e-05,
      "loss": 0.5886,
      "num_input_tokens_seen": 1035232,
      "step": 1575
    },
    {
      "epoch": 0.8280922431865828,
      "grad_norm": 12.625,
      "learning_rate": 4.137840670859539e-05,
      "loss": 0.5328,
      "num_input_tokens_seen": 1038176,
      "step": 1580
    },
    {
      "epoch": 0.8307127882599581,
      "grad_norm": 14.6875,
      "learning_rate": 4.150943396226415e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 1041216,
      "step": 1585
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 8.875,
      "learning_rate": 4.164046121593291e-05,
      "loss": 0.562,
      "num_input_tokens_seen": 1043840,
      "step": 1590
    },
    {
      "epoch": 0.8359538784067087,
      "grad_norm": 5.0625,
      "learning_rate": 4.177148846960168e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 1046656,
      "step": 1595
    },
    {
      "epoch": 0.8385744234800838,
      "grad_norm": 5.0,
      "learning_rate": 4.1902515723270444e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 1051616,
      "step": 1600
    },
    {
      "epoch": 0.8411949685534591,
      "grad_norm": 10.9375,
      "learning_rate": 4.2033542976939204e-05,
      "loss": 0.6511,
      "num_input_tokens_seen": 1053824,
      "step": 1605
    },
    {
      "epoch": 0.8438155136268344,
      "grad_norm": 7.75,
      "learning_rate": 4.216457023060797e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 1056512,
      "step": 1610
    },
    {
      "epoch": 0.8464360587002097,
      "grad_norm": 5.90625,
      "learning_rate": 4.2295597484276736e-05,
      "loss": 0.5484,
      "num_input_tokens_seen": 1060160,
      "step": 1615
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 10.4375,
      "learning_rate": 4.2426624737945495e-05,
      "loss": 0.6811,
      "num_input_tokens_seen": 1062880,
      "step": 1620
    },
    {
      "epoch": 0.8516771488469602,
      "grad_norm": 12.3125,
      "learning_rate": 4.2557651991614255e-05,
      "loss": 0.6266,
      "num_input_tokens_seen": 1066528,
      "step": 1625
    },
    {
      "epoch": 0.8542976939203354,
      "grad_norm": 8.125,
      "learning_rate": 4.268867924528302e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 1069472,
      "step": 1630
    },
    {
      "epoch": 0.8569182389937107,
      "grad_norm": 16.125,
      "learning_rate": 4.281970649895179e-05,
      "loss": 0.5581,
      "num_input_tokens_seen": 1074304,
      "step": 1635
    },
    {
      "epoch": 0.859538784067086,
      "grad_norm": 13.75,
      "learning_rate": 4.2950733752620546e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 1077376,
      "step": 1640
    },
    {
      "epoch": 0.8621593291404612,
      "grad_norm": 5.78125,
      "learning_rate": 4.308176100628931e-05,
      "loss": 0.5395,
      "num_input_tokens_seen": 1080672,
      "step": 1645
    },
    {
      "epoch": 0.8647798742138365,
      "grad_norm": 7.125,
      "learning_rate": 4.321278825995808e-05,
      "loss": 0.5723,
      "num_input_tokens_seen": 1084224,
      "step": 1650
    },
    {
      "epoch": 0.8674004192872118,
      "grad_norm": 12.8125,
      "learning_rate": 4.334381551362683e-05,
      "loss": 0.7138,
      "num_input_tokens_seen": 1086912,
      "step": 1655
    },
    {
      "epoch": 0.870020964360587,
      "grad_norm": 11.75,
      "learning_rate": 4.34748427672956e-05,
      "loss": 0.6483,
      "num_input_tokens_seen": 1089696,
      "step": 1660
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 11.875,
      "learning_rate": 4.3605870020964364e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 1092608,
      "step": 1665
    },
    {
      "epoch": 0.8752620545073375,
      "grad_norm": 9.125,
      "learning_rate": 4.373689727463312e-05,
      "loss": 0.704,
      "num_input_tokens_seen": 1095808,
      "step": 1670
    },
    {
      "epoch": 0.8778825995807128,
      "grad_norm": 6.78125,
      "learning_rate": 4.386792452830189e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 1098816,
      "step": 1675
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 4.5625,
      "learning_rate": 4.3998951781970655e-05,
      "loss": 0.6164,
      "num_input_tokens_seen": 1103136,
      "step": 1680
    },
    {
      "epoch": 0.8831236897274634,
      "grad_norm": 3.984375,
      "learning_rate": 4.4129979035639415e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 1107872,
      "step": 1685
    },
    {
      "epoch": 0.8857442348008385,
      "grad_norm": 5.90625,
      "learning_rate": 4.4261006289308174e-05,
      "loss": 0.6635,
      "num_input_tokens_seen": 1111072,
      "step": 1690
    },
    {
      "epoch": 0.8883647798742138,
      "grad_norm": 18.0,
      "learning_rate": 4.439203354297694e-05,
      "loss": 0.5985,
      "num_input_tokens_seen": 1113792,
      "step": 1695
    },
    {
      "epoch": 0.8909853249475891,
      "grad_norm": 16.125,
      "learning_rate": 4.4523060796645706e-05,
      "loss": 0.5043,
      "num_input_tokens_seen": 1116384,
      "step": 1700
    },
    {
      "epoch": 0.8936058700209644,
      "grad_norm": 7.34375,
      "learning_rate": 4.4654088050314466e-05,
      "loss": 0.5092,
      "num_input_tokens_seen": 1119296,
      "step": 1705
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 5.03125,
      "learning_rate": 4.478511530398323e-05,
      "loss": 0.6479,
      "num_input_tokens_seen": 1122720,
      "step": 1710
    },
    {
      "epoch": 0.8988469601677149,
      "grad_norm": 33.5,
      "learning_rate": 4.491614255765199e-05,
      "loss": 0.6388,
      "num_input_tokens_seen": 1125088,
      "step": 1715
    },
    {
      "epoch": 0.9014675052410901,
      "grad_norm": 9.125,
      "learning_rate": 4.504716981132076e-05,
      "loss": 0.6734,
      "num_input_tokens_seen": 1128288,
      "step": 1720
    },
    {
      "epoch": 0.9040880503144654,
      "grad_norm": 21.125,
      "learning_rate": 4.517819706498952e-05,
      "loss": 0.6219,
      "num_input_tokens_seen": 1131040,
      "step": 1725
    },
    {
      "epoch": 0.9067085953878407,
      "grad_norm": 4.9375,
      "learning_rate": 4.530922431865828e-05,
      "loss": 0.5586,
      "num_input_tokens_seen": 1135136,
      "step": 1730
    },
    {
      "epoch": 0.9093291404612159,
      "grad_norm": 6.0625,
      "learning_rate": 4.544025157232705e-05,
      "loss": 0.6771,
      "num_input_tokens_seen": 1139008,
      "step": 1735
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 5.625,
      "learning_rate": 4.557127882599581e-05,
      "loss": 0.5191,
      "num_input_tokens_seen": 1141920,
      "step": 1740
    },
    {
      "epoch": 0.9145702306079665,
      "grad_norm": 7.65625,
      "learning_rate": 4.570230607966457e-05,
      "loss": 0.5269,
      "num_input_tokens_seen": 1144960,
      "step": 1745
    },
    {
      "epoch": 0.9171907756813418,
      "grad_norm": 6.0625,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.596,
      "num_input_tokens_seen": 1147712,
      "step": 1750
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 5.75,
      "learning_rate": 4.59643605870021e-05,
      "loss": 0.5903,
      "num_input_tokens_seen": 1150528,
      "step": 1755
    },
    {
      "epoch": 0.9224318658280922,
      "grad_norm": 27.5,
      "learning_rate": 4.609538784067086e-05,
      "loss": 0.6195,
      "num_input_tokens_seen": 1154272,
      "step": 1760
    },
    {
      "epoch": 0.9250524109014675,
      "grad_norm": 13.75,
      "learning_rate": 4.6226415094339625e-05,
      "loss": 0.4946,
      "num_input_tokens_seen": 1157088,
      "step": 1765
    },
    {
      "epoch": 0.9276729559748428,
      "grad_norm": 8.9375,
      "learning_rate": 4.635744234800839e-05,
      "loss": 0.6689,
      "num_input_tokens_seen": 1160192,
      "step": 1770
    },
    {
      "epoch": 0.9302935010482181,
      "grad_norm": 7.125,
      "learning_rate": 4.648846960167715e-05,
      "loss": 0.5772,
      "num_input_tokens_seen": 1163424,
      "step": 1775
    },
    {
      "epoch": 0.9329140461215933,
      "grad_norm": 5.46875,
      "learning_rate": 4.661949685534591e-05,
      "loss": 0.6197,
      "num_input_tokens_seen": 1167104,
      "step": 1780
    },
    {
      "epoch": 0.9355345911949685,
      "grad_norm": 12.625,
      "learning_rate": 4.6750524109014677e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 1170368,
      "step": 1785
    },
    {
      "epoch": 0.9381551362683438,
      "grad_norm": 5.90625,
      "learning_rate": 4.688155136268344e-05,
      "loss": 0.6611,
      "num_input_tokens_seen": 1173888,
      "step": 1790
    },
    {
      "epoch": 0.9407756813417191,
      "grad_norm": 4.84375,
      "learning_rate": 4.70125786163522e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 1177408,
      "step": 1795
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 11.625,
      "learning_rate": 4.714360587002097e-05,
      "loss": 0.529,
      "num_input_tokens_seen": 1181056,
      "step": 1800
    },
    {
      "epoch": 0.9460167714884696,
      "grad_norm": 6.375,
      "learning_rate": 4.7274633123689734e-05,
      "loss": 0.7111,
      "num_input_tokens_seen": 1184256,
      "step": 1805
    },
    {
      "epoch": 0.9486373165618449,
      "grad_norm": 5.875,
      "learning_rate": 4.7405660377358494e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 1187168,
      "step": 1810
    },
    {
      "epoch": 0.9512578616352201,
      "grad_norm": 5.78125,
      "learning_rate": 4.753668763102725e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 1190784,
      "step": 1815
    },
    {
      "epoch": 0.9538784067085954,
      "grad_norm": 5.28125,
      "learning_rate": 4.766771488469602e-05,
      "loss": 0.626,
      "num_input_tokens_seen": 1193312,
      "step": 1820
    },
    {
      "epoch": 0.9564989517819706,
      "grad_norm": 4.5,
      "learning_rate": 4.7798742138364785e-05,
      "loss": 0.5499,
      "num_input_tokens_seen": 1196608,
      "step": 1825
    },
    {
      "epoch": 0.9591194968553459,
      "grad_norm": 15.6875,
      "learning_rate": 4.7929769392033545e-05,
      "loss": 0.6033,
      "num_input_tokens_seen": 1200160,
      "step": 1830
    },
    {
      "epoch": 0.9617400419287212,
      "grad_norm": 25.125,
      "learning_rate": 4.806079664570231e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 1203040,
      "step": 1835
    },
    {
      "epoch": 0.9643605870020965,
      "grad_norm": 5.6875,
      "learning_rate": 4.819182389937107e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 1205408,
      "step": 1840
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 24.75,
      "learning_rate": 4.8322851153039836e-05,
      "loss": 0.4213,
      "num_input_tokens_seen": 1208896,
      "step": 1845
    },
    {
      "epoch": 0.9696016771488469,
      "grad_norm": 6.875,
      "learning_rate": 4.8453878406708596e-05,
      "loss": 0.5276,
      "num_input_tokens_seen": 1212576,
      "step": 1850
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 9.0,
      "learning_rate": 4.858490566037736e-05,
      "loss": 0.5295,
      "num_input_tokens_seen": 1215872,
      "step": 1855
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 9.375,
      "learning_rate": 4.871593291404613e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 1219968,
      "step": 1860
    },
    {
      "epoch": 0.9774633123689728,
      "grad_norm": 4.03125,
      "learning_rate": 4.884696016771489e-05,
      "loss": 0.5459,
      "num_input_tokens_seen": 1222048,
      "step": 1865
    },
    {
      "epoch": 0.980083857442348,
      "grad_norm": 7.4375,
      "learning_rate": 4.897798742138365e-05,
      "loss": 0.5908,
      "num_input_tokens_seen": 1225120,
      "step": 1870
    },
    {
      "epoch": 0.9827044025157232,
      "grad_norm": 4.40625,
      "learning_rate": 4.910901467505241e-05,
      "loss": 0.5416,
      "num_input_tokens_seen": 1229056,
      "step": 1875
    },
    {
      "epoch": 0.9853249475890985,
      "grad_norm": 8.1875,
      "learning_rate": 4.924004192872117e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 1231776,
      "step": 1880
    },
    {
      "epoch": 0.9879454926624738,
      "grad_norm": 3.75,
      "learning_rate": 4.937106918238994e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 1235040,
      "step": 1885
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 10.0,
      "learning_rate": 4.9502096436058705e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 1238048,
      "step": 1890
    },
    {
      "epoch": 0.9931865828092243,
      "grad_norm": 5.09375,
      "learning_rate": 4.963312368972747e-05,
      "loss": 0.5785,
      "num_input_tokens_seen": 1241344,
      "step": 1895
    },
    {
      "epoch": 0.9958071278825996,
      "grad_norm": 6.9375,
      "learning_rate": 4.976415094339622e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 1243840,
      "step": 1900
    },
    {
      "epoch": 0.9984276729559748,
      "grad_norm": 5.1875,
      "learning_rate": 4.989517819706499e-05,
      "loss": 0.642,
      "num_input_tokens_seen": 1246816,
      "step": 1905
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5417815446853638,
      "eval_runtime": 24.1858,
      "eval_samples_per_second": 35.062,
      "eval_steps_per_second": 8.765,
      "num_input_tokens_seen": 1248304,
      "step": 1908
    },
    {
      "epoch": 1.00104821802935,
      "grad_norm": 8.75,
      "learning_rate": 4.9999999581622816e-05,
      "loss": 0.5297,
      "num_input_tokens_seen": 1249200,
      "step": 1910
    },
    {
      "epoch": 1.0036687631027255,
      "grad_norm": 11.4375,
      "learning_rate": 4.999998493842267e-05,
      "loss": 0.5581,
      "num_input_tokens_seen": 1251664,
      "step": 1915
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 7.375,
      "learning_rate": 4.999994937637709e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 1254928,
      "step": 1920
    },
    {
      "epoch": 1.0089098532494758,
      "grad_norm": 25.125,
      "learning_rate": 4.999989289551581e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 1257744,
      "step": 1925
    },
    {
      "epoch": 1.0115303983228512,
      "grad_norm": 16.625,
      "learning_rate": 4.999981549588612e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 1261904,
      "step": 1930
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 4.71875,
      "learning_rate": 4.9999717177552764e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 1264432,
      "step": 1935
    },
    {
      "epoch": 1.0167714884696017,
      "grad_norm": 9.8125,
      "learning_rate": 4.999959794059801e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 1268400,
      "step": 1940
    },
    {
      "epoch": 1.019392033542977,
      "grad_norm": 4.96875,
      "learning_rate": 4.999945778512164e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 1272752,
      "step": 1945
    },
    {
      "epoch": 1.0220125786163523,
      "grad_norm": 7.59375,
      "learning_rate": 4.999929671124093e-05,
      "loss": 0.6146,
      "num_input_tokens_seen": 1275568,
      "step": 1950
    },
    {
      "epoch": 1.0246331236897275,
      "grad_norm": 4.71875,
      "learning_rate": 4.9999114719090645e-05,
      "loss": 0.5919,
      "num_input_tokens_seen": 1278160,
      "step": 1955
    },
    {
      "epoch": 1.0272536687631026,
      "grad_norm": 8.75,
      "learning_rate": 4.999891180882308e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 1281424,
      "step": 1960
    },
    {
      "epoch": 1.029874213836478,
      "grad_norm": 9.75,
      "learning_rate": 4.9998687980608014e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 1283920,
      "step": 1965
    },
    {
      "epoch": 1.0324947589098532,
      "grad_norm": 15.75,
      "learning_rate": 4.9998443234632744e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 1287920,
      "step": 1970
    },
    {
      "epoch": 1.0351153039832286,
      "grad_norm": 14.5625,
      "learning_rate": 4.999817757110206e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 1290960,
      "step": 1975
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 5.34375,
      "learning_rate": 4.999789099023826e-05,
      "loss": 0.4116,
      "num_input_tokens_seen": 1294544,
      "step": 1980
    },
    {
      "epoch": 1.040356394129979,
      "grad_norm": 4.3125,
      "learning_rate": 4.9997583492281126e-05,
      "loss": 0.6765,
      "num_input_tokens_seen": 1298160,
      "step": 1985
    },
    {
      "epoch": 1.0429769392033543,
      "grad_norm": 4.96875,
      "learning_rate": 4.999725507748798e-05,
      "loss": 0.5551,
      "num_input_tokens_seen": 1300368,
      "step": 1990
    },
    {
      "epoch": 1.0455974842767295,
      "grad_norm": 3.40625,
      "learning_rate": 4.9996905746133606e-05,
      "loss": 0.4548,
      "num_input_tokens_seen": 1307184,
      "step": 1995
    },
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 8.4375,
      "learning_rate": 4.999653549851032e-05,
      "loss": 0.8381,
      "num_input_tokens_seen": 1309840,
      "step": 2000
    },
    {
      "epoch": 1.05083857442348,
      "grad_norm": 10.875,
      "learning_rate": 4.999614433492792e-05,
      "loss": 0.5026,
      "num_input_tokens_seen": 1313648,
      "step": 2005
    },
    {
      "epoch": 1.0534591194968554,
      "grad_norm": 4.96875,
      "learning_rate": 4.9995732255713725e-05,
      "loss": 0.5063,
      "num_input_tokens_seen": 1316496,
      "step": 2010
    },
    {
      "epoch": 1.0560796645702306,
      "grad_norm": 26.25,
      "learning_rate": 4.9995299261212536e-05,
      "loss": 0.5598,
      "num_input_tokens_seen": 1319952,
      "step": 2015
    },
    {
      "epoch": 1.0587002096436058,
      "grad_norm": 5.625,
      "learning_rate": 4.999484535178667e-05,
      "loss": 0.4857,
      "num_input_tokens_seen": 1323344,
      "step": 2020
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 8.375,
      "learning_rate": 4.9994370527815925e-05,
      "loss": 0.5875,
      "num_input_tokens_seen": 1326704,
      "step": 2025
    },
    {
      "epoch": 1.0639412997903563,
      "grad_norm": 3.90625,
      "learning_rate": 4.999387478969762e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 1330128,
      "step": 2030
    },
    {
      "epoch": 1.0665618448637317,
      "grad_norm": 7.90625,
      "learning_rate": 4.999335813784657e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 1333040,
      "step": 2035
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 12.75,
      "learning_rate": 4.999282057269508e-05,
      "loss": 0.5565,
      "num_input_tokens_seen": 1335504,
      "step": 2040
    },
    {
      "epoch": 1.0718029350104823,
      "grad_norm": 8.875,
      "learning_rate": 4.999226209469295e-05,
      "loss": 0.5534,
      "num_input_tokens_seen": 1339280,
      "step": 2045
    },
    {
      "epoch": 1.0744234800838575,
      "grad_norm": 6.65625,
      "learning_rate": 4.999168270430752e-05,
      "loss": 0.5924,
      "num_input_tokens_seen": 1342576,
      "step": 2050
    },
    {
      "epoch": 1.0770440251572326,
      "grad_norm": 9.1875,
      "learning_rate": 4.999108240202356e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 1346256,
      "step": 2055
    },
    {
      "epoch": 1.079664570230608,
      "grad_norm": 9.375,
      "learning_rate": 4.999046118834341e-05,
      "loss": 0.616,
      "num_input_tokens_seen": 1349008,
      "step": 2060
    },
    {
      "epoch": 1.0822851153039832,
      "grad_norm": 4.53125,
      "learning_rate": 4.998981906378684e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 1353552,
      "step": 2065
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 7.71875,
      "learning_rate": 4.998915602889117e-05,
      "loss": 0.4948,
      "num_input_tokens_seen": 1356400,
      "step": 2070
    },
    {
      "epoch": 1.0875262054507338,
      "grad_norm": 6.5625,
      "learning_rate": 4.9988472084211203e-05,
      "loss": 0.5678,
      "num_input_tokens_seen": 1361008,
      "step": 2075
    },
    {
      "epoch": 1.090146750524109,
      "grad_norm": 17.375,
      "learning_rate": 4.9987767230319215e-05,
      "loss": 0.6093,
      "num_input_tokens_seen": 1363184,
      "step": 2080
    },
    {
      "epoch": 1.0927672955974843,
      "grad_norm": 22.125,
      "learning_rate": 4.998704146780501e-05,
      "loss": 0.5017,
      "num_input_tokens_seen": 1365872,
      "step": 2085
    },
    {
      "epoch": 1.0953878406708595,
      "grad_norm": 4.6875,
      "learning_rate": 4.9986294797275857e-05,
      "loss": 0.5845,
      "num_input_tokens_seen": 1369168,
      "step": 2090
    },
    {
      "epoch": 1.0980083857442349,
      "grad_norm": 6.625,
      "learning_rate": 4.9985527219356554e-05,
      "loss": 0.488,
      "num_input_tokens_seen": 1372208,
      "step": 2095
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 5.96875,
      "learning_rate": 4.998473873468937e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 1377360,
      "step": 2100
    },
    {
      "epoch": 1.1032494758909852,
      "grad_norm": 15.0625,
      "learning_rate": 4.998392934393407e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 1380400,
      "step": 2105
    },
    {
      "epoch": 1.1058700209643606,
      "grad_norm": 12.4375,
      "learning_rate": 4.9983099047767905e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 1383056,
      "step": 2110
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 7.875,
      "learning_rate": 4.9982247846885644e-05,
      "loss": 0.5988,
      "num_input_tokens_seen": 1386160,
      "step": 2115
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 7.3125,
      "learning_rate": 4.9981375741999534e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 1389232,
      "step": 2120
    },
    {
      "epoch": 1.1137316561844863,
      "grad_norm": 8.25,
      "learning_rate": 4.99804827338393e-05,
      "loss": 0.6219,
      "num_input_tokens_seen": 1393328,
      "step": 2125
    },
    {
      "epoch": 1.1163522012578617,
      "grad_norm": 6.53125,
      "learning_rate": 4.997956882315218e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 1396816,
      "step": 2130
    },
    {
      "epoch": 1.118972746331237,
      "grad_norm": 8.3125,
      "learning_rate": 4.997863401070289e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 1400176,
      "step": 2135
    },
    {
      "epoch": 1.121593291404612,
      "grad_norm": 11.125,
      "learning_rate": 4.9977678297273634e-05,
      "loss": 0.6283,
      "num_input_tokens_seen": 1403408,
      "step": 2140
    },
    {
      "epoch": 1.1242138364779874,
      "grad_norm": 5.75,
      "learning_rate": 4.997670168366412e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 1406448,
      "step": 2145
    },
    {
      "epoch": 1.1268343815513626,
      "grad_norm": 29.625,
      "learning_rate": 4.997570417069152e-05,
      "loss": 0.5325,
      "num_input_tokens_seen": 1409584,
      "step": 2150
    },
    {
      "epoch": 1.129454926624738,
      "grad_norm": 8.9375,
      "learning_rate": 4.997468575919052e-05,
      "loss": 0.5124,
      "num_input_tokens_seen": 1413872,
      "step": 2155
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 8.375,
      "learning_rate": 4.9973646450013275e-05,
      "loss": 0.6733,
      "num_input_tokens_seen": 1416528,
      "step": 2160
    },
    {
      "epoch": 1.1346960167714886,
      "grad_norm": 12.5,
      "learning_rate": 4.997258624402943e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 1418896,
      "step": 2165
    },
    {
      "epoch": 1.1373165618448637,
      "grad_norm": 11.75,
      "learning_rate": 4.997150514212611e-05,
      "loss": 0.5799,
      "num_input_tokens_seen": 1421712,
      "step": 2170
    },
    {
      "epoch": 1.139937106918239,
      "grad_norm": 4.71875,
      "learning_rate": 4.997040314520795e-05,
      "loss": 0.5145,
      "num_input_tokens_seen": 1425072,
      "step": 2175
    },
    {
      "epoch": 1.1425576519916143,
      "grad_norm": 6.09375,
      "learning_rate": 4.9969280254197035e-05,
      "loss": 0.7095,
      "num_input_tokens_seen": 1428848,
      "step": 2180
    },
    {
      "epoch": 1.1451781970649895,
      "grad_norm": 11.375,
      "learning_rate": 4.996813647003296e-05,
      "loss": 0.5932,
      "num_input_tokens_seen": 1432432,
      "step": 2185
    },
    {
      "epoch": 1.1477987421383649,
      "grad_norm": 5.84375,
      "learning_rate": 4.9966971793672784e-05,
      "loss": 0.6819,
      "num_input_tokens_seen": 1436048,
      "step": 2190
    },
    {
      "epoch": 1.15041928721174,
      "grad_norm": 17.375,
      "learning_rate": 4.9965786226091054e-05,
      "loss": 0.6815,
      "num_input_tokens_seen": 1440144,
      "step": 2195
    },
    {
      "epoch": 1.1530398322851152,
      "grad_norm": 5.90625,
      "learning_rate": 4.9964579768279803e-05,
      "loss": 0.6124,
      "num_input_tokens_seen": 1443760,
      "step": 2200
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 7.46875,
      "learning_rate": 4.996335242124854e-05,
      "loss": 0.8067,
      "num_input_tokens_seen": 1446864,
      "step": 2205
    },
    {
      "epoch": 1.1582809224318658,
      "grad_norm": 9.75,
      "learning_rate": 4.996210418602425e-05,
      "loss": 0.3701,
      "num_input_tokens_seen": 1449456,
      "step": 2210
    },
    {
      "epoch": 1.1609014675052411,
      "grad_norm": 6.6875,
      "learning_rate": 4.99608350636514e-05,
      "loss": 0.5973,
      "num_input_tokens_seen": 1452944,
      "step": 2215
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 4.78125,
      "learning_rate": 4.995954505519193e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 1455952,
      "step": 2220
    },
    {
      "epoch": 1.1661425576519917,
      "grad_norm": 6.625,
      "learning_rate": 4.995823416172527e-05,
      "loss": 0.6233,
      "num_input_tokens_seen": 1459376,
      "step": 2225
    },
    {
      "epoch": 1.1687631027253669,
      "grad_norm": 5.28125,
      "learning_rate": 4.995690238434831e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 1462352,
      "step": 2230
    },
    {
      "epoch": 1.171383647798742,
      "grad_norm": 9.125,
      "learning_rate": 4.995554972417541e-05,
      "loss": 0.6016,
      "num_input_tokens_seen": 1465104,
      "step": 2235
    },
    {
      "epoch": 1.1740041928721174,
      "grad_norm": 2.609375,
      "learning_rate": 4.995417618233844e-05,
      "loss": 0.5611,
      "num_input_tokens_seen": 1468624,
      "step": 2240
    },
    {
      "epoch": 1.1766247379454926,
      "grad_norm": 9.1875,
      "learning_rate": 4.9952781759986694e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 1475184,
      "step": 2245
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 3.96875,
      "learning_rate": 4.995136645828697e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 1478640,
      "step": 2250
    },
    {
      "epoch": 1.1818658280922432,
      "grad_norm": 5.78125,
      "learning_rate": 4.994993027842353e-05,
      "loss": 0.5792,
      "num_input_tokens_seen": 1481552,
      "step": 2255
    },
    {
      "epoch": 1.1844863731656186,
      "grad_norm": 6.84375,
      "learning_rate": 4.9948473221598094e-05,
      "loss": 0.5168,
      "num_input_tokens_seen": 1485104,
      "step": 2260
    },
    {
      "epoch": 1.1871069182389937,
      "grad_norm": 8.125,
      "learning_rate": 4.994699528902987e-05,
      "loss": 0.5662,
      "num_input_tokens_seen": 1487728,
      "step": 2265
    },
    {
      "epoch": 1.189727463312369,
      "grad_norm": 10.6875,
      "learning_rate": 4.994549648195552e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 1491312,
      "step": 2270
    },
    {
      "epoch": 1.1923480083857443,
      "grad_norm": 16.0,
      "learning_rate": 4.994397680162918e-05,
      "loss": 0.6018,
      "num_input_tokens_seen": 1494192,
      "step": 2275
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 7.625,
      "learning_rate": 4.9942436249322444e-05,
      "loss": 0.6782,
      "num_input_tokens_seen": 1497168,
      "step": 2280
    },
    {
      "epoch": 1.1975890985324948,
      "grad_norm": 6.90625,
      "learning_rate": 4.994087482632438e-05,
      "loss": 0.5463,
      "num_input_tokens_seen": 1501936,
      "step": 2285
    },
    {
      "epoch": 1.20020964360587,
      "grad_norm": 5.5,
      "learning_rate": 4.993929253394152e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 1505040,
      "step": 2290
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 9.9375,
      "learning_rate": 4.993768937349784e-05,
      "loss": 0.661,
      "num_input_tokens_seen": 1507856,
      "step": 2295
    },
    {
      "epoch": 1.2054507337526206,
      "grad_norm": 8.625,
      "learning_rate": 4.993606534633481e-05,
      "loss": 0.5129,
      "num_input_tokens_seen": 1510768,
      "step": 2300
    },
    {
      "epoch": 1.2080712788259957,
      "grad_norm": 9.625,
      "learning_rate": 4.9934420453811334e-05,
      "loss": 0.6451,
      "num_input_tokens_seen": 1513808,
      "step": 2305
    },
    {
      "epoch": 1.2106918238993711,
      "grad_norm": 4.5625,
      "learning_rate": 4.993275469730377e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 1517392,
      "step": 2310
    },
    {
      "epoch": 1.2133123689727463,
      "grad_norm": 3.546875,
      "learning_rate": 4.993106807820597e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 1521648,
      "step": 2315
    },
    {
      "epoch": 1.2159329140461215,
      "grad_norm": 7.34375,
      "learning_rate": 4.99293605979292e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 1524496,
      "step": 2320
    },
    {
      "epoch": 1.2185534591194969,
      "grad_norm": 6.6875,
      "learning_rate": 4.992763225790221e-05,
      "loss": 0.5275,
      "num_input_tokens_seen": 1528144,
      "step": 2325
    },
    {
      "epoch": 1.221174004192872,
      "grad_norm": 4.6875,
      "learning_rate": 4.992588305957119e-05,
      "loss": 0.515,
      "num_input_tokens_seen": 1531664,
      "step": 2330
    },
    {
      "epoch": 1.2237945492662474,
      "grad_norm": 11.6875,
      "learning_rate": 4.99241130043998e-05,
      "loss": 0.6284,
      "num_input_tokens_seen": 1535408,
      "step": 2335
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 6.125,
      "learning_rate": 4.992232209386914e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 1538448,
      "step": 2340
    },
    {
      "epoch": 1.229035639412998,
      "grad_norm": 11.25,
      "learning_rate": 4.9920510329477756e-05,
      "loss": 0.5695,
      "num_input_tokens_seen": 1541968,
      "step": 2345
    },
    {
      "epoch": 1.2316561844863732,
      "grad_norm": 6.125,
      "learning_rate": 4.9918677712741644e-05,
      "loss": 0.5698,
      "num_input_tokens_seen": 1544560,
      "step": 2350
    },
    {
      "epoch": 1.2342767295597485,
      "grad_norm": 9.625,
      "learning_rate": 4.991682424519427e-05,
      "loss": 0.5374,
      "num_input_tokens_seen": 1548080,
      "step": 2355
    },
    {
      "epoch": 1.2368972746331237,
      "grad_norm": 9.125,
      "learning_rate": 4.9914949928386524e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 1551472,
      "step": 2360
    },
    {
      "epoch": 1.2395178197064989,
      "grad_norm": 6.625,
      "learning_rate": 4.991305476388673e-05,
      "loss": 0.5977,
      "num_input_tokens_seen": 1553904,
      "step": 2365
    },
    {
      "epoch": 1.2421383647798743,
      "grad_norm": 7.21875,
      "learning_rate": 4.991113875328072e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 1558352,
      "step": 2370
    },
    {
      "epoch": 1.2447589098532494,
      "grad_norm": 9.25,
      "learning_rate": 4.9909201898171676e-05,
      "loss": 0.4038,
      "num_input_tokens_seen": 1561616,
      "step": 2375
    },
    {
      "epoch": 1.2473794549266248,
      "grad_norm": 6.875,
      "learning_rate": 4.9907244200180295e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 1566000,
      "step": 2380
    },
    {
      "epoch": 1.25,
      "grad_norm": 3.03125,
      "learning_rate": 4.990526566094469e-05,
      "loss": 0.5236,
      "num_input_tokens_seen": 1569776,
      "step": 2385
    },
    {
      "epoch": 1.2526205450733752,
      "grad_norm": 10.8125,
      "learning_rate": 4.99032662821204e-05,
      "loss": 0.6199,
      "num_input_tokens_seen": 1572112,
      "step": 2390
    },
    {
      "epoch": 1.2552410901467506,
      "grad_norm": 2.703125,
      "learning_rate": 4.990124606538042e-05,
      "loss": 0.5328,
      "num_input_tokens_seen": 1575312,
      "step": 2395
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 3.640625,
      "learning_rate": 4.9899205012415184e-05,
      "loss": 0.603,
      "num_input_tokens_seen": 1578256,
      "step": 2400
    },
    {
      "epoch": 1.2604821802935011,
      "grad_norm": 3.609375,
      "learning_rate": 4.9897143124932547e-05,
      "loss": 0.7297,
      "num_input_tokens_seen": 1581328,
      "step": 2405
    },
    {
      "epoch": 1.2631027253668763,
      "grad_norm": 4.34375,
      "learning_rate": 4.9895060404657786e-05,
      "loss": 0.5915,
      "num_input_tokens_seen": 1583952,
      "step": 2410
    },
    {
      "epoch": 1.2657232704402515,
      "grad_norm": 10.875,
      "learning_rate": 4.9892956853333644e-05,
      "loss": 0.5586,
      "num_input_tokens_seen": 1587472,
      "step": 2415
    },
    {
      "epoch": 1.2683438155136268,
      "grad_norm": 11.6875,
      "learning_rate": 4.989083247272027e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 1590384,
      "step": 2420
    },
    {
      "epoch": 1.270964360587002,
      "grad_norm": 5.625,
      "learning_rate": 4.988868726459526e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 1593264,
      "step": 2425
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 4.75,
      "learning_rate": 4.988652123075361e-05,
      "loss": 0.4712,
      "num_input_tokens_seen": 1597680,
      "step": 2430
    },
    {
      "epoch": 1.2762054507337526,
      "grad_norm": 6.375,
      "learning_rate": 4.988433437300776e-05,
      "loss": 0.5674,
      "num_input_tokens_seen": 1600912,
      "step": 2435
    },
    {
      "epoch": 1.2788259958071277,
      "grad_norm": 5.46875,
      "learning_rate": 4.988212669318758e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 1603472,
      "step": 2440
    },
    {
      "epoch": 1.2814465408805031,
      "grad_norm": 11.3125,
      "learning_rate": 4.987989819314036e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 1606384,
      "step": 2445
    },
    {
      "epoch": 1.2840670859538785,
      "grad_norm": 12.1875,
      "learning_rate": 4.98776488747308e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 1609360,
      "step": 2450
    },
    {
      "epoch": 1.2866876310272537,
      "grad_norm": 6.875,
      "learning_rate": 4.9875378739841016e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 1612336,
      "step": 2455
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 5.59375,
      "learning_rate": 4.9873087790370576e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 1616048,
      "step": 2460
    },
    {
      "epoch": 1.2919287211740043,
      "grad_norm": 10.0,
      "learning_rate": 4.9870776028236424e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 1618960,
      "step": 2465
    },
    {
      "epoch": 1.2945492662473794,
      "grad_norm": 3.4375,
      "learning_rate": 4.9868443455372945e-05,
      "loss": 0.3938,
      "num_input_tokens_seen": 1622544,
      "step": 2470
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 5.34375,
      "learning_rate": 4.986609007373193e-05,
      "loss": 0.506,
      "num_input_tokens_seen": 1625872,
      "step": 2475
    },
    {
      "epoch": 1.29979035639413,
      "grad_norm": 11.4375,
      "learning_rate": 4.986371588528257e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 1629456,
      "step": 2480
    },
    {
      "epoch": 1.3024109014675052,
      "grad_norm": 3.671875,
      "learning_rate": 4.98613208920115e-05,
      "loss": 0.6577,
      "num_input_tokens_seen": 1632432,
      "step": 2485
    },
    {
      "epoch": 1.3050314465408805,
      "grad_norm": 3.75,
      "learning_rate": 4.985890509592271e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 1636080,
      "step": 2490
    },
    {
      "epoch": 1.3076519916142557,
      "grad_norm": 4.34375,
      "learning_rate": 4.985646849903766e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 1638960,
      "step": 2495
    },
    {
      "epoch": 1.310272536687631,
      "grad_norm": 2.53125,
      "learning_rate": 4.985401110339517e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 1643120,
      "step": 2500
    },
    {
      "epoch": 1.3128930817610063,
      "grad_norm": 5.6875,
      "learning_rate": 4.985153291105146e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 1646768,
      "step": 2505
    },
    {
      "epoch": 1.3155136268343814,
      "grad_norm": 4.03125,
      "learning_rate": 4.984903392408019e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 1650256,
      "step": 2510
    },
    {
      "epoch": 1.3181341719077568,
      "grad_norm": 7.1875,
      "learning_rate": 4.984651414457239e-05,
      "loss": 0.5702,
      "num_input_tokens_seen": 1652912,
      "step": 2515
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 3.46875,
      "learning_rate": 4.98439735746365e-05,
      "loss": 0.5279,
      "num_input_tokens_seen": 1656432,
      "step": 2520
    },
    {
      "epoch": 1.3233752620545074,
      "grad_norm": 5.875,
      "learning_rate": 4.984141221639835e-05,
      "loss": 0.438,
      "num_input_tokens_seen": 1659472,
      "step": 2525
    },
    {
      "epoch": 1.3259958071278826,
      "grad_norm": 8.375,
      "learning_rate": 4.9838830072001165e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 1662480,
      "step": 2530
    },
    {
      "epoch": 1.3286163522012577,
      "grad_norm": 6.90625,
      "learning_rate": 4.983622714360557e-05,
      "loss": 0.5578,
      "num_input_tokens_seen": 1665424,
      "step": 2535
    },
    {
      "epoch": 1.3312368972746331,
      "grad_norm": 6.34375,
      "learning_rate": 4.9833603433389576e-05,
      "loss": 0.5896,
      "num_input_tokens_seen": 1667760,
      "step": 2540
    },
    {
      "epoch": 1.3338574423480085,
      "grad_norm": 5.875,
      "learning_rate": 4.983095894354858e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 1670800,
      "step": 2545
    },
    {
      "epoch": 1.3364779874213837,
      "grad_norm": 5.625,
      "learning_rate": 4.982829367629537e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 1675760,
      "step": 2550
    },
    {
      "epoch": 1.3390985324947589,
      "grad_norm": 15.1875,
      "learning_rate": 4.982560763386013e-05,
      "loss": 0.5877,
      "num_input_tokens_seen": 1677936,
      "step": 2555
    },
    {
      "epoch": 1.3417190775681342,
      "grad_norm": 4.0,
      "learning_rate": 4.9822900818490404e-05,
      "loss": 0.6328,
      "num_input_tokens_seen": 1685008,
      "step": 2560
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 6.46875,
      "learning_rate": 4.982017323245114e-05,
      "loss": 0.6342,
      "num_input_tokens_seen": 1688240,
      "step": 2565
    },
    {
      "epoch": 1.3469601677148848,
      "grad_norm": 5.8125,
      "learning_rate": 4.981742487802466e-05,
      "loss": 0.6836,
      "num_input_tokens_seen": 1690832,
      "step": 2570
    },
    {
      "epoch": 1.34958071278826,
      "grad_norm": 3.625,
      "learning_rate": 4.9814655757510644e-05,
      "loss": 0.6471,
      "num_input_tokens_seen": 1695248,
      "step": 2575
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 3.65625,
      "learning_rate": 4.981186587322619e-05,
      "loss": 0.53,
      "num_input_tokens_seen": 1698896,
      "step": 2580
    },
    {
      "epoch": 1.3548218029350105,
      "grad_norm": 12.4375,
      "learning_rate": 4.980905522750573e-05,
      "loss": 0.6734,
      "num_input_tokens_seen": 1701296,
      "step": 2585
    },
    {
      "epoch": 1.3574423480083857,
      "grad_norm": 9.1875,
      "learning_rate": 4.980622382270108e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 1705136,
      "step": 2590
    },
    {
      "epoch": 1.360062893081761,
      "grad_norm": 5.625,
      "learning_rate": 4.9803371661181456e-05,
      "loss": 0.4978,
      "num_input_tokens_seen": 1708016,
      "step": 2595
    },
    {
      "epoch": 1.3626834381551363,
      "grad_norm": 30.125,
      "learning_rate": 4.980049874533338e-05,
      "loss": 0.6686,
      "num_input_tokens_seen": 1710864,
      "step": 2600
    },
    {
      "epoch": 1.3653039832285114,
      "grad_norm": 3.84375,
      "learning_rate": 4.979760507756081e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 1713552,
      "step": 2605
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 7.1875,
      "learning_rate": 4.979469066028502e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 1716528,
      "step": 2610
    },
    {
      "epoch": 1.370545073375262,
      "grad_norm": 29.75,
      "learning_rate": 4.9791755495944645e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 1719216,
      "step": 2615
    },
    {
      "epoch": 1.3731656184486374,
      "grad_norm": 4.71875,
      "learning_rate": 4.978879958699573e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 1722384,
      "step": 2620
    },
    {
      "epoch": 1.3757861635220126,
      "grad_norm": 5.53125,
      "learning_rate": 4.978582293591162e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 1725104,
      "step": 2625
    },
    {
      "epoch": 1.3784067085953877,
      "grad_norm": 9.0,
      "learning_rate": 4.978282554518305e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 1728112,
      "step": 2630
    },
    {
      "epoch": 1.381027253668763,
      "grad_norm": 7.34375,
      "learning_rate": 4.9779807417318096e-05,
      "loss": 0.5726,
      "num_input_tokens_seen": 1731536,
      "step": 2635
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 4.03125,
      "learning_rate": 4.977676855484219e-05,
      "loss": 0.588,
      "num_input_tokens_seen": 1734896,
      "step": 2640
    },
    {
      "epoch": 1.3862683438155137,
      "grad_norm": 11.75,
      "learning_rate": 4.977370896029812e-05,
      "loss": 0.5549,
      "num_input_tokens_seen": 1737584,
      "step": 2645
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 8.875,
      "learning_rate": 4.977062863624601e-05,
      "loss": 0.5991,
      "num_input_tokens_seen": 1740720,
      "step": 2650
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 5.875,
      "learning_rate": 4.976752758526333e-05,
      "loss": 0.3995,
      "num_input_tokens_seen": 1743216,
      "step": 2655
    },
    {
      "epoch": 1.3941299790356394,
      "grad_norm": 7.625,
      "learning_rate": 4.9764405809944906e-05,
      "loss": 0.5976,
      "num_input_tokens_seen": 1746000,
      "step": 2660
    },
    {
      "epoch": 1.3967505241090148,
      "grad_norm": 13.1875,
      "learning_rate": 4.9761263312902895e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 1748560,
      "step": 2665
    },
    {
      "epoch": 1.39937106918239,
      "grad_norm": 4.59375,
      "learning_rate": 4.9758100096766786e-05,
      "loss": 0.6213,
      "num_input_tokens_seen": 1751504,
      "step": 2670
    },
    {
      "epoch": 1.4019916142557651,
      "grad_norm": 5.5625,
      "learning_rate": 4.975491616418342e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 1754864,
      "step": 2675
    },
    {
      "epoch": 1.4046121593291405,
      "grad_norm": 14.25,
      "learning_rate": 4.975171151781698e-05,
      "loss": 0.5484,
      "num_input_tokens_seen": 1757648,
      "step": 2680
    },
    {
      "epoch": 1.4072327044025157,
      "grad_norm": 4.71875,
      "learning_rate": 4.974848616034894e-05,
      "loss": 0.5085,
      "num_input_tokens_seen": 1760432,
      "step": 2685
    },
    {
      "epoch": 1.409853249475891,
      "grad_norm": 3.1875,
      "learning_rate": 4.974524009447815e-05,
      "loss": 0.5285,
      "num_input_tokens_seen": 1763664,
      "step": 2690
    },
    {
      "epoch": 1.4124737945492662,
      "grad_norm": 4.6875,
      "learning_rate": 4.974197332292078e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 1766640,
      "step": 2695
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 4.90625,
      "learning_rate": 4.973868584841028e-05,
      "loss": 0.5254,
      "num_input_tokens_seen": 1769776,
      "step": 2700
    },
    {
      "epoch": 1.4177148846960168,
      "grad_norm": 6.75,
      "learning_rate": 4.973537767369749e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 1772112,
      "step": 2705
    },
    {
      "epoch": 1.420335429769392,
      "grad_norm": 6.1875,
      "learning_rate": 4.973204880155053e-05,
      "loss": 0.4854,
      "num_input_tokens_seen": 1775312,
      "step": 2710
    },
    {
      "epoch": 1.4229559748427674,
      "grad_norm": 3.765625,
      "learning_rate": 4.972869923475485e-05,
      "loss": 0.6498,
      "num_input_tokens_seen": 1779504,
      "step": 2715
    },
    {
      "epoch": 1.4255765199161425,
      "grad_norm": 3.703125,
      "learning_rate": 4.972532897611321e-05,
      "loss": 0.4558,
      "num_input_tokens_seen": 1783248,
      "step": 2720
    },
    {
      "epoch": 1.4281970649895177,
      "grad_norm": 10.25,
      "learning_rate": 4.972193802844569e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 1789168,
      "step": 2725
    },
    {
      "epoch": 1.430817610062893,
      "grad_norm": 6.5,
      "learning_rate": 4.971852639458968e-05,
      "loss": 0.4929,
      "num_input_tokens_seen": 1792592,
      "step": 2730
    },
    {
      "epoch": 1.4334381551362683,
      "grad_norm": 6.0,
      "learning_rate": 4.971509407739988e-05,
      "loss": 0.5455,
      "num_input_tokens_seen": 1795536,
      "step": 2735
    },
    {
      "epoch": 1.4360587002096437,
      "grad_norm": 3.59375,
      "learning_rate": 4.971164107974831e-05,
      "loss": 0.4469,
      "num_input_tokens_seen": 1798064,
      "step": 2740
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 4.75,
      "learning_rate": 4.970816740452425e-05,
      "loss": 0.5376,
      "num_input_tokens_seen": 1801392,
      "step": 2745
    },
    {
      "epoch": 1.441299790356394,
      "grad_norm": 7.40625,
      "learning_rate": 4.9704673054634335e-05,
      "loss": 0.5621,
      "num_input_tokens_seen": 1804368,
      "step": 2750
    },
    {
      "epoch": 1.4439203354297694,
      "grad_norm": 9.9375,
      "learning_rate": 4.970115803300247e-05,
      "loss": 0.5758,
      "num_input_tokens_seen": 1807504,
      "step": 2755
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 9.0,
      "learning_rate": 4.969762234256987e-05,
      "loss": 0.4022,
      "num_input_tokens_seen": 1810160,
      "step": 2760
    },
    {
      "epoch": 1.44916142557652,
      "grad_norm": 7.9375,
      "learning_rate": 4.969406598629503e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 1813616,
      "step": 2765
    },
    {
      "epoch": 1.4517819706498951,
      "grad_norm": 6.75,
      "learning_rate": 4.969048896715376e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 1816752,
      "step": 2770
    },
    {
      "epoch": 1.4544025157232705,
      "grad_norm": 5.8125,
      "learning_rate": 4.968689128813914e-05,
      "loss": 0.4932,
      "num_input_tokens_seen": 1820784,
      "step": 2775
    },
    {
      "epoch": 1.4570230607966457,
      "grad_norm": 6.59375,
      "learning_rate": 4.968327295226153e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 1823024,
      "step": 2780
    },
    {
      "epoch": 1.459643605870021,
      "grad_norm": 5.3125,
      "learning_rate": 4.967963396254861e-05,
      "loss": 0.5567,
      "num_input_tokens_seen": 1828016,
      "step": 2785
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 4.1875,
      "learning_rate": 4.967597432204531e-05,
      "loss": 0.5549,
      "num_input_tokens_seen": 1831280,
      "step": 2790
    },
    {
      "epoch": 1.4648846960167714,
      "grad_norm": 3.828125,
      "learning_rate": 4.9672294033813846e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 1834896,
      "step": 2795
    },
    {
      "epoch": 1.4675052410901468,
      "grad_norm": 3.71875,
      "learning_rate": 4.966859310093372e-05,
      "loss": 0.5104,
      "num_input_tokens_seen": 1837488,
      "step": 2800
    },
    {
      "epoch": 1.470125786163522,
      "grad_norm": 4.03125,
      "learning_rate": 4.966487152650171e-05,
      "loss": 0.5723,
      "num_input_tokens_seen": 1841584,
      "step": 2805
    },
    {
      "epoch": 1.4727463312368974,
      "grad_norm": 6.28125,
      "learning_rate": 4.966112931363185e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 1844688,
      "step": 2810
    },
    {
      "epoch": 1.4753668763102725,
      "grad_norm": 6.90625,
      "learning_rate": 4.965736646545546e-05,
      "loss": 0.5274,
      "num_input_tokens_seen": 1847472,
      "step": 2815
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 6.21875,
      "learning_rate": 4.96535829851211e-05,
      "loss": 0.7566,
      "num_input_tokens_seen": 1851088,
      "step": 2820
    },
    {
      "epoch": 1.480607966457023,
      "grad_norm": 3.890625,
      "learning_rate": 4.964977887579464e-05,
      "loss": 0.5209,
      "num_input_tokens_seen": 1855216,
      "step": 2825
    },
    {
      "epoch": 1.4832285115303983,
      "grad_norm": 2.8125,
      "learning_rate": 4.964595414065918e-05,
      "loss": 0.4375,
      "num_input_tokens_seen": 1857744,
      "step": 2830
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 4.6875,
      "learning_rate": 4.9642108782915066e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 1860400,
      "step": 2835
    },
    {
      "epoch": 1.4884696016771488,
      "grad_norm": 7.59375,
      "learning_rate": 4.963824280577993e-05,
      "loss": 0.5748,
      "num_input_tokens_seen": 1863248,
      "step": 2840
    },
    {
      "epoch": 1.491090146750524,
      "grad_norm": 5.375,
      "learning_rate": 4.963435621248865e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 1866576,
      "step": 2845
    },
    {
      "epoch": 1.4937106918238994,
      "grad_norm": 5.1875,
      "learning_rate": 4.9630449006293345e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 1869840,
      "step": 2850
    },
    {
      "epoch": 1.4963312368972748,
      "grad_norm": 4.71875,
      "learning_rate": 4.9626521190463375e-05,
      "loss": 0.6332,
      "num_input_tokens_seen": 1872624,
      "step": 2855
    },
    {
      "epoch": 1.49895178197065,
      "grad_norm": 4.375,
      "learning_rate": 4.9622572768285377e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 1876016,
      "step": 2860
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.5221880078315735,
      "eval_runtime": 24.1839,
      "eval_samples_per_second": 35.065,
      "eval_steps_per_second": 8.766,
      "num_input_tokens_seen": 1877040,
      "step": 2862
    },
    {
      "epoch": 1.501572327044025,
      "grad_norm": 3.765625,
      "learning_rate": 4.96186037430632e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 1880976,
      "step": 2865
    },
    {
      "epoch": 1.5041928721174003,
      "grad_norm": 15.9375,
      "learning_rate": 4.9614614118117934e-05,
      "loss": 0.6048,
      "num_input_tokens_seen": 1883568,
      "step": 2870
    },
    {
      "epoch": 1.5068134171907757,
      "grad_norm": 7.78125,
      "learning_rate": 4.961060389678793e-05,
      "loss": 0.4779,
      "num_input_tokens_seen": 1886800,
      "step": 2875
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 6.34375,
      "learning_rate": 4.9606573082428754e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 1890672,
      "step": 2880
    },
    {
      "epoch": 1.5120545073375262,
      "grad_norm": 7.59375,
      "learning_rate": 4.9602521678413206e-05,
      "loss": 0.3944,
      "num_input_tokens_seen": 1893456,
      "step": 2885
    },
    {
      "epoch": 1.5146750524109014,
      "grad_norm": 11.9375,
      "learning_rate": 4.959844968813132e-05,
      "loss": 0.5458,
      "num_input_tokens_seen": 1896752,
      "step": 2890
    },
    {
      "epoch": 1.5172955974842768,
      "grad_norm": 4.625,
      "learning_rate": 4.959435711499034e-05,
      "loss": 0.4432,
      "num_input_tokens_seen": 1899664,
      "step": 2895
    },
    {
      "epoch": 1.519916142557652,
      "grad_norm": 4.6875,
      "learning_rate": 4.959024396241475e-05,
      "loss": 0.4177,
      "num_input_tokens_seen": 1902320,
      "step": 2900
    },
    {
      "epoch": 1.5225366876310273,
      "grad_norm": 2.96875,
      "learning_rate": 4.958611023384626e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 1905488,
      "step": 2905
    },
    {
      "epoch": 1.5251572327044025,
      "grad_norm": 6.9375,
      "learning_rate": 4.958195593274376e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 1908144,
      "step": 2910
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 2.609375,
      "learning_rate": 4.957778106258341e-05,
      "loss": 0.4925,
      "num_input_tokens_seen": 1911536,
      "step": 2915
    },
    {
      "epoch": 1.530398322851153,
      "grad_norm": 3.265625,
      "learning_rate": 4.957358562685852e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 1916176,
      "step": 2920
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 3.171875,
      "learning_rate": 4.956936962907966e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 1920592,
      "step": 2925
    },
    {
      "epoch": 1.5356394129979036,
      "grad_norm": 5.59375,
      "learning_rate": 4.9565133072774585e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 1923152,
      "step": 2930
    },
    {
      "epoch": 1.5382599580712788,
      "grad_norm": 4.875,
      "learning_rate": 4.956087596148824e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 1926000,
      "step": 2935
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 7.28125,
      "learning_rate": 4.955659829878279e-05,
      "loss": 0.5936,
      "num_input_tokens_seen": 1929072,
      "step": 2940
    },
    {
      "epoch": 1.5435010482180294,
      "grad_norm": 2.65625,
      "learning_rate": 4.955230008823758e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 1932208,
      "step": 2945
    },
    {
      "epoch": 1.5461215932914047,
      "grad_norm": 5.46875,
      "learning_rate": 4.954798133344916e-05,
      "loss": 0.6909,
      "num_input_tokens_seen": 1935152,
      "step": 2950
    },
    {
      "epoch": 1.54874213836478,
      "grad_norm": 5.15625,
      "learning_rate": 4.954364203803127e-05,
      "loss": 0.5576,
      "num_input_tokens_seen": 1938704,
      "step": 2955
    },
    {
      "epoch": 1.551362683438155,
      "grad_norm": 4.5625,
      "learning_rate": 4.953928220561482e-05,
      "loss": 0.6054,
      "num_input_tokens_seen": 1942192,
      "step": 2960
    },
    {
      "epoch": 1.5539832285115303,
      "grad_norm": 12.375,
      "learning_rate": 4.953490183984795e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 1945744,
      "step": 2965
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 5.0,
      "learning_rate": 4.953050094439591e-05,
      "loss": 0.5016,
      "num_input_tokens_seen": 1948944,
      "step": 2970
    },
    {
      "epoch": 1.559224318658281,
      "grad_norm": 8.4375,
      "learning_rate": 4.95260795229412e-05,
      "loss": 0.4794,
      "num_input_tokens_seen": 1951792,
      "step": 2975
    },
    {
      "epoch": 1.5618448637316562,
      "grad_norm": 3.859375,
      "learning_rate": 4.952163757918344e-05,
      "loss": 0.5568,
      "num_input_tokens_seen": 1955376,
      "step": 2980
    },
    {
      "epoch": 1.5644654088050314,
      "grad_norm": 5.15625,
      "learning_rate": 4.951717511683947e-05,
      "loss": 0.5189,
      "num_input_tokens_seen": 1959120,
      "step": 2985
    },
    {
      "epoch": 1.5670859538784065,
      "grad_norm": 3.9375,
      "learning_rate": 4.9512692139643264e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 1964304,
      "step": 2990
    },
    {
      "epoch": 1.569706498951782,
      "grad_norm": 7.28125,
      "learning_rate": 4.950818865134596e-05,
      "loss": 0.6378,
      "num_input_tokens_seen": 1966832,
      "step": 2995
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 5.21875,
      "learning_rate": 4.9503664655715885e-05,
      "loss": 0.5912,
      "num_input_tokens_seen": 1970224,
      "step": 3000
    },
    {
      "epoch": 1.5749475890985325,
      "grad_norm": 3.34375,
      "learning_rate": 4.9499120156538516e-05,
      "loss": 0.5316,
      "num_input_tokens_seen": 1974288,
      "step": 3005
    },
    {
      "epoch": 1.5775681341719077,
      "grad_norm": 4.5625,
      "learning_rate": 4.949455515761647e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 1978064,
      "step": 3010
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 11.3125,
      "learning_rate": 4.948996966276953e-05,
      "loss": 0.5424,
      "num_input_tokens_seen": 1981744,
      "step": 3015
    },
    {
      "epoch": 1.5828092243186582,
      "grad_norm": 10.4375,
      "learning_rate": 4.948536367583464e-05,
      "loss": 0.5496,
      "num_input_tokens_seen": 1984016,
      "step": 3020
    },
    {
      "epoch": 1.5854297693920336,
      "grad_norm": 7.5625,
      "learning_rate": 4.948073720066587e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 1987056,
      "step": 3025
    },
    {
      "epoch": 1.5880503144654088,
      "grad_norm": 3.796875,
      "learning_rate": 4.947609024113444e-05,
      "loss": 0.5495,
      "num_input_tokens_seen": 1992048,
      "step": 3030
    },
    {
      "epoch": 1.590670859538784,
      "grad_norm": 3.90625,
      "learning_rate": 4.947142280112873e-05,
      "loss": 0.6609,
      "num_input_tokens_seen": 1995472,
      "step": 3035
    },
    {
      "epoch": 1.5932914046121593,
      "grad_norm": 5.03125,
      "learning_rate": 4.946673488455422e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 1999152,
      "step": 3040
    },
    {
      "epoch": 1.5959119496855347,
      "grad_norm": 6.1875,
      "learning_rate": 4.946202649533356e-05,
      "loss": 0.557,
      "num_input_tokens_seen": 2001744,
      "step": 3045
    },
    {
      "epoch": 1.59853249475891,
      "grad_norm": 5.5,
      "learning_rate": 4.9457297637406506e-05,
      "loss": 0.6279,
      "num_input_tokens_seen": 2004848,
      "step": 3050
    },
    {
      "epoch": 1.601153039832285,
      "grad_norm": 6.625,
      "learning_rate": 4.9452548314729965e-05,
      "loss": 0.5527,
      "num_input_tokens_seen": 2007920,
      "step": 3055
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 4.8125,
      "learning_rate": 4.944777853127793e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 2011152,
      "step": 3060
    },
    {
      "epoch": 1.6063941299790356,
      "grad_norm": 9.625,
      "learning_rate": 4.9442988291041545e-05,
      "loss": 0.6174,
      "num_input_tokens_seen": 2013968,
      "step": 3065
    },
    {
      "epoch": 1.609014675052411,
      "grad_norm": 12.4375,
      "learning_rate": 4.943817759802908e-05,
      "loss": 0.5346,
      "num_input_tokens_seen": 2017936,
      "step": 3070
    },
    {
      "epoch": 1.6116352201257862,
      "grad_norm": 3.796875,
      "learning_rate": 4.94333464562659e-05,
      "loss": 0.494,
      "num_input_tokens_seen": 2020528,
      "step": 3075
    },
    {
      "epoch": 1.6142557651991614,
      "grad_norm": 10.0625,
      "learning_rate": 4.942849486979446e-05,
      "loss": 0.6877,
      "num_input_tokens_seen": 2023792,
      "step": 3080
    },
    {
      "epoch": 1.6168763102725365,
      "grad_norm": 5.96875,
      "learning_rate": 4.9423622842674366e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 2027152,
      "step": 3085
    },
    {
      "epoch": 1.619496855345912,
      "grad_norm": 6.21875,
      "learning_rate": 4.9418730378982304e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 2029968,
      "step": 3090
    },
    {
      "epoch": 1.6221174004192873,
      "grad_norm": 8.875,
      "learning_rate": 4.9413817482812064e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 2033488,
      "step": 3095
    },
    {
      "epoch": 1.6247379454926625,
      "grad_norm": 16.875,
      "learning_rate": 4.9408884158274534e-05,
      "loss": 0.5821,
      "num_input_tokens_seen": 2037136,
      "step": 3100
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 4.34375,
      "learning_rate": 4.940393040949769e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 2039888,
      "step": 3105
    },
    {
      "epoch": 1.629979035639413,
      "grad_norm": 6.03125,
      "learning_rate": 4.939895624062661e-05,
      "loss": 0.6969,
      "num_input_tokens_seen": 2043024,
      "step": 3110
    },
    {
      "epoch": 1.6325995807127882,
      "grad_norm": 6.78125,
      "learning_rate": 4.9393961655823454e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 2047056,
      "step": 3115
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 10.75,
      "learning_rate": 4.9388946659267444e-05,
      "loss": 0.5718,
      "num_input_tokens_seen": 2050000,
      "step": 3120
    },
    {
      "epoch": 1.6378406708595388,
      "grad_norm": 4.3125,
      "learning_rate": 4.9383911255154916e-05,
      "loss": 0.5119,
      "num_input_tokens_seen": 2054032,
      "step": 3125
    },
    {
      "epoch": 1.640461215932914,
      "grad_norm": 5.8125,
      "learning_rate": 4.9378855447699264e-05,
      "loss": 0.5275,
      "num_input_tokens_seen": 2057072,
      "step": 3130
    },
    {
      "epoch": 1.6430817610062893,
      "grad_norm": 5.0,
      "learning_rate": 4.9373779241130955e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 2060656,
      "step": 3135
    },
    {
      "epoch": 1.6457023060796647,
      "grad_norm": 14.3125,
      "learning_rate": 4.936868263969752e-05,
      "loss": 0.5402,
      "num_input_tokens_seen": 2063568,
      "step": 3140
    },
    {
      "epoch": 1.64832285115304,
      "grad_norm": 11.0,
      "learning_rate": 4.936356564766358e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 2066960,
      "step": 3145
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 5.65625,
      "learning_rate": 4.935842826931078e-05,
      "loss": 0.5106,
      "num_input_tokens_seen": 2070096,
      "step": 3150
    },
    {
      "epoch": 1.6535639412997902,
      "grad_norm": 3.59375,
      "learning_rate": 4.9353270508937854e-05,
      "loss": 0.5208,
      "num_input_tokens_seen": 2073072,
      "step": 3155
    },
    {
      "epoch": 1.6561844863731656,
      "grad_norm": 5.59375,
      "learning_rate": 4.934809237086059e-05,
      "loss": 0.4565,
      "num_input_tokens_seen": 2075632,
      "step": 3160
    },
    {
      "epoch": 1.658805031446541,
      "grad_norm": 11.6875,
      "learning_rate": 4.934289385941179e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 2079024,
      "step": 3165
    },
    {
      "epoch": 1.6614255765199162,
      "grad_norm": 5.03125,
      "learning_rate": 4.9337674978941364e-05,
      "loss": 0.6495,
      "num_input_tokens_seen": 2081744,
      "step": 3170
    },
    {
      "epoch": 1.6640461215932913,
      "grad_norm": 3.203125,
      "learning_rate": 4.9332435733816204e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 2084784,
      "step": 3175
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 6.75,
      "learning_rate": 4.932717612842028e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 2087984,
      "step": 3180
    },
    {
      "epoch": 1.669287211740042,
      "grad_norm": 3.859375,
      "learning_rate": 4.93218961671546e-05,
      "loss": 0.5018,
      "num_input_tokens_seen": 2091344,
      "step": 3185
    },
    {
      "epoch": 1.6719077568134173,
      "grad_norm": 2.796875,
      "learning_rate": 4.931659585443719e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 2095312,
      "step": 3190
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 3.78125,
      "learning_rate": 4.931127519470311e-05,
      "loss": 0.4307,
      "num_input_tokens_seen": 2098832,
      "step": 3195
    },
    {
      "epoch": 1.6771488469601676,
      "grad_norm": 9.1875,
      "learning_rate": 4.9305934192404426e-05,
      "loss": 0.6336,
      "num_input_tokens_seen": 2101168,
      "step": 3200
    },
    {
      "epoch": 1.679769392033543,
      "grad_norm": 6.28125,
      "learning_rate": 4.930057285201027e-05,
      "loss": 0.7414,
      "num_input_tokens_seen": 2104816,
      "step": 3205
    },
    {
      "epoch": 1.6823899371069182,
      "grad_norm": 7.15625,
      "learning_rate": 4.929519117800676e-05,
      "loss": 0.5132,
      "num_input_tokens_seen": 2107728,
      "step": 3210
    },
    {
      "epoch": 1.6850104821802936,
      "grad_norm": 14.75,
      "learning_rate": 4.928978917489703e-05,
      "loss": 0.7368,
      "num_input_tokens_seen": 2110992,
      "step": 3215
    },
    {
      "epoch": 1.6876310272536688,
      "grad_norm": 6.59375,
      "learning_rate": 4.928436684720122e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 2113840,
      "step": 3220
    },
    {
      "epoch": 1.690251572327044,
      "grad_norm": 4.0,
      "learning_rate": 4.927892419945651e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 2116784,
      "step": 3225
    },
    {
      "epoch": 1.6928721174004193,
      "grad_norm": 3.796875,
      "learning_rate": 4.927346123621705e-05,
      "loss": 0.6663,
      "num_input_tokens_seen": 2119696,
      "step": 3230
    },
    {
      "epoch": 1.6954926624737947,
      "grad_norm": 3.6875,
      "learning_rate": 4.926797796205399e-05,
      "loss": 0.5585,
      "num_input_tokens_seen": 2122608,
      "step": 3235
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 4.125,
      "learning_rate": 4.926247438155549e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 2125808,
      "step": 3240
    },
    {
      "epoch": 1.700733752620545,
      "grad_norm": 6.59375,
      "learning_rate": 4.9256950499326684e-05,
      "loss": 0.4318,
      "num_input_tokens_seen": 2129584,
      "step": 3245
    },
    {
      "epoch": 1.7033542976939202,
      "grad_norm": 7.15625,
      "learning_rate": 4.9251406319989725e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 2132208,
      "step": 3250
    },
    {
      "epoch": 1.7059748427672956,
      "grad_norm": 3.03125,
      "learning_rate": 4.9245841848183714e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 2135632,
      "step": 3255
    },
    {
      "epoch": 1.708595387840671,
      "grad_norm": 10.8125,
      "learning_rate": 4.924025708856475e-05,
      "loss": 0.7203,
      "num_input_tokens_seen": 2139760,
      "step": 3260
    },
    {
      "epoch": 1.7112159329140462,
      "grad_norm": 7.6875,
      "learning_rate": 4.9234652045805895e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 2143280,
      "step": 3265
    },
    {
      "epoch": 1.7138364779874213,
      "grad_norm": 4.75,
      "learning_rate": 4.922902672459722e-05,
      "loss": 0.4477,
      "num_input_tokens_seen": 2145648,
      "step": 3270
    },
    {
      "epoch": 1.7164570230607965,
      "grad_norm": 5.46875,
      "learning_rate": 4.9223381129645706e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 2149040,
      "step": 3275
    },
    {
      "epoch": 1.719077568134172,
      "grad_norm": 3.703125,
      "learning_rate": 4.921771526567535e-05,
      "loss": 0.5841,
      "num_input_tokens_seen": 2151792,
      "step": 3280
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 4.5625,
      "learning_rate": 4.921202913742707e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 2155600,
      "step": 3285
    },
    {
      "epoch": 1.7243186582809225,
      "grad_norm": 5.0625,
      "learning_rate": 4.920632274965878e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 2159088,
      "step": 3290
    },
    {
      "epoch": 1.7269392033542976,
      "grad_norm": 8.125,
      "learning_rate": 4.920059610714531e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 2162800,
      "step": 3295
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 9.0625,
      "learning_rate": 4.919484921467846e-05,
      "loss": 0.579,
      "num_input_tokens_seen": 2165872,
      "step": 3300
    },
    {
      "epoch": 1.7321802935010482,
      "grad_norm": 4.15625,
      "learning_rate": 4.9189082077066965e-05,
      "loss": 0.5915,
      "num_input_tokens_seen": 2169456,
      "step": 3305
    },
    {
      "epoch": 1.7348008385744236,
      "grad_norm": 6.21875,
      "learning_rate": 4.918329469913649e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 2172368,
      "step": 3310
    },
    {
      "epoch": 1.7374213836477987,
      "grad_norm": 3.203125,
      "learning_rate": 4.917748708572967e-05,
      "loss": 0.6261,
      "num_input_tokens_seen": 2175504,
      "step": 3315
    },
    {
      "epoch": 1.740041928721174,
      "grad_norm": 5.28125,
      "learning_rate": 4.917165924170604e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 2179440,
      "step": 3320
    },
    {
      "epoch": 1.7426624737945493,
      "grad_norm": 5.78125,
      "learning_rate": 4.9165811171942064e-05,
      "loss": 0.6104,
      "num_input_tokens_seen": 2181904,
      "step": 3325
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 5.21875,
      "learning_rate": 4.915994288133115e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 2184816,
      "step": 3330
    },
    {
      "epoch": 1.7479035639412999,
      "grad_norm": 13.3125,
      "learning_rate": 4.9154054374783624e-05,
      "loss": 0.6195,
      "num_input_tokens_seen": 2187920,
      "step": 3335
    },
    {
      "epoch": 1.750524109014675,
      "grad_norm": 2.859375,
      "learning_rate": 4.914814565722671e-05,
      "loss": 0.5422,
      "num_input_tokens_seen": 2191216,
      "step": 3340
    },
    {
      "epoch": 1.7531446540880502,
      "grad_norm": 6.71875,
      "learning_rate": 4.914221673360455e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 2193744,
      "step": 3345
    },
    {
      "epoch": 1.7557651991614256,
      "grad_norm": 6.15625,
      "learning_rate": 4.91362676088782e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 2196368,
      "step": 3350
    },
    {
      "epoch": 1.758385744234801,
      "grad_norm": 4.4375,
      "learning_rate": 4.913029828802561e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 2199632,
      "step": 3355
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 3.953125,
      "learning_rate": 4.912430877604165e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 2202704,
      "step": 3360
    },
    {
      "epoch": 1.7636268343815513,
      "grad_norm": 4.625,
      "learning_rate": 4.9118299077938054e-05,
      "loss": 0.6993,
      "num_input_tokens_seen": 2205616,
      "step": 3365
    },
    {
      "epoch": 1.7662473794549265,
      "grad_norm": 3.609375,
      "learning_rate": 4.911226919874347e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 2209008,
      "step": 3370
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 6.0625,
      "learning_rate": 4.910621914350343e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 2211824,
      "step": 3375
    },
    {
      "epoch": 1.7714884696016773,
      "grad_norm": 3.640625,
      "learning_rate": 4.910014891728033e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 2215632,
      "step": 3380
    },
    {
      "epoch": 1.7741090146750524,
      "grad_norm": 5.4375,
      "learning_rate": 4.9094058525153475e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 2218512,
      "step": 3385
    },
    {
      "epoch": 1.7767295597484276,
      "grad_norm": 3.234375,
      "learning_rate": 4.908794797221902e-05,
      "loss": 0.3981,
      "num_input_tokens_seen": 2221776,
      "step": 3390
    },
    {
      "epoch": 1.7793501048218028,
      "grad_norm": 4.0,
      "learning_rate": 4.908181726358999e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 2224784,
      "step": 3395
    },
    {
      "epoch": 1.7819706498951782,
      "grad_norm": 8.75,
      "learning_rate": 4.907566640439628e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 2227408,
      "step": 3400
    },
    {
      "epoch": 1.7845911949685536,
      "grad_norm": 5.78125,
      "learning_rate": 4.906949539978467e-05,
      "loss": 0.4652,
      "num_input_tokens_seen": 2230064,
      "step": 3405
    },
    {
      "epoch": 1.7872117400419287,
      "grad_norm": 4.78125,
      "learning_rate": 4.906330425491875e-05,
      "loss": 0.3846,
      "num_input_tokens_seen": 2234288,
      "step": 3410
    },
    {
      "epoch": 1.789832285115304,
      "grad_norm": 3.59375,
      "learning_rate": 4.9057092974979e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 2237072,
      "step": 3415
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 4.40625,
      "learning_rate": 4.905086156516273e-05,
      "loss": 0.3886,
      "num_input_tokens_seen": 2240016,
      "step": 3420
    },
    {
      "epoch": 1.7950733752620545,
      "grad_norm": 4.75,
      "learning_rate": 4.904461003068411e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 2242800,
      "step": 3425
    },
    {
      "epoch": 1.7976939203354299,
      "grad_norm": 7.90625,
      "learning_rate": 4.9038338376774124e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 2245680,
      "step": 3430
    },
    {
      "epoch": 1.800314465408805,
      "grad_norm": 4.75,
      "learning_rate": 4.9032046608680613e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 2248272,
      "step": 3435
    },
    {
      "epoch": 1.8029350104821802,
      "grad_norm": 18.875,
      "learning_rate": 4.902573473166824e-05,
      "loss": 0.3577,
      "num_input_tokens_seen": 2251376,
      "step": 3440
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 18.125,
      "learning_rate": 4.9019402751018496e-05,
      "loss": 0.5685,
      "num_input_tokens_seen": 2254256,
      "step": 3445
    },
    {
      "epoch": 1.808176100628931,
      "grad_norm": 7.5,
      "learning_rate": 4.901305067202969e-05,
      "loss": 0.591,
      "num_input_tokens_seen": 2257104,
      "step": 3450
    },
    {
      "epoch": 1.8107966457023061,
      "grad_norm": 5.8125,
      "learning_rate": 4.900667850001696e-05,
      "loss": 0.5705,
      "num_input_tokens_seen": 2259920,
      "step": 3455
    },
    {
      "epoch": 1.8134171907756813,
      "grad_norm": 4.15625,
      "learning_rate": 4.900028624031223e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 2262736,
      "step": 3460
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 2.75,
      "learning_rate": 4.899387389826427e-05,
      "loss": 0.4541,
      "num_input_tokens_seen": 2266864,
      "step": 3465
    },
    {
      "epoch": 1.8186582809224319,
      "grad_norm": 21.25,
      "learning_rate": 4.898744147923863e-05,
      "loss": 0.6929,
      "num_input_tokens_seen": 2269584,
      "step": 3470
    },
    {
      "epoch": 1.8212788259958073,
      "grad_norm": 3.46875,
      "learning_rate": 4.898098898861766e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 2272720,
      "step": 3475
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 4.375,
      "learning_rate": 4.897451643180051e-05,
      "loss": 0.5908,
      "num_input_tokens_seen": 2276880,
      "step": 3480
    },
    {
      "epoch": 1.8265199161425576,
      "grad_norm": 28.75,
      "learning_rate": 4.896802381420313e-05,
      "loss": 0.4585,
      "num_input_tokens_seen": 2280016,
      "step": 3485
    },
    {
      "epoch": 1.8291404612159328,
      "grad_norm": 4.75,
      "learning_rate": 4.896151114125823e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 2282640,
      "step": 3490
    },
    {
      "epoch": 1.8317610062893082,
      "grad_norm": 3.96875,
      "learning_rate": 4.895497841841533e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 2286032,
      "step": 3495
    },
    {
      "epoch": 1.8343815513626835,
      "grad_norm": 4.46875,
      "learning_rate": 4.8948425651140704e-05,
      "loss": 0.5759,
      "num_input_tokens_seen": 2289136,
      "step": 3500
    },
    {
      "epoch": 1.8370020964360587,
      "grad_norm": 5.46875,
      "learning_rate": 4.894185284491742e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 2292944,
      "step": 3505
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 6.78125,
      "learning_rate": 4.893526000524529e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 2296240,
      "step": 3510
    },
    {
      "epoch": 1.8422431865828093,
      "grad_norm": 8.0625,
      "learning_rate": 4.892864713764091e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 2299088,
      "step": 3515
    },
    {
      "epoch": 1.8448637316561844,
      "grad_norm": 6.71875,
      "learning_rate": 4.892201424763762e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 2302736,
      "step": 3520
    },
    {
      "epoch": 1.8474842767295598,
      "grad_norm": 1.8125,
      "learning_rate": 4.891536134078553e-05,
      "loss": 0.5975,
      "num_input_tokens_seen": 2306640,
      "step": 3525
    },
    {
      "epoch": 1.850104821802935,
      "grad_norm": 4.84375,
      "learning_rate": 4.8908688422651465e-05,
      "loss": 0.5665,
      "num_input_tokens_seen": 2309840,
      "step": 3530
    },
    {
      "epoch": 1.8527253668763102,
      "grad_norm": 10.9375,
      "learning_rate": 4.8901995498819044e-05,
      "loss": 0.5333,
      "num_input_tokens_seen": 2312208,
      "step": 3535
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 14.625,
      "learning_rate": 4.8895282574888576e-05,
      "loss": 0.621,
      "num_input_tokens_seen": 2315088,
      "step": 3540
    },
    {
      "epoch": 1.857966457023061,
      "grad_norm": 15.125,
      "learning_rate": 4.888854965647716e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 2318224,
      "step": 3545
    },
    {
      "epoch": 1.8605870020964361,
      "grad_norm": 1.921875,
      "learning_rate": 4.8881796749218564e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 2322800,
      "step": 3550
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 8.0,
      "learning_rate": 4.8875023858763335e-05,
      "loss": 0.3502,
      "num_input_tokens_seen": 2329008,
      "step": 3555
    },
    {
      "epoch": 1.8658280922431865,
      "grad_norm": 10.25,
      "learning_rate": 4.88682309907787e-05,
      "loss": 0.6685,
      "num_input_tokens_seen": 2331344,
      "step": 3560
    },
    {
      "epoch": 1.8684486373165619,
      "grad_norm": 4.0625,
      "learning_rate": 4.886141815094863e-05,
      "loss": 0.4898,
      "num_input_tokens_seen": 2334352,
      "step": 3565
    },
    {
      "epoch": 1.8710691823899372,
      "grad_norm": 3.984375,
      "learning_rate": 4.88545853449738e-05,
      "loss": 0.5722,
      "num_input_tokens_seen": 2338512,
      "step": 3570
    },
    {
      "epoch": 1.8736897274633124,
      "grad_norm": 7.15625,
      "learning_rate": 4.8847732578571585e-05,
      "loss": 0.5944,
      "num_input_tokens_seen": 2342224,
      "step": 3575
    },
    {
      "epoch": 1.8763102725366876,
      "grad_norm": 16.25,
      "learning_rate": 4.8840859857476074e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 2345808,
      "step": 3580
    },
    {
      "epoch": 1.8789308176100628,
      "grad_norm": 3.546875,
      "learning_rate": 4.8833967187438034e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 2348560,
      "step": 3585
    },
    {
      "epoch": 1.8815513626834381,
      "grad_norm": 5.59375,
      "learning_rate": 4.882705457422495e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 2351824,
      "step": 3590
    },
    {
      "epoch": 1.8841719077568135,
      "grad_norm": 6.5,
      "learning_rate": 4.8820122023620975e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 2355696,
      "step": 3595
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 8.6875,
      "learning_rate": 4.881316954142694e-05,
      "loss": 0.4398,
      "num_input_tokens_seen": 2358352,
      "step": 3600
    },
    {
      "epoch": 1.8894129979035639,
      "grad_norm": 2.5625,
      "learning_rate": 4.880619713346039e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 2361776,
      "step": 3605
    },
    {
      "epoch": 1.892033542976939,
      "grad_norm": 4.71875,
      "learning_rate": 4.879920480555549e-05,
      "loss": 0.5691,
      "num_input_tokens_seen": 2365264,
      "step": 3610
    },
    {
      "epoch": 1.8946540880503144,
      "grad_norm": 6.28125,
      "learning_rate": 4.8792192563563114e-05,
      "loss": 0.7132,
      "num_input_tokens_seen": 2368080,
      "step": 3615
    },
    {
      "epoch": 1.8972746331236898,
      "grad_norm": 2.765625,
      "learning_rate": 4.8785160413350797e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 2371696,
      "step": 3620
    },
    {
      "epoch": 1.899895178197065,
      "grad_norm": 4.1875,
      "learning_rate": 4.877810836080269e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 2374160,
      "step": 3625
    },
    {
      "epoch": 1.9025157232704402,
      "grad_norm": 7.8125,
      "learning_rate": 4.8771036411819656e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 2376656,
      "step": 3630
    },
    {
      "epoch": 1.9051362683438156,
      "grad_norm": 5.375,
      "learning_rate": 4.876394457231917e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 2379856,
      "step": 3635
    },
    {
      "epoch": 1.9077568134171907,
      "grad_norm": 2.46875,
      "learning_rate": 4.875683284823537e-05,
      "loss": 0.5938,
      "num_input_tokens_seen": 2382736,
      "step": 3640
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 7.71875,
      "learning_rate": 4.8749701245519e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 2386672,
      "step": 3645
    },
    {
      "epoch": 1.9129979035639413,
      "grad_norm": 6.84375,
      "learning_rate": 4.874254977013747e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 2389296,
      "step": 3650
    },
    {
      "epoch": 1.9156184486373165,
      "grad_norm": 5.71875,
      "learning_rate": 4.8735378428074806e-05,
      "loss": 0.5786,
      "num_input_tokens_seen": 2392080,
      "step": 3655
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 2.8125,
      "learning_rate": 4.8728187225331665e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 2395952,
      "step": 3660
    },
    {
      "epoch": 1.9208595387840672,
      "grad_norm": 4.34375,
      "learning_rate": 4.872097616792532e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 2398512,
      "step": 3665
    },
    {
      "epoch": 1.9234800838574424,
      "grad_norm": 4.125,
      "learning_rate": 4.871374526188964e-05,
      "loss": 0.5604,
      "num_input_tokens_seen": 2401904,
      "step": 3670
    },
    {
      "epoch": 1.9261006289308176,
      "grad_norm": 3.328125,
      "learning_rate": 4.8706494513275134e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 2405584,
      "step": 3675
    },
    {
      "epoch": 1.9287211740041927,
      "grad_norm": 2.703125,
      "learning_rate": 4.869922392814889e-05,
      "loss": 0.5969,
      "num_input_tokens_seen": 2409712,
      "step": 3680
    },
    {
      "epoch": 1.9313417190775681,
      "grad_norm": 6.875,
      "learning_rate": 4.869193351259459e-05,
      "loss": 0.6339,
      "num_input_tokens_seen": 2412240,
      "step": 3685
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 5.125,
      "learning_rate": 4.868462327271254e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 2415056,
      "step": 3690
    },
    {
      "epoch": 1.9365828092243187,
      "grad_norm": 6.53125,
      "learning_rate": 4.86772932146196e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 2419184,
      "step": 3695
    },
    {
      "epoch": 1.9392033542976939,
      "grad_norm": 5.28125,
      "learning_rate": 4.866994334444923e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 2422512,
      "step": 3700
    },
    {
      "epoch": 1.941823899371069,
      "grad_norm": 5.90625,
      "learning_rate": 4.866257366835147e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 2425936,
      "step": 3705
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 7.125,
      "learning_rate": 4.865518419249294e-05,
      "loss": 0.6088,
      "num_input_tokens_seen": 2428944,
      "step": 3710
    },
    {
      "epoch": 1.9470649895178198,
      "grad_norm": 3.796875,
      "learning_rate": 4.864777492305679e-05,
      "loss": 0.4055,
      "num_input_tokens_seen": 2432080,
      "step": 3715
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 5.09375,
      "learning_rate": 4.864034586624277e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 2434864,
      "step": 3720
    },
    {
      "epoch": 1.9523060796645701,
      "grad_norm": 9.375,
      "learning_rate": 4.863289702826719e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 2437104,
      "step": 3725
    },
    {
      "epoch": 1.9549266247379455,
      "grad_norm": 7.3125,
      "learning_rate": 4.862542841536288e-05,
      "loss": 0.6748,
      "num_input_tokens_seen": 2439792,
      "step": 3730
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 3.6875,
      "learning_rate": 4.861794003377923e-05,
      "loss": 0.6529,
      "num_input_tokens_seen": 2442320,
      "step": 3735
    },
    {
      "epoch": 1.960167714884696,
      "grad_norm": 5.875,
      "learning_rate": 4.8610431889782195e-05,
      "loss": 0.4518,
      "num_input_tokens_seen": 2445136,
      "step": 3740
    },
    {
      "epoch": 1.9627882599580713,
      "grad_norm": 3.859375,
      "learning_rate": 4.8602903989654224e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 2448080,
      "step": 3745
    },
    {
      "epoch": 1.9654088050314464,
      "grad_norm": 6.3125,
      "learning_rate": 4.859535633969434e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 2451376,
      "step": 3750
    },
    {
      "epoch": 1.9680293501048218,
      "grad_norm": 4.0625,
      "learning_rate": 4.858778894621807e-05,
      "loss": 0.6031,
      "num_input_tokens_seen": 2454672,
      "step": 3755
    },
    {
      "epoch": 1.9706498951781972,
      "grad_norm": 6.28125,
      "learning_rate": 4.858020181555745e-05,
      "loss": 0.4429,
      "num_input_tokens_seen": 2457712,
      "step": 3760
    },
    {
      "epoch": 1.9732704402515724,
      "grad_norm": 5.46875,
      "learning_rate": 4.857259495406105e-05,
      "loss": 0.4803,
      "num_input_tokens_seen": 2461936,
      "step": 3765
    },
    {
      "epoch": 1.9758909853249476,
      "grad_norm": 4.9375,
      "learning_rate": 4.856496836809394e-05,
      "loss": 0.4323,
      "num_input_tokens_seen": 2465296,
      "step": 3770
    },
    {
      "epoch": 1.9785115303983227,
      "grad_norm": 8.1875,
      "learning_rate": 4.8557322064037714e-05,
      "loss": 0.5203,
      "num_input_tokens_seen": 2469872,
      "step": 3775
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 4.875,
      "learning_rate": 4.854965604829044e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 2473840,
      "step": 3780
    },
    {
      "epoch": 1.9837526205450735,
      "grad_norm": 5.09375,
      "learning_rate": 4.8541970327266685e-05,
      "loss": 0.5126,
      "num_input_tokens_seen": 2476944,
      "step": 3785
    },
    {
      "epoch": 1.9863731656184487,
      "grad_norm": 5.4375,
      "learning_rate": 4.853426490739751e-05,
      "loss": 0.4844,
      "num_input_tokens_seen": 2479792,
      "step": 3790
    },
    {
      "epoch": 1.9889937106918238,
      "grad_norm": 6.1875,
      "learning_rate": 4.852653979513047e-05,
      "loss": 0.6113,
      "num_input_tokens_seen": 2484464,
      "step": 3795
    },
    {
      "epoch": 1.991614255765199,
      "grad_norm": 4.875,
      "learning_rate": 4.851879499692958e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 2487312,
      "step": 3800
    },
    {
      "epoch": 1.9942348008385744,
      "grad_norm": 9.25,
      "learning_rate": 4.851103051927532e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 2490288,
      "step": 3805
    },
    {
      "epoch": 1.9968553459119498,
      "grad_norm": 3.21875,
      "learning_rate": 4.850324636866468e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 2494128,
      "step": 3810
    },
    {
      "epoch": 1.999475890985325,
      "grad_norm": 5.875,
      "learning_rate": 4.849544255161106e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 2496848,
      "step": 3815
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5130470395088196,
      "eval_runtime": 24.1978,
      "eval_samples_per_second": 35.044,
      "eval_steps_per_second": 8.761,
      "num_input_tokens_seen": 2497016,
      "step": 3816
    },
    {
      "epoch": 2.0020964360587,
      "grad_norm": 5.53125,
      "learning_rate": 4.848761907464433e-05,
      "loss": 0.4101,
      "num_input_tokens_seen": 2499256,
      "step": 3820
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 5.90625,
      "learning_rate": 4.847977594431084e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 2502200,
      "step": 3825
    },
    {
      "epoch": 2.007337526205451,
      "grad_norm": 9.8125,
      "learning_rate": 4.847191316717335e-05,
      "loss": 0.6563,
      "num_input_tokens_seen": 2505080,
      "step": 3830
    },
    {
      "epoch": 2.009958071278826,
      "grad_norm": 5.71875,
      "learning_rate": 4.846403074981107e-05,
      "loss": 0.5142,
      "num_input_tokens_seen": 2508216,
      "step": 3835
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 5.3125,
      "learning_rate": 4.845612869881967e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 2511128,
      "step": 3840
    },
    {
      "epoch": 2.0151991614255764,
      "grad_norm": 4.09375,
      "learning_rate": 4.8448207020811194e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 2514776,
      "step": 3845
    },
    {
      "epoch": 2.0178197064989516,
      "grad_norm": 6.21875,
      "learning_rate": 4.8440265722414155e-05,
      "loss": 0.68,
      "num_input_tokens_seen": 2517528,
      "step": 3850
    },
    {
      "epoch": 2.020440251572327,
      "grad_norm": 3.328125,
      "learning_rate": 4.843230481027347e-05,
      "loss": 0.5495,
      "num_input_tokens_seen": 2523032,
      "step": 3855
    },
    {
      "epoch": 2.0230607966457024,
      "grad_norm": 4.84375,
      "learning_rate": 4.8424324291050464e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 2526200,
      "step": 3860
    },
    {
      "epoch": 2.0256813417190775,
      "grad_norm": 4.96875,
      "learning_rate": 4.841632417142287e-05,
      "loss": 0.5953,
      "num_input_tokens_seen": 2532760,
      "step": 3865
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 4.65625,
      "learning_rate": 4.840830445808483e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 2535640,
      "step": 3870
    },
    {
      "epoch": 2.030922431865828,
      "grad_norm": 3.75,
      "learning_rate": 4.840026515774686e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 2539288,
      "step": 3875
    },
    {
      "epoch": 2.0335429769392035,
      "grad_norm": 7.75,
      "learning_rate": 4.8392206277135896e-05,
      "loss": 0.5757,
      "num_input_tokens_seen": 2541880,
      "step": 3880
    },
    {
      "epoch": 2.0361635220125787,
      "grad_norm": 4.03125,
      "learning_rate": 4.8384127822995227e-05,
      "loss": 0.489,
      "num_input_tokens_seen": 2545080,
      "step": 3885
    },
    {
      "epoch": 2.038784067085954,
      "grad_norm": 3.046875,
      "learning_rate": 4.8376029802084546e-05,
      "loss": 0.3837,
      "num_input_tokens_seen": 2548376,
      "step": 3890
    },
    {
      "epoch": 2.041404612159329,
      "grad_norm": 4.75,
      "learning_rate": 4.836791222117989e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 2550616,
      "step": 3895
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 4.90625,
      "learning_rate": 4.83597750870737e-05,
      "loss": 0.603,
      "num_input_tokens_seen": 2555000,
      "step": 3900
    },
    {
      "epoch": 2.04664570230608,
      "grad_norm": 7.03125,
      "learning_rate": 4.8351618406574746e-05,
      "loss": 0.5226,
      "num_input_tokens_seen": 2558328,
      "step": 3905
    },
    {
      "epoch": 2.049266247379455,
      "grad_norm": 6.5,
      "learning_rate": 4.834344218650817e-05,
      "loss": 0.5849,
      "num_input_tokens_seen": 2561592,
      "step": 3910
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 3.390625,
      "learning_rate": 4.833524643371545e-05,
      "loss": 0.3651,
      "num_input_tokens_seen": 2564472,
      "step": 3915
    },
    {
      "epoch": 2.0545073375262053,
      "grad_norm": 9.625,
      "learning_rate": 4.8327031155054434e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 2567352,
      "step": 3920
    },
    {
      "epoch": 2.057127882599581,
      "grad_norm": 4.75,
      "learning_rate": 4.831879635739929e-05,
      "loss": 0.375,
      "num_input_tokens_seen": 2571448,
      "step": 3925
    },
    {
      "epoch": 2.059748427672956,
      "grad_norm": 7.40625,
      "learning_rate": 4.83105420476405e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 2574264,
      "step": 3930
    },
    {
      "epoch": 2.0623689727463312,
      "grad_norm": 5.21875,
      "learning_rate": 4.830226823268491e-05,
      "loss": 0.5984,
      "num_input_tokens_seen": 2577592,
      "step": 3935
    },
    {
      "epoch": 2.0649895178197064,
      "grad_norm": 6.5625,
      "learning_rate": 4.829397491945568e-05,
      "loss": 0.5156,
      "num_input_tokens_seen": 2581624,
      "step": 3940
    },
    {
      "epoch": 2.0676100628930816,
      "grad_norm": 10.875,
      "learning_rate": 4.828566211489225e-05,
      "loss": 0.5279,
      "num_input_tokens_seen": 2584888,
      "step": 3945
    },
    {
      "epoch": 2.070230607966457,
      "grad_norm": 10.5625,
      "learning_rate": 4.827732982595041e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 2587672,
      "step": 3950
    },
    {
      "epoch": 2.0728511530398324,
      "grad_norm": 6.1875,
      "learning_rate": 4.826897805960224e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 2590136,
      "step": 3955
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 5.09375,
      "learning_rate": 4.8260606822836116e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 2593976,
      "step": 3960
    },
    {
      "epoch": 2.0780922431865827,
      "grad_norm": 7.1875,
      "learning_rate": 4.8252216122656716e-05,
      "loss": 1.0014,
      "num_input_tokens_seen": 2597208,
      "step": 3965
    },
    {
      "epoch": 2.080712788259958,
      "grad_norm": 6.1875,
      "learning_rate": 4.824380596608497e-05,
      "loss": 0.522,
      "num_input_tokens_seen": 2600536,
      "step": 3970
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 5.0625,
      "learning_rate": 4.823537636015812e-05,
      "loss": 0.4355,
      "num_input_tokens_seen": 2603608,
      "step": 3975
    },
    {
      "epoch": 2.0859538784067087,
      "grad_norm": 7.625,
      "learning_rate": 4.822692731192969e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 2607288,
      "step": 3980
    },
    {
      "epoch": 2.088574423480084,
      "grad_norm": 4.34375,
      "learning_rate": 4.8218458828469445e-05,
      "loss": 0.6259,
      "num_input_tokens_seen": 2610392,
      "step": 3985
    },
    {
      "epoch": 2.091194968553459,
      "grad_norm": 3.703125,
      "learning_rate": 4.820997091686343e-05,
      "loss": 0.6229,
      "num_input_tokens_seen": 2613208,
      "step": 3990
    },
    {
      "epoch": 2.0938155136268346,
      "grad_norm": 4.40625,
      "learning_rate": 4.8201463584213946e-05,
      "loss": 0.6334,
      "num_input_tokens_seen": 2616472,
      "step": 3995
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 5.625,
      "learning_rate": 4.819293683763954e-05,
      "loss": 0.4964,
      "num_input_tokens_seen": 2620504,
      "step": 4000
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 3.859375,
      "learning_rate": 4.818439068427498e-05,
      "loss": 0.5624,
      "num_input_tokens_seen": 2624632,
      "step": 4005
    },
    {
      "epoch": 2.10167714884696,
      "grad_norm": 4.3125,
      "learning_rate": 4.817582513127133e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 2627608,
      "step": 4010
    },
    {
      "epoch": 2.1042976939203353,
      "grad_norm": 3.328125,
      "learning_rate": 4.8167240185795835e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 2630232,
      "step": 4015
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 2.359375,
      "learning_rate": 4.8158635855032e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 2633624,
      "step": 4020
    },
    {
      "epoch": 2.109538784067086,
      "grad_norm": 18.75,
      "learning_rate": 4.8150012146179514e-05,
      "loss": 0.5333,
      "num_input_tokens_seen": 2636280,
      "step": 4025
    },
    {
      "epoch": 2.1121593291404612,
      "grad_norm": 4.15625,
      "learning_rate": 4.814136906645431e-05,
      "loss": 0.5419,
      "num_input_tokens_seen": 2640152,
      "step": 4030
    },
    {
      "epoch": 2.1147798742138364,
      "grad_norm": 2.890625,
      "learning_rate": 4.813270662308854e-05,
      "loss": 0.4965,
      "num_input_tokens_seen": 2643800,
      "step": 4035
    },
    {
      "epoch": 2.1174004192872116,
      "grad_norm": 3.359375,
      "learning_rate": 4.812402482333052e-05,
      "loss": 0.4667,
      "num_input_tokens_seen": 2647608,
      "step": 4040
    },
    {
      "epoch": 2.120020964360587,
      "grad_norm": 3.796875,
      "learning_rate": 4.811532367444479e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 2650200,
      "step": 4045
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 4.875,
      "learning_rate": 4.810660318371208e-05,
      "loss": 0.6246,
      "num_input_tokens_seen": 2653784,
      "step": 4050
    },
    {
      "epoch": 2.1252620545073375,
      "grad_norm": 2.703125,
      "learning_rate": 4.809786335842929e-05,
      "loss": 0.5261,
      "num_input_tokens_seen": 2658104,
      "step": 4055
    },
    {
      "epoch": 2.1278825995807127,
      "grad_norm": 7.75,
      "learning_rate": 4.8089104205909506e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 2661080,
      "step": 4060
    },
    {
      "epoch": 2.130503144654088,
      "grad_norm": 3.484375,
      "learning_rate": 4.8080325733482004e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 2664312,
      "step": 4065
    },
    {
      "epoch": 2.1331236897274635,
      "grad_norm": 12.875,
      "learning_rate": 4.8071527948492176e-05,
      "loss": 0.5761,
      "num_input_tokens_seen": 2667704,
      "step": 4070
    },
    {
      "epoch": 2.1357442348008386,
      "grad_norm": 5.4375,
      "learning_rate": 4.806271085830164e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 2671160,
      "step": 4075
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 3.03125,
      "learning_rate": 4.805387447028812e-05,
      "loss": 0.5323,
      "num_input_tokens_seen": 2675544,
      "step": 4080
    },
    {
      "epoch": 2.140985324947589,
      "grad_norm": 6.65625,
      "learning_rate": 4.80450187918455e-05,
      "loss": 0.4791,
      "num_input_tokens_seen": 2678264,
      "step": 4085
    },
    {
      "epoch": 2.1436058700209646,
      "grad_norm": 6.8125,
      "learning_rate": 4.8036143830383807e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 2684120,
      "step": 4090
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 5.1875,
      "learning_rate": 4.8027249593329206e-05,
      "loss": 0.6785,
      "num_input_tokens_seen": 2686872,
      "step": 4095
    },
    {
      "epoch": 2.148846960167715,
      "grad_norm": 4.3125,
      "learning_rate": 4.8018336088123986e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 2690456,
      "step": 4100
    },
    {
      "epoch": 2.15146750524109,
      "grad_norm": 3.78125,
      "learning_rate": 4.800940332222656e-05,
      "loss": 0.4895,
      "num_input_tokens_seen": 2692824,
      "step": 4105
    },
    {
      "epoch": 2.1540880503144653,
      "grad_norm": 4.90625,
      "learning_rate": 4.8000451303111474e-05,
      "loss": 0.6011,
      "num_input_tokens_seen": 2695704,
      "step": 4110
    },
    {
      "epoch": 2.156708595387841,
      "grad_norm": 4.3125,
      "learning_rate": 4.799148003826936e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 2699480,
      "step": 4115
    },
    {
      "epoch": 2.159329140461216,
      "grad_norm": 5.71875,
      "learning_rate": 4.798248953520694e-05,
      "loss": 0.5521,
      "num_input_tokens_seen": 2703960,
      "step": 4120
    },
    {
      "epoch": 2.161949685534591,
      "grad_norm": 5.15625,
      "learning_rate": 4.7973479801447084e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 2706552,
      "step": 4125
    },
    {
      "epoch": 2.1645702306079664,
      "grad_norm": 4.75,
      "learning_rate": 4.796445084452871e-05,
      "loss": 0.5469,
      "num_input_tokens_seen": 2709368,
      "step": 4130
    },
    {
      "epoch": 2.1671907756813416,
      "grad_norm": 5.34375,
      "learning_rate": 4.7955402672006854e-05,
      "loss": 0.5383,
      "num_input_tokens_seen": 2712248,
      "step": 4135
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 3.84375,
      "learning_rate": 4.794633529145259e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 2715448,
      "step": 4140
    },
    {
      "epoch": 2.1724318658280923,
      "grad_norm": 2.71875,
      "learning_rate": 4.793724871045312e-05,
      "loss": 0.402,
      "num_input_tokens_seen": 2719064,
      "step": 4145
    },
    {
      "epoch": 2.1750524109014675,
      "grad_norm": 3.546875,
      "learning_rate": 4.792814293661164e-05,
      "loss": 0.4645,
      "num_input_tokens_seen": 2722488,
      "step": 4150
    },
    {
      "epoch": 2.1776729559748427,
      "grad_norm": 5.46875,
      "learning_rate": 4.791901797754748e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 2725528,
      "step": 4155
    },
    {
      "epoch": 2.180293501048218,
      "grad_norm": 2.34375,
      "learning_rate": 4.790987384089597e-05,
      "loss": 0.4426,
      "num_input_tokens_seen": 2728824,
      "step": 4160
    },
    {
      "epoch": 2.1829140461215935,
      "grad_norm": 4.875,
      "learning_rate": 4.790071053430851e-05,
      "loss": 0.6082,
      "num_input_tokens_seen": 2732472,
      "step": 4165
    },
    {
      "epoch": 2.1855345911949686,
      "grad_norm": 3.171875,
      "learning_rate": 4.7891528065452544e-05,
      "loss": 0.5059,
      "num_input_tokens_seen": 2735544,
      "step": 4170
    },
    {
      "epoch": 2.188155136268344,
      "grad_norm": 4.8125,
      "learning_rate": 4.788232644201153e-05,
      "loss": 0.5145,
      "num_input_tokens_seen": 2739032,
      "step": 4175
    },
    {
      "epoch": 2.190775681341719,
      "grad_norm": 3.9375,
      "learning_rate": 4.787310567168498e-05,
      "loss": 0.6504,
      "num_input_tokens_seen": 2742456,
      "step": 4180
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 7.625,
      "learning_rate": 4.78638657621884e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 2744888,
      "step": 4185
    },
    {
      "epoch": 2.1960167714884697,
      "grad_norm": 4.78125,
      "learning_rate": 4.785460672125332e-05,
      "loss": 0.6102,
      "num_input_tokens_seen": 2747512,
      "step": 4190
    },
    {
      "epoch": 2.198637316561845,
      "grad_norm": 13.0,
      "learning_rate": 4.7845328556627306e-05,
      "loss": 0.5433,
      "num_input_tokens_seen": 2750648,
      "step": 4195
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 5.3125,
      "learning_rate": 4.783603127607388e-05,
      "loss": 0.404,
      "num_input_tokens_seen": 2753368,
      "step": 4200
    },
    {
      "epoch": 2.2038784067085953,
      "grad_norm": 6.6875,
      "learning_rate": 4.78267148873726e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 2756696,
      "step": 4205
    },
    {
      "epoch": 2.2064989517819704,
      "grad_norm": 9.8125,
      "learning_rate": 4.781737939831898e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 2759256,
      "step": 4210
    },
    {
      "epoch": 2.209119496855346,
      "grad_norm": 8.9375,
      "learning_rate": 4.7808024816724536e-05,
      "loss": 0.551,
      "num_input_tokens_seen": 2762296,
      "step": 4215
    },
    {
      "epoch": 2.211740041928721,
      "grad_norm": 5.9375,
      "learning_rate": 4.7798651150416754e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 2765400,
      "step": 4220
    },
    {
      "epoch": 2.2143605870020964,
      "grad_norm": 6.40625,
      "learning_rate": 4.778925840723909e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 2768120,
      "step": 4225
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 5.28125,
      "learning_rate": 4.777984659505096e-05,
      "loss": 0.5421,
      "num_input_tokens_seen": 2771096,
      "step": 4230
    },
    {
      "epoch": 2.219601677148847,
      "grad_norm": 5.46875,
      "learning_rate": 4.777041572172774e-05,
      "loss": 0.5678,
      "num_input_tokens_seen": 2774328,
      "step": 4235
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 3.015625,
      "learning_rate": 4.776096579516076e-05,
      "loss": 0.5539,
      "num_input_tokens_seen": 2778008,
      "step": 4240
    },
    {
      "epoch": 2.2248427672955975,
      "grad_norm": 38.75,
      "learning_rate": 4.775149682325728e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 2780216,
      "step": 4245
    },
    {
      "epoch": 2.2274633123689727,
      "grad_norm": 7.3125,
      "learning_rate": 4.77420088139405e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 2784088,
      "step": 4250
    },
    {
      "epoch": 2.230083857442348,
      "grad_norm": 6.46875,
      "learning_rate": 4.7732501775149564e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 2786488,
      "step": 4255
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 6.0,
      "learning_rate": 4.7722975714839526e-05,
      "loss": 0.4165,
      "num_input_tokens_seen": 2788920,
      "step": 4260
    },
    {
      "epoch": 2.2353249475890986,
      "grad_norm": 6.09375,
      "learning_rate": 4.7713430640981346e-05,
      "loss": 0.6698,
      "num_input_tokens_seen": 2791640,
      "step": 4265
    },
    {
      "epoch": 2.237945492662474,
      "grad_norm": 6.1875,
      "learning_rate": 4.7703866561561915e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 2794904,
      "step": 4270
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 4.78125,
      "learning_rate": 4.769428348458402e-05,
      "loss": 0.4361,
      "num_input_tokens_seen": 2798840,
      "step": 4275
    },
    {
      "epoch": 2.243186582809224,
      "grad_norm": 3.234375,
      "learning_rate": 4.7684681418066334e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 2802616,
      "step": 4280
    },
    {
      "epoch": 2.2458071278825997,
      "grad_norm": 8.5,
      "learning_rate": 4.767506037004344e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 2805432,
      "step": 4285
    },
    {
      "epoch": 2.248427672955975,
      "grad_norm": 5.96875,
      "learning_rate": 4.766542034856577e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 2808696,
      "step": 4290
    },
    {
      "epoch": 2.25104821802935,
      "grad_norm": 4.1875,
      "learning_rate": 4.7655761361699676e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 2811640,
      "step": 4295
    },
    {
      "epoch": 2.2536687631027252,
      "grad_norm": 6.40625,
      "learning_rate": 4.7646083417527345e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 2815224,
      "step": 4300
    },
    {
      "epoch": 2.2562893081761004,
      "grad_norm": 5.15625,
      "learning_rate": 4.7636386524146846e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 2817688,
      "step": 4305
    },
    {
      "epoch": 2.258909853249476,
      "grad_norm": 4.9375,
      "learning_rate": 4.7626670689672095e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 2821016,
      "step": 4310
    },
    {
      "epoch": 2.261530398322851,
      "grad_norm": 9.0,
      "learning_rate": 4.761693592223285e-05,
      "loss": 0.5853,
      "num_input_tokens_seen": 2823608,
      "step": 4315
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 2.234375,
      "learning_rate": 4.760718222997472e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 2828248,
      "step": 4320
    },
    {
      "epoch": 2.2667714884696015,
      "grad_norm": 3.671875,
      "learning_rate": 4.7597409621059164e-05,
      "loss": 0.5348,
      "num_input_tokens_seen": 2832344,
      "step": 4325
    },
    {
      "epoch": 2.269392033542977,
      "grad_norm": 4.40625,
      "learning_rate": 4.7587618103663444e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 2836408,
      "step": 4330
    },
    {
      "epoch": 2.2720125786163523,
      "grad_norm": 3.484375,
      "learning_rate": 4.757780768598066e-05,
      "loss": 0.6164,
      "num_input_tokens_seen": 2839320,
      "step": 4335
    },
    {
      "epoch": 2.2746331236897275,
      "grad_norm": 9.9375,
      "learning_rate": 4.756797837621971e-05,
      "loss": 0.5889,
      "num_input_tokens_seen": 2842456,
      "step": 4340
    },
    {
      "epoch": 2.2772536687631026,
      "grad_norm": 12.25,
      "learning_rate": 4.755813018260532e-05,
      "loss": 0.6463,
      "num_input_tokens_seen": 2845432,
      "step": 4345
    },
    {
      "epoch": 2.279874213836478,
      "grad_norm": 3.953125,
      "learning_rate": 4.754826311337801e-05,
      "loss": 0.5626,
      "num_input_tokens_seen": 2848536,
      "step": 4350
    },
    {
      "epoch": 2.2824947589098534,
      "grad_norm": 2.484375,
      "learning_rate": 4.753837717679409e-05,
      "loss": 0.5942,
      "num_input_tokens_seen": 2851544,
      "step": 4355
    },
    {
      "epoch": 2.2851153039832286,
      "grad_norm": 4.5,
      "learning_rate": 4.7528472381125653e-05,
      "loss": 0.4852,
      "num_input_tokens_seen": 2854392,
      "step": 4360
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 3.609375,
      "learning_rate": 4.75185487346606e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 2858424,
      "step": 4365
    },
    {
      "epoch": 2.290356394129979,
      "grad_norm": 4.28125,
      "learning_rate": 4.750860624570256e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 2861368,
      "step": 4370
    },
    {
      "epoch": 2.2929769392033545,
      "grad_norm": 6.75,
      "learning_rate": 4.7498644922570966e-05,
      "loss": 0.5311,
      "num_input_tokens_seen": 2863960,
      "step": 4375
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 8.125,
      "learning_rate": 4.7488664773601004e-05,
      "loss": 0.5903,
      "num_input_tokens_seen": 2867352,
      "step": 4380
    },
    {
      "epoch": 2.298218029350105,
      "grad_norm": 5.78125,
      "learning_rate": 4.7478665807143605e-05,
      "loss": 0.5526,
      "num_input_tokens_seen": 2871992,
      "step": 4385
    },
    {
      "epoch": 2.30083857442348,
      "grad_norm": 6.875,
      "learning_rate": 4.7468648031565434e-05,
      "loss": 0.4712,
      "num_input_tokens_seen": 2874680,
      "step": 4390
    },
    {
      "epoch": 2.3034591194968552,
      "grad_norm": 6.4375,
      "learning_rate": 4.745861145524892e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 2877688,
      "step": 4395
    },
    {
      "epoch": 2.3060796645702304,
      "grad_norm": 6.59375,
      "learning_rate": 4.74485560865922e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 2885528,
      "step": 4400
    },
    {
      "epoch": 2.308700209643606,
      "grad_norm": 2.8125,
      "learning_rate": 4.743848193400917e-05,
      "loss": 0.6239,
      "num_input_tokens_seen": 2889464,
      "step": 4405
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 4.71875,
      "learning_rate": 4.7428389005929405e-05,
      "loss": 0.3769,
      "num_input_tokens_seen": 2893464,
      "step": 4410
    },
    {
      "epoch": 2.3139412997903563,
      "grad_norm": 5.84375,
      "learning_rate": 4.74182773107982e-05,
      "loss": 0.557,
      "num_input_tokens_seen": 2896792,
      "step": 4415
    },
    {
      "epoch": 2.3165618448637315,
      "grad_norm": 7.5,
      "learning_rate": 4.7408146857076566e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 2900664,
      "step": 4420
    },
    {
      "epoch": 2.319182389937107,
      "grad_norm": 5.34375,
      "learning_rate": 4.739799765324121e-05,
      "loss": 0.5687,
      "num_input_tokens_seen": 2904152,
      "step": 4425
    },
    {
      "epoch": 2.3218029350104823,
      "grad_norm": 3.84375,
      "learning_rate": 4.738782970778452e-05,
      "loss": 0.5208,
      "num_input_tokens_seen": 2906776,
      "step": 4430
    },
    {
      "epoch": 2.3244234800838575,
      "grad_norm": 3.1875,
      "learning_rate": 4.737764302921456e-05,
      "loss": 0.5292,
      "num_input_tokens_seen": 2909720,
      "step": 4435
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 5.21875,
      "learning_rate": 4.7367437626055087e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 2916536,
      "step": 4440
    },
    {
      "epoch": 2.329664570230608,
      "grad_norm": 7.09375,
      "learning_rate": 4.735721350684551e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 2919608,
      "step": 4445
    },
    {
      "epoch": 2.3322851153039834,
      "grad_norm": 3.859375,
      "learning_rate": 4.734697068014091e-05,
      "loss": 0.6359,
      "num_input_tokens_seen": 2923352,
      "step": 4450
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 8.3125,
      "learning_rate": 4.733670915451202e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 2926232,
      "step": 4455
    },
    {
      "epoch": 2.3375262054507338,
      "grad_norm": 4.96875,
      "learning_rate": 4.732642893854519e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 2929656,
      "step": 4460
    },
    {
      "epoch": 2.340146750524109,
      "grad_norm": 6.59375,
      "learning_rate": 4.7316130040842466e-05,
      "loss": 0.5841,
      "num_input_tokens_seen": 2932344,
      "step": 4465
    },
    {
      "epoch": 2.342767295597484,
      "grad_norm": 2.421875,
      "learning_rate": 4.730581247002148e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 2936152,
      "step": 4470
    },
    {
      "epoch": 2.3453878406708597,
      "grad_norm": 4.53125,
      "learning_rate": 4.7295476234715516e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 2939224,
      "step": 4475
    },
    {
      "epoch": 2.348008385744235,
      "grad_norm": 4.09375,
      "learning_rate": 4.728512134357345e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 2941912,
      "step": 4480
    },
    {
      "epoch": 2.35062893081761,
      "grad_norm": 4.125,
      "learning_rate": 4.727474780525979e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 2944568,
      "step": 4485
    },
    {
      "epoch": 2.353249475890985,
      "grad_norm": 3.796875,
      "learning_rate": 4.7264355628454636e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 2947480,
      "step": 4490
    },
    {
      "epoch": 2.3558700209643604,
      "grad_norm": 7.03125,
      "learning_rate": 4.7253944821853685e-05,
      "loss": 0.5529,
      "num_input_tokens_seen": 2951192,
      "step": 4495
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 5.8125,
      "learning_rate": 4.724351539416822e-05,
      "loss": 0.5199,
      "num_input_tokens_seen": 2954648,
      "step": 4500
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 3.765625,
      "learning_rate": 4.7233067354125125e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 2958552,
      "step": 4505
    },
    {
      "epoch": 2.3637316561844863,
      "grad_norm": 2.421875,
      "learning_rate": 4.722260071046683e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 2962040,
      "step": 4510
    },
    {
      "epoch": 2.3663522012578615,
      "grad_norm": 3.390625,
      "learning_rate": 4.721211547195136e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 2964824,
      "step": 4515
    },
    {
      "epoch": 2.368972746331237,
      "grad_norm": 2.671875,
      "learning_rate": 4.7201611647352264e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 2967448,
      "step": 4520
    },
    {
      "epoch": 2.3715932914046123,
      "grad_norm": 2.671875,
      "learning_rate": 4.719108924545866e-05,
      "loss": 0.5865,
      "num_input_tokens_seen": 2971000,
      "step": 4525
    },
    {
      "epoch": 2.3742138364779874,
      "grad_norm": 4.71875,
      "learning_rate": 4.718054827507524e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 2973592,
      "step": 4530
    },
    {
      "epoch": 2.3768343815513626,
      "grad_norm": 4.1875,
      "learning_rate": 4.716998874502218e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 2977560,
      "step": 4535
    },
    {
      "epoch": 2.379454926624738,
      "grad_norm": 5.28125,
      "learning_rate": 4.7159410664135225e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 2980280,
      "step": 4540
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 4.96875,
      "learning_rate": 4.714881404126563e-05,
      "loss": 0.5003,
      "num_input_tokens_seen": 2983352,
      "step": 4545
    },
    {
      "epoch": 2.3846960167714886,
      "grad_norm": 2.390625,
      "learning_rate": 4.713819888528016e-05,
      "loss": 0.5341,
      "num_input_tokens_seen": 2986648,
      "step": 4550
    },
    {
      "epoch": 2.3873165618448637,
      "grad_norm": 6.75,
      "learning_rate": 4.7127565205061096e-05,
      "loss": 0.4813,
      "num_input_tokens_seen": 2989880,
      "step": 4555
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 4.21875,
      "learning_rate": 4.711691300950622e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 2992536,
      "step": 4560
    },
    {
      "epoch": 2.392557651991614,
      "grad_norm": 17.875,
      "learning_rate": 4.710624230752879e-05,
      "loss": 0.5195,
      "num_input_tokens_seen": 2996152,
      "step": 4565
    },
    {
      "epoch": 2.3951781970649897,
      "grad_norm": 12.6875,
      "learning_rate": 4.709555310805758e-05,
      "loss": 0.494,
      "num_input_tokens_seen": 2999288,
      "step": 4570
    },
    {
      "epoch": 2.397798742138365,
      "grad_norm": 5.8125,
      "learning_rate": 4.7084845420036805e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 3002520,
      "step": 4575
    },
    {
      "epoch": 2.40041928721174,
      "grad_norm": 4.125,
      "learning_rate": 4.7074119252426175e-05,
      "loss": 0.5929,
      "num_input_tokens_seen": 3005752,
      "step": 4580
    },
    {
      "epoch": 2.403039832285115,
      "grad_norm": 4.8125,
      "learning_rate": 4.7063374614200866e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 3009432,
      "step": 4585
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 4.46875,
      "learning_rate": 4.7052611514351495e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 3014136,
      "step": 4590
    },
    {
      "epoch": 2.408280922431866,
      "grad_norm": 5.21875,
      "learning_rate": 4.704182996188413e-05,
      "loss": 0.5324,
      "num_input_tokens_seen": 3017144,
      "step": 4595
    },
    {
      "epoch": 2.410901467505241,
      "grad_norm": 45.75,
      "learning_rate": 4.703102996582028e-05,
      "loss": 0.6111,
      "num_input_tokens_seen": 3019448,
      "step": 4600
    },
    {
      "epoch": 2.4135220125786163,
      "grad_norm": 5.90625,
      "learning_rate": 4.70202115351969e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 3022104,
      "step": 4605
    },
    {
      "epoch": 2.4161425576519915,
      "grad_norm": 2.890625,
      "learning_rate": 4.700937467906634e-05,
      "loss": 0.546,
      "num_input_tokens_seen": 3025688,
      "step": 4610
    },
    {
      "epoch": 2.418763102725367,
      "grad_norm": 6.03125,
      "learning_rate": 4.69985194064964e-05,
      "loss": 0.3645,
      "num_input_tokens_seen": 3029816,
      "step": 4615
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 3.421875,
      "learning_rate": 4.698764572657029e-05,
      "loss": 0.3671,
      "num_input_tokens_seen": 3032824,
      "step": 4620
    },
    {
      "epoch": 2.4240041928721174,
      "grad_norm": 4.0,
      "learning_rate": 4.697675364838657e-05,
      "loss": 0.6176,
      "num_input_tokens_seen": 3036632,
      "step": 4625
    },
    {
      "epoch": 2.4266247379454926,
      "grad_norm": 6.625,
      "learning_rate": 4.6965843181059264e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 3039384,
      "step": 4630
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 5.25,
      "learning_rate": 4.695491433371774e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 3042296,
      "step": 4635
    },
    {
      "epoch": 2.431865828092243,
      "grad_norm": 4.40625,
      "learning_rate": 4.694396711550676e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 3045304,
      "step": 4640
    },
    {
      "epoch": 2.4344863731656186,
      "grad_norm": 5.15625,
      "learning_rate": 4.693300153558646e-05,
      "loss": 0.4005,
      "num_input_tokens_seen": 3048824,
      "step": 4645
    },
    {
      "epoch": 2.4371069182389937,
      "grad_norm": 3.9375,
      "learning_rate": 4.692201760313233e-05,
      "loss": 0.5492,
      "num_input_tokens_seen": 3052120,
      "step": 4650
    },
    {
      "epoch": 2.439727463312369,
      "grad_norm": 4.84375,
      "learning_rate": 4.691101532733524e-05,
      "loss": 0.4975,
      "num_input_tokens_seen": 3055384,
      "step": 4655
    },
    {
      "epoch": 2.442348008385744,
      "grad_norm": 3.890625,
      "learning_rate": 4.689999471740137e-05,
      "loss": 0.391,
      "num_input_tokens_seen": 3058328,
      "step": 4660
    },
    {
      "epoch": 2.4449685534591197,
      "grad_norm": 4.34375,
      "learning_rate": 4.6888955782552274e-05,
      "loss": 0.5582,
      "num_input_tokens_seen": 3063512,
      "step": 4665
    },
    {
      "epoch": 2.447589098532495,
      "grad_norm": 3.796875,
      "learning_rate": 4.6877898532024825e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 3068056,
      "step": 4670
    },
    {
      "epoch": 2.45020964360587,
      "grad_norm": 7.03125,
      "learning_rate": 4.686682297507123e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 3071608,
      "step": 4675
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 4.9375,
      "learning_rate": 4.6855729120959e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 3074872,
      "step": 4680
    },
    {
      "epoch": 2.4554507337526204,
      "grad_norm": 3.71875,
      "learning_rate": 4.684461697897098e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 3077720,
      "step": 4685
    },
    {
      "epoch": 2.458071278825996,
      "grad_norm": 5.4375,
      "learning_rate": 4.683348655840529e-05,
      "loss": 0.6065,
      "num_input_tokens_seen": 3080664,
      "step": 4690
    },
    {
      "epoch": 2.460691823899371,
      "grad_norm": 9.125,
      "learning_rate": 4.682233786857536e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 3084408,
      "step": 4695
    },
    {
      "epoch": 2.4633123689727463,
      "grad_norm": 2.734375,
      "learning_rate": 4.681117091880991e-05,
      "loss": 0.5469,
      "num_input_tokens_seen": 3087640,
      "step": 4700
    },
    {
      "epoch": 2.4659329140461215,
      "grad_norm": 5.5,
      "learning_rate": 4.679998571845293e-05,
      "loss": 0.5708,
      "num_input_tokens_seen": 3090520,
      "step": 4705
    },
    {
      "epoch": 2.468553459119497,
      "grad_norm": 6.1875,
      "learning_rate": 4.678878227686368e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 3093624,
      "step": 4710
    },
    {
      "epoch": 2.4711740041928723,
      "grad_norm": 4.53125,
      "learning_rate": 4.677756060341669e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 3096184,
      "step": 4715
    },
    {
      "epoch": 2.4737945492662474,
      "grad_norm": 3.96875,
      "learning_rate": 4.676632070750175e-05,
      "loss": 0.6024,
      "num_input_tokens_seen": 3099928,
      "step": 4720
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 9.0,
      "learning_rate": 4.6755062598523894e-05,
      "loss": 0.5137,
      "num_input_tokens_seen": 3102456,
      "step": 4725
    },
    {
      "epoch": 2.4790356394129978,
      "grad_norm": 22.375,
      "learning_rate": 4.674378628590338e-05,
      "loss": 0.6002,
      "num_input_tokens_seen": 3104728,
      "step": 4730
    },
    {
      "epoch": 2.481656184486373,
      "grad_norm": 7.8125,
      "learning_rate": 4.673249177907571e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 3107704,
      "step": 4735
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 3.46875,
      "learning_rate": 4.672117908749164e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 3110328,
      "step": 4740
    },
    {
      "epoch": 2.4868972746331237,
      "grad_norm": 4.65625,
      "learning_rate": 4.670984822061708e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 3113464,
      "step": 4745
    },
    {
      "epoch": 2.489517819706499,
      "grad_norm": 2.859375,
      "learning_rate": 4.6698499187933196e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 3117112,
      "step": 4750
    },
    {
      "epoch": 2.492138364779874,
      "grad_norm": 4.375,
      "learning_rate": 4.668713199893635e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 3120280,
      "step": 4755
    },
    {
      "epoch": 2.4947589098532497,
      "grad_norm": 4.21875,
      "learning_rate": 4.6675746663138066e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 3123928,
      "step": 4760
    },
    {
      "epoch": 2.497379454926625,
      "grad_norm": 4.40625,
      "learning_rate": 4.666434319006508e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 3126584,
      "step": 4765
    },
    {
      "epoch": 2.5,
      "grad_norm": 8.8125,
      "learning_rate": 4.66529215892593e-05,
      "loss": 0.6237,
      "num_input_tokens_seen": 3129368,
      "step": 4770
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.507256031036377,
      "eval_runtime": 24.1997,
      "eval_samples_per_second": 35.042,
      "eval_steps_per_second": 8.76,
      "num_input_tokens_seen": 3129368,
      "step": 4770
    },
    {
      "epoch": 2.502620545073375,
      "grad_norm": 13.375,
      "learning_rate": 4.664148187027781e-05,
      "loss": 0.5804,
      "num_input_tokens_seen": 3132248,
      "step": 4775
    },
    {
      "epoch": 2.5052410901467503,
      "grad_norm": 3.6875,
      "learning_rate": 4.663002404269283e-05,
      "loss": 0.381,
      "num_input_tokens_seen": 3135320,
      "step": 4780
    },
    {
      "epoch": 2.507861635220126,
      "grad_norm": 5.375,
      "learning_rate": 4.661854811609174e-05,
      "loss": 0.7114,
      "num_input_tokens_seen": 3139032,
      "step": 4785
    },
    {
      "epoch": 2.510482180293501,
      "grad_norm": 2.890625,
      "learning_rate": 4.6607054100077096e-05,
      "loss": 0.6451,
      "num_input_tokens_seen": 3141944,
      "step": 4790
    },
    {
      "epoch": 2.5131027253668763,
      "grad_norm": 6.90625,
      "learning_rate": 4.6595542004266545e-05,
      "loss": 0.5459,
      "num_input_tokens_seen": 3145400,
      "step": 4795
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 3.0,
      "learning_rate": 4.65840118382929e-05,
      "loss": 0.5898,
      "num_input_tokens_seen": 3149496,
      "step": 4800
    },
    {
      "epoch": 2.518343815513627,
      "grad_norm": 4.03125,
      "learning_rate": 4.657246361180405e-05,
      "loss": 0.6125,
      "num_input_tokens_seen": 3152344,
      "step": 4805
    },
    {
      "epoch": 2.5209643605870022,
      "grad_norm": 4.375,
      "learning_rate": 4.656089733446305e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 3155192,
      "step": 4810
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 4.5625,
      "learning_rate": 4.6549313015948025e-05,
      "loss": 0.4002,
      "num_input_tokens_seen": 3157688,
      "step": 4815
    },
    {
      "epoch": 2.5262054507337526,
      "grad_norm": 2.015625,
      "learning_rate": 4.653771066595219e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 3161592,
      "step": 4820
    },
    {
      "epoch": 2.5288259958071277,
      "grad_norm": 9.375,
      "learning_rate": 4.652609029418389e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 3164024,
      "step": 4825
    },
    {
      "epoch": 2.531446540880503,
      "grad_norm": 7.53125,
      "learning_rate": 4.65144519103665e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 3166424,
      "step": 4830
    },
    {
      "epoch": 2.5340670859538785,
      "grad_norm": 9.25,
      "learning_rate": 4.650279552423849e-05,
      "loss": 0.6924,
      "num_input_tokens_seen": 3169240,
      "step": 4835
    },
    {
      "epoch": 2.5366876310272537,
      "grad_norm": 6.15625,
      "learning_rate": 4.6491121145553386e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 3172088,
      "step": 4840
    },
    {
      "epoch": 2.539308176100629,
      "grad_norm": 5.3125,
      "learning_rate": 4.6479428784079796e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 3175480,
      "step": 4845
    },
    {
      "epoch": 2.541928721174004,
      "grad_norm": 3.75,
      "learning_rate": 4.6467718449601326e-05,
      "loss": 0.5292,
      "num_input_tokens_seen": 3179480,
      "step": 4850
    },
    {
      "epoch": 2.5445492662473796,
      "grad_norm": 3.734375,
      "learning_rate": 4.645599015191667e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 3182520,
      "step": 4855
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 5.3125,
      "learning_rate": 4.6444243900839525e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 3185560,
      "step": 4860
    },
    {
      "epoch": 2.54979035639413,
      "grad_norm": 3.359375,
      "learning_rate": 4.643247970619862e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 3189592,
      "step": 4865
    },
    {
      "epoch": 2.552410901467505,
      "grad_norm": 11.625,
      "learning_rate": 4.642069757783769e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 3192760,
      "step": 4870
    },
    {
      "epoch": 2.5550314465408803,
      "grad_norm": 10.1875,
      "learning_rate": 4.640889752561549e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 3195992,
      "step": 4875
    },
    {
      "epoch": 2.5576519916142555,
      "grad_norm": 5.28125,
      "learning_rate": 4.639707955940575e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 3198200,
      "step": 4880
    },
    {
      "epoch": 2.560272536687631,
      "grad_norm": 4.71875,
      "learning_rate": 4.6385243689097226e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 3201848,
      "step": 4885
    },
    {
      "epoch": 2.5628930817610063,
      "grad_norm": 16.375,
      "learning_rate": 4.6373389924593615e-05,
      "loss": 0.6568,
      "num_input_tokens_seen": 3204504,
      "step": 4890
    },
    {
      "epoch": 2.5655136268343814,
      "grad_norm": 6.53125,
      "learning_rate": 4.6361518275813615e-05,
      "loss": 0.6171,
      "num_input_tokens_seen": 3207320,
      "step": 4895
    },
    {
      "epoch": 2.568134171907757,
      "grad_norm": 7.3125,
      "learning_rate": 4.6349628752690876e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 3210584,
      "step": 4900
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 3.75,
      "learning_rate": 4.633772136517401e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 3213496,
      "step": 4905
    },
    {
      "epoch": 2.5733752620545074,
      "grad_norm": 2.890625,
      "learning_rate": 4.6325796123226575e-05,
      "loss": 0.4853,
      "num_input_tokens_seen": 3217016,
      "step": 4910
    },
    {
      "epoch": 2.5759958071278826,
      "grad_norm": 3.390625,
      "learning_rate": 4.6313853036827057e-05,
      "loss": 0.624,
      "num_input_tokens_seen": 3219800,
      "step": 4915
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 5.28125,
      "learning_rate": 4.630189211596891e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 3222424,
      "step": 4920
    },
    {
      "epoch": 2.581236897274633,
      "grad_norm": 6.28125,
      "learning_rate": 4.628991337066047e-05,
      "loss": 0.5074,
      "num_input_tokens_seen": 3225272,
      "step": 4925
    },
    {
      "epoch": 2.5838574423480085,
      "grad_norm": 3.734375,
      "learning_rate": 4.627791681092499e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 3228088,
      "step": 4930
    },
    {
      "epoch": 2.5864779874213837,
      "grad_norm": 3.234375,
      "learning_rate": 4.626590244680068e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 3231096,
      "step": 4935
    },
    {
      "epoch": 2.589098532494759,
      "grad_norm": 5.40625,
      "learning_rate": 4.625387028834057e-05,
      "loss": 0.5894,
      "num_input_tokens_seen": 3234552,
      "step": 4940
    },
    {
      "epoch": 2.591719077568134,
      "grad_norm": 12.8125,
      "learning_rate": 4.6241820345612654e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 3238072,
      "step": 4945
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 6.15625,
      "learning_rate": 4.622975262869976e-05,
      "loss": 0.5699,
      "num_input_tokens_seen": 3241624,
      "step": 4950
    },
    {
      "epoch": 2.596960167714885,
      "grad_norm": 5.65625,
      "learning_rate": 4.62176671476996e-05,
      "loss": 0.7879,
      "num_input_tokens_seen": 3245080,
      "step": 4955
    },
    {
      "epoch": 2.59958071278826,
      "grad_norm": 5.53125,
      "learning_rate": 4.620556391272476e-05,
      "loss": 0.5479,
      "num_input_tokens_seen": 3247800,
      "step": 4960
    },
    {
      "epoch": 2.602201257861635,
      "grad_norm": 3.828125,
      "learning_rate": 4.619344293390266e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 3251000,
      "step": 4965
    },
    {
      "epoch": 2.6048218029350103,
      "grad_norm": 4.4375,
      "learning_rate": 4.61813042213756e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 3254872,
      "step": 4970
    },
    {
      "epoch": 2.6074423480083855,
      "grad_norm": 5.09375,
      "learning_rate": 4.6169147785300685e-05,
      "loss": 0.5269,
      "num_input_tokens_seen": 3258520,
      "step": 4975
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 3.921875,
      "learning_rate": 4.6156973635849864e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 3261944,
      "step": 4980
    },
    {
      "epoch": 2.6126834381551363,
      "grad_norm": 3.84375,
      "learning_rate": 4.614478178320993e-05,
      "loss": 0.6069,
      "num_input_tokens_seen": 3265432,
      "step": 4985
    },
    {
      "epoch": 2.6153039832285114,
      "grad_norm": 7.34375,
      "learning_rate": 4.613257223758245e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 3268632,
      "step": 4990
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 10.875,
      "learning_rate": 4.612034500918381e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 3271416,
      "step": 4995
    },
    {
      "epoch": 2.620545073375262,
      "grad_norm": 5.65625,
      "learning_rate": 4.610810010824522e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 3274424,
      "step": 5000
    },
    {
      "epoch": 2.6231656184486374,
      "grad_norm": 5.5,
      "learning_rate": 4.609583754501263e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 3277688,
      "step": 5005
    },
    {
      "epoch": 2.6257861635220126,
      "grad_norm": 9.25,
      "learning_rate": 4.6083557329746805e-05,
      "loss": 0.3878,
      "num_input_tokens_seen": 3280440,
      "step": 5010
    },
    {
      "epoch": 2.6284067085953877,
      "grad_norm": 8.0625,
      "learning_rate": 4.607125947272326e-05,
      "loss": 0.677,
      "num_input_tokens_seen": 3283704,
      "step": 5015
    },
    {
      "epoch": 2.631027253668763,
      "grad_norm": 4.25,
      "learning_rate": 4.6058943984232286e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 3286808,
      "step": 5020
    },
    {
      "epoch": 2.6336477987421385,
      "grad_norm": 2.75,
      "learning_rate": 4.604661087457893e-05,
      "loss": 0.5945,
      "num_input_tokens_seen": 3289592,
      "step": 5025
    },
    {
      "epoch": 2.6362683438155137,
      "grad_norm": 3.03125,
      "learning_rate": 4.6034260154082955e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 3292696,
      "step": 5030
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 3.765625,
      "learning_rate": 4.602189183307889e-05,
      "loss": 0.347,
      "num_input_tokens_seen": 3296632,
      "step": 5035
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 5.25,
      "learning_rate": 4.600950592191599e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 3300536,
      "step": 5040
    },
    {
      "epoch": 2.6441299790356396,
      "grad_norm": 4.25,
      "learning_rate": 4.599710243095819e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 3304728,
      "step": 5045
    },
    {
      "epoch": 2.646750524109015,
      "grad_norm": 5.375,
      "learning_rate": 4.59846813705842e-05,
      "loss": 0.5649,
      "num_input_tokens_seen": 3307928,
      "step": 5050
    },
    {
      "epoch": 2.64937106918239,
      "grad_norm": 5.875,
      "learning_rate": 4.597224275118738e-05,
      "loss": 0.5867,
      "num_input_tokens_seen": 3310552,
      "step": 5055
    },
    {
      "epoch": 2.651991614255765,
      "grad_norm": 4.0,
      "learning_rate": 4.59597865831758e-05,
      "loss": 0.6181,
      "num_input_tokens_seen": 3314904,
      "step": 5060
    },
    {
      "epoch": 2.6546121593291403,
      "grad_norm": 4.03125,
      "learning_rate": 4.5947312876972214e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 3320184,
      "step": 5065
    },
    {
      "epoch": 2.6572327044025155,
      "grad_norm": 4.0,
      "learning_rate": 4.5934821643014034e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 3322904,
      "step": 5070
    },
    {
      "epoch": 2.659853249475891,
      "grad_norm": 5.25,
      "learning_rate": 4.5922312891753385e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 3325784,
      "step": 5075
    },
    {
      "epoch": 2.6624737945492662,
      "grad_norm": 3.59375,
      "learning_rate": 4.590978663365699e-05,
      "loss": 0.5617,
      "num_input_tokens_seen": 3328760,
      "step": 5080
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 5.90625,
      "learning_rate": 4.589724287920627e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 3331576,
      "step": 5085
    },
    {
      "epoch": 2.667714884696017,
      "grad_norm": 8.4375,
      "learning_rate": 4.5884681638897246e-05,
      "loss": 0.563,
      "num_input_tokens_seen": 3333880,
      "step": 5090
    },
    {
      "epoch": 2.670335429769392,
      "grad_norm": 5.59375,
      "learning_rate": 4.587210292324061e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 3337880,
      "step": 5095
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 5.75,
      "learning_rate": 4.585950674276164e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 3340920,
      "step": 5100
    },
    {
      "epoch": 2.6755765199161425,
      "grad_norm": 5.21875,
      "learning_rate": 4.5846893108000256e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 3344312,
      "step": 5105
    },
    {
      "epoch": 2.6781970649895177,
      "grad_norm": 4.09375,
      "learning_rate": 4.5834262029510965e-05,
      "loss": 0.4412,
      "num_input_tokens_seen": 3347672,
      "step": 5110
    },
    {
      "epoch": 2.680817610062893,
      "grad_norm": 3.34375,
      "learning_rate": 4.5821613517862883e-05,
      "loss": 0.5192,
      "num_input_tokens_seen": 3351032,
      "step": 5115
    },
    {
      "epoch": 2.6834381551362685,
      "grad_norm": 7.0,
      "learning_rate": 4.5808947583639693e-05,
      "loss": 0.6647,
      "num_input_tokens_seen": 3353752,
      "step": 5120
    },
    {
      "epoch": 2.6860587002096437,
      "grad_norm": 6.21875,
      "learning_rate": 4.579626423743969e-05,
      "loss": 0.4576,
      "num_input_tokens_seen": 3356600,
      "step": 5125
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 7.625,
      "learning_rate": 4.57835634898757e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 3360280,
      "step": 5130
    },
    {
      "epoch": 2.691299790356394,
      "grad_norm": 4.15625,
      "learning_rate": 4.577084535157514e-05,
      "loss": 0.6214,
      "num_input_tokens_seen": 3363768,
      "step": 5135
    },
    {
      "epoch": 2.6939203354297696,
      "grad_norm": 4.03125,
      "learning_rate": 4.5758109833179963e-05,
      "loss": 0.4343,
      "num_input_tokens_seen": 3367480,
      "step": 5140
    },
    {
      "epoch": 2.6965408805031448,
      "grad_norm": 12.125,
      "learning_rate": 4.5745356945346676e-05,
      "loss": 0.5133,
      "num_input_tokens_seen": 3370520,
      "step": 5145
    },
    {
      "epoch": 2.69916142557652,
      "grad_norm": 15.0625,
      "learning_rate": 4.57325866987463e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 3372984,
      "step": 5150
    },
    {
      "epoch": 2.701781970649895,
      "grad_norm": 17.375,
      "learning_rate": 4.571979910406441e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 3375288,
      "step": 5155
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 5.375,
      "learning_rate": 4.570699417200106e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 3378488,
      "step": 5160
    },
    {
      "epoch": 2.7070230607966455,
      "grad_norm": 3.578125,
      "learning_rate": 4.569417191327086e-05,
      "loss": 0.4561,
      "num_input_tokens_seen": 3381368,
      "step": 5165
    },
    {
      "epoch": 2.709643605870021,
      "grad_norm": 6.15625,
      "learning_rate": 4.5681332338602864e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 3385016,
      "step": 5170
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 3.046875,
      "learning_rate": 4.5668475458740654e-05,
      "loss": 0.5312,
      "num_input_tokens_seen": 3387992,
      "step": 5175
    },
    {
      "epoch": 2.7148846960167714,
      "grad_norm": 5.71875,
      "learning_rate": 4.5655601284442276e-05,
      "loss": 0.534,
      "num_input_tokens_seen": 3390872,
      "step": 5180
    },
    {
      "epoch": 2.717505241090147,
      "grad_norm": 4.25,
      "learning_rate": 4.5642709826480256e-05,
      "loss": 0.5573,
      "num_input_tokens_seen": 3393496,
      "step": 5185
    },
    {
      "epoch": 2.720125786163522,
      "grad_norm": 4.28125,
      "learning_rate": 4.562980109564158e-05,
      "loss": 0.7422,
      "num_input_tokens_seen": 3395992,
      "step": 5190
    },
    {
      "epoch": 2.7227463312368974,
      "grad_norm": 4.65625,
      "learning_rate": 4.561687510272767e-05,
      "loss": 0.5933,
      "num_input_tokens_seen": 3399160,
      "step": 5195
    },
    {
      "epoch": 2.7253668763102725,
      "grad_norm": 3.953125,
      "learning_rate": 4.5603931858554415e-05,
      "loss": 0.5085,
      "num_input_tokens_seen": 3402072,
      "step": 5200
    },
    {
      "epoch": 2.7279874213836477,
      "grad_norm": 9.75,
      "learning_rate": 4.559097137395214e-05,
      "loss": 0.4961,
      "num_input_tokens_seen": 3404344,
      "step": 5205
    },
    {
      "epoch": 2.730607966457023,
      "grad_norm": 4.59375,
      "learning_rate": 4.5577993659765574e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 3407640,
      "step": 5210
    },
    {
      "epoch": 2.7332285115303985,
      "grad_norm": 6.09375,
      "learning_rate": 4.556499872685387e-05,
      "loss": 0.3878,
      "num_input_tokens_seen": 3410200,
      "step": 5215
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 8.5,
      "learning_rate": 4.555198658609061e-05,
      "loss": 0.419,
      "num_input_tokens_seen": 3412856,
      "step": 5220
    },
    {
      "epoch": 2.738469601677149,
      "grad_norm": 2.640625,
      "learning_rate": 4.5538957248363756e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 3415864,
      "step": 5225
    },
    {
      "epoch": 2.741090146750524,
      "grad_norm": 3.1875,
      "learning_rate": 4.552591072457565e-05,
      "loss": 0.5167,
      "num_input_tokens_seen": 3419224,
      "step": 5230
    },
    {
      "epoch": 2.7437106918238996,
      "grad_norm": 6.5625,
      "learning_rate": 4.551284702564304e-05,
      "loss": 0.4515,
      "num_input_tokens_seen": 3422520,
      "step": 5235
    },
    {
      "epoch": 2.7463312368972748,
      "grad_norm": 7.09375,
      "learning_rate": 4.5499766162497025e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 3425976,
      "step": 5240
    },
    {
      "epoch": 2.74895178197065,
      "grad_norm": 4.5625,
      "learning_rate": 4.548666814608308e-05,
      "loss": 0.5013,
      "num_input_tokens_seen": 3429368,
      "step": 5245
    },
    {
      "epoch": 2.751572327044025,
      "grad_norm": 5.90625,
      "learning_rate": 4.5473552987361024e-05,
      "loss": 0.5381,
      "num_input_tokens_seen": 3434680,
      "step": 5250
    },
    {
      "epoch": 2.7541928721174003,
      "grad_norm": 8.125,
      "learning_rate": 4.5460420697305024e-05,
      "loss": 0.5936,
      "num_input_tokens_seen": 3437752,
      "step": 5255
    },
    {
      "epoch": 2.7568134171907754,
      "grad_norm": 3.671875,
      "learning_rate": 4.544727128690358e-05,
      "loss": 0.5992,
      "num_input_tokens_seen": 3440760,
      "step": 5260
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 3.59375,
      "learning_rate": 4.543410476715951e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 3443736,
      "step": 5265
    },
    {
      "epoch": 2.762054507337526,
      "grad_norm": 7.9375,
      "learning_rate": 4.542092114908997e-05,
      "loss": 0.6294,
      "num_input_tokens_seen": 3446904,
      "step": 5270
    },
    {
      "epoch": 2.7646750524109014,
      "grad_norm": 2.828125,
      "learning_rate": 4.54077204437264e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 3450104,
      "step": 5275
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 3.9375,
      "learning_rate": 4.5394502662114555e-05,
      "loss": 0.6576,
      "num_input_tokens_seen": 3453592,
      "step": 5280
    },
    {
      "epoch": 2.769916142557652,
      "grad_norm": 10.125,
      "learning_rate": 4.538126781531446e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 3456824,
      "step": 5285
    },
    {
      "epoch": 2.7725366876310273,
      "grad_norm": 9.0,
      "learning_rate": 4.536801591440044e-05,
      "loss": 0.5442,
      "num_input_tokens_seen": 3460792,
      "step": 5290
    },
    {
      "epoch": 2.7751572327044025,
      "grad_norm": 4.46875,
      "learning_rate": 4.535474697046107e-05,
      "loss": 0.5709,
      "num_input_tokens_seen": 3464280,
      "step": 5295
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 9.1875,
      "learning_rate": 4.534146099459921e-05,
      "loss": 0.6306,
      "num_input_tokens_seen": 3467320,
      "step": 5300
    },
    {
      "epoch": 2.780398322851153,
      "grad_norm": 5.03125,
      "learning_rate": 4.5328157997931955e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 3470520,
      "step": 5305
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 4.46875,
      "learning_rate": 4.531483799159062e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 3474008,
      "step": 5310
    },
    {
      "epoch": 2.7856394129979036,
      "grad_norm": 15.0625,
      "learning_rate": 4.5301500986720816e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 3477208,
      "step": 5315
    },
    {
      "epoch": 2.788259958071279,
      "grad_norm": 5.09375,
      "learning_rate": 4.528814699448232e-05,
      "loss": 0.5516,
      "num_input_tokens_seen": 3480088,
      "step": 5320
    },
    {
      "epoch": 2.790880503144654,
      "grad_norm": 2.109375,
      "learning_rate": 4.527477602604914e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 3484504,
      "step": 5325
    },
    {
      "epoch": 2.7935010482180296,
      "grad_norm": 5.375,
      "learning_rate": 4.52613880926095e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 3487000,
      "step": 5330
    },
    {
      "epoch": 2.7961215932914047,
      "grad_norm": 4.4375,
      "learning_rate": 4.5247983205365806e-05,
      "loss": 0.4546,
      "num_input_tokens_seen": 3490712,
      "step": 5335
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 4.84375,
      "learning_rate": 4.5234561375534655e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 3494296,
      "step": 5340
    },
    {
      "epoch": 2.801362683438155,
      "grad_norm": 3.234375,
      "learning_rate": 4.5221122614346823e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 3498232,
      "step": 5345
    },
    {
      "epoch": 2.8039832285115303,
      "grad_norm": 5.40625,
      "learning_rate": 4.520766693304725e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 3501080,
      "step": 5350
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 9.0,
      "learning_rate": 4.519419434289504e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 3504024,
      "step": 5355
    },
    {
      "epoch": 2.809224318658281,
      "grad_norm": 3.453125,
      "learning_rate": 4.518070485516344e-05,
      "loss": 0.3919,
      "num_input_tokens_seen": 3507256,
      "step": 5360
    },
    {
      "epoch": 2.811844863731656,
      "grad_norm": 4.125,
      "learning_rate": 4.5167198481139825e-05,
      "loss": 0.546,
      "num_input_tokens_seen": 3510392,
      "step": 5365
    },
    {
      "epoch": 2.8144654088050314,
      "grad_norm": 5.9375,
      "learning_rate": 4.515367523212573e-05,
      "loss": 0.5478,
      "num_input_tokens_seen": 3513272,
      "step": 5370
    },
    {
      "epoch": 2.8170859538784065,
      "grad_norm": 4.03125,
      "learning_rate": 4.5140135119436776e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 3516312,
      "step": 5375
    },
    {
      "epoch": 2.819706498951782,
      "grad_norm": 6.875,
      "learning_rate": 4.512657815440273e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 3519032,
      "step": 5380
    },
    {
      "epoch": 2.8223270440251573,
      "grad_norm": 4.375,
      "learning_rate": 4.511300434836743e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 3521720,
      "step": 5385
    },
    {
      "epoch": 2.8249475890985325,
      "grad_norm": 6.40625,
      "learning_rate": 4.5099413712688805e-05,
      "loss": 0.4708,
      "num_input_tokens_seen": 3524888,
      "step": 5390
    },
    {
      "epoch": 2.8275681341719077,
      "grad_norm": 3.546875,
      "learning_rate": 4.5085806258738896e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 3529688,
      "step": 5395
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 5.8125,
      "learning_rate": 4.507218199790379e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 3532440,
      "step": 5400
    },
    {
      "epoch": 2.832809224318658,
      "grad_norm": 5.53125,
      "learning_rate": 4.505854094158365e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 3535288,
      "step": 5405
    },
    {
      "epoch": 2.8354297693920336,
      "grad_norm": 6.21875,
      "learning_rate": 4.5044883101192695e-05,
      "loss": 0.5292,
      "num_input_tokens_seen": 3538840,
      "step": 5410
    },
    {
      "epoch": 2.838050314465409,
      "grad_norm": 4.78125,
      "learning_rate": 4.503120848815916e-05,
      "loss": 0.3844,
      "num_input_tokens_seen": 3541880,
      "step": 5415
    },
    {
      "epoch": 2.840670859538784,
      "grad_norm": 5.0,
      "learning_rate": 4.501751711392536e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 3545208,
      "step": 5420
    },
    {
      "epoch": 2.8432914046121596,
      "grad_norm": 8.375,
      "learning_rate": 4.5003808989947605e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 3548120,
      "step": 5425
    },
    {
      "epoch": 2.8459119496855347,
      "grad_norm": 2.5,
      "learning_rate": 4.499008412769622e-05,
      "loss": 0.5174,
      "num_input_tokens_seen": 3551512,
      "step": 5430
    },
    {
      "epoch": 2.84853249475891,
      "grad_norm": 8.875,
      "learning_rate": 4.4976342538655546e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 3554520,
      "step": 5435
    },
    {
      "epoch": 2.851153039832285,
      "grad_norm": 3.828125,
      "learning_rate": 4.4962584234323925e-05,
      "loss": 0.5848,
      "num_input_tokens_seen": 3557208,
      "step": 5440
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 4.40625,
      "learning_rate": 4.4948809226213664e-05,
      "loss": 0.6789,
      "num_input_tokens_seen": 3560600,
      "step": 5445
    },
    {
      "epoch": 2.8563941299790354,
      "grad_norm": 3.875,
      "learning_rate": 4.4935017525851067e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 3565752,
      "step": 5450
    },
    {
      "epoch": 2.859014675052411,
      "grad_norm": 4.875,
      "learning_rate": 4.4921209144776414e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 3569400,
      "step": 5455
    },
    {
      "epoch": 2.861635220125786,
      "grad_norm": 9.75,
      "learning_rate": 4.490738409454389e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 3572088,
      "step": 5460
    },
    {
      "epoch": 2.8642557651991614,
      "grad_norm": 2.703125,
      "learning_rate": 4.48935423867217e-05,
      "loss": 0.5882,
      "num_input_tokens_seen": 3575032,
      "step": 5465
    },
    {
      "epoch": 2.8668763102725365,
      "grad_norm": 6.5625,
      "learning_rate": 4.487968403289195e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 3578232,
      "step": 5470
    },
    {
      "epoch": 2.869496855345912,
      "grad_norm": 5.625,
      "learning_rate": 4.4865809044650655e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 3581304,
      "step": 5475
    },
    {
      "epoch": 2.8721174004192873,
      "grad_norm": 5.46875,
      "learning_rate": 4.48519174336078e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 3584664,
      "step": 5480
    },
    {
      "epoch": 2.8747379454926625,
      "grad_norm": 5.78125,
      "learning_rate": 4.483800921138722e-05,
      "loss": 0.5833,
      "num_input_tokens_seen": 3587544,
      "step": 5485
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 11.6875,
      "learning_rate": 4.4824084389626705e-05,
      "loss": 0.5448,
      "num_input_tokens_seen": 3590360,
      "step": 5490
    },
    {
      "epoch": 2.879979035639413,
      "grad_norm": 5.1875,
      "learning_rate": 4.48101429799779e-05,
      "loss": 0.4307,
      "num_input_tokens_seen": 3597208,
      "step": 5495
    },
    {
      "epoch": 2.882599580712788,
      "grad_norm": 7.03125,
      "learning_rate": 4.479618499410634e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 3600568,
      "step": 5500
    },
    {
      "epoch": 2.8852201257861636,
      "grad_norm": 26.875,
      "learning_rate": 4.478221044369143e-05,
      "loss": 0.522,
      "num_input_tokens_seen": 3603480,
      "step": 5505
    },
    {
      "epoch": 2.8878406708595388,
      "grad_norm": 4.59375,
      "learning_rate": 4.476821934042644e-05,
      "loss": 0.5655,
      "num_input_tokens_seen": 3606360,
      "step": 5510
    },
    {
      "epoch": 2.890461215932914,
      "grad_norm": 6.96875,
      "learning_rate": 4.4754211696018475e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 3608856,
      "step": 5515
    },
    {
      "epoch": 2.8930817610062896,
      "grad_norm": 3.3125,
      "learning_rate": 4.47401875221885e-05,
      "loss": 0.4192,
      "num_input_tokens_seen": 3611384,
      "step": 5520
    },
    {
      "epoch": 2.8957023060796647,
      "grad_norm": 4.53125,
      "learning_rate": 4.4726146830671304e-05,
      "loss": 0.7124,
      "num_input_tokens_seen": 3614488,
      "step": 5525
    },
    {
      "epoch": 2.89832285115304,
      "grad_norm": 4.25,
      "learning_rate": 4.47120896332155e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 3617080,
      "step": 5530
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 11.8125,
      "learning_rate": 4.4698015941583494e-05,
      "loss": 0.6342,
      "num_input_tokens_seen": 3620728,
      "step": 5535
    },
    {
      "epoch": 2.9035639412997902,
      "grad_norm": 3.765625,
      "learning_rate": 4.4683925767551525e-05,
      "loss": 0.5489,
      "num_input_tokens_seen": 3624184,
      "step": 5540
    },
    {
      "epoch": 2.9061844863731654,
      "grad_norm": 2.796875,
      "learning_rate": 4.466981912290959e-05,
      "loss": 0.4178,
      "num_input_tokens_seen": 3628856,
      "step": 5545
    },
    {
      "epoch": 2.908805031446541,
      "grad_norm": 4.46875,
      "learning_rate": 4.46556960194615e-05,
      "loss": 0.578,
      "num_input_tokens_seen": 3631800,
      "step": 5550
    },
    {
      "epoch": 2.911425576519916,
      "grad_norm": 12.25,
      "learning_rate": 4.464155646902482e-05,
      "loss": 0.4847,
      "num_input_tokens_seen": 3634680,
      "step": 5555
    },
    {
      "epoch": 2.9140461215932913,
      "grad_norm": 4.0,
      "learning_rate": 4.462740048343087e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 3638264,
      "step": 5560
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 10.9375,
      "learning_rate": 4.461322807452475e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 3641240,
      "step": 5565
    },
    {
      "epoch": 2.919287211740042,
      "grad_norm": 5.1875,
      "learning_rate": 4.4599039254165264e-05,
      "loss": 0.5699,
      "num_input_tokens_seen": 3643640,
      "step": 5570
    },
    {
      "epoch": 2.9219077568134173,
      "grad_norm": 2.3125,
      "learning_rate": 4.458483403422498e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 3648088,
      "step": 5575
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 3.015625,
      "learning_rate": 4.457061242659018e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 3651352,
      "step": 5580
    },
    {
      "epoch": 2.9271488469601676,
      "grad_norm": 3.0625,
      "learning_rate": 4.455637444316085e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 3655288,
      "step": 5585
    },
    {
      "epoch": 2.929769392033543,
      "grad_norm": 5.125,
      "learning_rate": 4.454212009585068e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 3657624,
      "step": 5590
    },
    {
      "epoch": 2.932389937106918,
      "grad_norm": 7.4375,
      "learning_rate": 4.4527849396587065e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 3660472,
      "step": 5595
    },
    {
      "epoch": 2.9350104821802936,
      "grad_norm": 3.78125,
      "learning_rate": 4.4513562357311074e-05,
      "loss": 0.5674,
      "num_input_tokens_seen": 3663992,
      "step": 5600
    },
    {
      "epoch": 2.9376310272536688,
      "grad_norm": 6.5,
      "learning_rate": 4.449925898997744e-05,
      "loss": 0.5562,
      "num_input_tokens_seen": 3666424,
      "step": 5605
    },
    {
      "epoch": 2.940251572327044,
      "grad_norm": 5.09375,
      "learning_rate": 4.4484939306554585e-05,
      "loss": 0.5567,
      "num_input_tokens_seen": 3669400,
      "step": 5610
    },
    {
      "epoch": 2.9428721174004195,
      "grad_norm": 4.96875,
      "learning_rate": 4.4470603319024554e-05,
      "loss": 0.6383,
      "num_input_tokens_seen": 3672440,
      "step": 5615
    },
    {
      "epoch": 2.9454926624737947,
      "grad_norm": 18.25,
      "learning_rate": 4.445625103938304e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 3675704,
      "step": 5620
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 8.3125,
      "learning_rate": 4.4441882479639375e-05,
      "loss": 0.6497,
      "num_input_tokens_seen": 3679160,
      "step": 5625
    },
    {
      "epoch": 2.950733752620545,
      "grad_norm": 4.1875,
      "learning_rate": 4.442749765181653e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 3682328,
      "step": 5630
    },
    {
      "epoch": 2.95335429769392,
      "grad_norm": 7.4375,
      "learning_rate": 4.441309656795106e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 3685240,
      "step": 5635
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 8.375,
      "learning_rate": 4.4398679240093144e-05,
      "loss": 0.6536,
      "num_input_tokens_seen": 3687960,
      "step": 5640
    },
    {
      "epoch": 2.958595387840671,
      "grad_norm": 6.53125,
      "learning_rate": 4.438424568030652e-05,
      "loss": 0.4084,
      "num_input_tokens_seen": 3691704,
      "step": 5645
    },
    {
      "epoch": 2.961215932914046,
      "grad_norm": 4.8125,
      "learning_rate": 4.436979590066857e-05,
      "loss": 0.5403,
      "num_input_tokens_seen": 3695416,
      "step": 5650
    },
    {
      "epoch": 2.9638364779874213,
      "grad_norm": 8.4375,
      "learning_rate": 4.435532991327017e-05,
      "loss": 0.4699,
      "num_input_tokens_seen": 3698168,
      "step": 5655
    },
    {
      "epoch": 2.9664570230607965,
      "grad_norm": 8.375,
      "learning_rate": 4.434084773021582e-05,
      "loss": 0.6453,
      "num_input_tokens_seen": 3700984,
      "step": 5660
    },
    {
      "epoch": 2.969077568134172,
      "grad_norm": 6.0625,
      "learning_rate": 4.432634936362354e-05,
      "loss": 0.6827,
      "num_input_tokens_seen": 3704472,
      "step": 5665
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 5.0,
      "learning_rate": 4.431183482562491e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 3707960,
      "step": 5670
    },
    {
      "epoch": 2.9743186582809225,
      "grad_norm": 10.625,
      "learning_rate": 4.429730412836503e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 3710520,
      "step": 5675
    },
    {
      "epoch": 2.9769392033542976,
      "grad_norm": 4.90625,
      "learning_rate": 4.4282757284002515e-05,
      "loss": 0.5106,
      "num_input_tokens_seen": 3713688,
      "step": 5680
    },
    {
      "epoch": 2.979559748427673,
      "grad_norm": 3.890625,
      "learning_rate": 4.426819430470951e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 3717304,
      "step": 5685
    },
    {
      "epoch": 2.982180293501048,
      "grad_norm": 4.3125,
      "learning_rate": 4.425361520267165e-05,
      "loss": 0.584,
      "num_input_tokens_seen": 3720888,
      "step": 5690
    },
    {
      "epoch": 2.9848008385744236,
      "grad_norm": 4.1875,
      "learning_rate": 4.423901999008805e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 3724888,
      "step": 5695
    },
    {
      "epoch": 2.9874213836477987,
      "grad_norm": 8.0,
      "learning_rate": 4.4224408679171324e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 3727288,
      "step": 5700
    },
    {
      "epoch": 2.990041928721174,
      "grad_norm": 6.90625,
      "learning_rate": 4.4209781282147555e-05,
      "loss": 0.5778,
      "num_input_tokens_seen": 3730328,
      "step": 5705
    },
    {
      "epoch": 2.9926624737945495,
      "grad_norm": 7.0,
      "learning_rate": 4.419513781125628e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 3733912,
      "step": 5710
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 4.625,
      "learning_rate": 4.418047827875048e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 3737464,
      "step": 5715
    },
    {
      "epoch": 2.9979035639413,
      "grad_norm": 4.09375,
      "learning_rate": 4.416580269689658e-05,
      "loss": 0.443,
      "num_input_tokens_seen": 3740152,
      "step": 5720
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.505059540271759,
      "eval_runtime": 24.2479,
      "eval_samples_per_second": 34.972,
      "eval_steps_per_second": 8.743,
      "num_input_tokens_seen": 3742552,
      "step": 5724
    },
    {
      "epoch": 3.000524109014675,
      "grad_norm": 9.375,
      "learning_rate": 4.415111107797445e-05,
      "loss": 0.4581,
      "num_input_tokens_seen": 3742968,
      "step": 5725
    },
    {
      "epoch": 3.00314465408805,
      "grad_norm": 4.125,
      "learning_rate": 4.4136403434277364e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 3745880,
      "step": 5730
    },
    {
      "epoch": 3.0057651991614254,
      "grad_norm": 3.078125,
      "learning_rate": 4.412167977811199e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 3749464,
      "step": 5735
    },
    {
      "epoch": 3.008385744234801,
      "grad_norm": 10.9375,
      "learning_rate": 4.4106940121798424e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 3752248,
      "step": 5740
    },
    {
      "epoch": 3.011006289308176,
      "grad_norm": 5.125,
      "learning_rate": 4.409218447767013e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 3755480,
      "step": 5745
    },
    {
      "epoch": 3.0136268343815513,
      "grad_norm": 4.5,
      "learning_rate": 4.4077412858073966e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 3758040,
      "step": 5750
    },
    {
      "epoch": 3.0162473794549265,
      "grad_norm": 10.25,
      "learning_rate": 4.406262527537014e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 3761368,
      "step": 5755
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 5.96875,
      "learning_rate": 4.404782174193223e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 3764824,
      "step": 5760
    },
    {
      "epoch": 3.0214884696016773,
      "grad_norm": 4.1875,
      "learning_rate": 4.403300227014716e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 3768056,
      "step": 5765
    },
    {
      "epoch": 3.0241090146750524,
      "grad_norm": 6.6875,
      "learning_rate": 4.4018166872415176e-05,
      "loss": 0.561,
      "num_input_tokens_seen": 3770168,
      "step": 5770
    },
    {
      "epoch": 3.0267295597484276,
      "grad_norm": 3.734375,
      "learning_rate": 4.4003315561149875e-05,
      "loss": 0.5201,
      "num_input_tokens_seen": 3773112,
      "step": 5775
    },
    {
      "epoch": 3.029350104821803,
      "grad_norm": 4.53125,
      "learning_rate": 4.398844834877815e-05,
      "loss": 0.5632,
      "num_input_tokens_seen": 3776568,
      "step": 5780
    },
    {
      "epoch": 3.0319706498951784,
      "grad_norm": 3.21875,
      "learning_rate": 4.39735652477402e-05,
      "loss": 0.6783,
      "num_input_tokens_seen": 3780920,
      "step": 5785
    },
    {
      "epoch": 3.0345911949685536,
      "grad_norm": 5.71875,
      "learning_rate": 4.395866627048953e-05,
      "loss": 0.5726,
      "num_input_tokens_seen": 3783960,
      "step": 5790
    },
    {
      "epoch": 3.0372117400419287,
      "grad_norm": 5.90625,
      "learning_rate": 4.3943751429492925e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 3786968,
      "step": 5795
    },
    {
      "epoch": 3.039832285115304,
      "grad_norm": 3.140625,
      "learning_rate": 4.392882073723043e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 3790808,
      "step": 5800
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 2.9375,
      "learning_rate": 4.391387420619539e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 3794264,
      "step": 5805
    },
    {
      "epoch": 3.0450733752620547,
      "grad_norm": 4.65625,
      "learning_rate": 4.389891184889435e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 3797496,
      "step": 5810
    },
    {
      "epoch": 3.04769392033543,
      "grad_norm": 5.1875,
      "learning_rate": 4.3883933677847154e-05,
      "loss": 0.6375,
      "num_input_tokens_seen": 3800888,
      "step": 5815
    },
    {
      "epoch": 3.050314465408805,
      "grad_norm": 4.1875,
      "learning_rate": 4.3868939705586844e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 3803512,
      "step": 5820
    },
    {
      "epoch": 3.05293501048218,
      "grad_norm": 10.125,
      "learning_rate": 4.385392994465968e-05,
      "loss": 0.6079,
      "num_input_tokens_seen": 3806552,
      "step": 5825
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 2.71875,
      "learning_rate": 4.383890440762515e-05,
      "loss": 0.5795,
      "num_input_tokens_seen": 3810648,
      "step": 5830
    },
    {
      "epoch": 3.058176100628931,
      "grad_norm": 4.6875,
      "learning_rate": 4.3823863107055936e-05,
      "loss": 0.4047,
      "num_input_tokens_seen": 3813976,
      "step": 5835
    },
    {
      "epoch": 3.060796645702306,
      "grad_norm": 6.8125,
      "learning_rate": 4.380880605553792e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 3816824,
      "step": 5840
    },
    {
      "epoch": 3.0634171907756813,
      "grad_norm": 1.8984375,
      "learning_rate": 4.3793733265670147e-05,
      "loss": 0.6144,
      "num_input_tokens_seen": 3820376,
      "step": 5845
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 11.0625,
      "learning_rate": 4.3778644750064834e-05,
      "loss": 0.4523,
      "num_input_tokens_seen": 3823736,
      "step": 5850
    },
    {
      "epoch": 3.068658280922432,
      "grad_norm": 5.5,
      "learning_rate": 4.376354052134738e-05,
      "loss": 0.6596,
      "num_input_tokens_seen": 3826104,
      "step": 5855
    },
    {
      "epoch": 3.0712788259958073,
      "grad_norm": 3.59375,
      "learning_rate": 4.374842059215629e-05,
      "loss": 0.5758,
      "num_input_tokens_seen": 3828632,
      "step": 5860
    },
    {
      "epoch": 3.0738993710691824,
      "grad_norm": 4.1875,
      "learning_rate": 4.373328497514325e-05,
      "loss": 0.3431,
      "num_input_tokens_seen": 3831576,
      "step": 5865
    },
    {
      "epoch": 3.0765199161425576,
      "grad_norm": 3.140625,
      "learning_rate": 4.371813368297304e-05,
      "loss": 0.3699,
      "num_input_tokens_seen": 3834392,
      "step": 5870
    },
    {
      "epoch": 3.0791404612159328,
      "grad_norm": 6.78125,
      "learning_rate": 4.370296672832358e-05,
      "loss": 0.5669,
      "num_input_tokens_seen": 3837144,
      "step": 5875
    },
    {
      "epoch": 3.0817610062893084,
      "grad_norm": 5.28125,
      "learning_rate": 4.3687784123885875e-05,
      "loss": 0.5613,
      "num_input_tokens_seen": 3840504,
      "step": 5880
    },
    {
      "epoch": 3.0843815513626835,
      "grad_norm": 4.15625,
      "learning_rate": 4.3672585882364045e-05,
      "loss": 0.5359,
      "num_input_tokens_seen": 3844408,
      "step": 5885
    },
    {
      "epoch": 3.0870020964360587,
      "grad_norm": 4.0,
      "learning_rate": 4.3657372016475275e-05,
      "loss": 0.5481,
      "num_input_tokens_seen": 3847736,
      "step": 5890
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 4.15625,
      "learning_rate": 4.364214253894983e-05,
      "loss": 0.5625,
      "num_input_tokens_seen": 3850360,
      "step": 5895
    },
    {
      "epoch": 3.092243186582809,
      "grad_norm": 3.46875,
      "learning_rate": 4.3626897462531054e-05,
      "loss": 0.4868,
      "num_input_tokens_seen": 3853624,
      "step": 5900
    },
    {
      "epoch": 3.0948637316561847,
      "grad_norm": 4.4375,
      "learning_rate": 4.361163679997532e-05,
      "loss": 0.4115,
      "num_input_tokens_seen": 3856664,
      "step": 5905
    },
    {
      "epoch": 3.09748427672956,
      "grad_norm": 4.09375,
      "learning_rate": 4.359636056405206e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 3859640,
      "step": 5910
    },
    {
      "epoch": 3.100104821802935,
      "grad_norm": 7.65625,
      "learning_rate": 4.3581068767543724e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 3862712,
      "step": 5915
    },
    {
      "epoch": 3.10272536687631,
      "grad_norm": 4.28125,
      "learning_rate": 4.35657614232458e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 3866328,
      "step": 5920
    },
    {
      "epoch": 3.1053459119496853,
      "grad_norm": 5.25,
      "learning_rate": 4.355043854396677e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 3869592,
      "step": 5925
    },
    {
      "epoch": 3.107966457023061,
      "grad_norm": 5.65625,
      "learning_rate": 4.35351001425281e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 3875032,
      "step": 5930
    },
    {
      "epoch": 3.110587002096436,
      "grad_norm": 3.09375,
      "learning_rate": 4.351974623176429e-05,
      "loss": 0.5854,
      "num_input_tokens_seen": 3879288,
      "step": 5935
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 4.40625,
      "learning_rate": 4.3504376824522787e-05,
      "loss": 0.3904,
      "num_input_tokens_seen": 3882936,
      "step": 5940
    },
    {
      "epoch": 3.1158280922431865,
      "grad_norm": 2.734375,
      "learning_rate": 4.348899193366399e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 3886296,
      "step": 5945
    },
    {
      "epoch": 3.1184486373165616,
      "grad_norm": 9.75,
      "learning_rate": 4.34735915720613e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 3889816,
      "step": 5950
    },
    {
      "epoch": 3.1210691823899372,
      "grad_norm": 5.25,
      "learning_rate": 4.345817575260101e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 3892888,
      "step": 5955
    },
    {
      "epoch": 3.1236897274633124,
      "grad_norm": 12.5625,
      "learning_rate": 4.3442744488182395e-05,
      "loss": 0.5867,
      "num_input_tokens_seen": 3896152,
      "step": 5960
    },
    {
      "epoch": 3.1263102725366876,
      "grad_norm": 4.40625,
      "learning_rate": 4.342729779171761e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 3898776,
      "step": 5965
    },
    {
      "epoch": 3.1289308176100628,
      "grad_norm": 5.8125,
      "learning_rate": 4.341183567613177e-05,
      "loss": 0.5443,
      "num_input_tokens_seen": 3901912,
      "step": 5970
    },
    {
      "epoch": 3.131551362683438,
      "grad_norm": 3.390625,
      "learning_rate": 4.339635815436286e-05,
      "loss": 0.6008,
      "num_input_tokens_seen": 3905688,
      "step": 5975
    },
    {
      "epoch": 3.1341719077568135,
      "grad_norm": 5.8125,
      "learning_rate": 4.3380865239361754e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 3908888,
      "step": 5980
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 7.1875,
      "learning_rate": 4.336535694409222e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 3912280,
      "step": 5985
    },
    {
      "epoch": 3.139412997903564,
      "grad_norm": 7.40625,
      "learning_rate": 4.334983328153088e-05,
      "loss": 0.4712,
      "num_input_tokens_seen": 3915256,
      "step": 5990
    },
    {
      "epoch": 3.142033542976939,
      "grad_norm": 4.09375,
      "learning_rate": 4.3334294264667255e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 3917912,
      "step": 5995
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 3.578125,
      "learning_rate": 4.3318739906503655e-05,
      "loss": 0.553,
      "num_input_tokens_seen": 3920824,
      "step": 6000
    },
    {
      "epoch": 3.14727463312369,
      "grad_norm": 7.25,
      "learning_rate": 4.3303170220055264e-05,
      "loss": 0.4221,
      "num_input_tokens_seen": 3923736,
      "step": 6005
    },
    {
      "epoch": 3.149895178197065,
      "grad_norm": 5.78125,
      "learning_rate": 4.32875852183501e-05,
      "loss": 0.5475,
      "num_input_tokens_seen": 3926584,
      "step": 6010
    },
    {
      "epoch": 3.15251572327044,
      "grad_norm": 5.5625,
      "learning_rate": 4.3271984914428965e-05,
      "loss": 0.6348,
      "num_input_tokens_seen": 3929432,
      "step": 6015
    },
    {
      "epoch": 3.1551362683438153,
      "grad_norm": 5.0,
      "learning_rate": 4.325636932134548e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 3932088,
      "step": 6020
    },
    {
      "epoch": 3.157756813417191,
      "grad_norm": 5.34375,
      "learning_rate": 4.324073845216606e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 3935384,
      "step": 6025
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 5.96875,
      "learning_rate": 4.322509231996992e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 3938424,
      "step": 6030
    },
    {
      "epoch": 3.1629979035639413,
      "grad_norm": 5.25,
      "learning_rate": 4.320943093784901e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 3940984,
      "step": 6035
    },
    {
      "epoch": 3.1656184486373165,
      "grad_norm": 6.09375,
      "learning_rate": 4.319375431890806e-05,
      "loss": 0.4845,
      "num_input_tokens_seen": 3945144,
      "step": 6040
    },
    {
      "epoch": 3.1682389937106916,
      "grad_norm": 3.59375,
      "learning_rate": 4.317806247626456e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 3948344,
      "step": 6045
    },
    {
      "epoch": 3.1708595387840672,
      "grad_norm": 3.78125,
      "learning_rate": 4.316235542304872e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 3951000,
      "step": 6050
    },
    {
      "epoch": 3.1734800838574424,
      "grad_norm": 4.53125,
      "learning_rate": 4.314663317240348e-05,
      "loss": 0.489,
      "num_input_tokens_seen": 3954584,
      "step": 6055
    },
    {
      "epoch": 3.1761006289308176,
      "grad_norm": 4.6875,
      "learning_rate": 4.313089573748451e-05,
      "loss": 0.3635,
      "num_input_tokens_seen": 3957944,
      "step": 6060
    },
    {
      "epoch": 3.1787211740041927,
      "grad_norm": 4.03125,
      "learning_rate": 4.311514313146018e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 3961816,
      "step": 6065
    },
    {
      "epoch": 3.181341719077568,
      "grad_norm": 5.15625,
      "learning_rate": 4.309937536751153e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 3964824,
      "step": 6070
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 4.1875,
      "learning_rate": 4.3083592458832327e-05,
      "loss": 0.3907,
      "num_input_tokens_seen": 3968056,
      "step": 6075
    },
    {
      "epoch": 3.1865828092243187,
      "grad_norm": 6.125,
      "learning_rate": 4.3067794418628976e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 3970424,
      "step": 6080
    },
    {
      "epoch": 3.189203354297694,
      "grad_norm": 11.1875,
      "learning_rate": 4.305198126012057e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 3973208,
      "step": 6085
    },
    {
      "epoch": 3.191823899371069,
      "grad_norm": 6.90625,
      "learning_rate": 4.303615299653881e-05,
      "loss": 0.5395,
      "num_input_tokens_seen": 3976216,
      "step": 6090
    },
    {
      "epoch": 3.1944444444444446,
      "grad_norm": 7.90625,
      "learning_rate": 4.30203096411281e-05,
      "loss": 0.5616,
      "num_input_tokens_seen": 3980824,
      "step": 6095
    },
    {
      "epoch": 3.19706498951782,
      "grad_norm": 3.796875,
      "learning_rate": 4.30044512071454e-05,
      "loss": 0.5568,
      "num_input_tokens_seen": 3987480,
      "step": 6100
    },
    {
      "epoch": 3.199685534591195,
      "grad_norm": 4.5625,
      "learning_rate": 4.2988577707860346e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 3991288,
      "step": 6105
    },
    {
      "epoch": 3.20230607966457,
      "grad_norm": 6.03125,
      "learning_rate": 4.2972689156555154e-05,
      "loss": 0.575,
      "num_input_tokens_seen": 3993848,
      "step": 6110
    },
    {
      "epoch": 3.2049266247379453,
      "grad_norm": 3.09375,
      "learning_rate": 4.295678556652464e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 3997848,
      "step": 6115
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 6.09375,
      "learning_rate": 4.294086695107619e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 4001784,
      "step": 6120
    },
    {
      "epoch": 3.210167714884696,
      "grad_norm": 3.34375,
      "learning_rate": 4.292493332352978e-05,
      "loss": 0.5254,
      "num_input_tokens_seen": 4005048,
      "step": 6125
    },
    {
      "epoch": 3.2127882599580713,
      "grad_norm": 17.5,
      "learning_rate": 4.290898469721795e-05,
      "loss": 0.5312,
      "num_input_tokens_seen": 4007864,
      "step": 6130
    },
    {
      "epoch": 3.2154088050314464,
      "grad_norm": 6.0625,
      "learning_rate": 4.2893021085485765e-05,
      "loss": 0.5571,
      "num_input_tokens_seen": 4011832,
      "step": 6135
    },
    {
      "epoch": 3.2180293501048216,
      "grad_norm": 3.390625,
      "learning_rate": 4.287704250169086e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 4015032,
      "step": 6140
    },
    {
      "epoch": 3.220649895178197,
      "grad_norm": 6.46875,
      "learning_rate": 4.2861048959203386e-05,
      "loss": 0.6216,
      "num_input_tokens_seen": 4017432,
      "step": 6145
    },
    {
      "epoch": 3.2232704402515724,
      "grad_norm": 9.8125,
      "learning_rate": 4.284504047140599e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 4020216,
      "step": 6150
    },
    {
      "epoch": 3.2258909853249476,
      "grad_norm": 12.5625,
      "learning_rate": 4.282901705169387e-05,
      "loss": 0.5587,
      "num_input_tokens_seen": 4023096,
      "step": 6155
    },
    {
      "epoch": 3.2285115303983227,
      "grad_norm": 7.875,
      "learning_rate": 4.281297871347468e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 4026712,
      "step": 6160
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 7.6875,
      "learning_rate": 4.279692547016856e-05,
      "loss": 0.5416,
      "num_input_tokens_seen": 4030872,
      "step": 6165
    },
    {
      "epoch": 3.2337526205450735,
      "grad_norm": 5.9375,
      "learning_rate": 4.278085733520814e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 4034744,
      "step": 6170
    },
    {
      "epoch": 3.2363731656184487,
      "grad_norm": 5.6875,
      "learning_rate": 4.2764774322038494e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 4037752,
      "step": 6175
    },
    {
      "epoch": 3.238993710691824,
      "grad_norm": 16.875,
      "learning_rate": 4.2748676444117156e-05,
      "loss": 0.4609,
      "num_input_tokens_seen": 4041048,
      "step": 6180
    },
    {
      "epoch": 3.241614255765199,
      "grad_norm": 6.6875,
      "learning_rate": 4.27325637149141e-05,
      "loss": 0.5948,
      "num_input_tokens_seen": 4043832,
      "step": 6185
    },
    {
      "epoch": 3.2442348008385746,
      "grad_norm": 4.03125,
      "learning_rate": 4.271643614791172e-05,
      "loss": 0.4477,
      "num_input_tokens_seen": 4047512,
      "step": 6190
    },
    {
      "epoch": 3.24685534591195,
      "grad_norm": 4.78125,
      "learning_rate": 4.2700293756604824e-05,
      "loss": 0.4928,
      "num_input_tokens_seen": 4050584,
      "step": 6195
    },
    {
      "epoch": 3.249475890985325,
      "grad_norm": 4.75,
      "learning_rate": 4.268413655450064e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 4053752,
      "step": 6200
    },
    {
      "epoch": 3.2520964360587,
      "grad_norm": 6.03125,
      "learning_rate": 4.266796455511875e-05,
      "loss": 0.5371,
      "num_input_tokens_seen": 4056824,
      "step": 6205
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 16.375,
      "learning_rate": 4.2651777771991176e-05,
      "loss": 0.6742,
      "num_input_tokens_seen": 4059864,
      "step": 6210
    },
    {
      "epoch": 3.257337526205451,
      "grad_norm": 5.9375,
      "learning_rate": 4.2635576218662257e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 4063416,
      "step": 6215
    },
    {
      "epoch": 3.259958071278826,
      "grad_norm": 4.0,
      "learning_rate": 4.261935990868871e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 4066264,
      "step": 6220
    },
    {
      "epoch": 3.2625786163522013,
      "grad_norm": 3.796875,
      "learning_rate": 4.260312885563962e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 4068664,
      "step": 6225
    },
    {
      "epoch": 3.2651991614255764,
      "grad_norm": 5.90625,
      "learning_rate": 4.2586883073096386e-05,
      "loss": 0.7138,
      "num_input_tokens_seen": 4072088,
      "step": 6230
    },
    {
      "epoch": 3.2678197064989516,
      "grad_norm": 12.8125,
      "learning_rate": 4.257062257465272e-05,
      "loss": 0.6692,
      "num_input_tokens_seen": 4075064,
      "step": 6235
    },
    {
      "epoch": 3.270440251572327,
      "grad_norm": 6.25,
      "learning_rate": 4.255434737391469e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 4077752,
      "step": 6240
    },
    {
      "epoch": 3.2730607966457024,
      "grad_norm": 5.21875,
      "learning_rate": 4.2538057484500624e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 4080472,
      "step": 6245
    },
    {
      "epoch": 3.2756813417190775,
      "grad_norm": 5.46875,
      "learning_rate": 4.2521752920041155e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 4083384,
      "step": 6250
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 3.03125,
      "learning_rate": 4.2505433694179216e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 4086488,
      "step": 6255
    },
    {
      "epoch": 3.280922431865828,
      "grad_norm": 4.21875,
      "learning_rate": 4.2489099820569974e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 4089464,
      "step": 6260
    },
    {
      "epoch": 3.2835429769392035,
      "grad_norm": 4.53125,
      "learning_rate": 4.247275131288086e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 4092088,
      "step": 6265
    },
    {
      "epoch": 3.2861635220125787,
      "grad_norm": 5.21875,
      "learning_rate": 4.2456388184791584e-05,
      "loss": 0.4831,
      "num_input_tokens_seen": 4095544,
      "step": 6270
    },
    {
      "epoch": 3.288784067085954,
      "grad_norm": 3.9375,
      "learning_rate": 4.2440010449994054e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 4098264,
      "step": 6275
    },
    {
      "epoch": 3.291404612159329,
      "grad_norm": 3.828125,
      "learning_rate": 4.24236181221924e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 4100792,
      "step": 6280
    },
    {
      "epoch": 3.2940251572327046,
      "grad_norm": 3.859375,
      "learning_rate": 4.240721121510298e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 4103896,
      "step": 6285
    },
    {
      "epoch": 3.29664570230608,
      "grad_norm": 4.21875,
      "learning_rate": 4.2390789742454354e-05,
      "loss": 0.3888,
      "num_input_tokens_seen": 4106968,
      "step": 6290
    },
    {
      "epoch": 3.299266247379455,
      "grad_norm": 2.84375,
      "learning_rate": 4.2374353717987244e-05,
      "loss": 0.5586,
      "num_input_tokens_seen": 4110040,
      "step": 6295
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 4.5625,
      "learning_rate": 4.235790315545457e-05,
      "loss": 0.4555,
      "num_input_tokens_seen": 4112664,
      "step": 6300
    },
    {
      "epoch": 3.3045073375262053,
      "grad_norm": 8.0625,
      "learning_rate": 4.234143806862141e-05,
      "loss": 0.4941,
      "num_input_tokens_seen": 4115832,
      "step": 6305
    },
    {
      "epoch": 3.307127882599581,
      "grad_norm": 5.9375,
      "learning_rate": 4.2324958471265006e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 4119160,
      "step": 6310
    },
    {
      "epoch": 3.309748427672956,
      "grad_norm": 5.28125,
      "learning_rate": 4.230846437717472e-05,
      "loss": 0.853,
      "num_input_tokens_seen": 4122520,
      "step": 6315
    },
    {
      "epoch": 3.3123689727463312,
      "grad_norm": 6.3125,
      "learning_rate": 4.2291955800152063e-05,
      "loss": 0.4779,
      "num_input_tokens_seen": 4124888,
      "step": 6320
    },
    {
      "epoch": 3.3149895178197064,
      "grad_norm": 6.34375,
      "learning_rate": 4.2275432754010663e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 4127832,
      "step": 6325
    },
    {
      "epoch": 3.3176100628930816,
      "grad_norm": 4.8125,
      "learning_rate": 4.225889525257624e-05,
      "loss": 0.6314,
      "num_input_tokens_seen": 4130232,
      "step": 6330
    },
    {
      "epoch": 3.320230607966457,
      "grad_norm": 4.1875,
      "learning_rate": 4.224234330968663e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 4133080,
      "step": 6335
    },
    {
      "epoch": 3.3228511530398324,
      "grad_norm": 2.578125,
      "learning_rate": 4.222577693919173e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 4136536,
      "step": 6340
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 4.28125,
      "learning_rate": 4.2209196154953536e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 4139608,
      "step": 6345
    },
    {
      "epoch": 3.3280922431865827,
      "grad_norm": 10.6875,
      "learning_rate": 4.219260097084608e-05,
      "loss": 0.6319,
      "num_input_tokens_seen": 4142200,
      "step": 6350
    },
    {
      "epoch": 3.330712788259958,
      "grad_norm": 8.0625,
      "learning_rate": 4.217599140075546e-05,
      "loss": 0.8329,
      "num_input_tokens_seen": 4146200,
      "step": 6355
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.40625,
      "learning_rate": 4.2159367458579793e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 4149464,
      "step": 6360
    },
    {
      "epoch": 3.3359538784067087,
      "grad_norm": 4.0625,
      "learning_rate": 4.2142729158229256e-05,
      "loss": 0.5786,
      "num_input_tokens_seen": 4151992,
      "step": 6365
    },
    {
      "epoch": 3.338574423480084,
      "grad_norm": 3.578125,
      "learning_rate": 4.2126076513626004e-05,
      "loss": 0.5055,
      "num_input_tokens_seen": 4155576,
      "step": 6370
    },
    {
      "epoch": 3.341194968553459,
      "grad_norm": 6.1875,
      "learning_rate": 4.210940953870422e-05,
      "loss": 0.617,
      "num_input_tokens_seen": 4159672,
      "step": 6375
    },
    {
      "epoch": 3.3438155136268346,
      "grad_norm": 4.46875,
      "learning_rate": 4.209272824741005e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 4162360,
      "step": 6380
    },
    {
      "epoch": 3.3464360587002098,
      "grad_norm": 7.1875,
      "learning_rate": 4.207603265370166e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 4165304,
      "step": 6385
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 5.5625,
      "learning_rate": 4.205932277154914e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 4168728,
      "step": 6390
    },
    {
      "epoch": 3.35167714884696,
      "grad_norm": 6.96875,
      "learning_rate": 4.204259861493457e-05,
      "loss": 0.3903,
      "num_input_tokens_seen": 4171960,
      "step": 6395
    },
    {
      "epoch": 3.3542976939203353,
      "grad_norm": 2.6875,
      "learning_rate": 4.202586019785194e-05,
      "loss": 0.4897,
      "num_input_tokens_seen": 4175544,
      "step": 6400
    },
    {
      "epoch": 3.3569182389937104,
      "grad_norm": 4.625,
      "learning_rate": 4.2009107534307214e-05,
      "loss": 0.4074,
      "num_input_tokens_seen": 4178264,
      "step": 6405
    },
    {
      "epoch": 3.359538784067086,
      "grad_norm": 7.84375,
      "learning_rate": 4.199234063831825e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 4180984,
      "step": 6410
    },
    {
      "epoch": 3.3621593291404612,
      "grad_norm": 4.5,
      "learning_rate": 4.197555952391482e-05,
      "loss": 0.6093,
      "num_input_tokens_seen": 4183768,
      "step": 6415
    },
    {
      "epoch": 3.3647798742138364,
      "grad_norm": 4.40625,
      "learning_rate": 4.195876420513859e-05,
      "loss": 0.4592,
      "num_input_tokens_seen": 4187000,
      "step": 6420
    },
    {
      "epoch": 3.3674004192872116,
      "grad_norm": 9.1875,
      "learning_rate": 4.194195469604312e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 4189304,
      "step": 6425
    },
    {
      "epoch": 3.370020964360587,
      "grad_norm": 2.78125,
      "learning_rate": 4.192513101069383e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 4193080,
      "step": 6430
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 6.25,
      "learning_rate": 4.190829316316803e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 4195608,
      "step": 6435
    },
    {
      "epoch": 3.3752620545073375,
      "grad_norm": 6.21875,
      "learning_rate": 4.189144116755485e-05,
      "loss": 0.4021,
      "num_input_tokens_seen": 4198744,
      "step": 6440
    },
    {
      "epoch": 3.3778825995807127,
      "grad_norm": 7.375,
      "learning_rate": 4.187457503795527e-05,
      "loss": 0.7412,
      "num_input_tokens_seen": 4202680,
      "step": 6445
    },
    {
      "epoch": 3.380503144654088,
      "grad_norm": 6.6875,
      "learning_rate": 4.1857694788482094e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 4205144,
      "step": 6450
    },
    {
      "epoch": 3.3831236897274635,
      "grad_norm": 4.84375,
      "learning_rate": 4.184080043325995e-05,
      "loss": 0.4794,
      "num_input_tokens_seen": 4209112,
      "step": 6455
    },
    {
      "epoch": 3.3857442348008386,
      "grad_norm": 5.78125,
      "learning_rate": 4.1823891986425256e-05,
      "loss": 0.4987,
      "num_input_tokens_seen": 4213048,
      "step": 6460
    },
    {
      "epoch": 3.388364779874214,
      "grad_norm": 2.703125,
      "learning_rate": 4.180696946212624e-05,
      "loss": 0.4488,
      "num_input_tokens_seen": 4216472,
      "step": 6465
    },
    {
      "epoch": 3.390985324947589,
      "grad_norm": 5.46875,
      "learning_rate": 4.1790032874522885e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 4219672,
      "step": 6470
    },
    {
      "epoch": 3.3936058700209646,
      "grad_norm": 6.125,
      "learning_rate": 4.177308223778696e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 4222872,
      "step": 6475
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 3.4375,
      "learning_rate": 4.175611756610198e-05,
      "loss": 0.5446,
      "num_input_tokens_seen": 4225528,
      "step": 6480
    },
    {
      "epoch": 3.398846960167715,
      "grad_norm": 4.71875,
      "learning_rate": 4.173913887366322e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 4229144,
      "step": 6485
    },
    {
      "epoch": 3.40146750524109,
      "grad_norm": 5.75,
      "learning_rate": 4.172214617467765e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 4232792,
      "step": 6490
    },
    {
      "epoch": 3.4040880503144653,
      "grad_norm": 3.8125,
      "learning_rate": 4.1705139483364e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 4236600,
      "step": 6495
    },
    {
      "epoch": 3.4067085953878404,
      "grad_norm": 5.65625,
      "learning_rate": 4.1688118813952706e-05,
      "loss": 0.6633,
      "num_input_tokens_seen": 4239896,
      "step": 6500
    },
    {
      "epoch": 3.409329140461216,
      "grad_norm": 3.421875,
      "learning_rate": 4.167108418068585e-05,
      "loss": 0.5172,
      "num_input_tokens_seen": 4243640,
      "step": 6505
    },
    {
      "epoch": 3.411949685534591,
      "grad_norm": 3.671875,
      "learning_rate": 4.165403559781727e-05,
      "loss": 0.5813,
      "num_input_tokens_seen": 4247160,
      "step": 6510
    },
    {
      "epoch": 3.4145702306079664,
      "grad_norm": 4.96875,
      "learning_rate": 4.163697307961242e-05,
      "loss": 0.5987,
      "num_input_tokens_seen": 4251032,
      "step": 6515
    },
    {
      "epoch": 3.4171907756813416,
      "grad_norm": 8.625,
      "learning_rate": 4.1619896640348445e-05,
      "loss": 0.5672,
      "num_input_tokens_seen": 4254680,
      "step": 6520
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 4.46875,
      "learning_rate": 4.160280629431413e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 4257720,
      "step": 6525
    },
    {
      "epoch": 3.4224318658280923,
      "grad_norm": 4.0625,
      "learning_rate": 4.158570205580989e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 4260696,
      "step": 6530
    },
    {
      "epoch": 3.4250524109014675,
      "grad_norm": 5.71875,
      "learning_rate": 4.156858393914779e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 4263416,
      "step": 6535
    },
    {
      "epoch": 3.4276729559748427,
      "grad_norm": 4.4375,
      "learning_rate": 4.1551451958651455e-05,
      "loss": 0.6098,
      "num_input_tokens_seen": 4266232,
      "step": 6540
    },
    {
      "epoch": 3.430293501048218,
      "grad_norm": 16.0,
      "learning_rate": 4.153430612865616e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 4269400,
      "step": 6545
    },
    {
      "epoch": 3.4329140461215935,
      "grad_norm": 6.53125,
      "learning_rate": 4.151714646350876e-05,
      "loss": 0.4936,
      "num_input_tokens_seen": 4271896,
      "step": 6550
    },
    {
      "epoch": 3.4355345911949686,
      "grad_norm": 4.90625,
      "learning_rate": 4.149997297756767e-05,
      "loss": 0.6816,
      "num_input_tokens_seen": 4275416,
      "step": 6555
    },
    {
      "epoch": 3.438155136268344,
      "grad_norm": 3.640625,
      "learning_rate": 4.148278568520289e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 4278936,
      "step": 6560
    },
    {
      "epoch": 3.440775681341719,
      "grad_norm": 4.75,
      "learning_rate": 4.146558460079595e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 4282808,
      "step": 6565
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 5.0,
      "learning_rate": 4.1448369738739923e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 4285400,
      "step": 6570
    },
    {
      "epoch": 3.4460167714884697,
      "grad_norm": 8.8125,
      "learning_rate": 4.143114111343944e-05,
      "loss": 0.4101,
      "num_input_tokens_seen": 4288280,
      "step": 6575
    },
    {
      "epoch": 3.448637316561845,
      "grad_norm": 3.734375,
      "learning_rate": 4.1413898739310605e-05,
      "loss": 0.4945,
      "num_input_tokens_seen": 4291576,
      "step": 6580
    },
    {
      "epoch": 3.45125786163522,
      "grad_norm": 3.96875,
      "learning_rate": 4.1396642630781076e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 4295288,
      "step": 6585
    },
    {
      "epoch": 3.4538784067085953,
      "grad_norm": 5.15625,
      "learning_rate": 4.137937280228996e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 4298808,
      "step": 6590
    },
    {
      "epoch": 3.4564989517819704,
      "grad_norm": 3.4375,
      "learning_rate": 4.136208926828786e-05,
      "loss": 0.3802,
      "num_input_tokens_seen": 4304280,
      "step": 6595
    },
    {
      "epoch": 3.459119496855346,
      "grad_norm": 6.25,
      "learning_rate": 4.134479204323685e-05,
      "loss": 0.7073,
      "num_input_tokens_seen": 4306616,
      "step": 6600
    },
    {
      "epoch": 3.461740041928721,
      "grad_norm": 7.34375,
      "learning_rate": 4.132748114161046e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 4309112,
      "step": 6605
    },
    {
      "epoch": 3.4643605870020964,
      "grad_norm": 12.5625,
      "learning_rate": 4.131015657789365e-05,
      "loss": 0.572,
      "num_input_tokens_seen": 4311832,
      "step": 6610
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 1.1953125,
      "learning_rate": 4.129281836658285e-05,
      "loss": 0.4818,
      "num_input_tokens_seen": 4318328,
      "step": 6615
    },
    {
      "epoch": 3.469601677148847,
      "grad_norm": 7.65625,
      "learning_rate": 4.127546652218586e-05,
      "loss": 0.4718,
      "num_input_tokens_seen": 4321848,
      "step": 6620
    },
    {
      "epoch": 3.4722222222222223,
      "grad_norm": 2.34375,
      "learning_rate": 4.1258101059221914e-05,
      "loss": 0.3578,
      "num_input_tokens_seen": 4324984,
      "step": 6625
    },
    {
      "epoch": 3.4748427672955975,
      "grad_norm": 7.21875,
      "learning_rate": 4.124072199222165e-05,
      "loss": 0.4948,
      "num_input_tokens_seen": 4327704,
      "step": 6630
    },
    {
      "epoch": 3.4774633123689727,
      "grad_norm": 4.65625,
      "learning_rate": 4.122332933572707e-05,
      "loss": 0.4373,
      "num_input_tokens_seen": 4330904,
      "step": 6635
    },
    {
      "epoch": 3.480083857442348,
      "grad_norm": 7.34375,
      "learning_rate": 4.120592310429154e-05,
      "loss": 0.5442,
      "num_input_tokens_seen": 4333592,
      "step": 6640
    },
    {
      "epoch": 3.4827044025157234,
      "grad_norm": 7.5625,
      "learning_rate": 4.118850331247982e-05,
      "loss": 0.4597,
      "num_input_tokens_seen": 4336696,
      "step": 6645
    },
    {
      "epoch": 3.4853249475890986,
      "grad_norm": 17.375,
      "learning_rate": 4.1171069974868e-05,
      "loss": 0.7559,
      "num_input_tokens_seen": 4340088,
      "step": 6650
    },
    {
      "epoch": 3.487945492662474,
      "grad_norm": 4.9375,
      "learning_rate": 4.115362310604347e-05,
      "loss": 0.5666,
      "num_input_tokens_seen": 4343640,
      "step": 6655
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 2.6875,
      "learning_rate": 4.113616272060501e-05,
      "loss": 0.4592,
      "num_input_tokens_seen": 4348568,
      "step": 6660
    },
    {
      "epoch": 3.4931865828092246,
      "grad_norm": 4.3125,
      "learning_rate": 4.111868883316266e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 4352376,
      "step": 6665
    },
    {
      "epoch": 3.4958071278825997,
      "grad_norm": 4.09375,
      "learning_rate": 4.110120145833775e-05,
      "loss": 0.5341,
      "num_input_tokens_seen": 4356120,
      "step": 6670
    },
    {
      "epoch": 3.498427672955975,
      "grad_norm": 3.421875,
      "learning_rate": 4.108370061076294e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 4359448,
      "step": 6675
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.5015654563903809,
      "eval_runtime": 24.1746,
      "eval_samples_per_second": 35.078,
      "eval_steps_per_second": 8.77,
      "num_input_tokens_seen": 4361944,
      "step": 6678
    },
    {
      "epoch": 3.50104821802935,
      "grad_norm": 4.6875,
      "learning_rate": 4.106618630508213e-05,
      "loss": 0.3697,
      "num_input_tokens_seen": 4362936,
      "step": 6680
    },
    {
      "epoch": 3.5036687631027252,
      "grad_norm": 5.8125,
      "learning_rate": 4.10486585559505e-05,
      "loss": 0.449,
      "num_input_tokens_seen": 4366968,
      "step": 6685
    },
    {
      "epoch": 3.5062893081761004,
      "grad_norm": 4.21875,
      "learning_rate": 4.103111737803446e-05,
      "loss": 0.3755,
      "num_input_tokens_seen": 4370264,
      "step": 6690
    },
    {
      "epoch": 3.508909853249476,
      "grad_norm": 4.40625,
      "learning_rate": 4.101356278601167e-05,
      "loss": 0.3547,
      "num_input_tokens_seen": 4373432,
      "step": 6695
    },
    {
      "epoch": 3.511530398322851,
      "grad_norm": 3.734375,
      "learning_rate": 4.0995994794571015e-05,
      "loss": 0.6424,
      "num_input_tokens_seen": 4376824,
      "step": 6700
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 3.109375,
      "learning_rate": 4.0978413418412574e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 4380504,
      "step": 6705
    },
    {
      "epoch": 3.5167714884696015,
      "grad_norm": 4.3125,
      "learning_rate": 4.0960818672247656e-05,
      "loss": 0.4705,
      "num_input_tokens_seen": 4383704,
      "step": 6710
    },
    {
      "epoch": 3.519392033542977,
      "grad_norm": 6.21875,
      "learning_rate": 4.094321057079874e-05,
      "loss": 0.6087,
      "num_input_tokens_seen": 4387128,
      "step": 6715
    },
    {
      "epoch": 3.5220125786163523,
      "grad_norm": 4.75,
      "learning_rate": 4.092558912879947e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 4390296,
      "step": 6720
    },
    {
      "epoch": 3.5246331236897275,
      "grad_norm": 5.3125,
      "learning_rate": 4.090795436099466e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 4393624,
      "step": 6725
    },
    {
      "epoch": 3.5272536687631026,
      "grad_norm": 5.34375,
      "learning_rate": 4.089030628214029e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 4397944,
      "step": 6730
    },
    {
      "epoch": 3.529874213836478,
      "grad_norm": 4.84375,
      "learning_rate": 4.0872644907003476e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 4400696,
      "step": 6735
    },
    {
      "epoch": 3.532494758909853,
      "grad_norm": 8.5,
      "learning_rate": 4.0854970250362426e-05,
      "loss": 0.5025,
      "num_input_tokens_seen": 4403416,
      "step": 6740
    },
    {
      "epoch": 3.5351153039832286,
      "grad_norm": 3.75,
      "learning_rate": 4.0837282327006495e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 4407320,
      "step": 6745
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 5.21875,
      "learning_rate": 4.081958115173614e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 4409848,
      "step": 6750
    },
    {
      "epoch": 3.540356394129979,
      "grad_norm": 4.4375,
      "learning_rate": 4.080186673936288e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 4412920,
      "step": 6755
    },
    {
      "epoch": 3.5429769392033545,
      "grad_norm": 8.0,
      "learning_rate": 4.078413910470934e-05,
      "loss": 0.422,
      "num_input_tokens_seen": 4415384,
      "step": 6760
    },
    {
      "epoch": 3.5455974842767297,
      "grad_norm": 6.9375,
      "learning_rate": 4.076639826260919e-05,
      "loss": 0.4219,
      "num_input_tokens_seen": 4418040,
      "step": 6765
    },
    {
      "epoch": 3.548218029350105,
      "grad_norm": 4.1875,
      "learning_rate": 4.074864422790714e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 4421464,
      "step": 6770
    },
    {
      "epoch": 3.55083857442348,
      "grad_norm": 5.1875,
      "learning_rate": 4.073087701545897e-05,
      "loss": 0.5137,
      "num_input_tokens_seen": 4423672,
      "step": 6775
    },
    {
      "epoch": 3.5534591194968552,
      "grad_norm": 7.96875,
      "learning_rate": 4.071309664013148e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 4426648,
      "step": 6780
    },
    {
      "epoch": 3.5560796645702304,
      "grad_norm": 7.9375,
      "learning_rate": 4.069530311680247e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 4429752,
      "step": 6785
    },
    {
      "epoch": 3.558700209643606,
      "grad_norm": 5.9375,
      "learning_rate": 4.0677496460360734e-05,
      "loss": 0.6168,
      "num_input_tokens_seen": 4433784,
      "step": 6790
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 5.8125,
      "learning_rate": 4.0659676685706084e-05,
      "loss": 0.5262,
      "num_input_tokens_seen": 4436376,
      "step": 6795
    },
    {
      "epoch": 3.5639412997903563,
      "grad_norm": 10.625,
      "learning_rate": 4.064184380774929e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 4439384,
      "step": 6800
    },
    {
      "epoch": 3.5665618448637315,
      "grad_norm": 13.125,
      "learning_rate": 4.062399784141209e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 4442712,
      "step": 6805
    },
    {
      "epoch": 3.569182389937107,
      "grad_norm": 4.25,
      "learning_rate": 4.060613880162717e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 4446936,
      "step": 6810
    },
    {
      "epoch": 3.5718029350104823,
      "grad_norm": 5.84375,
      "learning_rate": 4.0588266703338164e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 4450328,
      "step": 6815
    },
    {
      "epoch": 3.5744234800838575,
      "grad_norm": 3.265625,
      "learning_rate": 4.057038156149961e-05,
      "loss": 0.452,
      "num_input_tokens_seen": 4453816,
      "step": 6820
    },
    {
      "epoch": 3.5770440251572326,
      "grad_norm": 2.90625,
      "learning_rate": 4.055248339107701e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 4458072,
      "step": 6825
    },
    {
      "epoch": 3.579664570230608,
      "grad_norm": 3.671875,
      "learning_rate": 4.053457220704671e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 4461144,
      "step": 6830
    },
    {
      "epoch": 3.582285115303983,
      "grad_norm": 4.78125,
      "learning_rate": 4.0516648024395974e-05,
      "loss": 0.3802,
      "num_input_tokens_seen": 4463928,
      "step": 6835
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 7.03125,
      "learning_rate": 4.049871085812295e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 4466488,
      "step": 6840
    },
    {
      "epoch": 3.5875262054507338,
      "grad_norm": 3.6875,
      "learning_rate": 4.0480760723236633e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 4470104,
      "step": 6845
    },
    {
      "epoch": 3.590146750524109,
      "grad_norm": 5.46875,
      "learning_rate": 4.046279763475687e-05,
      "loss": 0.5985,
      "num_input_tokens_seen": 4472184,
      "step": 6850
    },
    {
      "epoch": 3.5927672955974845,
      "grad_norm": 6.5625,
      "learning_rate": 4.0444821607714366e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 4474840,
      "step": 6855
    },
    {
      "epoch": 3.5953878406708597,
      "grad_norm": 8.6875,
      "learning_rate": 4.042683265715063e-05,
      "loss": 0.573,
      "num_input_tokens_seen": 4477816,
      "step": 6860
    },
    {
      "epoch": 3.598008385744235,
      "grad_norm": 5.15625,
      "learning_rate": 4.040883079811799e-05,
      "loss": 0.5735,
      "num_input_tokens_seen": 4481816,
      "step": 6865
    },
    {
      "epoch": 3.60062893081761,
      "grad_norm": 4.6875,
      "learning_rate": 4.039081604567959e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 4485464,
      "step": 6870
    },
    {
      "epoch": 3.603249475890985,
      "grad_norm": 4.125,
      "learning_rate": 4.037278841490933e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 4490392,
      "step": 6875
    },
    {
      "epoch": 3.6058700209643604,
      "grad_norm": 7.59375,
      "learning_rate": 4.0354747920891954e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 4494712,
      "step": 6880
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 4.34375,
      "learning_rate": 4.033669457872288e-05,
      "loss": 0.5447,
      "num_input_tokens_seen": 4497528,
      "step": 6885
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 2.984375,
      "learning_rate": 4.0318628403508336e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 4500536,
      "step": 6890
    },
    {
      "epoch": 3.6137316561844863,
      "grad_norm": 4.125,
      "learning_rate": 4.0300549410365276e-05,
      "loss": 0.521,
      "num_input_tokens_seen": 4504056,
      "step": 6895
    },
    {
      "epoch": 3.6163522012578615,
      "grad_norm": 6.0,
      "learning_rate": 4.0282457614421364e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 4506936,
      "step": 6900
    },
    {
      "epoch": 3.618972746331237,
      "grad_norm": 3.96875,
      "learning_rate": 4.0264353030814996e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 4509560,
      "step": 6905
    },
    {
      "epoch": 3.6215932914046123,
      "grad_norm": 5.875,
      "learning_rate": 4.0246235674695255e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 4513240,
      "step": 6910
    },
    {
      "epoch": 3.6242138364779874,
      "grad_norm": 5.21875,
      "learning_rate": 4.022810556122193e-05,
      "loss": 0.4845,
      "num_input_tokens_seen": 4515640,
      "step": 6915
    },
    {
      "epoch": 3.6268343815513626,
      "grad_norm": 2.4375,
      "learning_rate": 4.020996270556546e-05,
      "loss": 0.4307,
      "num_input_tokens_seen": 4520824,
      "step": 6920
    },
    {
      "epoch": 3.629454926624738,
      "grad_norm": 4.15625,
      "learning_rate": 4.0191807122906964e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 4523896,
      "step": 6925
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 5.5,
      "learning_rate": 4.01736388284382e-05,
      "loss": 0.6313,
      "num_input_tokens_seen": 4527000,
      "step": 6930
    },
    {
      "epoch": 3.6346960167714886,
      "grad_norm": 7.5625,
      "learning_rate": 4.015545783736157e-05,
      "loss": 0.6038,
      "num_input_tokens_seen": 4529912,
      "step": 6935
    },
    {
      "epoch": 3.6373165618448637,
      "grad_norm": 8.625,
      "learning_rate": 4.013726416489009e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 4533336,
      "step": 6940
    },
    {
      "epoch": 3.639937106918239,
      "grad_norm": 4.84375,
      "learning_rate": 4.01190578262474e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 4536344,
      "step": 6945
    },
    {
      "epoch": 3.6425576519916145,
      "grad_norm": 3.390625,
      "learning_rate": 4.0100838836667735e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 4540152,
      "step": 6950
    },
    {
      "epoch": 3.6451781970649897,
      "grad_norm": 4.5625,
      "learning_rate": 4.0082607211395904e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 4543032,
      "step": 6955
    },
    {
      "epoch": 3.647798742138365,
      "grad_norm": 6.15625,
      "learning_rate": 4.006436296568731e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 4547064,
      "step": 6960
    },
    {
      "epoch": 3.65041928721174,
      "grad_norm": 3.8125,
      "learning_rate": 4.00461061148079e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 4551000,
      "step": 6965
    },
    {
      "epoch": 3.653039832285115,
      "grad_norm": 2.546875,
      "learning_rate": 4.0027836674034174e-05,
      "loss": 0.5634,
      "num_input_tokens_seen": 4555000,
      "step": 6970
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 3.15625,
      "learning_rate": 4.000955465865316e-05,
      "loss": 0.427,
      "num_input_tokens_seen": 4558680,
      "step": 6975
    },
    {
      "epoch": 3.658280922431866,
      "grad_norm": 4.6875,
      "learning_rate": 3.999126008396242e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 4561336,
      "step": 6980
    },
    {
      "epoch": 3.660901467505241,
      "grad_norm": 4.78125,
      "learning_rate": 3.9972952965270006e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 4564568,
      "step": 6985
    },
    {
      "epoch": 3.6635220125786163,
      "grad_norm": 5.25,
      "learning_rate": 3.9954633317894496e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 4567448,
      "step": 6990
    },
    {
      "epoch": 3.6661425576519915,
      "grad_norm": 4.28125,
      "learning_rate": 3.9936301157164926e-05,
      "loss": 0.6044,
      "num_input_tokens_seen": 4570136,
      "step": 6995
    },
    {
      "epoch": 3.668763102725367,
      "grad_norm": 3.421875,
      "learning_rate": 3.99179564984208e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 4573624,
      "step": 7000
    },
    {
      "epoch": 3.6713836477987423,
      "grad_norm": 21.75,
      "learning_rate": 3.989959935701211e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 4577048,
      "step": 7005
    },
    {
      "epoch": 3.6740041928721174,
      "grad_norm": 5.03125,
      "learning_rate": 3.988122974829926e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 4580088,
      "step": 7010
    },
    {
      "epoch": 3.6766247379454926,
      "grad_norm": 4.0625,
      "learning_rate": 3.9862847687653116e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 4583480,
      "step": 7015
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 7.125,
      "learning_rate": 3.9844453190454924e-05,
      "loss": 0.6525,
      "num_input_tokens_seen": 4586168,
      "step": 7020
    },
    {
      "epoch": 3.681865828092243,
      "grad_norm": 3.78125,
      "learning_rate": 3.982604627209637e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 4589592,
      "step": 7025
    },
    {
      "epoch": 3.6844863731656186,
      "grad_norm": 5.96875,
      "learning_rate": 3.980762694797953e-05,
      "loss": 0.5559,
      "num_input_tokens_seen": 4592440,
      "step": 7030
    },
    {
      "epoch": 3.6871069182389937,
      "grad_norm": 6.1875,
      "learning_rate": 3.978919523351684e-05,
      "loss": 0.3709,
      "num_input_tokens_seen": 4594904,
      "step": 7035
    },
    {
      "epoch": 3.689727463312369,
      "grad_norm": 5.90625,
      "learning_rate": 3.977075114413112e-05,
      "loss": 0.5557,
      "num_input_tokens_seen": 4597624,
      "step": 7040
    },
    {
      "epoch": 3.6923480083857445,
      "grad_norm": 9.3125,
      "learning_rate": 3.9752294695255545e-05,
      "loss": 0.5824,
      "num_input_tokens_seen": 4599928,
      "step": 7045
    },
    {
      "epoch": 3.6949685534591197,
      "grad_norm": 12.25,
      "learning_rate": 3.973382590233362e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 4602328,
      "step": 7050
    },
    {
      "epoch": 3.697589098532495,
      "grad_norm": 4.6875,
      "learning_rate": 3.9715344780819205e-05,
      "loss": 0.5108,
      "num_input_tokens_seen": 4606328,
      "step": 7055
    },
    {
      "epoch": 3.70020964360587,
      "grad_norm": 4.75,
      "learning_rate": 3.9696851346176445e-05,
      "loss": 0.418,
      "num_input_tokens_seen": 4609176,
      "step": 7060
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 3.453125,
      "learning_rate": 3.9678345613879796e-05,
      "loss": 0.5771,
      "num_input_tokens_seen": 4612056,
      "step": 7065
    },
    {
      "epoch": 3.7054507337526204,
      "grad_norm": 7.375,
      "learning_rate": 3.965982759941403e-05,
      "loss": 0.6218,
      "num_input_tokens_seen": 4615320,
      "step": 7070
    },
    {
      "epoch": 3.708071278825996,
      "grad_norm": 4.78125,
      "learning_rate": 3.964129731827415e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 4618200,
      "step": 7075
    },
    {
      "epoch": 3.710691823899371,
      "grad_norm": 4.71875,
      "learning_rate": 3.9622754785965474e-05,
      "loss": 0.3621,
      "num_input_tokens_seen": 4621368,
      "step": 7080
    },
    {
      "epoch": 3.7133123689727463,
      "grad_norm": 3.40625,
      "learning_rate": 3.9604200018003525e-05,
      "loss": 0.5277,
      "num_input_tokens_seen": 4624088,
      "step": 7085
    },
    {
      "epoch": 3.7159329140461215,
      "grad_norm": 4.75,
      "learning_rate": 3.95856330299141e-05,
      "loss": 0.6184,
      "num_input_tokens_seen": 4628568,
      "step": 7090
    },
    {
      "epoch": 3.718553459119497,
      "grad_norm": 4.90625,
      "learning_rate": 3.956705383723319e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 4631288,
      "step": 7095
    },
    {
      "epoch": 3.7211740041928723,
      "grad_norm": 3.546875,
      "learning_rate": 3.954846245550704e-05,
      "loss": 0.3999,
      "num_input_tokens_seen": 4635224,
      "step": 7100
    },
    {
      "epoch": 3.7237945492662474,
      "grad_norm": 4.25,
      "learning_rate": 3.952985890029205e-05,
      "loss": 0.708,
      "num_input_tokens_seen": 4638648,
      "step": 7105
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 11.9375,
      "learning_rate": 3.951124318715482e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 4641848,
      "step": 7110
    },
    {
      "epoch": 3.7290356394129978,
      "grad_norm": 4.75,
      "learning_rate": 3.9492615331672145e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 4644408,
      "step": 7115
    },
    {
      "epoch": 3.731656184486373,
      "grad_norm": 4.0625,
      "learning_rate": 3.947397534943096e-05,
      "loss": 0.6413,
      "num_input_tokens_seen": 4647096,
      "step": 7120
    },
    {
      "epoch": 3.7342767295597485,
      "grad_norm": 4.75,
      "learning_rate": 3.9455323256028344e-05,
      "loss": 0.5044,
      "num_input_tokens_seen": 4649176,
      "step": 7125
    },
    {
      "epoch": 3.7368972746331237,
      "grad_norm": 5.375,
      "learning_rate": 3.943665906707153e-05,
      "loss": 0.5205,
      "num_input_tokens_seen": 4653144,
      "step": 7130
    },
    {
      "epoch": 3.739517819706499,
      "grad_norm": 4.8125,
      "learning_rate": 3.9417982798177834e-05,
      "loss": 0.3733,
      "num_input_tokens_seen": 4655864,
      "step": 7135
    },
    {
      "epoch": 3.742138364779874,
      "grad_norm": 5.375,
      "learning_rate": 3.939929446497472e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 4659000,
      "step": 7140
    },
    {
      "epoch": 3.7447589098532497,
      "grad_norm": 2.484375,
      "learning_rate": 3.938059408309974e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 4661720,
      "step": 7145
    },
    {
      "epoch": 3.747379454926625,
      "grad_norm": 9.125,
      "learning_rate": 3.936188166820051e-05,
      "loss": 0.4875,
      "num_input_tokens_seen": 4666584,
      "step": 7150
    },
    {
      "epoch": 3.75,
      "grad_norm": 5.5625,
      "learning_rate": 3.9343157235934714e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 4669272,
      "step": 7155
    },
    {
      "epoch": 3.752620545073375,
      "grad_norm": 6.25,
      "learning_rate": 3.932442080197012e-05,
      "loss": 0.5039,
      "num_input_tokens_seen": 4672280,
      "step": 7160
    },
    {
      "epoch": 3.7552410901467503,
      "grad_norm": 4.40625,
      "learning_rate": 3.930567238198451e-05,
      "loss": 0.379,
      "num_input_tokens_seen": 4675032,
      "step": 7165
    },
    {
      "epoch": 3.757861635220126,
      "grad_norm": 6.25,
      "learning_rate": 3.928691199166571e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 4678360,
      "step": 7170
    },
    {
      "epoch": 3.760482180293501,
      "grad_norm": 11.8125,
      "learning_rate": 3.926813964671156e-05,
      "loss": 0.5248,
      "num_input_tokens_seen": 4681208,
      "step": 7175
    },
    {
      "epoch": 3.7631027253668763,
      "grad_norm": 3.640625,
      "learning_rate": 3.9249355362829884e-05,
      "loss": 0.4852,
      "num_input_tokens_seen": 4683768,
      "step": 7180
    },
    {
      "epoch": 3.7657232704402515,
      "grad_norm": 10.0,
      "learning_rate": 3.923055915573853e-05,
      "loss": 0.4928,
      "num_input_tokens_seen": 4686872,
      "step": 7185
    },
    {
      "epoch": 3.768343815513627,
      "grad_norm": 3.875,
      "learning_rate": 3.921175104116531e-05,
      "loss": 0.4881,
      "num_input_tokens_seen": 4690424,
      "step": 7190
    },
    {
      "epoch": 3.7709643605870022,
      "grad_norm": 6.25,
      "learning_rate": 3.9192931034847966e-05,
      "loss": 0.5756,
      "num_input_tokens_seen": 4692824,
      "step": 7195
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 5.75,
      "learning_rate": 3.917409915253426e-05,
      "loss": 0.5435,
      "num_input_tokens_seen": 4696440,
      "step": 7200
    },
    {
      "epoch": 3.7762054507337526,
      "grad_norm": 6.78125,
      "learning_rate": 3.915525540998182e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 4700440,
      "step": 7205
    },
    {
      "epoch": 3.7788259958071277,
      "grad_norm": 4.5,
      "learning_rate": 3.9136399822958235e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 4703320,
      "step": 7210
    },
    {
      "epoch": 3.781446540880503,
      "grad_norm": 3.359375,
      "learning_rate": 3.911753240724101e-05,
      "loss": 0.5871,
      "num_input_tokens_seen": 4706552,
      "step": 7215
    },
    {
      "epoch": 3.7840670859538785,
      "grad_norm": 2.890625,
      "learning_rate": 3.909865317861753e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 4709592,
      "step": 7220
    },
    {
      "epoch": 3.7866876310272537,
      "grad_norm": 4.34375,
      "learning_rate": 3.907976215288507e-05,
      "loss": 0.5989,
      "num_input_tokens_seen": 4712344,
      "step": 7225
    },
    {
      "epoch": 3.789308176100629,
      "grad_norm": 5.28125,
      "learning_rate": 3.9060859345850774e-05,
      "loss": 0.5591,
      "num_input_tokens_seen": 4715000,
      "step": 7230
    },
    {
      "epoch": 3.791928721174004,
      "grad_norm": 3.609375,
      "learning_rate": 3.904194477333166e-05,
      "loss": 0.4524,
      "num_input_tokens_seen": 4718904,
      "step": 7235
    },
    {
      "epoch": 3.7945492662473796,
      "grad_norm": 6.90625,
      "learning_rate": 3.902301845115456e-05,
      "loss": 0.5908,
      "num_input_tokens_seen": 4721656,
      "step": 7240
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 20.0,
      "learning_rate": 3.900408039515617e-05,
      "loss": 0.5856,
      "num_input_tokens_seen": 4726264,
      "step": 7245
    },
    {
      "epoch": 3.79979035639413,
      "grad_norm": 4.34375,
      "learning_rate": 3.8985130621182985e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 4729112,
      "step": 7250
    },
    {
      "epoch": 3.802410901467505,
      "grad_norm": 4.1875,
      "learning_rate": 3.896616914509131e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 4731832,
      "step": 7255
    },
    {
      "epoch": 3.8050314465408803,
      "grad_norm": 5.15625,
      "learning_rate": 3.894719598274725e-05,
      "loss": 0.6052,
      "num_input_tokens_seen": 4738136,
      "step": 7260
    },
    {
      "epoch": 3.8076519916142555,
      "grad_norm": 5.5625,
      "learning_rate": 3.892821115002667e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 4741080,
      "step": 7265
    },
    {
      "epoch": 3.810272536687631,
      "grad_norm": 6.0,
      "learning_rate": 3.8909214662815216e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 4744472,
      "step": 7270
    },
    {
      "epoch": 3.8128930817610063,
      "grad_norm": 3.375,
      "learning_rate": 3.889020653700828e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 4748664,
      "step": 7275
    },
    {
      "epoch": 3.8155136268343814,
      "grad_norm": 7.96875,
      "learning_rate": 3.8871186788511e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 4751704,
      "step": 7280
    },
    {
      "epoch": 3.818134171907757,
      "grad_norm": 19.0,
      "learning_rate": 3.8852155433238214e-05,
      "loss": 0.5555,
      "num_input_tokens_seen": 4754104,
      "step": 7285
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 4.78125,
      "learning_rate": 3.8833112487114505e-05,
      "loss": 0.4606,
      "num_input_tokens_seen": 4757528,
      "step": 7290
    },
    {
      "epoch": 3.8233752620545074,
      "grad_norm": 2.921875,
      "learning_rate": 3.881405796607414e-05,
      "loss": 0.6261,
      "num_input_tokens_seen": 4761208,
      "step": 7295
    },
    {
      "epoch": 3.8259958071278826,
      "grad_norm": 5.46875,
      "learning_rate": 3.879499188606107e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 4764088,
      "step": 7300
    },
    {
      "epoch": 3.8286163522012577,
      "grad_norm": 3.28125,
      "learning_rate": 3.877591426302892e-05,
      "loss": 0.5149,
      "num_input_tokens_seen": 4767352,
      "step": 7305
    },
    {
      "epoch": 3.831236897274633,
      "grad_norm": 4.75,
      "learning_rate": 3.8756825112940964e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 4770648,
      "step": 7310
    },
    {
      "epoch": 3.8338574423480085,
      "grad_norm": 1.2890625,
      "learning_rate": 3.873772445177015e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 4777624,
      "step": 7315
    },
    {
      "epoch": 3.8364779874213837,
      "grad_norm": 6.59375,
      "learning_rate": 3.8718612295499036e-05,
      "loss": 0.5616,
      "num_input_tokens_seen": 4780376,
      "step": 7320
    },
    {
      "epoch": 3.839098532494759,
      "grad_norm": 3.453125,
      "learning_rate": 3.8699488660119784e-05,
      "loss": 0.4948,
      "num_input_tokens_seen": 4783096,
      "step": 7325
    },
    {
      "epoch": 3.841719077568134,
      "grad_norm": 3.078125,
      "learning_rate": 3.868035356163419e-05,
      "loss": 0.3482,
      "num_input_tokens_seen": 4787000,
      "step": 7330
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 6.65625,
      "learning_rate": 3.866120701605363e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 4790168,
      "step": 7335
    },
    {
      "epoch": 3.846960167714885,
      "grad_norm": 3.265625,
      "learning_rate": 3.8642049039399054e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 4793944,
      "step": 7340
    },
    {
      "epoch": 3.84958071278826,
      "grad_norm": 5.6875,
      "learning_rate": 3.862287964770099e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 4797016,
      "step": 7345
    },
    {
      "epoch": 3.852201257861635,
      "grad_norm": 3.34375,
      "learning_rate": 3.86036988569995e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 4800216,
      "step": 7350
    },
    {
      "epoch": 3.8548218029350103,
      "grad_norm": 4.3125,
      "learning_rate": 3.8584506683344216e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 4803992,
      "step": 7355
    },
    {
      "epoch": 3.8574423480083855,
      "grad_norm": 7.1875,
      "learning_rate": 3.8565303142794234e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 4806872,
      "step": 7360
    },
    {
      "epoch": 3.860062893081761,
      "grad_norm": 5.5,
      "learning_rate": 3.8546088251418224e-05,
      "loss": 0.3172,
      "num_input_tokens_seen": 4809912,
      "step": 7365
    },
    {
      "epoch": 3.8626834381551363,
      "grad_norm": 2.546875,
      "learning_rate": 3.8526862025294336e-05,
      "loss": 0.3353,
      "num_input_tokens_seen": 4816984,
      "step": 7370
    },
    {
      "epoch": 3.8653039832285114,
      "grad_norm": 5.375,
      "learning_rate": 3.8507624480510186e-05,
      "loss": 0.5549,
      "num_input_tokens_seen": 4819608,
      "step": 7375
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 2.671875,
      "learning_rate": 3.848837563316287e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 4823448,
      "step": 7380
    },
    {
      "epoch": 3.870545073375262,
      "grad_norm": 6.78125,
      "learning_rate": 3.8469115499358945e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 4826328,
      "step": 7385
    },
    {
      "epoch": 3.8731656184486374,
      "grad_norm": 7.28125,
      "learning_rate": 3.844984409521442e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 4829048,
      "step": 7390
    },
    {
      "epoch": 3.8757861635220126,
      "grad_norm": 6.21875,
      "learning_rate": 3.843056143685472e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 4833048,
      "step": 7395
    },
    {
      "epoch": 3.8784067085953877,
      "grad_norm": 3.984375,
      "learning_rate": 3.841126754041468e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 4836184,
      "step": 7400
    },
    {
      "epoch": 3.881027253668763,
      "grad_norm": 9.25,
      "learning_rate": 3.839196242203859e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 4839896,
      "step": 7405
    },
    {
      "epoch": 3.8836477987421385,
      "grad_norm": 12.125,
      "learning_rate": 3.837264609788005e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 4842936,
      "step": 7410
    },
    {
      "epoch": 3.8862683438155137,
      "grad_norm": 4.46875,
      "learning_rate": 3.8353318584102096e-05,
      "loss": 0.6722,
      "num_input_tokens_seen": 4846200,
      "step": 7415
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 4.5625,
      "learning_rate": 3.83339798968771e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 4849336,
      "step": 7420
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 6.75,
      "learning_rate": 3.83146300523868e-05,
      "loss": 0.5026,
      "num_input_tokens_seen": 4852632,
      "step": 7425
    },
    {
      "epoch": 3.8941299790356396,
      "grad_norm": 5.59375,
      "learning_rate": 3.829526906682227e-05,
      "loss": 0.5655,
      "num_input_tokens_seen": 4855384,
      "step": 7430
    },
    {
      "epoch": 3.896750524109015,
      "grad_norm": 2.875,
      "learning_rate": 3.827589695638388e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 4858936,
      "step": 7435
    },
    {
      "epoch": 3.89937106918239,
      "grad_norm": 5.15625,
      "learning_rate": 3.825651373728133e-05,
      "loss": 0.5172,
      "num_input_tokens_seen": 4862488,
      "step": 7440
    },
    {
      "epoch": 3.901991614255765,
      "grad_norm": 3.78125,
      "learning_rate": 3.8237119425733625e-05,
      "loss": 0.4558,
      "num_input_tokens_seen": 4865240,
      "step": 7445
    },
    {
      "epoch": 3.9046121593291403,
      "grad_norm": 3.015625,
      "learning_rate": 3.8217714037969035e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 4868152,
      "step": 7450
    },
    {
      "epoch": 3.9072327044025155,
      "grad_norm": 6.5,
      "learning_rate": 3.8198297590225095e-05,
      "loss": 0.365,
      "num_input_tokens_seen": 4870808,
      "step": 7455
    },
    {
      "epoch": 3.909853249475891,
      "grad_norm": 4.1875,
      "learning_rate": 3.817887009874861e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 4873944,
      "step": 7460
    },
    {
      "epoch": 3.9124737945492662,
      "grad_norm": 6.625,
      "learning_rate": 3.815943157979561e-05,
      "loss": 0.5702,
      "num_input_tokens_seen": 4877528,
      "step": 7465
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 3.375,
      "learning_rate": 3.813998204963136e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 4880504,
      "step": 7470
    },
    {
      "epoch": 3.917714884696017,
      "grad_norm": 4.1875,
      "learning_rate": 3.812052152453035e-05,
      "loss": 0.4863,
      "num_input_tokens_seen": 4884312,
      "step": 7475
    },
    {
      "epoch": 3.920335429769392,
      "grad_norm": 7.28125,
      "learning_rate": 3.8101050020776244e-05,
      "loss": 0.3973,
      "num_input_tokens_seen": 4887032,
      "step": 7480
    },
    {
      "epoch": 3.9229559748427674,
      "grad_norm": 2.609375,
      "learning_rate": 3.808156755466191e-05,
      "loss": 0.5303,
      "num_input_tokens_seen": 4889976,
      "step": 7485
    },
    {
      "epoch": 3.9255765199161425,
      "grad_norm": 4.34375,
      "learning_rate": 3.806207414248939e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 4892600,
      "step": 7490
    },
    {
      "epoch": 3.9281970649895177,
      "grad_norm": 15.125,
      "learning_rate": 3.804256980056988e-05,
      "loss": 0.4701,
      "num_input_tokens_seen": 4895640,
      "step": 7495
    },
    {
      "epoch": 3.930817610062893,
      "grad_norm": 4.53125,
      "learning_rate": 3.8023054545223723e-05,
      "loss": 0.404,
      "num_input_tokens_seen": 4898200,
      "step": 7500
    },
    {
      "epoch": 3.9334381551362685,
      "grad_norm": 9.3125,
      "learning_rate": 3.8003528392780385e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 4901080,
      "step": 7505
    },
    {
      "epoch": 3.9360587002096437,
      "grad_norm": 3.5,
      "learning_rate": 3.798399135957847e-05,
      "loss": 0.522,
      "num_input_tokens_seen": 4904280,
      "step": 7510
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 7.34375,
      "learning_rate": 3.7964443461965674e-05,
      "loss": 0.7054,
      "num_input_tokens_seen": 4907832,
      "step": 7515
    },
    {
      "epoch": 3.941299790356394,
      "grad_norm": 5.6875,
      "learning_rate": 3.794488471629878e-05,
      "loss": 0.628,
      "num_input_tokens_seen": 4911320,
      "step": 7520
    },
    {
      "epoch": 3.9439203354297696,
      "grad_norm": 4.0625,
      "learning_rate": 3.7925315138943655e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 4914072,
      "step": 7525
    },
    {
      "epoch": 3.9465408805031448,
      "grad_norm": 5.125,
      "learning_rate": 3.790573474627522e-05,
      "loss": 0.525,
      "num_input_tokens_seen": 4918584,
      "step": 7530
    },
    {
      "epoch": 3.94916142557652,
      "grad_norm": 6.78125,
      "learning_rate": 3.7886143554677466e-05,
      "loss": 0.4899,
      "num_input_tokens_seen": 4921496,
      "step": 7535
    },
    {
      "epoch": 3.951781970649895,
      "grad_norm": 5.3125,
      "learning_rate": 3.7866541580543405e-05,
      "loss": 0.5307,
      "num_input_tokens_seen": 4924600,
      "step": 7540
    },
    {
      "epoch": 3.9544025157232703,
      "grad_norm": 3.109375,
      "learning_rate": 3.7846928840275056e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 4927512,
      "step": 7545
    },
    {
      "epoch": 3.9570230607966455,
      "grad_norm": 7.21875,
      "learning_rate": 3.782730535028348e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 4930360,
      "step": 7550
    },
    {
      "epoch": 3.959643605870021,
      "grad_norm": 3.75,
      "learning_rate": 3.780767112698872e-05,
      "loss": 0.3992,
      "num_input_tokens_seen": 4933496,
      "step": 7555
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 4.8125,
      "learning_rate": 3.77880261868198e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 4936152,
      "step": 7560
    },
    {
      "epoch": 3.9648846960167714,
      "grad_norm": 5.03125,
      "learning_rate": 3.7768370546214685e-05,
      "loss": 0.7134,
      "num_input_tokens_seen": 4939832,
      "step": 7565
    },
    {
      "epoch": 3.967505241090147,
      "grad_norm": 3.0625,
      "learning_rate": 3.774870422162034e-05,
      "loss": 0.5295,
      "num_input_tokens_seen": 4943192,
      "step": 7570
    },
    {
      "epoch": 3.970125786163522,
      "grad_norm": 9.9375,
      "learning_rate": 3.7729027229492645e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 4946392,
      "step": 7575
    },
    {
      "epoch": 3.9727463312368974,
      "grad_norm": 4.375,
      "learning_rate": 3.770933958629639e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 4949464,
      "step": 7580
    },
    {
      "epoch": 3.9753668763102725,
      "grad_norm": 3.28125,
      "learning_rate": 3.768964130850532e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 4955288,
      "step": 7585
    },
    {
      "epoch": 3.9779874213836477,
      "grad_norm": 4.875,
      "learning_rate": 3.766993241260204e-05,
      "loss": 0.3998,
      "num_input_tokens_seen": 4958008,
      "step": 7590
    },
    {
      "epoch": 3.980607966457023,
      "grad_norm": 3.0625,
      "learning_rate": 3.765021291507805e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 4961400,
      "step": 7595
    },
    {
      "epoch": 3.9832285115303985,
      "grad_norm": 8.375,
      "learning_rate": 3.763048283243374e-05,
      "loss": 0.5963,
      "num_input_tokens_seen": 4964536,
      "step": 7600
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 4.59375,
      "learning_rate": 3.7610742181178325e-05,
      "loss": 0.5952,
      "num_input_tokens_seen": 4968280,
      "step": 7605
    },
    {
      "epoch": 3.988469601677149,
      "grad_norm": 2.734375,
      "learning_rate": 3.75909909778299e-05,
      "loss": 0.5623,
      "num_input_tokens_seen": 4971640,
      "step": 7610
    },
    {
      "epoch": 3.991090146750524,
      "grad_norm": 3.796875,
      "learning_rate": 3.757122923891534e-05,
      "loss": 0.5074,
      "num_input_tokens_seen": 4975512,
      "step": 7615
    },
    {
      "epoch": 3.9937106918238996,
      "grad_norm": 5.6875,
      "learning_rate": 3.75514569809704e-05,
      "loss": 0.5784,
      "num_input_tokens_seen": 4978520,
      "step": 7620
    },
    {
      "epoch": 3.9963312368972748,
      "grad_norm": 4.75,
      "learning_rate": 3.7531674220539584e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 4981944,
      "step": 7625
    },
    {
      "epoch": 3.99895178197065,
      "grad_norm": 7.6875,
      "learning_rate": 3.751188097417619e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 4984632,
      "step": 7630
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.4977997839450836,
      "eval_runtime": 24.2705,
      "eval_samples_per_second": 34.94,
      "eval_steps_per_second": 8.735,
      "num_input_tokens_seen": 4985200,
      "step": 7632
    },
    {
      "epoch": 4.001572327044025,
      "grad_norm": 4.65625,
      "learning_rate": 3.749207725844234e-05,
      "loss": 0.4667,
      "num_input_tokens_seen": 4987024,
      "step": 7635
    },
    {
      "epoch": 4.0041928721174,
      "grad_norm": 6.03125,
      "learning_rate": 3.747226308990884e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 4989552,
      "step": 7640
    },
    {
      "epoch": 4.006813417190775,
      "grad_norm": 5.53125,
      "learning_rate": 3.74524384851553e-05,
      "loss": 0.7042,
      "num_input_tokens_seen": 4992080,
      "step": 7645
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 10.25,
      "learning_rate": 3.743260346077004e-05,
      "loss": 0.688,
      "num_input_tokens_seen": 4994512,
      "step": 7650
    },
    {
      "epoch": 4.012054507337526,
      "grad_norm": 4.03125,
      "learning_rate": 3.741275803335011e-05,
      "loss": 0.3288,
      "num_input_tokens_seen": 4998000,
      "step": 7655
    },
    {
      "epoch": 4.014675052410902,
      "grad_norm": 6.53125,
      "learning_rate": 3.7392902219501234e-05,
      "loss": 0.4624,
      "num_input_tokens_seen": 5002000,
      "step": 7660
    },
    {
      "epoch": 4.017295597484277,
      "grad_norm": 3.96875,
      "learning_rate": 3.737303603583788e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 5005648,
      "step": 7665
    },
    {
      "epoch": 4.019916142557652,
      "grad_norm": 5.25,
      "learning_rate": 3.735315949898314e-05,
      "loss": 0.6248,
      "num_input_tokens_seen": 5009104,
      "step": 7670
    },
    {
      "epoch": 4.022536687631027,
      "grad_norm": 4.375,
      "learning_rate": 3.7333272625568804e-05,
      "loss": 0.3995,
      "num_input_tokens_seen": 5012304,
      "step": 7675
    },
    {
      "epoch": 4.0251572327044025,
      "grad_norm": 19.75,
      "learning_rate": 3.7313375432235295e-05,
      "loss": 0.5595,
      "num_input_tokens_seen": 5014608,
      "step": 7680
    },
    {
      "epoch": 4.027777777777778,
      "grad_norm": 7.84375,
      "learning_rate": 3.729346793563167e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 5017776,
      "step": 7685
    },
    {
      "epoch": 4.030398322851153,
      "grad_norm": 4.78125,
      "learning_rate": 3.7273550152415635e-05,
      "loss": 0.6423,
      "num_input_tokens_seen": 5020432,
      "step": 7690
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 14.75,
      "learning_rate": 3.725362209925346e-05,
      "loss": 0.3481,
      "num_input_tokens_seen": 5023824,
      "step": 7695
    },
    {
      "epoch": 4.035639412997903,
      "grad_norm": 2.703125,
      "learning_rate": 3.7233683792820036e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 5027344,
      "step": 7700
    },
    {
      "epoch": 4.038259958071279,
      "grad_norm": 7.84375,
      "learning_rate": 3.721373524979883e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 5030928,
      "step": 7705
    },
    {
      "epoch": 4.040880503144654,
      "grad_norm": 4.5625,
      "learning_rate": 3.7193776486881854e-05,
      "loss": 0.3951,
      "num_input_tokens_seen": 5034320,
      "step": 7710
    },
    {
      "epoch": 4.04350104821803,
      "grad_norm": 3.328125,
      "learning_rate": 3.717380752076971e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 5037456,
      "step": 7715
    },
    {
      "epoch": 4.046121593291405,
      "grad_norm": 4.0625,
      "learning_rate": 3.715382836817152e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 5040784,
      "step": 7720
    },
    {
      "epoch": 4.04874213836478,
      "grad_norm": 5.71875,
      "learning_rate": 3.7133839045804906e-05,
      "loss": 0.5945,
      "num_input_tokens_seen": 5044176,
      "step": 7725
    },
    {
      "epoch": 4.051362683438155,
      "grad_norm": 5.78125,
      "learning_rate": 3.711383957039602e-05,
      "loss": 0.5789,
      "num_input_tokens_seen": 5047376,
      "step": 7730
    },
    {
      "epoch": 4.05398322851153,
      "grad_norm": 6.28125,
      "learning_rate": 3.709382995867954e-05,
      "loss": 0.4846,
      "num_input_tokens_seen": 5053712,
      "step": 7735
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 3.328125,
      "learning_rate": 3.707381022739856e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 5057968,
      "step": 7740
    },
    {
      "epoch": 4.059224318658281,
      "grad_norm": 11.9375,
      "learning_rate": 3.7053780393304705e-05,
      "loss": 0.6801,
      "num_input_tokens_seen": 5061328,
      "step": 7745
    },
    {
      "epoch": 4.061844863731656,
      "grad_norm": 4.46875,
      "learning_rate": 3.7033740473158e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 5064464,
      "step": 7750
    },
    {
      "epoch": 4.064465408805032,
      "grad_norm": 5.5,
      "learning_rate": 3.701369048372695e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 5067344,
      "step": 7755
    },
    {
      "epoch": 4.067085953878407,
      "grad_norm": 4.9375,
      "learning_rate": 3.699363044178847e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 5070512,
      "step": 7760
    },
    {
      "epoch": 4.069706498951782,
      "grad_norm": 10.0625,
      "learning_rate": 3.697356036412788e-05,
      "loss": 0.4208,
      "num_input_tokens_seen": 5073488,
      "step": 7765
    },
    {
      "epoch": 4.072327044025157,
      "grad_norm": 5.28125,
      "learning_rate": 3.695348026753891e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 5079984,
      "step": 7770
    },
    {
      "epoch": 4.0749475890985325,
      "grad_norm": 3.40625,
      "learning_rate": 3.6933390168823655e-05,
      "loss": 0.4945,
      "num_input_tokens_seen": 5083056,
      "step": 7775
    },
    {
      "epoch": 4.077568134171908,
      "grad_norm": 5.84375,
      "learning_rate": 3.6913290084792616e-05,
      "loss": 0.551,
      "num_input_tokens_seen": 5085904,
      "step": 7780
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 5.28125,
      "learning_rate": 3.689318003226461e-05,
      "loss": 0.4698,
      "num_input_tokens_seen": 5088368,
      "step": 7785
    },
    {
      "epoch": 4.082809224318658,
      "grad_norm": 3.375,
      "learning_rate": 3.687306002806681e-05,
      "loss": 0.4612,
      "num_input_tokens_seen": 5091088,
      "step": 7790
    },
    {
      "epoch": 4.085429769392033,
      "grad_norm": 3.171875,
      "learning_rate": 3.685293008903471e-05,
      "loss": 0.5122,
      "num_input_tokens_seen": 5094960,
      "step": 7795
    },
    {
      "epoch": 4.088050314465409,
      "grad_norm": 4.25,
      "learning_rate": 3.683279023201213e-05,
      "loss": 0.4946,
      "num_input_tokens_seen": 5098352,
      "step": 7800
    },
    {
      "epoch": 4.090670859538784,
      "grad_norm": 15.75,
      "learning_rate": 3.681264047385119e-05,
      "loss": 0.5155,
      "num_input_tokens_seen": 5102416,
      "step": 7805
    },
    {
      "epoch": 4.09329140461216,
      "grad_norm": 2.9375,
      "learning_rate": 3.6792480831412293e-05,
      "loss": 0.6301,
      "num_input_tokens_seen": 5105072,
      "step": 7810
    },
    {
      "epoch": 4.095911949685535,
      "grad_norm": 8.25,
      "learning_rate": 3.677231132156408e-05,
      "loss": 0.5271,
      "num_input_tokens_seen": 5108528,
      "step": 7815
    },
    {
      "epoch": 4.09853249475891,
      "grad_norm": 4.53125,
      "learning_rate": 3.675213196118349e-05,
      "loss": 0.6192,
      "num_input_tokens_seen": 5112176,
      "step": 7820
    },
    {
      "epoch": 4.101153039832285,
      "grad_norm": 3.921875,
      "learning_rate": 3.67319427671557e-05,
      "loss": 0.4815,
      "num_input_tokens_seen": 5116848,
      "step": 7825
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 4.03125,
      "learning_rate": 3.6711743756374103e-05,
      "loss": 0.4756,
      "num_input_tokens_seen": 5120464,
      "step": 7830
    },
    {
      "epoch": 4.106394129979035,
      "grad_norm": 3.140625,
      "learning_rate": 3.6691534945740284e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 5123312,
      "step": 7835
    },
    {
      "epoch": 4.109014675052411,
      "grad_norm": 2.609375,
      "learning_rate": 3.667131635216408e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 5127088,
      "step": 7840
    },
    {
      "epoch": 4.111635220125786,
      "grad_norm": 3.4375,
      "learning_rate": 3.665108799256348e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 5130384,
      "step": 7845
    },
    {
      "epoch": 4.114255765199162,
      "grad_norm": 4.15625,
      "learning_rate": 3.663084988386464e-05,
      "loss": 0.4776,
      "num_input_tokens_seen": 5134288,
      "step": 7850
    },
    {
      "epoch": 4.116876310272537,
      "grad_norm": 15.1875,
      "learning_rate": 3.6610602043001894e-05,
      "loss": 0.6389,
      "num_input_tokens_seen": 5137392,
      "step": 7855
    },
    {
      "epoch": 4.119496855345912,
      "grad_norm": 6.6875,
      "learning_rate": 3.659034448691771e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 5139824,
      "step": 7860
    },
    {
      "epoch": 4.122117400419287,
      "grad_norm": 3.34375,
      "learning_rate": 3.657007723256268e-05,
      "loss": 0.5707,
      "num_input_tokens_seen": 5143280,
      "step": 7865
    },
    {
      "epoch": 4.1247379454926625,
      "grad_norm": 3.640625,
      "learning_rate": 3.654980029689553e-05,
      "loss": 0.5603,
      "num_input_tokens_seen": 5145584,
      "step": 7870
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 5.03125,
      "learning_rate": 3.6529513696883075e-05,
      "loss": 0.4548,
      "num_input_tokens_seen": 5148240,
      "step": 7875
    },
    {
      "epoch": 4.129979035639413,
      "grad_norm": 8.875,
      "learning_rate": 3.650921744950019e-05,
      "loss": 0.5857,
      "num_input_tokens_seen": 5151632,
      "step": 7880
    },
    {
      "epoch": 4.132599580712788,
      "grad_norm": 16.75,
      "learning_rate": 3.6488911571729864e-05,
      "loss": 0.5869,
      "num_input_tokens_seen": 5154576,
      "step": 7885
    },
    {
      "epoch": 4.135220125786163,
      "grad_norm": 4.9375,
      "learning_rate": 3.6468596080563134e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 5158224,
      "step": 7890
    },
    {
      "epoch": 4.137840670859539,
      "grad_norm": 5.625,
      "learning_rate": 3.6448270992999065e-05,
      "loss": 0.3889,
      "num_input_tokens_seen": 5161360,
      "step": 7895
    },
    {
      "epoch": 4.140461215932914,
      "grad_norm": 7.65625,
      "learning_rate": 3.6427936326044756e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 5164656,
      "step": 7900
    },
    {
      "epoch": 4.1430817610062896,
      "grad_norm": 6.15625,
      "learning_rate": 3.6407592096715345e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 5168336,
      "step": 7905
    },
    {
      "epoch": 4.145702306079665,
      "grad_norm": 5.875,
      "learning_rate": 3.638723832203396e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 5170736,
      "step": 7910
    },
    {
      "epoch": 4.14832285115304,
      "grad_norm": 6.28125,
      "learning_rate": 3.6366875019031676e-05,
      "loss": 0.3026,
      "num_input_tokens_seen": 5173424,
      "step": 7915
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 4.59375,
      "learning_rate": 3.6346502204747596e-05,
      "loss": 0.3564,
      "num_input_tokens_seen": 5175856,
      "step": 7920
    },
    {
      "epoch": 4.15356394129979,
      "grad_norm": 5.6875,
      "learning_rate": 3.6326119896228766e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 5178672,
      "step": 7925
    },
    {
      "epoch": 4.156184486373165,
      "grad_norm": 3.78125,
      "learning_rate": 3.630572811053016e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 5184496,
      "step": 7930
    },
    {
      "epoch": 4.158805031446541,
      "grad_norm": 3.078125,
      "learning_rate": 3.62853268647147e-05,
      "loss": 0.5434,
      "num_input_tokens_seen": 5188144,
      "step": 7935
    },
    {
      "epoch": 4.161425576519916,
      "grad_norm": 5.4375,
      "learning_rate": 3.6264916175853204e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 5192112,
      "step": 7940
    },
    {
      "epoch": 4.164046121593292,
      "grad_norm": 9.75,
      "learning_rate": 3.624449606102441e-05,
      "loss": 0.5239,
      "num_input_tokens_seen": 5194320,
      "step": 7945
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 10.4375,
      "learning_rate": 3.622406653731495e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 5197456,
      "step": 7950
    },
    {
      "epoch": 4.169287211740042,
      "grad_norm": 3.5,
      "learning_rate": 3.620362762181931e-05,
      "loss": 0.5258,
      "num_input_tokens_seen": 5199792,
      "step": 7955
    },
    {
      "epoch": 4.171907756813417,
      "grad_norm": 5.75,
      "learning_rate": 3.6183179331639825e-05,
      "loss": 0.6377,
      "num_input_tokens_seen": 5202672,
      "step": 7960
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 6.4375,
      "learning_rate": 3.616272168388671e-05,
      "loss": 0.6036,
      "num_input_tokens_seen": 5205808,
      "step": 7965
    },
    {
      "epoch": 4.177148846960168,
      "grad_norm": 5.40625,
      "learning_rate": 3.614225469567798e-05,
      "loss": 0.3565,
      "num_input_tokens_seen": 5208176,
      "step": 7970
    },
    {
      "epoch": 4.179769392033543,
      "grad_norm": 1.3984375,
      "learning_rate": 3.612177838413948e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 5215376,
      "step": 7975
    },
    {
      "epoch": 4.182389937106918,
      "grad_norm": 4.90625,
      "learning_rate": 3.6101292766404854e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 5218064,
      "step": 7980
    },
    {
      "epoch": 4.185010482180293,
      "grad_norm": 6.8125,
      "learning_rate": 3.608079785961552e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 5221328,
      "step": 7985
    },
    {
      "epoch": 4.187631027253669,
      "grad_norm": 5.34375,
      "learning_rate": 3.60602936809207e-05,
      "loss": 0.4951,
      "num_input_tokens_seen": 5224432,
      "step": 7990
    },
    {
      "epoch": 4.190251572327044,
      "grad_norm": 6.0,
      "learning_rate": 3.603978024747733e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 5226896,
      "step": 7995
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 5.75,
      "learning_rate": 3.601925757645013e-05,
      "loss": 0.4703,
      "num_input_tokens_seen": 5231056,
      "step": 8000
    },
    {
      "epoch": 4.195492662473795,
      "grad_norm": 3.796875,
      "learning_rate": 3.599872568501152e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 5234480,
      "step": 8005
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 12.125,
      "learning_rate": 3.5978184590341676e-05,
      "loss": 0.6737,
      "num_input_tokens_seen": 5236816,
      "step": 8010
    },
    {
      "epoch": 4.200733752620545,
      "grad_norm": 19.75,
      "learning_rate": 3.5957634309628424e-05,
      "loss": 0.5041,
      "num_input_tokens_seen": 5239600,
      "step": 8015
    },
    {
      "epoch": 4.20335429769392,
      "grad_norm": 5.96875,
      "learning_rate": 3.59370748600673e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 5242896,
      "step": 8020
    },
    {
      "epoch": 4.205974842767295,
      "grad_norm": 4.5625,
      "learning_rate": 3.591650625886152e-05,
      "loss": 0.5695,
      "num_input_tokens_seen": 5245904,
      "step": 8025
    },
    {
      "epoch": 4.2085953878406706,
      "grad_norm": 7.03125,
      "learning_rate": 3.5895928523221955e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 5249968,
      "step": 8030
    },
    {
      "epoch": 4.211215932914046,
      "grad_norm": 3.171875,
      "learning_rate": 3.58753416703671e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 5253680,
      "step": 8035
    },
    {
      "epoch": 4.213836477987422,
      "grad_norm": 7.5625,
      "learning_rate": 3.585474571752311e-05,
      "loss": 0.6677,
      "num_input_tokens_seen": 5256144,
      "step": 8040
    },
    {
      "epoch": 4.216457023060797,
      "grad_norm": 4.375,
      "learning_rate": 3.583414068192372e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 5261456,
      "step": 8045
    },
    {
      "epoch": 4.219077568134172,
      "grad_norm": 5.4375,
      "learning_rate": 3.58135265808103e-05,
      "loss": 0.572,
      "num_input_tokens_seen": 5264688,
      "step": 8050
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 4.0,
      "learning_rate": 3.5792903431431775e-05,
      "loss": 0.6713,
      "num_input_tokens_seen": 5267184,
      "step": 8055
    },
    {
      "epoch": 4.2243186582809225,
      "grad_norm": 10.375,
      "learning_rate": 3.577227125104466e-05,
      "loss": 0.5415,
      "num_input_tokens_seen": 5269776,
      "step": 8060
    },
    {
      "epoch": 4.226939203354298,
      "grad_norm": 4.03125,
      "learning_rate": 3.575163005691302e-05,
      "loss": 0.4924,
      "num_input_tokens_seen": 5273296,
      "step": 8065
    },
    {
      "epoch": 4.229559748427673,
      "grad_norm": 4.40625,
      "learning_rate": 3.573097986630845e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 5276432,
      "step": 8070
    },
    {
      "epoch": 4.232180293501048,
      "grad_norm": 6.0625,
      "learning_rate": 3.5710320696510114e-05,
      "loss": 0.418,
      "num_input_tokens_seen": 5279696,
      "step": 8075
    },
    {
      "epoch": 4.234800838574423,
      "grad_norm": 4.40625,
      "learning_rate": 3.5689652564804646e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 5283312,
      "step": 8080
    },
    {
      "epoch": 4.237421383647799,
      "grad_norm": 6.875,
      "learning_rate": 3.566897548848619e-05,
      "loss": 0.3972,
      "num_input_tokens_seen": 5285712,
      "step": 8085
    },
    {
      "epoch": 4.240041928721174,
      "grad_norm": 3.828125,
      "learning_rate": 3.564828948485639e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 5289136,
      "step": 8090
    },
    {
      "epoch": 4.2426624737945495,
      "grad_norm": 4.03125,
      "learning_rate": 3.562759457122434e-05,
      "loss": 0.3307,
      "num_input_tokens_seen": 5292528,
      "step": 8095
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 4.9375,
      "learning_rate": 3.5606890764906603e-05,
      "loss": 0.5441,
      "num_input_tokens_seen": 5295984,
      "step": 8100
    },
    {
      "epoch": 4.2479035639413,
      "grad_norm": 5.5,
      "learning_rate": 3.5586178083227175e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 5299120,
      "step": 8105
    },
    {
      "epoch": 4.250524109014675,
      "grad_norm": 4.78125,
      "learning_rate": 3.556545654351749e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 5302192,
      "step": 8110
    },
    {
      "epoch": 4.25314465408805,
      "grad_norm": 8.5,
      "learning_rate": 3.554472616311638e-05,
      "loss": 0.82,
      "num_input_tokens_seen": 5304880,
      "step": 8115
    },
    {
      "epoch": 4.255765199161425,
      "grad_norm": 4.0625,
      "learning_rate": 3.552398695937007e-05,
      "loss": 0.5629,
      "num_input_tokens_seen": 5308304,
      "step": 8120
    },
    {
      "epoch": 4.2583857442348005,
      "grad_norm": 5.90625,
      "learning_rate": 3.55032389496322e-05,
      "loss": 0.5637,
      "num_input_tokens_seen": 5310960,
      "step": 8125
    },
    {
      "epoch": 4.261006289308176,
      "grad_norm": 4.6875,
      "learning_rate": 3.548248215126374e-05,
      "loss": 0.4621,
      "num_input_tokens_seen": 5313264,
      "step": 8130
    },
    {
      "epoch": 4.263626834381552,
      "grad_norm": 7.0,
      "learning_rate": 3.546171658163304e-05,
      "loss": 0.5876,
      "num_input_tokens_seen": 5315600,
      "step": 8135
    },
    {
      "epoch": 4.266247379454927,
      "grad_norm": 3.0,
      "learning_rate": 3.544094225811577e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 5319024,
      "step": 8140
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 6.8125,
      "learning_rate": 3.542015919809495e-05,
      "loss": 0.4697,
      "num_input_tokens_seen": 5322160,
      "step": 8145
    },
    {
      "epoch": 4.271488469601677,
      "grad_norm": 2.40625,
      "learning_rate": 3.539936741896088e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 5327152,
      "step": 8150
    },
    {
      "epoch": 4.274109014675052,
      "grad_norm": 5.1875,
      "learning_rate": 3.537856693811118e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 5330256,
      "step": 8155
    },
    {
      "epoch": 4.276729559748428,
      "grad_norm": 11.9375,
      "learning_rate": 3.5357757772950746e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 5333392,
      "step": 8160
    },
    {
      "epoch": 4.279350104821803,
      "grad_norm": 6.875,
      "learning_rate": 3.533693994089173e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 5336720,
      "step": 8165
    },
    {
      "epoch": 4.281970649895178,
      "grad_norm": 4.5,
      "learning_rate": 3.531611345935353e-05,
      "loss": 0.3173,
      "num_input_tokens_seen": 5339696,
      "step": 8170
    },
    {
      "epoch": 4.284591194968553,
      "grad_norm": 6.59375,
      "learning_rate": 3.529527834576282e-05,
      "loss": 0.575,
      "num_input_tokens_seen": 5342896,
      "step": 8175
    },
    {
      "epoch": 4.287211740041929,
      "grad_norm": 6.65625,
      "learning_rate": 3.527443461755346e-05,
      "loss": 0.5328,
      "num_input_tokens_seen": 5345392,
      "step": 8180
    },
    {
      "epoch": 4.289832285115304,
      "grad_norm": 4.75,
      "learning_rate": 3.525358229216653e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 5349392,
      "step": 8185
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 6.6875,
      "learning_rate": 3.52327213870503e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 5353008,
      "step": 8190
    },
    {
      "epoch": 4.295073375262055,
      "grad_norm": 8.0625,
      "learning_rate": 3.521185191966022e-05,
      "loss": 0.749,
      "num_input_tokens_seen": 5356464,
      "step": 8195
    },
    {
      "epoch": 4.29769392033543,
      "grad_norm": 4.0,
      "learning_rate": 3.5190973907458924e-05,
      "loss": 0.4186,
      "num_input_tokens_seen": 5359440,
      "step": 8200
    },
    {
      "epoch": 4.300314465408805,
      "grad_norm": 9.4375,
      "learning_rate": 3.517008736791616e-05,
      "loss": 0.62,
      "num_input_tokens_seen": 5361968,
      "step": 8205
    },
    {
      "epoch": 4.30293501048218,
      "grad_norm": 5.46875,
      "learning_rate": 3.514919231850885e-05,
      "loss": 0.6559,
      "num_input_tokens_seen": 5365744,
      "step": 8210
    },
    {
      "epoch": 4.305555555555555,
      "grad_norm": 6.1875,
      "learning_rate": 3.512828877672099e-05,
      "loss": 0.3931,
      "num_input_tokens_seen": 5368720,
      "step": 8215
    },
    {
      "epoch": 4.3081761006289305,
      "grad_norm": 12.875,
      "learning_rate": 3.510737676004372e-05,
      "loss": 0.5406,
      "num_input_tokens_seen": 5371408,
      "step": 8220
    },
    {
      "epoch": 4.310796645702306,
      "grad_norm": 7.5625,
      "learning_rate": 3.5086456285975274e-05,
      "loss": 0.5789,
      "num_input_tokens_seen": 5374256,
      "step": 8225
    },
    {
      "epoch": 4.313417190775682,
      "grad_norm": 17.75,
      "learning_rate": 3.5065527372020935e-05,
      "loss": 0.5592,
      "num_input_tokens_seen": 5376624,
      "step": 8230
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 9.125,
      "learning_rate": 3.504459003569306e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 5380048,
      "step": 8235
    },
    {
      "epoch": 4.318658280922432,
      "grad_norm": 6.03125,
      "learning_rate": 3.5023644294511074e-05,
      "loss": 0.412,
      "num_input_tokens_seen": 5382416,
      "step": 8240
    },
    {
      "epoch": 4.321278825995807,
      "grad_norm": 5.4375,
      "learning_rate": 3.50026901660014e-05,
      "loss": 0.6629,
      "num_input_tokens_seen": 5385520,
      "step": 8245
    },
    {
      "epoch": 4.323899371069182,
      "grad_norm": 5.9375,
      "learning_rate": 3.4981727667697497e-05,
      "loss": 0.6967,
      "num_input_tokens_seen": 5388368,
      "step": 8250
    },
    {
      "epoch": 4.326519916142558,
      "grad_norm": 5.5,
      "learning_rate": 3.4960756817139825e-05,
      "loss": 0.4401,
      "num_input_tokens_seen": 5391920,
      "step": 8255
    },
    {
      "epoch": 4.329140461215933,
      "grad_norm": 4.71875,
      "learning_rate": 3.493977763187584e-05,
      "loss": 0.3901,
      "num_input_tokens_seen": 5394864,
      "step": 8260
    },
    {
      "epoch": 4.331761006289308,
      "grad_norm": 4.53125,
      "learning_rate": 3.4918790129459975e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 5398448,
      "step": 8265
    },
    {
      "epoch": 4.334381551362683,
      "grad_norm": 4.4375,
      "learning_rate": 3.4897794327453586e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 5401904,
      "step": 8270
    },
    {
      "epoch": 4.337002096436059,
      "grad_norm": 7.21875,
      "learning_rate": 3.487679024342502e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 5404432,
      "step": 8275
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 7.875,
      "learning_rate": 3.4855777894949536e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 5407376,
      "step": 8280
    },
    {
      "epoch": 4.3422431865828095,
      "grad_norm": 9.8125,
      "learning_rate": 3.4834757299609306e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 5410768,
      "step": 8285
    },
    {
      "epoch": 4.344863731656185,
      "grad_norm": 4.15625,
      "learning_rate": 3.48137284749934e-05,
      "loss": 0.5303,
      "num_input_tokens_seen": 5414256,
      "step": 8290
    },
    {
      "epoch": 4.34748427672956,
      "grad_norm": 6.4375,
      "learning_rate": 3.479269143869777e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 5417296,
      "step": 8295
    },
    {
      "epoch": 4.350104821802935,
      "grad_norm": 9.25,
      "learning_rate": 3.477164620832527e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 5420560,
      "step": 8300
    },
    {
      "epoch": 4.35272536687631,
      "grad_norm": 8.5,
      "learning_rate": 3.4750592801485564e-05,
      "loss": 0.4394,
      "num_input_tokens_seen": 5424496,
      "step": 8305
    },
    {
      "epoch": 4.355345911949685,
      "grad_norm": 2.625,
      "learning_rate": 3.47295312357952e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 5428944,
      "step": 8310
    },
    {
      "epoch": 4.3579664570230605,
      "grad_norm": 4.40625,
      "learning_rate": 3.4708461528877514e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 5432624,
      "step": 8315
    },
    {
      "epoch": 4.360587002096436,
      "grad_norm": 10.75,
      "learning_rate": 3.468738369836269e-05,
      "loss": 0.5241,
      "num_input_tokens_seen": 5435152,
      "step": 8320
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 5.5625,
      "learning_rate": 3.466629776188769e-05,
      "loss": 0.5548,
      "num_input_tokens_seen": 5438640,
      "step": 8325
    },
    {
      "epoch": 4.365828092243187,
      "grad_norm": 4.59375,
      "learning_rate": 3.464520373709627e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 5441264,
      "step": 8330
    },
    {
      "epoch": 4.368448637316562,
      "grad_norm": 4.28125,
      "learning_rate": 3.462410164163893e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 5444752,
      "step": 8335
    },
    {
      "epoch": 4.371069182389937,
      "grad_norm": 5.0625,
      "learning_rate": 3.460299149317294e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 5447632,
      "step": 8340
    },
    {
      "epoch": 4.373689727463312,
      "grad_norm": 5.15625,
      "learning_rate": 3.4581873309362326e-05,
      "loss": 0.6184,
      "num_input_tokens_seen": 5451344,
      "step": 8345
    },
    {
      "epoch": 4.376310272536688,
      "grad_norm": 6.46875,
      "learning_rate": 3.456074710787781e-05,
      "loss": 0.5831,
      "num_input_tokens_seen": 5454736,
      "step": 8350
    },
    {
      "epoch": 4.378930817610063,
      "grad_norm": 3.234375,
      "learning_rate": 3.453961290639683e-05,
      "loss": 0.5265,
      "num_input_tokens_seen": 5458096,
      "step": 8355
    },
    {
      "epoch": 4.381551362683438,
      "grad_norm": 5.125,
      "learning_rate": 3.451847072260351e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 5460560,
      "step": 8360
    },
    {
      "epoch": 4.384171907756813,
      "grad_norm": 17.375,
      "learning_rate": 3.4497320574188694e-05,
      "loss": 0.4703,
      "num_input_tokens_seen": 5463792,
      "step": 8365
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 3.640625,
      "learning_rate": 3.447616247884983e-05,
      "loss": 0.4659,
      "num_input_tokens_seen": 5467216,
      "step": 8370
    },
    {
      "epoch": 4.389412997903564,
      "grad_norm": 2.75,
      "learning_rate": 3.445499645429107e-05,
      "loss": 0.4688,
      "num_input_tokens_seen": 5470320,
      "step": 8375
    },
    {
      "epoch": 4.3920335429769395,
      "grad_norm": 3.59375,
      "learning_rate": 3.443382251822315e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 5473232,
      "step": 8380
    },
    {
      "epoch": 4.394654088050315,
      "grad_norm": 5.75,
      "learning_rate": 3.4412640688363475e-05,
      "loss": 0.617,
      "num_input_tokens_seen": 5476432,
      "step": 8385
    },
    {
      "epoch": 4.39727463312369,
      "grad_norm": 8.4375,
      "learning_rate": 3.439145098243601e-05,
      "loss": 0.5749,
      "num_input_tokens_seen": 5478928,
      "step": 8390
    },
    {
      "epoch": 4.399895178197065,
      "grad_norm": 5.65625,
      "learning_rate": 3.437025341817137e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 5481776,
      "step": 8395
    },
    {
      "epoch": 4.40251572327044,
      "grad_norm": 4.65625,
      "learning_rate": 3.434904801330667e-05,
      "loss": 0.4862,
      "num_input_tokens_seen": 5485008,
      "step": 8400
    },
    {
      "epoch": 4.405136268343815,
      "grad_norm": 2.625,
      "learning_rate": 3.432783478558564e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 5492496,
      "step": 8405
    },
    {
      "epoch": 4.4077568134171905,
      "grad_norm": 6.65625,
      "learning_rate": 3.430661375275854e-05,
      "loss": 0.4707,
      "num_input_tokens_seen": 5495664,
      "step": 8410
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 9.0625,
      "learning_rate": 3.4285384932582175e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 5499792,
      "step": 8415
    },
    {
      "epoch": 4.412997903563941,
      "grad_norm": 5.6875,
      "learning_rate": 3.426414834281982e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 5502736,
      "step": 8420
    },
    {
      "epoch": 4.415618448637317,
      "grad_norm": 10.0625,
      "learning_rate": 3.424290400124131e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 5505328,
      "step": 8425
    },
    {
      "epoch": 4.418238993710692,
      "grad_norm": 5.9375,
      "learning_rate": 3.422165192562293e-05,
      "loss": 0.5168,
      "num_input_tokens_seen": 5507952,
      "step": 8430
    },
    {
      "epoch": 4.420859538784067,
      "grad_norm": 16.75,
      "learning_rate": 3.420039213374745e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 5510544,
      "step": 8435
    },
    {
      "epoch": 4.423480083857442,
      "grad_norm": 6.53125,
      "learning_rate": 3.4179124643404084e-05,
      "loss": 0.5761,
      "num_input_tokens_seen": 5512976,
      "step": 8440
    },
    {
      "epoch": 4.426100628930818,
      "grad_norm": 3.21875,
      "learning_rate": 3.41578494723885e-05,
      "loss": 0.5527,
      "num_input_tokens_seen": 5516496,
      "step": 8445
    },
    {
      "epoch": 4.428721174004193,
      "grad_norm": 5.03125,
      "learning_rate": 3.4136566638502795e-05,
      "loss": 0.3744,
      "num_input_tokens_seen": 5520720,
      "step": 8450
    },
    {
      "epoch": 4.431341719077568,
      "grad_norm": 3.359375,
      "learning_rate": 3.4115276159555464e-05,
      "loss": 0.4737,
      "num_input_tokens_seen": 5524336,
      "step": 8455
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 3.53125,
      "learning_rate": 3.409397805336142e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 5528112,
      "step": 8460
    },
    {
      "epoch": 4.436582809224318,
      "grad_norm": 2.4375,
      "learning_rate": 3.407267233774193e-05,
      "loss": 0.5493,
      "num_input_tokens_seen": 5531664,
      "step": 8465
    },
    {
      "epoch": 4.439203354297694,
      "grad_norm": 7.4375,
      "learning_rate": 3.4051359030524654e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 5536368,
      "step": 8470
    },
    {
      "epoch": 4.4418238993710695,
      "grad_norm": 5.3125,
      "learning_rate": 3.4030038149543594e-05,
      "loss": 0.4771,
      "num_input_tokens_seen": 5539600,
      "step": 8475
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 5.09375,
      "learning_rate": 3.4008709712639084e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 5542544,
      "step": 8480
    },
    {
      "epoch": 4.44706498951782,
      "grad_norm": 5.09375,
      "learning_rate": 3.398737373765779e-05,
      "loss": 0.4847,
      "num_input_tokens_seen": 5546704,
      "step": 8485
    },
    {
      "epoch": 4.449685534591195,
      "grad_norm": 4.96875,
      "learning_rate": 3.396603024245267e-05,
      "loss": 0.4318,
      "num_input_tokens_seen": 5549520,
      "step": 8490
    },
    {
      "epoch": 4.45230607966457,
      "grad_norm": 12.375,
      "learning_rate": 3.3944679244883e-05,
      "loss": 0.4965,
      "num_input_tokens_seen": 5553040,
      "step": 8495
    },
    {
      "epoch": 4.454926624737945,
      "grad_norm": 3.265625,
      "learning_rate": 3.392332076281433e-05,
      "loss": 0.4936,
      "num_input_tokens_seen": 5556880,
      "step": 8500
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 5.6875,
      "learning_rate": 3.390195481411842e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 5559760,
      "step": 8505
    },
    {
      "epoch": 4.460167714884696,
      "grad_norm": 3.96875,
      "learning_rate": 3.3880581416673366e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 5564688,
      "step": 8510
    },
    {
      "epoch": 4.462788259958071,
      "grad_norm": 8.0625,
      "learning_rate": 3.385920058836342e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 5568112,
      "step": 8515
    },
    {
      "epoch": 4.465408805031447,
      "grad_norm": 3.53125,
      "learning_rate": 3.38378123470791e-05,
      "loss": 0.598,
      "num_input_tokens_seen": 5570928,
      "step": 8520
    },
    {
      "epoch": 4.468029350104822,
      "grad_norm": 7.625,
      "learning_rate": 3.381641671071709e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 5574480,
      "step": 8525
    },
    {
      "epoch": 4.470649895178197,
      "grad_norm": 9.75,
      "learning_rate": 3.379501369718031e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 5577296,
      "step": 8530
    },
    {
      "epoch": 4.473270440251572,
      "grad_norm": 4.0625,
      "learning_rate": 3.377360332437781e-05,
      "loss": 0.395,
      "num_input_tokens_seen": 5580784,
      "step": 8535
    },
    {
      "epoch": 4.475890985324948,
      "grad_norm": 4.3125,
      "learning_rate": 3.37521856102248e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 5583856,
      "step": 8540
    },
    {
      "epoch": 4.478511530398323,
      "grad_norm": 5.71875,
      "learning_rate": 3.373076057264266e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 5586256,
      "step": 8545
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 3.703125,
      "learning_rate": 3.370932822955888e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 5589200,
      "step": 8550
    },
    {
      "epoch": 4.483752620545073,
      "grad_norm": 4.125,
      "learning_rate": 3.368788859890706e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 5592976,
      "step": 8555
    },
    {
      "epoch": 4.486373165618448,
      "grad_norm": 7.0625,
      "learning_rate": 3.3666441698626906e-05,
      "loss": 0.5179,
      "num_input_tokens_seen": 5595696,
      "step": 8560
    },
    {
      "epoch": 4.488993710691824,
      "grad_norm": 5.71875,
      "learning_rate": 3.364498754666421e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 5598288,
      "step": 8565
    },
    {
      "epoch": 4.4916142557651995,
      "grad_norm": 4.40625,
      "learning_rate": 3.362352616097082e-05,
      "loss": 0.5543,
      "num_input_tokens_seen": 5601296,
      "step": 8570
    },
    {
      "epoch": 4.494234800838575,
      "grad_norm": 11.0,
      "learning_rate": 3.360205755950464e-05,
      "loss": 0.3924,
      "num_input_tokens_seen": 5604752,
      "step": 8575
    },
    {
      "epoch": 4.49685534591195,
      "grad_norm": 4.5625,
      "learning_rate": 3.358058176022963e-05,
      "loss": 0.5443,
      "num_input_tokens_seen": 5607824,
      "step": 8580
    },
    {
      "epoch": 4.499475890985325,
      "grad_norm": 8.25,
      "learning_rate": 3.355909878111574e-05,
      "loss": 0.5341,
      "num_input_tokens_seen": 5611248,
      "step": 8585
    },
    {
      "epoch": 4.5,
      "eval_loss": 0.4971226453781128,
      "eval_runtime": 24.2064,
      "eval_samples_per_second": 35.032,
      "eval_steps_per_second": 8.758,
      "num_input_tokens_seen": 5611760,
      "step": 8586
    },
    {
      "epoch": 4.5020964360587,
      "grad_norm": 8.75,
      "learning_rate": 3.3537608640138954e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 5614192,
      "step": 8590
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 4.8125,
      "learning_rate": 3.351611135528125e-05,
      "loss": 0.5472,
      "num_input_tokens_seen": 5616944,
      "step": 8595
    },
    {
      "epoch": 4.5073375262054505,
      "grad_norm": 5.875,
      "learning_rate": 3.349460694453056e-05,
      "loss": 0.5439,
      "num_input_tokens_seen": 5619344,
      "step": 8600
    },
    {
      "epoch": 4.509958071278826,
      "grad_norm": 3.15625,
      "learning_rate": 3.3473095425880796e-05,
      "loss": 0.4203,
      "num_input_tokens_seen": 5622608,
      "step": 8605
    },
    {
      "epoch": 4.512578616352201,
      "grad_norm": 9.625,
      "learning_rate": 3.345157681733181e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 5625296,
      "step": 8610
    },
    {
      "epoch": 4.515199161425577,
      "grad_norm": 5.90625,
      "learning_rate": 3.3430051136889404e-05,
      "loss": 0.4307,
      "num_input_tokens_seen": 5628272,
      "step": 8615
    },
    {
      "epoch": 4.517819706498952,
      "grad_norm": 4.0625,
      "learning_rate": 3.3408518402565276e-05,
      "loss": 0.449,
      "num_input_tokens_seen": 5632944,
      "step": 8620
    },
    {
      "epoch": 4.520440251572327,
      "grad_norm": 8.25,
      "learning_rate": 3.338697863237703e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 5636016,
      "step": 8625
    },
    {
      "epoch": 4.523060796645702,
      "grad_norm": 7.03125,
      "learning_rate": 3.336543184434817e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 5642064,
      "step": 8630
    },
    {
      "epoch": 4.5256813417190775,
      "grad_norm": 7.03125,
      "learning_rate": 3.334387805650805e-05,
      "loss": 0.5379,
      "num_input_tokens_seen": 5645072,
      "step": 8635
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 3.859375,
      "learning_rate": 3.3322317286891913e-05,
      "loss": 0.5347,
      "num_input_tokens_seen": 5648304,
      "step": 8640
    },
    {
      "epoch": 4.530922431865828,
      "grad_norm": 5.46875,
      "learning_rate": 3.330074955354082e-05,
      "loss": 0.5963,
      "num_input_tokens_seen": 5651120,
      "step": 8645
    },
    {
      "epoch": 4.533542976939203,
      "grad_norm": 11.625,
      "learning_rate": 3.3279174874501664e-05,
      "loss": 0.4103,
      "num_input_tokens_seen": 5653712,
      "step": 8650
    },
    {
      "epoch": 4.536163522012579,
      "grad_norm": 3.875,
      "learning_rate": 3.325759326782715e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 5656624,
      "step": 8655
    },
    {
      "epoch": 4.538784067085954,
      "grad_norm": 4.90625,
      "learning_rate": 3.323600475157578e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 5658960,
      "step": 8660
    },
    {
      "epoch": 4.5414046121593294,
      "grad_norm": 6.75,
      "learning_rate": 3.321440934381184e-05,
      "loss": 0.495,
      "num_input_tokens_seen": 5661872,
      "step": 8665
    },
    {
      "epoch": 4.544025157232705,
      "grad_norm": 3.46875,
      "learning_rate": 3.319280706260538e-05,
      "loss": 0.5955,
      "num_input_tokens_seen": 5664560,
      "step": 8670
    },
    {
      "epoch": 4.54664570230608,
      "grad_norm": 4.84375,
      "learning_rate": 3.31711979260322e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 5667664,
      "step": 8675
    },
    {
      "epoch": 4.549266247379455,
      "grad_norm": 7.53125,
      "learning_rate": 3.3149581952173846e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 5670128,
      "step": 8680
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 5.53125,
      "learning_rate": 3.312795915911757e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 5673296,
      "step": 8685
    },
    {
      "epoch": 4.554507337526205,
      "grad_norm": 6.6875,
      "learning_rate": 3.310632956495634e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 5675920,
      "step": 8690
    },
    {
      "epoch": 4.5571278825995805,
      "grad_norm": 13.75,
      "learning_rate": 3.308469318778881e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 5678896,
      "step": 8695
    },
    {
      "epoch": 4.559748427672956,
      "grad_norm": 11.625,
      "learning_rate": 3.306305004571932e-05,
      "loss": 0.5391,
      "num_input_tokens_seen": 5681488,
      "step": 8700
    },
    {
      "epoch": 4.562368972746331,
      "grad_norm": 10.5,
      "learning_rate": 3.304140015685785e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 5684432,
      "step": 8705
    },
    {
      "epoch": 4.564989517819707,
      "grad_norm": 7.53125,
      "learning_rate": 3.301974353932005e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 5688112,
      "step": 8710
    },
    {
      "epoch": 4.567610062893082,
      "grad_norm": 4.0625,
      "learning_rate": 3.2998080211227185e-05,
      "loss": 0.5383,
      "num_input_tokens_seen": 5690416,
      "step": 8715
    },
    {
      "epoch": 4.570230607966457,
      "grad_norm": 4.59375,
      "learning_rate": 3.297641019070613e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 5693744,
      "step": 8720
    },
    {
      "epoch": 4.572851153039832,
      "grad_norm": 4.21875,
      "learning_rate": 3.2954733495889376e-05,
      "loss": 0.5118,
      "num_input_tokens_seen": 5697776,
      "step": 8725
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 7.8125,
      "learning_rate": 3.2933050144915e-05,
      "loss": 0.7916,
      "num_input_tokens_seen": 5700240,
      "step": 8730
    },
    {
      "epoch": 4.578092243186583,
      "grad_norm": 5.40625,
      "learning_rate": 3.2911360155926624e-05,
      "loss": 0.4448,
      "num_input_tokens_seen": 5702992,
      "step": 8735
    },
    {
      "epoch": 4.580712788259958,
      "grad_norm": 13.8125,
      "learning_rate": 3.2889663547073444e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 5706544,
      "step": 8740
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 6.6875,
      "learning_rate": 3.286796033651019e-05,
      "loss": 0.4869,
      "num_input_tokens_seen": 5709328,
      "step": 8745
    },
    {
      "epoch": 4.585953878406709,
      "grad_norm": 3.484375,
      "learning_rate": 3.284625054239714e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 5712720,
      "step": 8750
    },
    {
      "epoch": 4.588574423480084,
      "grad_norm": 3.5625,
      "learning_rate": 3.282453418290002e-05,
      "loss": 0.6001,
      "num_input_tokens_seen": 5715536,
      "step": 8755
    },
    {
      "epoch": 4.591194968553459,
      "grad_norm": 6.625,
      "learning_rate": 3.28028112761901e-05,
      "loss": 0.5125,
      "num_input_tokens_seen": 5718352,
      "step": 8760
    },
    {
      "epoch": 4.593815513626835,
      "grad_norm": 8.125,
      "learning_rate": 3.278108184044414e-05,
      "loss": 0.4934,
      "num_input_tokens_seen": 5722096,
      "step": 8765
    },
    {
      "epoch": 4.59643605870021,
      "grad_norm": 7.25,
      "learning_rate": 3.275934589384432e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 5724816,
      "step": 8770
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 3.15625,
      "learning_rate": 3.273760345457828e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 5729040,
      "step": 8775
    },
    {
      "epoch": 4.60167714884696,
      "grad_norm": 3.625,
      "learning_rate": 3.2715854540839106e-05,
      "loss": 0.4722,
      "num_input_tokens_seen": 5732464,
      "step": 8780
    },
    {
      "epoch": 4.604297693920335,
      "grad_norm": 5.25,
      "learning_rate": 3.269409917082531e-05,
      "loss": 0.3635,
      "num_input_tokens_seen": 5735472,
      "step": 8785
    },
    {
      "epoch": 4.6069182389937104,
      "grad_norm": 4.25,
      "learning_rate": 3.2672337362740765e-05,
      "loss": 0.489,
      "num_input_tokens_seen": 5738896,
      "step": 8790
    },
    {
      "epoch": 4.609538784067086,
      "grad_norm": 6.0625,
      "learning_rate": 3.265056913479479e-05,
      "loss": 0.4263,
      "num_input_tokens_seen": 5742032,
      "step": 8795
    },
    {
      "epoch": 4.612159329140461,
      "grad_norm": 4.875,
      "learning_rate": 3.262879450520201e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 5745168,
      "step": 8800
    },
    {
      "epoch": 4.614779874213837,
      "grad_norm": 6.40625,
      "learning_rate": 3.260701349218248e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 5747920,
      "step": 8805
    },
    {
      "epoch": 4.617400419287212,
      "grad_norm": 4.4375,
      "learning_rate": 3.258522611396151e-05,
      "loss": 0.527,
      "num_input_tokens_seen": 5751120,
      "step": 8810
    },
    {
      "epoch": 4.620020964360587,
      "grad_norm": 5.53125,
      "learning_rate": 3.256343238876983e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 5754352,
      "step": 8815
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 24.125,
      "learning_rate": 3.2541632334843394e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 5757200,
      "step": 8820
    },
    {
      "epoch": 4.6252620545073375,
      "grad_norm": 6.09375,
      "learning_rate": 3.251982597042351e-05,
      "loss": 0.3614,
      "num_input_tokens_seen": 5760944,
      "step": 8825
    },
    {
      "epoch": 4.627882599580713,
      "grad_norm": 4.96875,
      "learning_rate": 3.249801331375675e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 5764848,
      "step": 8830
    },
    {
      "epoch": 4.630503144654088,
      "grad_norm": 7.03125,
      "learning_rate": 3.2476194383094946e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 5767984,
      "step": 8835
    },
    {
      "epoch": 4.633123689727463,
      "grad_norm": 4.34375,
      "learning_rate": 3.245436919669517e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 5771472,
      "step": 8840
    },
    {
      "epoch": 4.635744234800838,
      "grad_norm": 7.3125,
      "learning_rate": 3.243253777281977e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 5775536,
      "step": 8845
    },
    {
      "epoch": 4.638364779874214,
      "grad_norm": 3.75,
      "learning_rate": 3.241070012973625e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 5778704,
      "step": 8850
    },
    {
      "epoch": 4.640985324947589,
      "grad_norm": 7.96875,
      "learning_rate": 3.238885628571738e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 5781104,
      "step": 8855
    },
    {
      "epoch": 4.643605870020965,
      "grad_norm": 4.03125,
      "learning_rate": 3.236700625904107e-05,
      "loss": 0.5195,
      "num_input_tokens_seen": 5784528,
      "step": 8860
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 4.46875,
      "learning_rate": 3.234515006799045e-05,
      "loss": 0.7418,
      "num_input_tokens_seen": 5787184,
      "step": 8865
    },
    {
      "epoch": 4.648846960167715,
      "grad_norm": 10.3125,
      "learning_rate": 3.232328773085375e-05,
      "loss": 0.5301,
      "num_input_tokens_seen": 5790032,
      "step": 8870
    },
    {
      "epoch": 4.65146750524109,
      "grad_norm": 5.25,
      "learning_rate": 3.2301419265924395e-05,
      "loss": 0.4975,
      "num_input_tokens_seen": 5793104,
      "step": 8875
    },
    {
      "epoch": 4.654088050314465,
      "grad_norm": 8.3125,
      "learning_rate": 3.2279544691500915e-05,
      "loss": 0.5827,
      "num_input_tokens_seen": 5796368,
      "step": 8880
    },
    {
      "epoch": 4.65670859538784,
      "grad_norm": 8.0,
      "learning_rate": 3.2257664025886956e-05,
      "loss": 0.6626,
      "num_input_tokens_seen": 5799248,
      "step": 8885
    },
    {
      "epoch": 4.659329140461216,
      "grad_norm": 3.34375,
      "learning_rate": 3.2235777287391256e-05,
      "loss": 0.5769,
      "num_input_tokens_seen": 5802448,
      "step": 8890
    },
    {
      "epoch": 4.661949685534591,
      "grad_norm": 6.6875,
      "learning_rate": 3.221388449432764e-05,
      "loss": 0.5384,
      "num_input_tokens_seen": 5805936,
      "step": 8895
    },
    {
      "epoch": 4.664570230607967,
      "grad_norm": 6.3125,
      "learning_rate": 3.219198566501499e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 5809744,
      "step": 8900
    },
    {
      "epoch": 4.667190775681342,
      "grad_norm": 5.21875,
      "learning_rate": 3.217008081777726e-05,
      "loss": 0.5943,
      "num_input_tokens_seen": 5812656,
      "step": 8905
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 9.125,
      "learning_rate": 3.214816997094341e-05,
      "loss": 0.6207,
      "num_input_tokens_seen": 5816048,
      "step": 8910
    },
    {
      "epoch": 4.672431865828092,
      "grad_norm": 3.5,
      "learning_rate": 3.2126253142847454e-05,
      "loss": 0.5978,
      "num_input_tokens_seen": 5820080,
      "step": 8915
    },
    {
      "epoch": 4.6750524109014675,
      "grad_norm": 3.734375,
      "learning_rate": 3.2104330351828374e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 5823344,
      "step": 8920
    },
    {
      "epoch": 4.677672955974843,
      "grad_norm": 3.671875,
      "learning_rate": 3.208240161623017e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 5826448,
      "step": 8925
    },
    {
      "epoch": 4.680293501048218,
      "grad_norm": 7.3125,
      "learning_rate": 3.20604669544018e-05,
      "loss": 0.4531,
      "num_input_tokens_seen": 5830000,
      "step": 8930
    },
    {
      "epoch": 4.682914046121593,
      "grad_norm": 4.8125,
      "learning_rate": 3.2038526384697204e-05,
      "loss": 0.3851,
      "num_input_tokens_seen": 5832464,
      "step": 8935
    },
    {
      "epoch": 4.685534591194968,
      "grad_norm": 5.5,
      "learning_rate": 3.201657992547523e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 5835568,
      "step": 8940
    },
    {
      "epoch": 4.688155136268344,
      "grad_norm": 3.328125,
      "learning_rate": 3.1994627595099674e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 5841360,
      "step": 8945
    },
    {
      "epoch": 4.690775681341719,
      "grad_norm": 6.78125,
      "learning_rate": 3.1972669411939256e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 5844240,
      "step": 8950
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 4.34375,
      "learning_rate": 3.195070539436757e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 5848144,
      "step": 8955
    },
    {
      "epoch": 4.69601677148847,
      "grad_norm": 8.875,
      "learning_rate": 3.19287355607631e-05,
      "loss": 0.5502,
      "num_input_tokens_seen": 5850800,
      "step": 8960
    },
    {
      "epoch": 4.698637316561845,
      "grad_norm": 6.03125,
      "learning_rate": 3.190675992950921e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 5853712,
      "step": 8965
    },
    {
      "epoch": 4.70125786163522,
      "grad_norm": 6.34375,
      "learning_rate": 3.18847785189941e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 5856400,
      "step": 8970
    },
    {
      "epoch": 4.703878406708595,
      "grad_norm": 5.40625,
      "learning_rate": 3.186279134761081e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 5859760,
      "step": 8975
    },
    {
      "epoch": 4.70649895178197,
      "grad_norm": 3.5,
      "learning_rate": 3.18407984337572e-05,
      "loss": 0.4103,
      "num_input_tokens_seen": 5862672,
      "step": 8980
    },
    {
      "epoch": 4.709119496855346,
      "grad_norm": 5.15625,
      "learning_rate": 3.181879979583593e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 5866288,
      "step": 8985
    },
    {
      "epoch": 4.711740041928721,
      "grad_norm": 3.34375,
      "learning_rate": 3.179679545225447e-05,
      "loss": 0.5648,
      "num_input_tokens_seen": 5869424,
      "step": 8990
    },
    {
      "epoch": 4.714360587002097,
      "grad_norm": 3.78125,
      "learning_rate": 3.177478542142503e-05,
      "loss": 0.3841,
      "num_input_tokens_seen": 5872240,
      "step": 8995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 4.96875,
      "learning_rate": 3.175276972176462e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 5874672,
      "step": 9000
    },
    {
      "epoch": 4.719601677148847,
      "grad_norm": 4.53125,
      "learning_rate": 3.173074837169495e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 5878544,
      "step": 9005
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 4.125,
      "learning_rate": 3.1708721389642495e-05,
      "loss": 0.5492,
      "num_input_tokens_seen": 5882096,
      "step": 9010
    },
    {
      "epoch": 4.7248427672955975,
      "grad_norm": 9.0,
      "learning_rate": 3.1686688794038436e-05,
      "loss": 0.5096,
      "num_input_tokens_seen": 5886512,
      "step": 9015
    },
    {
      "epoch": 4.727463312368973,
      "grad_norm": 5.25,
      "learning_rate": 3.1664650603318616e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 5889648,
      "step": 9020
    },
    {
      "epoch": 4.730083857442348,
      "grad_norm": 5.46875,
      "learning_rate": 3.1642606835923606e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 5892752,
      "step": 9025
    },
    {
      "epoch": 4.732704402515723,
      "grad_norm": 3.671875,
      "learning_rate": 3.1620557510298607e-05,
      "loss": 0.6061,
      "num_input_tokens_seen": 5896112,
      "step": 9030
    },
    {
      "epoch": 4.735324947589098,
      "grad_norm": 3.71875,
      "learning_rate": 3.159850264489351e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 5899440,
      "step": 9035
    },
    {
      "epoch": 4.737945492662474,
      "grad_norm": 6.375,
      "learning_rate": 3.157644225816281e-05,
      "loss": 0.5938,
      "num_input_tokens_seen": 5902096,
      "step": 9040
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 5.65625,
      "learning_rate": 3.1554376368565616e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 5904880,
      "step": 9045
    },
    {
      "epoch": 4.743186582809225,
      "grad_norm": 5.8125,
      "learning_rate": 3.153230499456568e-05,
      "loss": 0.5184,
      "num_input_tokens_seen": 5907632,
      "step": 9050
    },
    {
      "epoch": 4.7458071278826,
      "grad_norm": 5.28125,
      "learning_rate": 3.15102281546313e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 5911120,
      "step": 9055
    },
    {
      "epoch": 4.748427672955975,
      "grad_norm": 4.09375,
      "learning_rate": 3.148814586723537e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 5914928,
      "step": 9060
    },
    {
      "epoch": 4.75104821802935,
      "grad_norm": 4.875,
      "learning_rate": 3.146605815085536e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 5918192,
      "step": 9065
    },
    {
      "epoch": 4.753668763102725,
      "grad_norm": 5.5625,
      "learning_rate": 3.1443965023973245e-05,
      "loss": 0.3744,
      "num_input_tokens_seen": 5921232,
      "step": 9070
    },
    {
      "epoch": 4.7562893081761,
      "grad_norm": 2.328125,
      "learning_rate": 3.142186650507554e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 5925040,
      "step": 9075
    },
    {
      "epoch": 4.758909853249476,
      "grad_norm": 10.5625,
      "learning_rate": 3.1399762612653286e-05,
      "loss": 0.5546,
      "num_input_tokens_seen": 5927376,
      "step": 9080
    },
    {
      "epoch": 4.761530398322851,
      "grad_norm": 4.03125,
      "learning_rate": 3.137765336520201e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 5930576,
      "step": 9085
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 24.875,
      "learning_rate": 3.1355538781221705e-05,
      "loss": 0.3995,
      "num_input_tokens_seen": 5933328,
      "step": 9090
    },
    {
      "epoch": 4.766771488469602,
      "grad_norm": 4.6875,
      "learning_rate": 3.133341887921687e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 5936208,
      "step": 9095
    },
    {
      "epoch": 4.769392033542977,
      "grad_norm": 5.84375,
      "learning_rate": 3.1311293677696404e-05,
      "loss": 0.6121,
      "num_input_tokens_seen": 5938832,
      "step": 9100
    },
    {
      "epoch": 4.772012578616352,
      "grad_norm": 5.3125,
      "learning_rate": 3.1289163195173695e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 5942128,
      "step": 9105
    },
    {
      "epoch": 4.7746331236897275,
      "grad_norm": 3.546875,
      "learning_rate": 3.126702745016648e-05,
      "loss": 0.3088,
      "num_input_tokens_seen": 5945648,
      "step": 9110
    },
    {
      "epoch": 4.777253668763103,
      "grad_norm": 3.625,
      "learning_rate": 3.1244886461196976e-05,
      "loss": 0.4099,
      "num_input_tokens_seen": 5950352,
      "step": 9115
    },
    {
      "epoch": 4.779874213836478,
      "grad_norm": 8.0625,
      "learning_rate": 3.1222740246791734e-05,
      "loss": 0.6345,
      "num_input_tokens_seen": 5953488,
      "step": 9120
    },
    {
      "epoch": 4.782494758909853,
      "grad_norm": 6.25,
      "learning_rate": 3.12005888254817e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 5956784,
      "step": 9125
    },
    {
      "epoch": 4.785115303983228,
      "grad_norm": 5.53125,
      "learning_rate": 3.1178432215802155e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 5960080,
      "step": 9130
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 5.84375,
      "learning_rate": 3.115627043629277e-05,
      "loss": 0.4005,
      "num_input_tokens_seen": 5962864,
      "step": 9135
    },
    {
      "epoch": 4.790356394129979,
      "grad_norm": 4.96875,
      "learning_rate": 3.113410350549748e-05,
      "loss": 0.4843,
      "num_input_tokens_seen": 5966000,
      "step": 9140
    },
    {
      "epoch": 4.7929769392033545,
      "grad_norm": 4.65625,
      "learning_rate": 3.111193144196457e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 5969168,
      "step": 9145
    },
    {
      "epoch": 4.79559748427673,
      "grad_norm": 4.3125,
      "learning_rate": 3.1089754264246615e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 5972720,
      "step": 9150
    },
    {
      "epoch": 4.798218029350105,
      "grad_norm": 3.515625,
      "learning_rate": 3.106757199090046e-05,
      "loss": 0.3675,
      "num_input_tokens_seen": 5975760,
      "step": 9155
    },
    {
      "epoch": 4.80083857442348,
      "grad_norm": 6.0,
      "learning_rate": 3.104538464048721e-05,
      "loss": 0.3631,
      "num_input_tokens_seen": 5979120,
      "step": 9160
    },
    {
      "epoch": 4.803459119496855,
      "grad_norm": 4.40625,
      "learning_rate": 3.102319223157225e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 5983376,
      "step": 9165
    },
    {
      "epoch": 4.80607966457023,
      "grad_norm": 3.765625,
      "learning_rate": 3.100099478272515e-05,
      "loss": 0.5694,
      "num_input_tokens_seen": 5986416,
      "step": 9170
    },
    {
      "epoch": 4.808700209643606,
      "grad_norm": 7.46875,
      "learning_rate": 3.097879231251973e-05,
      "loss": 0.4124,
      "num_input_tokens_seen": 5989136,
      "step": 9175
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 8.5625,
      "learning_rate": 3.0956584839534006e-05,
      "loss": 0.6728,
      "num_input_tokens_seen": 5992112,
      "step": 9180
    },
    {
      "epoch": 4.813941299790356,
      "grad_norm": 4.03125,
      "learning_rate": 3.093437238235018e-05,
      "loss": 0.4375,
      "num_input_tokens_seen": 5995472,
      "step": 9185
    },
    {
      "epoch": 4.816561844863732,
      "grad_norm": 6.03125,
      "learning_rate": 3.0912154959554606e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 5998736,
      "step": 9190
    },
    {
      "epoch": 4.819182389937107,
      "grad_norm": 4.21875,
      "learning_rate": 3.088993258973782e-05,
      "loss": 0.6623,
      "num_input_tokens_seen": 6001552,
      "step": 9195
    },
    {
      "epoch": 4.821802935010482,
      "grad_norm": 7.15625,
      "learning_rate": 3.0867705291494486e-05,
      "loss": 0.531,
      "num_input_tokens_seen": 6005520,
      "step": 9200
    },
    {
      "epoch": 4.8244234800838575,
      "grad_norm": 4.625,
      "learning_rate": 3.0845473083423395e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 6009200,
      "step": 9205
    },
    {
      "epoch": 4.827044025157233,
      "grad_norm": 3.75,
      "learning_rate": 3.082323598412743e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 6012912,
      "step": 9210
    },
    {
      "epoch": 4.829664570230608,
      "grad_norm": 4.0,
      "learning_rate": 3.080099401221359e-05,
      "loss": 0.5739,
      "num_input_tokens_seen": 6017744,
      "step": 9215
    },
    {
      "epoch": 4.832285115303983,
      "grad_norm": 6.03125,
      "learning_rate": 3.0778747186292936e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 6020624,
      "step": 9220
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 6.3125,
      "learning_rate": 3.075649552498061e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 6023312,
      "step": 9225
    },
    {
      "epoch": 4.837526205450734,
      "grad_norm": 10.8125,
      "learning_rate": 3.073423904689577e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 6026032,
      "step": 9230
    },
    {
      "epoch": 4.840146750524109,
      "grad_norm": 7.84375,
      "learning_rate": 3.071197777066162e-05,
      "loss": 0.4109,
      "num_input_tokens_seen": 6029872,
      "step": 9235
    },
    {
      "epoch": 4.8427672955974845,
      "grad_norm": 3.234375,
      "learning_rate": 3.068971171490539e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 6033232,
      "step": 9240
    },
    {
      "epoch": 4.84538784067086,
      "grad_norm": 15.0625,
      "learning_rate": 3.066744089825829e-05,
      "loss": 0.6813,
      "num_input_tokens_seen": 6035888,
      "step": 9245
    },
    {
      "epoch": 4.848008385744235,
      "grad_norm": 4.40625,
      "learning_rate": 3.064516533935553e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 6039408,
      "step": 9250
    },
    {
      "epoch": 4.85062893081761,
      "grad_norm": 4.59375,
      "learning_rate": 3.062288505683626e-05,
      "loss": 0.5688,
      "num_input_tokens_seen": 6042736,
      "step": 9255
    },
    {
      "epoch": 4.853249475890985,
      "grad_norm": 5.8125,
      "learning_rate": 3.060060006934363e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 6045296,
      "step": 9260
    },
    {
      "epoch": 4.85587002096436,
      "grad_norm": 7.71875,
      "learning_rate": 3.057831039552469e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 6047920,
      "step": 9265
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 4.40625,
      "learning_rate": 3.0556016054030416e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 6052432,
      "step": 9270
    },
    {
      "epoch": 4.861111111111111,
      "grad_norm": 7.46875,
      "learning_rate": 3.053371706351569e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 6055472,
      "step": 9275
    },
    {
      "epoch": 4.863731656184486,
      "grad_norm": 5.28125,
      "learning_rate": 3.0511413442639296e-05,
      "loss": 0.5957,
      "num_input_tokens_seen": 6059184,
      "step": 9280
    },
    {
      "epoch": 4.866352201257862,
      "grad_norm": 6.375,
      "learning_rate": 3.048910521006389e-05,
      "loss": 0.6131,
      "num_input_tokens_seen": 6062192,
      "step": 9285
    },
    {
      "epoch": 4.868972746331237,
      "grad_norm": 7.78125,
      "learning_rate": 3.046679238445598e-05,
      "loss": 0.4949,
      "num_input_tokens_seen": 6065264,
      "step": 9290
    },
    {
      "epoch": 4.871593291404612,
      "grad_norm": 6.5,
      "learning_rate": 3.0444474984485905e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 6069008,
      "step": 9295
    },
    {
      "epoch": 4.8742138364779874,
      "grad_norm": 4.5625,
      "learning_rate": 3.042215302882786e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 6072624,
      "step": 9300
    },
    {
      "epoch": 4.876834381551363,
      "grad_norm": 4.09375,
      "learning_rate": 3.0399826536159836e-05,
      "loss": 0.5903,
      "num_input_tokens_seen": 6075280,
      "step": 9305
    },
    {
      "epoch": 4.879454926624738,
      "grad_norm": 9.625,
      "learning_rate": 3.0377495525163624e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 6078448,
      "step": 9310
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 6.09375,
      "learning_rate": 3.0355160014524786e-05,
      "loss": 0.3256,
      "num_input_tokens_seen": 6081008,
      "step": 9315
    },
    {
      "epoch": 4.884696016771488,
      "grad_norm": 4.96875,
      "learning_rate": 3.033282002293266e-05,
      "loss": 0.4932,
      "num_input_tokens_seen": 6085296,
      "step": 9320
    },
    {
      "epoch": 4.887316561844864,
      "grad_norm": 6.1875,
      "learning_rate": 3.0310475569080345e-05,
      "loss": 0.4757,
      "num_input_tokens_seen": 6087664,
      "step": 9325
    },
    {
      "epoch": 4.889937106918239,
      "grad_norm": 5.28125,
      "learning_rate": 3.0288126671664628e-05,
      "loss": 0.4897,
      "num_input_tokens_seen": 6090832,
      "step": 9330
    },
    {
      "epoch": 4.8925576519916145,
      "grad_norm": 7.4375,
      "learning_rate": 3.0265773349386078e-05,
      "loss": 0.5488,
      "num_input_tokens_seen": 6093616,
      "step": 9335
    },
    {
      "epoch": 4.89517819706499,
      "grad_norm": 5.875,
      "learning_rate": 3.024341562094891e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 6096592,
      "step": 9340
    },
    {
      "epoch": 4.897798742138365,
      "grad_norm": 2.5625,
      "learning_rate": 3.0221053505061063e-05,
      "loss": 0.3428,
      "num_input_tokens_seen": 6100144,
      "step": 9345
    },
    {
      "epoch": 4.90041928721174,
      "grad_norm": 5.96875,
      "learning_rate": 3.0198687020434142e-05,
      "loss": 0.5675,
      "num_input_tokens_seen": 6103536,
      "step": 9350
    },
    {
      "epoch": 4.903039832285115,
      "grad_norm": 3.84375,
      "learning_rate": 3.0176316185783383e-05,
      "loss": 0.4891,
      "num_input_tokens_seen": 6107408,
      "step": 9355
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 5.03125,
      "learning_rate": 3.015394101982768e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 6110480,
      "step": 9360
    },
    {
      "epoch": 4.9082809224318655,
      "grad_norm": 11.25,
      "learning_rate": 3.013156154128955e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 6113488,
      "step": 9365
    },
    {
      "epoch": 4.910901467505241,
      "grad_norm": 4.3125,
      "learning_rate": 3.010917776889513e-05,
      "loss": 0.4895,
      "num_input_tokens_seen": 6117648,
      "step": 9370
    },
    {
      "epoch": 4.913522012578616,
      "grad_norm": 9.4375,
      "learning_rate": 3.0086789721374137e-05,
      "loss": 0.4838,
      "num_input_tokens_seen": 6120592,
      "step": 9375
    },
    {
      "epoch": 4.916142557651992,
      "grad_norm": 6.5,
      "learning_rate": 3.006439741745985e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 6124400,
      "step": 9380
    },
    {
      "epoch": 4.918763102725367,
      "grad_norm": 4.4375,
      "learning_rate": 3.004200087588914e-05,
      "loss": 0.5348,
      "num_input_tokens_seen": 6128400,
      "step": 9385
    },
    {
      "epoch": 4.921383647798742,
      "grad_norm": 3.09375,
      "learning_rate": 3.00196001154024e-05,
      "loss": 0.4691,
      "num_input_tokens_seen": 6131536,
      "step": 9390
    },
    {
      "epoch": 4.924004192872117,
      "grad_norm": 4.71875,
      "learning_rate": 2.999719515474358e-05,
      "loss": 0.3438,
      "num_input_tokens_seen": 6135280,
      "step": 9395
    },
    {
      "epoch": 4.926624737945493,
      "grad_norm": 8.125,
      "learning_rate": 2.997478601266011e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 6137680,
      "step": 9400
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 6.59375,
      "learning_rate": 2.995237270790295e-05,
      "loss": 0.7049,
      "num_input_tokens_seen": 6141264,
      "step": 9405
    },
    {
      "epoch": 4.931865828092243,
      "grad_norm": 9.9375,
      "learning_rate": 2.9929955259226515e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 6145104,
      "step": 9410
    },
    {
      "epoch": 4.934486373165618,
      "grad_norm": 5.40625,
      "learning_rate": 2.990753368538872e-05,
      "loss": 0.4703,
      "num_input_tokens_seen": 6148976,
      "step": 9415
    },
    {
      "epoch": 4.937106918238994,
      "grad_norm": 8.875,
      "learning_rate": 2.9885108005150897e-05,
      "loss": 0.6315,
      "num_input_tokens_seen": 6152560,
      "step": 9420
    },
    {
      "epoch": 4.939727463312369,
      "grad_norm": 7.21875,
      "learning_rate": 2.986267823727784e-05,
      "loss": 0.6016,
      "num_input_tokens_seen": 6155024,
      "step": 9425
    },
    {
      "epoch": 4.9423480083857445,
      "grad_norm": 5.6875,
      "learning_rate": 2.9840244400537754e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 6157584,
      "step": 9430
    },
    {
      "epoch": 4.94496855345912,
      "grad_norm": 7.59375,
      "learning_rate": 2.9817806513702244e-05,
      "loss": 0.5124,
      "num_input_tokens_seen": 6160240,
      "step": 9435
    },
    {
      "epoch": 4.947589098532495,
      "grad_norm": 4.0,
      "learning_rate": 2.9795364595546315e-05,
      "loss": 0.5393,
      "num_input_tokens_seen": 6163696,
      "step": 9440
    },
    {
      "epoch": 4.95020964360587,
      "grad_norm": 5.71875,
      "learning_rate": 2.977291866484833e-05,
      "loss": 0.3554,
      "num_input_tokens_seen": 6168880,
      "step": 9445
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 8.875,
      "learning_rate": 2.975046874039003e-05,
      "loss": 0.5135,
      "num_input_tokens_seen": 6172208,
      "step": 9450
    },
    {
      "epoch": 4.95545073375262,
      "grad_norm": 4.0625,
      "learning_rate": 2.9728014840956488e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 6174928,
      "step": 9455
    },
    {
      "epoch": 4.9580712788259955,
      "grad_norm": 4.3125,
      "learning_rate": 2.9705556985336086e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 6178992,
      "step": 9460
    },
    {
      "epoch": 4.960691823899371,
      "grad_norm": 6.625,
      "learning_rate": 2.968309519232053e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 6182128,
      "step": 9465
    },
    {
      "epoch": 4.963312368972746,
      "grad_norm": 8.25,
      "learning_rate": 2.966062948070485e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 6185328,
      "step": 9470
    },
    {
      "epoch": 4.965932914046122,
      "grad_norm": 3.890625,
      "learning_rate": 2.9638159869287303e-05,
      "loss": 0.4959,
      "num_input_tokens_seen": 6188560,
      "step": 9475
    },
    {
      "epoch": 4.968553459119497,
      "grad_norm": 3.40625,
      "learning_rate": 2.9615686376869434e-05,
      "loss": 0.5786,
      "num_input_tokens_seen": 6193136,
      "step": 9480
    },
    {
      "epoch": 4.971174004192872,
      "grad_norm": 4.125,
      "learning_rate": 2.9593209022256046e-05,
      "loss": 0.6782,
      "num_input_tokens_seen": 6196816,
      "step": 9485
    },
    {
      "epoch": 4.973794549266247,
      "grad_norm": 3.515625,
      "learning_rate": 2.9570727824255163e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 6200272,
      "step": 9490
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 12.5,
      "learning_rate": 2.954824280167801e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 6203216,
      "step": 9495
    },
    {
      "epoch": 4.979035639412998,
      "grad_norm": 3.609375,
      "learning_rate": 2.9525753973339044e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 6206736,
      "step": 9500
    },
    {
      "epoch": 4.981656184486373,
      "grad_norm": 3.9375,
      "learning_rate": 2.9503261358055873e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 6209936,
      "step": 9505
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 3.59375,
      "learning_rate": 2.9480764974649305e-05,
      "loss": 0.4997,
      "num_input_tokens_seen": 6213424,
      "step": 9510
    },
    {
      "epoch": 4.986897274633124,
      "grad_norm": 7.6875,
      "learning_rate": 2.9458264841943272e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 6217360,
      "step": 9515
    },
    {
      "epoch": 4.989517819706499,
      "grad_norm": 6.96875,
      "learning_rate": 2.9435760978764874e-05,
      "loss": 0.5307,
      "num_input_tokens_seen": 6219984,
      "step": 9520
    },
    {
      "epoch": 4.9921383647798745,
      "grad_norm": 5.0,
      "learning_rate": 2.9413253403944297e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 6223664,
      "step": 9525
    },
    {
      "epoch": 4.99475890985325,
      "grad_norm": 3.921875,
      "learning_rate": 2.9390742136314863e-05,
      "loss": 0.4941,
      "num_input_tokens_seen": 6227440,
      "step": 9530
    },
    {
      "epoch": 4.997379454926625,
      "grad_norm": 5.8125,
      "learning_rate": 2.9368227194712978e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 6230608,
      "step": 9535
    },
    {
      "epoch": 5.0,
      "grad_norm": 18.5,
      "learning_rate": 2.9345708597978106e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 6233920,
      "step": 9540
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.4963121712207794,
      "eval_runtime": 24.2805,
      "eval_samples_per_second": 34.925,
      "eval_steps_per_second": 8.731,
      "num_input_tokens_seen": 6233920,
      "step": 9540
    },
    {
      "epoch": 5.002620545073375,
      "grad_norm": 5.125,
      "learning_rate": 2.932318636495278e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 6237312,
      "step": 9545
    },
    {
      "epoch": 5.00524109014675,
      "grad_norm": 5.6875,
      "learning_rate": 2.930066051448258e-05,
      "loss": 0.4196,
      "num_input_tokens_seen": 6241344,
      "step": 9550
    },
    {
      "epoch": 5.0078616352201255,
      "grad_norm": 4.78125,
      "learning_rate": 2.927813106541611e-05,
      "loss": 0.5716,
      "num_input_tokens_seen": 6244064,
      "step": 9555
    },
    {
      "epoch": 5.010482180293501,
      "grad_norm": 5.75,
      "learning_rate": 2.9255598036604982e-05,
      "loss": 0.6902,
      "num_input_tokens_seen": 6247008,
      "step": 9560
    },
    {
      "epoch": 5.013102725366877,
      "grad_norm": 5.65625,
      "learning_rate": 2.92330614469038e-05,
      "loss": 0.5819,
      "num_input_tokens_seen": 6250016,
      "step": 9565
    },
    {
      "epoch": 5.015723270440252,
      "grad_norm": 2.59375,
      "learning_rate": 2.921052131517016e-05,
      "loss": 0.6075,
      "num_input_tokens_seen": 6253504,
      "step": 9570
    },
    {
      "epoch": 5.018343815513627,
      "grad_norm": 3.375,
      "learning_rate": 2.9187977660264615e-05,
      "loss": 0.514,
      "num_input_tokens_seen": 6256256,
      "step": 9575
    },
    {
      "epoch": 5.020964360587002,
      "grad_norm": 4.3125,
      "learning_rate": 2.9165430501050657e-05,
      "loss": 0.6581,
      "num_input_tokens_seen": 6260160,
      "step": 9580
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 3.71875,
      "learning_rate": 2.9142879856394732e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 6264032,
      "step": 9585
    },
    {
      "epoch": 5.026205450733753,
      "grad_norm": 4.09375,
      "learning_rate": 2.9120325745166178e-05,
      "loss": 0.5216,
      "num_input_tokens_seen": 6267552,
      "step": 9590
    },
    {
      "epoch": 5.028825995807128,
      "grad_norm": 3.671875,
      "learning_rate": 2.909776818623725e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 6270272,
      "step": 9595
    },
    {
      "epoch": 5.031446540880503,
      "grad_norm": 4.6875,
      "learning_rate": 2.9075207198483084e-05,
      "loss": 0.3184,
      "num_input_tokens_seen": 6274624,
      "step": 9600
    },
    {
      "epoch": 5.034067085953878,
      "grad_norm": 5.84375,
      "learning_rate": 2.905264280078168e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 6277728,
      "step": 9605
    },
    {
      "epoch": 5.036687631027253,
      "grad_norm": 8.6875,
      "learning_rate": 2.9030075012013902e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 6281024,
      "step": 9610
    },
    {
      "epoch": 5.039308176100629,
      "grad_norm": 6.1875,
      "learning_rate": 2.9007503851063433e-05,
      "loss": 0.6068,
      "num_input_tokens_seen": 6284576,
      "step": 9615
    },
    {
      "epoch": 5.0419287211740045,
      "grad_norm": 3.5625,
      "learning_rate": 2.8984929336816807e-05,
      "loss": 0.5626,
      "num_input_tokens_seen": 6288096,
      "step": 9620
    },
    {
      "epoch": 5.04454926624738,
      "grad_norm": 4.71875,
      "learning_rate": 2.896235148816333e-05,
      "loss": 0.5376,
      "num_input_tokens_seen": 6292128,
      "step": 9625
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 7.09375,
      "learning_rate": 2.893977032399512e-05,
      "loss": 0.5201,
      "num_input_tokens_seen": 6298912,
      "step": 9630
    },
    {
      "epoch": 5.04979035639413,
      "grad_norm": 5.1875,
      "learning_rate": 2.8917185863207062e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 6301536,
      "step": 9635
    },
    {
      "epoch": 5.052410901467505,
      "grad_norm": 7.875,
      "learning_rate": 2.889459812469681e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 6303936,
      "step": 9640
    },
    {
      "epoch": 5.05503144654088,
      "grad_norm": 11.8125,
      "learning_rate": 2.8872007127364746e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 6306880,
      "step": 9645
    },
    {
      "epoch": 5.0576519916142555,
      "grad_norm": 5.09375,
      "learning_rate": 2.884941289011398e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 6309920,
      "step": 9650
    },
    {
      "epoch": 5.060272536687631,
      "grad_norm": 7.21875,
      "learning_rate": 2.882681543185034e-05,
      "loss": 0.4279,
      "num_input_tokens_seen": 6312480,
      "step": 9655
    },
    {
      "epoch": 5.062893081761007,
      "grad_norm": 3.09375,
      "learning_rate": 2.880421477148235e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 6316256,
      "step": 9660
    },
    {
      "epoch": 5.065513626834382,
      "grad_norm": 5.0,
      "learning_rate": 2.878161092792121e-05,
      "loss": 0.5477,
      "num_input_tokens_seen": 6319520,
      "step": 9665
    },
    {
      "epoch": 5.068134171907757,
      "grad_norm": 3.921875,
      "learning_rate": 2.8759003920080786e-05,
      "loss": 0.5928,
      "num_input_tokens_seen": 6323328,
      "step": 9670
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 4.25,
      "learning_rate": 2.8736393766877578e-05,
      "loss": 0.3158,
      "num_input_tokens_seen": 6326720,
      "step": 9675
    },
    {
      "epoch": 5.073375262054507,
      "grad_norm": 7.71875,
      "learning_rate": 2.871378048723074e-05,
      "loss": 0.5027,
      "num_input_tokens_seen": 6329280,
      "step": 9680
    },
    {
      "epoch": 5.075995807127883,
      "grad_norm": 3.25,
      "learning_rate": 2.8691164100062034e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 6332384,
      "step": 9685
    },
    {
      "epoch": 5.078616352201258,
      "grad_norm": 7.0,
      "learning_rate": 2.8668544624295814e-05,
      "loss": 0.3851,
      "num_input_tokens_seen": 6336096,
      "step": 9690
    },
    {
      "epoch": 5.081236897274633,
      "grad_norm": 8.0,
      "learning_rate": 2.864592207885902e-05,
      "loss": 0.5654,
      "num_input_tokens_seen": 6338432,
      "step": 9695
    },
    {
      "epoch": 5.083857442348008,
      "grad_norm": 4.53125,
      "learning_rate": 2.8623296482681166e-05,
      "loss": 0.4886,
      "num_input_tokens_seen": 6342016,
      "step": 9700
    },
    {
      "epoch": 5.086477987421383,
      "grad_norm": 13.5,
      "learning_rate": 2.8600667854694328e-05,
      "loss": 0.5881,
      "num_input_tokens_seen": 6344896,
      "step": 9705
    },
    {
      "epoch": 5.089098532494759,
      "grad_norm": 6.6875,
      "learning_rate": 2.857803621383311e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 6348096,
      "step": 9710
    },
    {
      "epoch": 5.0917190775681345,
      "grad_norm": 7.6875,
      "learning_rate": 2.8555401579034607e-05,
      "loss": 0.6144,
      "num_input_tokens_seen": 6351488,
      "step": 9715
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 6.75,
      "learning_rate": 2.853276396923848e-05,
      "loss": 0.5445,
      "num_input_tokens_seen": 6354464,
      "step": 9720
    },
    {
      "epoch": 5.096960167714885,
      "grad_norm": 4.46875,
      "learning_rate": 2.851012340338683e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 6358912,
      "step": 9725
    },
    {
      "epoch": 5.09958071278826,
      "grad_norm": 11.4375,
      "learning_rate": 2.8487479900424253e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 6362112,
      "step": 9730
    },
    {
      "epoch": 5.102201257861635,
      "grad_norm": 6.21875,
      "learning_rate": 2.8464833479297794e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 6364896,
      "step": 9735
    },
    {
      "epoch": 5.10482180293501,
      "grad_norm": 5.46875,
      "learning_rate": 2.8442184158956947e-05,
      "loss": 0.6706,
      "num_input_tokens_seen": 6368160,
      "step": 9740
    },
    {
      "epoch": 5.1074423480083855,
      "grad_norm": 7.75,
      "learning_rate": 2.8419531958353635e-05,
      "loss": 0.4027,
      "num_input_tokens_seen": 6371264,
      "step": 9745
    },
    {
      "epoch": 5.110062893081761,
      "grad_norm": 5.9375,
      "learning_rate": 2.839687689644217e-05,
      "loss": 0.5575,
      "num_input_tokens_seen": 6374144,
      "step": 9750
    },
    {
      "epoch": 5.112683438155136,
      "grad_norm": 5.65625,
      "learning_rate": 2.837421899217928e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 6377600,
      "step": 9755
    },
    {
      "epoch": 5.115303983228512,
      "grad_norm": 7.375,
      "learning_rate": 2.8351558264524076e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 6380608,
      "step": 9760
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 4.6875,
      "learning_rate": 2.8328894732437998e-05,
      "loss": 0.6185,
      "num_input_tokens_seen": 6384640,
      "step": 9765
    },
    {
      "epoch": 5.120545073375262,
      "grad_norm": 8.25,
      "learning_rate": 2.830622841488488e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 6387456,
      "step": 9770
    },
    {
      "epoch": 5.123165618448637,
      "grad_norm": 5.0,
      "learning_rate": 2.8283559330830834e-05,
      "loss": 0.467,
      "num_input_tokens_seen": 6390880,
      "step": 9775
    },
    {
      "epoch": 5.1257861635220126,
      "grad_norm": 7.71875,
      "learning_rate": 2.8260887499244333e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 6393696,
      "step": 9780
    },
    {
      "epoch": 5.128406708595388,
      "grad_norm": 3.078125,
      "learning_rate": 2.823821293909612e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 6397472,
      "step": 9785
    },
    {
      "epoch": 5.131027253668763,
      "grad_norm": 3.109375,
      "learning_rate": 2.821553566935924e-05,
      "loss": 0.4786,
      "num_input_tokens_seen": 6400576,
      "step": 9790
    },
    {
      "epoch": 5.133647798742138,
      "grad_norm": 3.515625,
      "learning_rate": 2.8192855709008985e-05,
      "loss": 0.5944,
      "num_input_tokens_seen": 6403168,
      "step": 9795
    },
    {
      "epoch": 5.136268343815513,
      "grad_norm": 4.8125,
      "learning_rate": 2.8170173077022915e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 6406144,
      "step": 9800
    },
    {
      "epoch": 5.138888888888889,
      "grad_norm": 8.0,
      "learning_rate": 2.8147487792380832e-05,
      "loss": 0.5756,
      "num_input_tokens_seen": 6409344,
      "step": 9805
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 8.0625,
      "learning_rate": 2.8124799874064733e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 6411776,
      "step": 9810
    },
    {
      "epoch": 5.14412997903564,
      "grad_norm": 3.09375,
      "learning_rate": 2.810210934105883e-05,
      "loss": 0.3794,
      "num_input_tokens_seen": 6414784,
      "step": 9815
    },
    {
      "epoch": 5.146750524109015,
      "grad_norm": 12.9375,
      "learning_rate": 2.8079416212349528e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 6417248,
      "step": 9820
    },
    {
      "epoch": 5.14937106918239,
      "grad_norm": 9.875,
      "learning_rate": 2.805672050692541e-05,
      "loss": 0.6054,
      "num_input_tokens_seen": 6420416,
      "step": 9825
    },
    {
      "epoch": 5.151991614255765,
      "grad_norm": 6.625,
      "learning_rate": 2.8034022243777197e-05,
      "loss": 0.4094,
      "num_input_tokens_seen": 6424288,
      "step": 9830
    },
    {
      "epoch": 5.15461215932914,
      "grad_norm": 6.03125,
      "learning_rate": 2.8011321441897754e-05,
      "loss": 0.5911,
      "num_input_tokens_seen": 6426784,
      "step": 9835
    },
    {
      "epoch": 5.1572327044025155,
      "grad_norm": 2.765625,
      "learning_rate": 2.7988618120282074e-05,
      "loss": 0.5409,
      "num_input_tokens_seen": 6429760,
      "step": 9840
    },
    {
      "epoch": 5.159853249475891,
      "grad_norm": 4.6875,
      "learning_rate": 2.7965912297927277e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 6432480,
      "step": 9845
    },
    {
      "epoch": 5.162473794549266,
      "grad_norm": 4.25,
      "learning_rate": 2.794320399383254e-05,
      "loss": 0.443,
      "num_input_tokens_seen": 6437312,
      "step": 9850
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 4.75,
      "learning_rate": 2.7920493226999143e-05,
      "loss": 0.5965,
      "num_input_tokens_seen": 6441664,
      "step": 9855
    },
    {
      "epoch": 5.167714884696017,
      "grad_norm": 3.109375,
      "learning_rate": 2.7897780016430414e-05,
      "loss": 0.3973,
      "num_input_tokens_seen": 6445120,
      "step": 9860
    },
    {
      "epoch": 5.170335429769392,
      "grad_norm": 5.59375,
      "learning_rate": 2.7875064381131733e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 6447968,
      "step": 9865
    },
    {
      "epoch": 5.172955974842767,
      "grad_norm": 6.25,
      "learning_rate": 2.7852346340110508e-05,
      "loss": 0.4597,
      "num_input_tokens_seen": 6450592,
      "step": 9870
    },
    {
      "epoch": 5.1755765199161425,
      "grad_norm": 9.8125,
      "learning_rate": 2.7829625912376163e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 6454496,
      "step": 9875
    },
    {
      "epoch": 5.178197064989518,
      "grad_norm": 8.25,
      "learning_rate": 2.7806903116940093e-05,
      "loss": 0.4038,
      "num_input_tokens_seen": 6457088,
      "step": 9880
    },
    {
      "epoch": 5.180817610062893,
      "grad_norm": 7.34375,
      "learning_rate": 2.778417797281571e-05,
      "loss": 0.556,
      "num_input_tokens_seen": 6461088,
      "step": 9885
    },
    {
      "epoch": 5.183438155136268,
      "grad_norm": 3.84375,
      "learning_rate": 2.7761450499018383e-05,
      "loss": 0.5227,
      "num_input_tokens_seen": 6463872,
      "step": 9890
    },
    {
      "epoch": 5.186058700209643,
      "grad_norm": 4.6875,
      "learning_rate": 2.7738720714565418e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 6467424,
      "step": 9895
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 6.5625,
      "learning_rate": 2.7715988638476055e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 6470688,
      "step": 9900
    },
    {
      "epoch": 5.191299790356394,
      "grad_norm": 3.703125,
      "learning_rate": 2.7693254289771454e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 6473440,
      "step": 9905
    },
    {
      "epoch": 5.19392033542977,
      "grad_norm": 12.6875,
      "learning_rate": 2.7670517687474697e-05,
      "loss": 0.6013,
      "num_input_tokens_seen": 6476352,
      "step": 9910
    },
    {
      "epoch": 5.196540880503145,
      "grad_norm": 3.640625,
      "learning_rate": 2.7647778850610723e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 6480064,
      "step": 9915
    },
    {
      "epoch": 5.19916142557652,
      "grad_norm": 6.375,
      "learning_rate": 2.7625037798206345e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 6483840,
      "step": 9920
    },
    {
      "epoch": 5.201781970649895,
      "grad_norm": 6.84375,
      "learning_rate": 2.7602294549290243e-05,
      "loss": 0.6888,
      "num_input_tokens_seen": 6486368,
      "step": 9925
    },
    {
      "epoch": 5.20440251572327,
      "grad_norm": 4.3125,
      "learning_rate": 2.757954912289294e-05,
      "loss": 0.4048,
      "num_input_tokens_seen": 6489152,
      "step": 9930
    },
    {
      "epoch": 5.2070230607966455,
      "grad_norm": 4.8125,
      "learning_rate": 2.755680153804675e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 6493088,
      "step": 9935
    },
    {
      "epoch": 5.209643605870021,
      "grad_norm": 7.9375,
      "learning_rate": 2.7534051813785834e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 6496256,
      "step": 9940
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 6.03125,
      "learning_rate": 2.75112999691461e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 6499296,
      "step": 9945
    },
    {
      "epoch": 5.214884696016772,
      "grad_norm": 6.25,
      "learning_rate": 2.7488546023165262e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 6502624,
      "step": 9950
    },
    {
      "epoch": 5.217505241090147,
      "grad_norm": 3.65625,
      "learning_rate": 2.7465789994882796e-05,
      "loss": 0.6734,
      "num_input_tokens_seen": 6506272,
      "step": 9955
    },
    {
      "epoch": 5.220125786163522,
      "grad_norm": 9.0625,
      "learning_rate": 2.7443031903339896e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 6509504,
      "step": 9960
    },
    {
      "epoch": 5.222746331236897,
      "grad_norm": 8.1875,
      "learning_rate": 2.742027176757948e-05,
      "loss": 0.5674,
      "num_input_tokens_seen": 6512512,
      "step": 9965
    },
    {
      "epoch": 5.2253668763102725,
      "grad_norm": 3.875,
      "learning_rate": 2.7397509606646204e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 6515456,
      "step": 9970
    },
    {
      "epoch": 5.227987421383648,
      "grad_norm": 6.21875,
      "learning_rate": 2.7374745439586414e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 6518880,
      "step": 9975
    },
    {
      "epoch": 5.230607966457023,
      "grad_norm": 11.0,
      "learning_rate": 2.735197928544811e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 6521664,
      "step": 9980
    },
    {
      "epoch": 5.233228511530398,
      "grad_norm": 7.8125,
      "learning_rate": 2.7329211163280972e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 6524160,
      "step": 9985
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 4.875,
      "learning_rate": 2.730644109213632e-05,
      "loss": 0.3846,
      "num_input_tokens_seen": 6527904,
      "step": 9990
    },
    {
      "epoch": 5.238469601677149,
      "grad_norm": 15.3125,
      "learning_rate": 2.7283669091067127e-05,
      "loss": 0.7208,
      "num_input_tokens_seen": 6530560,
      "step": 9995
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 7.0625,
      "learning_rate": 2.7260895179127944e-05,
      "loss": 0.53,
      "num_input_tokens_seen": 6534048,
      "step": 10000
    },
    {
      "epoch": 5.2437106918239,
      "grad_norm": 5.78125,
      "learning_rate": 2.7238119375374954e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 6536864,
      "step": 10005
    },
    {
      "epoch": 5.246331236897275,
      "grad_norm": 10.75,
      "learning_rate": 2.7215341698865904e-05,
      "loss": 0.5663,
      "num_input_tokens_seen": 6540608,
      "step": 10010
    },
    {
      "epoch": 5.24895178197065,
      "grad_norm": 6.75,
      "learning_rate": 2.7192562168660113e-05,
      "loss": 0.5439,
      "num_input_tokens_seen": 6543744,
      "step": 10015
    },
    {
      "epoch": 5.251572327044025,
      "grad_norm": 7.03125,
      "learning_rate": 2.7169780803818445e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 6546656,
      "step": 10020
    },
    {
      "epoch": 5.2541928721174,
      "grad_norm": 8.75,
      "learning_rate": 2.714699762340332e-05,
      "loss": 0.5514,
      "num_input_tokens_seen": 6549152,
      "step": 10025
    },
    {
      "epoch": 5.256813417190775,
      "grad_norm": 7.65625,
      "learning_rate": 2.7124212646478652e-05,
      "loss": 0.5647,
      "num_input_tokens_seen": 6552096,
      "step": 10030
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 4.84375,
      "learning_rate": 2.7101425892109865e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 6555232,
      "step": 10035
    },
    {
      "epoch": 5.262054507337526,
      "grad_norm": 7.21875,
      "learning_rate": 2.707863737936389e-05,
      "loss": 0.4293,
      "num_input_tokens_seen": 6557920,
      "step": 10040
    },
    {
      "epoch": 5.264675052410902,
      "grad_norm": 5.5,
      "learning_rate": 2.7055847127309107e-05,
      "loss": 0.4911,
      "num_input_tokens_seen": 6560928,
      "step": 10045
    },
    {
      "epoch": 5.267295597484277,
      "grad_norm": 5.4375,
      "learning_rate": 2.703305515501534e-05,
      "loss": 0.4572,
      "num_input_tokens_seen": 6563488,
      "step": 10050
    },
    {
      "epoch": 5.269916142557652,
      "grad_norm": 6.0625,
      "learning_rate": 2.70102614815539e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 6567520,
      "step": 10055
    },
    {
      "epoch": 5.272536687631027,
      "grad_norm": 3.78125,
      "learning_rate": 2.6987466125997475e-05,
      "loss": 0.5185,
      "num_input_tokens_seen": 6570528,
      "step": 10060
    },
    {
      "epoch": 5.2751572327044025,
      "grad_norm": 7.21875,
      "learning_rate": 2.696466910742018e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 6573824,
      "step": 10065
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 3.65625,
      "learning_rate": 2.694187044489751e-05,
      "loss": 0.5301,
      "num_input_tokens_seen": 6576896,
      "step": 10070
    },
    {
      "epoch": 5.280398322851153,
      "grad_norm": 15.375,
      "learning_rate": 2.691907015750636e-05,
      "loss": 0.5151,
      "num_input_tokens_seen": 6579136,
      "step": 10075
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 5.6875,
      "learning_rate": 2.6896268264324964e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 6582080,
      "step": 10080
    },
    {
      "epoch": 5.285639412997903,
      "grad_norm": 7.65625,
      "learning_rate": 2.6873464784432894e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 6584896,
      "step": 10085
    },
    {
      "epoch": 5.288259958071279,
      "grad_norm": 4.78125,
      "learning_rate": 2.6850659736911073e-05,
      "loss": 0.4906,
      "num_input_tokens_seen": 6588448,
      "step": 10090
    },
    {
      "epoch": 5.290880503144654,
      "grad_norm": 8.125,
      "learning_rate": 2.682785314084172e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 6592704,
      "step": 10095
    },
    {
      "epoch": 5.29350104821803,
      "grad_norm": 6.0,
      "learning_rate": 2.680504501530835e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 6596256,
      "step": 10100
    },
    {
      "epoch": 5.296121593291405,
      "grad_norm": 5.5625,
      "learning_rate": 2.6782235379395766e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 6598656,
      "step": 10105
    },
    {
      "epoch": 5.29874213836478,
      "grad_norm": 5.0625,
      "learning_rate": 2.675942425219002e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 6601504,
      "step": 10110
    },
    {
      "epoch": 5.301362683438155,
      "grad_norm": 15.375,
      "learning_rate": 2.673661165277843e-05,
      "loss": 0.5498,
      "num_input_tokens_seen": 6604640,
      "step": 10115
    },
    {
      "epoch": 5.30398322851153,
      "grad_norm": 4.71875,
      "learning_rate": 2.6713797600249536e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 6607520,
      "step": 10120
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 4.21875,
      "learning_rate": 2.6690982113693092e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 6611360,
      "step": 10125
    },
    {
      "epoch": 5.309224318658281,
      "grad_norm": 3.421875,
      "learning_rate": 2.6668165212200057e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 6616672,
      "step": 10130
    },
    {
      "epoch": 5.311844863731656,
      "grad_norm": 5.75,
      "learning_rate": 2.664534691486257e-05,
      "loss": 0.644,
      "num_input_tokens_seen": 6619424,
      "step": 10135
    },
    {
      "epoch": 5.314465408805032,
      "grad_norm": 7.53125,
      "learning_rate": 2.6622527240773942e-05,
      "loss": 0.3554,
      "num_input_tokens_seen": 6622656,
      "step": 10140
    },
    {
      "epoch": 5.317085953878407,
      "grad_norm": 7.9375,
      "learning_rate": 2.6599706209028634e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 6625568,
      "step": 10145
    },
    {
      "epoch": 5.319706498951782,
      "grad_norm": 21.75,
      "learning_rate": 2.657688383872224e-05,
      "loss": 0.6901,
      "num_input_tokens_seen": 6628896,
      "step": 10150
    },
    {
      "epoch": 5.322327044025157,
      "grad_norm": 5.53125,
      "learning_rate": 2.655406014895147e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 6632672,
      "step": 10155
    },
    {
      "epoch": 5.3249475890985325,
      "grad_norm": 4.21875,
      "learning_rate": 2.653123515881417e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 6637344,
      "step": 10160
    },
    {
      "epoch": 5.327568134171908,
      "grad_norm": 6.1875,
      "learning_rate": 2.650840888740923e-05,
      "loss": 0.6067,
      "num_input_tokens_seen": 6640416,
      "step": 10165
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 5.84375,
      "learning_rate": 2.6485581353836624e-05,
      "loss": 0.3551,
      "num_input_tokens_seen": 6642688,
      "step": 10170
    },
    {
      "epoch": 5.332809224318658,
      "grad_norm": 6.53125,
      "learning_rate": 2.6462752577197407e-05,
      "loss": 0.5249,
      "num_input_tokens_seen": 6645952,
      "step": 10175
    },
    {
      "epoch": 5.335429769392033,
      "grad_norm": 6.03125,
      "learning_rate": 2.643992257659365e-05,
      "loss": 0.4953,
      "num_input_tokens_seen": 6648576,
      "step": 10180
    },
    {
      "epoch": 5.338050314465409,
      "grad_norm": 6.5625,
      "learning_rate": 2.641709137112845e-05,
      "loss": 0.5802,
      "num_input_tokens_seen": 6651840,
      "step": 10185
    },
    {
      "epoch": 5.340670859538784,
      "grad_norm": 7.21875,
      "learning_rate": 2.639425897990593e-05,
      "loss": 0.636,
      "num_input_tokens_seen": 6654336,
      "step": 10190
    },
    {
      "epoch": 5.34329140461216,
      "grad_norm": 7.25,
      "learning_rate": 2.6371425422031172e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 6657600,
      "step": 10195
    },
    {
      "epoch": 5.345911949685535,
      "grad_norm": 5.40625,
      "learning_rate": 2.6348590716610273e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 6660768,
      "step": 10200
    },
    {
      "epoch": 5.34853249475891,
      "grad_norm": 11.1875,
      "learning_rate": 2.6325754882750252e-05,
      "loss": 0.5882,
      "num_input_tokens_seen": 6663808,
      "step": 10205
    },
    {
      "epoch": 5.351153039832285,
      "grad_norm": 3.171875,
      "learning_rate": 2.630291793955911e-05,
      "loss": 0.4971,
      "num_input_tokens_seen": 6666048,
      "step": 10210
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 4.34375,
      "learning_rate": 2.6280079906145756e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 6668672,
      "step": 10215
    },
    {
      "epoch": 5.356394129979035,
      "grad_norm": 10.25,
      "learning_rate": 2.6257240801620004e-05,
      "loss": 0.6369,
      "num_input_tokens_seen": 6671680,
      "step": 10220
    },
    {
      "epoch": 5.359014675052411,
      "grad_norm": 6.4375,
      "learning_rate": 2.623440064509258e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 6674592,
      "step": 10225
    },
    {
      "epoch": 5.361635220125786,
      "grad_norm": 3.890625,
      "learning_rate": 2.621155945567508e-05,
      "loss": 0.412,
      "num_input_tokens_seen": 6677632,
      "step": 10230
    },
    {
      "epoch": 5.364255765199162,
      "grad_norm": 6.875,
      "learning_rate": 2.6188717252479968e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 6680768,
      "step": 10235
    },
    {
      "epoch": 5.366876310272537,
      "grad_norm": 7.15625,
      "learning_rate": 2.6165874054620552e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 6683712,
      "step": 10240
    },
    {
      "epoch": 5.369496855345912,
      "grad_norm": 3.875,
      "learning_rate": 2.614302988121099e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 6688512,
      "step": 10245
    },
    {
      "epoch": 5.372117400419287,
      "grad_norm": 5.15625,
      "learning_rate": 2.6120184751366238e-05,
      "loss": 0.4571,
      "num_input_tokens_seen": 6691520,
      "step": 10250
    },
    {
      "epoch": 5.3747379454926625,
      "grad_norm": 3.46875,
      "learning_rate": 2.6097338684202043e-05,
      "loss": 0.5699,
      "num_input_tokens_seen": 6695360,
      "step": 10255
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 3.796875,
      "learning_rate": 2.607449169883497e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 6698688,
      "step": 10260
    },
    {
      "epoch": 5.379979035639413,
      "grad_norm": 7.0625,
      "learning_rate": 2.605164381438232e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 6701312,
      "step": 10265
    },
    {
      "epoch": 5.382599580712788,
      "grad_norm": 5.625,
      "learning_rate": 2.6028795049962167e-05,
      "loss": 0.4872,
      "num_input_tokens_seen": 6703840,
      "step": 10270
    },
    {
      "epoch": 5.385220125786163,
      "grad_norm": 4.875,
      "learning_rate": 2.600594542469331e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 6706720,
      "step": 10275
    },
    {
      "epoch": 5.387840670859539,
      "grad_norm": 6.6875,
      "learning_rate": 2.5983094957695263e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 6710240,
      "step": 10280
    },
    {
      "epoch": 5.390461215932914,
      "grad_norm": 5.34375,
      "learning_rate": 2.596024366808827e-05,
      "loss": 0.607,
      "num_input_tokens_seen": 6713632,
      "step": 10285
    },
    {
      "epoch": 5.3930817610062896,
      "grad_norm": 4.1875,
      "learning_rate": 2.5937391574993238e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 6716032,
      "step": 10290
    },
    {
      "epoch": 5.395702306079665,
      "grad_norm": 9.5,
      "learning_rate": 2.5914538697531755e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 6719136,
      "step": 10295
    },
    {
      "epoch": 5.39832285115304,
      "grad_norm": 12.625,
      "learning_rate": 2.5891685054826054e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 6721824,
      "step": 10300
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 3.5,
      "learning_rate": 2.586883066599904e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 6724864,
      "step": 10305
    },
    {
      "epoch": 5.40356394129979,
      "grad_norm": 6.34375,
      "learning_rate": 2.5845975550174206e-05,
      "loss": 0.3925,
      "num_input_tokens_seen": 6728064,
      "step": 10310
    },
    {
      "epoch": 5.406184486373165,
      "grad_norm": 5.9375,
      "learning_rate": 2.5823119726475682e-05,
      "loss": 0.541,
      "num_input_tokens_seen": 6730784,
      "step": 10315
    },
    {
      "epoch": 5.408805031446541,
      "grad_norm": 3.71875,
      "learning_rate": 2.5800263214028153e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 6733408,
      "step": 10320
    },
    {
      "epoch": 5.411425576519916,
      "grad_norm": 0.80078125,
      "learning_rate": 2.5777406031956935e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 6738912,
      "step": 10325
    },
    {
      "epoch": 5.414046121593292,
      "grad_norm": 6.71875,
      "learning_rate": 2.5754548199387863e-05,
      "loss": 0.4623,
      "num_input_tokens_seen": 6742784,
      "step": 10330
    },
    {
      "epoch": 5.416666666666667,
      "grad_norm": 5.03125,
      "learning_rate": 2.5731689735447317e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 6746080,
      "step": 10335
    },
    {
      "epoch": 5.419287211740042,
      "grad_norm": 7.90625,
      "learning_rate": 2.5708830659262218e-05,
      "loss": 0.6127,
      "num_input_tokens_seen": 6750368,
      "step": 10340
    },
    {
      "epoch": 5.421907756813417,
      "grad_norm": 7.875,
      "learning_rate": 2.5685970989960005e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 6753984,
      "step": 10345
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 6.0,
      "learning_rate": 2.5663110746668612e-05,
      "loss": 0.5879,
      "num_input_tokens_seen": 6756576,
      "step": 10350
    },
    {
      "epoch": 5.427148846960168,
      "grad_norm": 6.9375,
      "learning_rate": 2.564024994851642e-05,
      "loss": 0.4529,
      "num_input_tokens_seen": 6759904,
      "step": 10355
    },
    {
      "epoch": 5.429769392033543,
      "grad_norm": 4.84375,
      "learning_rate": 2.561738861463232e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 6762880,
      "step": 10360
    },
    {
      "epoch": 5.432389937106918,
      "grad_norm": 5.21875,
      "learning_rate": 2.559452676414564e-05,
      "loss": 0.6656,
      "num_input_tokens_seen": 6765728,
      "step": 10365
    },
    {
      "epoch": 5.435010482180293,
      "grad_norm": 9.375,
      "learning_rate": 2.5571664416186108e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 6768704,
      "step": 10370
    },
    {
      "epoch": 5.437631027253669,
      "grad_norm": 5.71875,
      "learning_rate": 2.5548801589883913e-05,
      "loss": 0.3641,
      "num_input_tokens_seen": 6772256,
      "step": 10375
    },
    {
      "epoch": 5.440251572327044,
      "grad_norm": 3.828125,
      "learning_rate": 2.5525938304369614e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 6775552,
      "step": 10380
    },
    {
      "epoch": 5.4428721174004195,
      "grad_norm": 9.375,
      "learning_rate": 2.5503074578774166e-05,
      "loss": 0.398,
      "num_input_tokens_seen": 6778176,
      "step": 10385
    },
    {
      "epoch": 5.445492662473795,
      "grad_norm": 5.46875,
      "learning_rate": 2.5480210432228886e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 6781888,
      "step": 10390
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 6.09375,
      "learning_rate": 2.5457345883865457e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 6785056,
      "step": 10395
    },
    {
      "epoch": 5.450733752620545,
      "grad_norm": 11.6875,
      "learning_rate": 2.5434480952815877e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 6788832,
      "step": 10400
    },
    {
      "epoch": 5.45335429769392,
      "grad_norm": 5.53125,
      "learning_rate": 2.5411615658212478e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 6792320,
      "step": 10405
    },
    {
      "epoch": 5.455974842767295,
      "grad_norm": 3.609375,
      "learning_rate": 2.5388750019187912e-05,
      "loss": 0.5886,
      "num_input_tokens_seen": 6796000,
      "step": 10410
    },
    {
      "epoch": 5.4585953878406706,
      "grad_norm": 6.09375,
      "learning_rate": 2.5365884054875084e-05,
      "loss": 0.4688,
      "num_input_tokens_seen": 6798400,
      "step": 10415
    },
    {
      "epoch": 5.461215932914046,
      "grad_norm": 10.4375,
      "learning_rate": 2.5343017784407184e-05,
      "loss": 0.5499,
      "num_input_tokens_seen": 6801440,
      "step": 10420
    },
    {
      "epoch": 5.463836477987422,
      "grad_norm": 5.09375,
      "learning_rate": 2.532015122691767e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 6807072,
      "step": 10425
    },
    {
      "epoch": 5.466457023060797,
      "grad_norm": 10.0625,
      "learning_rate": 2.5297284401540243e-05,
      "loss": 0.4902,
      "num_input_tokens_seen": 6810272,
      "step": 10430
    },
    {
      "epoch": 5.469077568134172,
      "grad_norm": 7.65625,
      "learning_rate": 2.5274417327408805e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 6813184,
      "step": 10435
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 6.875,
      "learning_rate": 2.5251550023657478e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 6817824,
      "step": 10440
    },
    {
      "epoch": 5.4743186582809225,
      "grad_norm": 4.8125,
      "learning_rate": 2.5228682509420582e-05,
      "loss": 0.4367,
      "num_input_tokens_seen": 6820480,
      "step": 10445
    },
    {
      "epoch": 5.476939203354298,
      "grad_norm": 8.375,
      "learning_rate": 2.5205814803832617e-05,
      "loss": 0.5121,
      "num_input_tokens_seen": 6823360,
      "step": 10450
    },
    {
      "epoch": 5.479559748427673,
      "grad_norm": 5.21875,
      "learning_rate": 2.518294692602821e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 6825856,
      "step": 10455
    },
    {
      "epoch": 5.482180293501048,
      "grad_norm": 7.84375,
      "learning_rate": 2.5160078895142186e-05,
      "loss": 0.5425,
      "num_input_tokens_seen": 6828896,
      "step": 10460
    },
    {
      "epoch": 5.484800838574423,
      "grad_norm": 8.3125,
      "learning_rate": 2.5137210730309447e-05,
      "loss": 0.5319,
      "num_input_tokens_seen": 6832032,
      "step": 10465
    },
    {
      "epoch": 5.487421383647799,
      "grad_norm": 11.5625,
      "learning_rate": 2.5114342450665034e-05,
      "loss": 0.5454,
      "num_input_tokens_seen": 6834624,
      "step": 10470
    },
    {
      "epoch": 5.490041928721174,
      "grad_norm": 6.21875,
      "learning_rate": 2.509147407534409e-05,
      "loss": 0.6736,
      "num_input_tokens_seen": 6837504,
      "step": 10475
    },
    {
      "epoch": 5.4926624737945495,
      "grad_norm": 5.65625,
      "learning_rate": 2.5068605623481816e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 6840224,
      "step": 10480
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 3.3125,
      "learning_rate": 2.5045737114213487e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 6843520,
      "step": 10485
    },
    {
      "epoch": 5.4979035639413,
      "grad_norm": 4.0625,
      "learning_rate": 2.502286856667443e-05,
      "loss": 0.4896,
      "num_input_tokens_seen": 6846464,
      "step": 10490
    },
    {
      "epoch": 5.5,
      "eval_loss": 0.494876503944397,
      "eval_runtime": 24.141,
      "eval_samples_per_second": 35.127,
      "eval_steps_per_second": 8.782,
      "num_input_tokens_seen": 6849184,
      "step": 10494
    },
    {
      "epoch": 5.500524109014675,
      "grad_norm": 7.9375,
      "learning_rate": 2.5e-05,
      "loss": 0.4978,
      "num_input_tokens_seen": 6849696,
      "step": 10495
    },
    {
      "epoch": 5.50314465408805,
      "grad_norm": 5.40625,
      "learning_rate": 2.497713143332557e-05,
      "loss": 0.6359,
      "num_input_tokens_seen": 6853216,
      "step": 10500
    },
    {
      "epoch": 5.505765199161425,
      "grad_norm": 5.75,
      "learning_rate": 2.495426288578652e-05,
      "loss": 0.466,
      "num_input_tokens_seen": 6855872,
      "step": 10505
    },
    {
      "epoch": 5.5083857442348005,
      "grad_norm": 6.75,
      "learning_rate": 2.493139437651819e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 6858688,
      "step": 10510
    },
    {
      "epoch": 5.511006289308176,
      "grad_norm": 7.46875,
      "learning_rate": 2.490852592465591e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 6861632,
      "step": 10515
    },
    {
      "epoch": 5.513626834381551,
      "grad_norm": 5.03125,
      "learning_rate": 2.488565754933497e-05,
      "loss": 0.6814,
      "num_input_tokens_seen": 6865696,
      "step": 10520
    },
    {
      "epoch": 5.516247379454927,
      "grad_norm": 5.28125,
      "learning_rate": 2.486278926969056e-05,
      "loss": 0.5101,
      "num_input_tokens_seen": 6869056,
      "step": 10525
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 13.8125,
      "learning_rate": 2.483992110485782e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 6871488,
      "step": 10530
    },
    {
      "epoch": 5.521488469601677,
      "grad_norm": 3.25,
      "learning_rate": 2.4817053073971792e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 6875008,
      "step": 10535
    },
    {
      "epoch": 5.524109014675052,
      "grad_norm": 10.375,
      "learning_rate": 2.4794185196167392e-05,
      "loss": 0.708,
      "num_input_tokens_seen": 6878336,
      "step": 10540
    },
    {
      "epoch": 5.526729559748428,
      "grad_norm": 2.96875,
      "learning_rate": 2.477131749057942e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 6882080,
      "step": 10545
    },
    {
      "epoch": 5.529350104821803,
      "grad_norm": 4.5,
      "learning_rate": 2.4748449976342524e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 6885504,
      "step": 10550
    },
    {
      "epoch": 5.531970649895178,
      "grad_norm": 5.15625,
      "learning_rate": 2.47255826725912e-05,
      "loss": 0.5229,
      "num_input_tokens_seen": 6888704,
      "step": 10555
    },
    {
      "epoch": 5.534591194968553,
      "grad_norm": 6.4375,
      "learning_rate": 2.4702715598459766e-05,
      "loss": 0.5822,
      "num_input_tokens_seen": 6891936,
      "step": 10560
    },
    {
      "epoch": 5.537211740041929,
      "grad_norm": 6.78125,
      "learning_rate": 2.467984877308233e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 6894848,
      "step": 10565
    },
    {
      "epoch": 5.539832285115304,
      "grad_norm": 7.125,
      "learning_rate": 2.4656982215592818e-05,
      "loss": 0.4702,
      "num_input_tokens_seen": 6897728,
      "step": 10570
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 10.8125,
      "learning_rate": 2.463411594512493e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 6900672,
      "step": 10575
    },
    {
      "epoch": 5.545073375262055,
      "grad_norm": 5.28125,
      "learning_rate": 2.4611249980812094e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 6905920,
      "step": 10580
    },
    {
      "epoch": 5.54769392033543,
      "grad_norm": 4.03125,
      "learning_rate": 2.4588384341787518e-05,
      "loss": 0.4863,
      "num_input_tokens_seen": 6909408,
      "step": 10585
    },
    {
      "epoch": 5.550314465408805,
      "grad_norm": 9.5,
      "learning_rate": 2.456551904718413e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 6912128,
      "step": 10590
    },
    {
      "epoch": 5.55293501048218,
      "grad_norm": 8.25,
      "learning_rate": 2.454265411613455e-05,
      "loss": 0.5271,
      "num_input_tokens_seen": 6915136,
      "step": 10595
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 4.8125,
      "learning_rate": 2.4519789567771116e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 6917856,
      "step": 10600
    },
    {
      "epoch": 5.5581761006289305,
      "grad_norm": 7.15625,
      "learning_rate": 2.4496925421225847e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 6920864,
      "step": 10605
    },
    {
      "epoch": 5.560796645702306,
      "grad_norm": 4.09375,
      "learning_rate": 2.4474061695630395e-05,
      "loss": 0.6767,
      "num_input_tokens_seen": 6924384,
      "step": 10610
    },
    {
      "epoch": 5.563417190775681,
      "grad_norm": 6.34375,
      "learning_rate": 2.4451198410116086e-05,
      "loss": 0.538,
      "num_input_tokens_seen": 6926848,
      "step": 10615
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 4.25,
      "learning_rate": 2.4428335583813898e-05,
      "loss": 0.3941,
      "num_input_tokens_seen": 6929536,
      "step": 10620
    },
    {
      "epoch": 5.568658280922432,
      "grad_norm": 11.125,
      "learning_rate": 2.4405473235854367e-05,
      "loss": 0.5488,
      "num_input_tokens_seen": 6932064,
      "step": 10625
    },
    {
      "epoch": 5.571278825995807,
      "grad_norm": 4.34375,
      "learning_rate": 2.4382611385367678e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 6935296,
      "step": 10630
    },
    {
      "epoch": 5.573899371069182,
      "grad_norm": 6.59375,
      "learning_rate": 2.4359750051483584e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 6937824,
      "step": 10635
    },
    {
      "epoch": 5.576519916142558,
      "grad_norm": 5.15625,
      "learning_rate": 2.4336889253331397e-05,
      "loss": 0.4726,
      "num_input_tokens_seen": 6941600,
      "step": 10640
    },
    {
      "epoch": 5.579140461215933,
      "grad_norm": 4.8125,
      "learning_rate": 2.4314029010040004e-05,
      "loss": 0.4755,
      "num_input_tokens_seen": 6944608,
      "step": 10645
    },
    {
      "epoch": 5.581761006289308,
      "grad_norm": 7.25,
      "learning_rate": 2.429116934073779e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 6948000,
      "step": 10650
    },
    {
      "epoch": 5.584381551362683,
      "grad_norm": 5.28125,
      "learning_rate": 2.426831026455269e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 6951424,
      "step": 10655
    },
    {
      "epoch": 5.587002096436059,
      "grad_norm": 4.53125,
      "learning_rate": 2.424545180061215e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 6956000,
      "step": 10660
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 7.1875,
      "learning_rate": 2.422259396804307e-05,
      "loss": 0.4936,
      "num_input_tokens_seen": 6958944,
      "step": 10665
    },
    {
      "epoch": 5.5922431865828095,
      "grad_norm": 3.296875,
      "learning_rate": 2.4199736785971846e-05,
      "loss": 0.4405,
      "num_input_tokens_seen": 6962304,
      "step": 10670
    },
    {
      "epoch": 5.594863731656185,
      "grad_norm": 6.59375,
      "learning_rate": 2.417688027352433e-05,
      "loss": 0.5969,
      "num_input_tokens_seen": 6965632,
      "step": 10675
    },
    {
      "epoch": 5.59748427672956,
      "grad_norm": 3.375,
      "learning_rate": 2.41540244498258e-05,
      "loss": 0.5814,
      "num_input_tokens_seen": 6969312,
      "step": 10680
    },
    {
      "epoch": 5.600104821802935,
      "grad_norm": 6.1875,
      "learning_rate": 2.4131169334000963e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 6972576,
      "step": 10685
    },
    {
      "epoch": 5.60272536687631,
      "grad_norm": 7.5,
      "learning_rate": 2.4108314945173955e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 6975872,
      "step": 10690
    },
    {
      "epoch": 5.605345911949685,
      "grad_norm": 3.875,
      "learning_rate": 2.4085461302468254e-05,
      "loss": 0.3815,
      "num_input_tokens_seen": 6978784,
      "step": 10695
    },
    {
      "epoch": 5.6079664570230605,
      "grad_norm": 7.28125,
      "learning_rate": 2.4062608425006765e-05,
      "loss": 0.6865,
      "num_input_tokens_seen": 6981408,
      "step": 10700
    },
    {
      "epoch": 5.610587002096436,
      "grad_norm": 8.9375,
      "learning_rate": 2.4039756331911737e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 6984800,
      "step": 10705
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 3.78125,
      "learning_rate": 2.401690504230474e-05,
      "loss": 0.487,
      "num_input_tokens_seen": 6988192,
      "step": 10710
    },
    {
      "epoch": 5.615828092243187,
      "grad_norm": 3.859375,
      "learning_rate": 2.3994054575306698e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 6992416,
      "step": 10715
    },
    {
      "epoch": 5.618448637316562,
      "grad_norm": 6.84375,
      "learning_rate": 2.397120495003784e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 6995520,
      "step": 10720
    },
    {
      "epoch": 5.621069182389937,
      "grad_norm": 7.0,
      "learning_rate": 2.394835618561768e-05,
      "loss": 0.459,
      "num_input_tokens_seen": 6998464,
      "step": 10725
    },
    {
      "epoch": 5.623689727463312,
      "grad_norm": 4.90625,
      "learning_rate": 2.3925508301165043e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 7001600,
      "step": 10730
    },
    {
      "epoch": 5.626310272536688,
      "grad_norm": 5.46875,
      "learning_rate": 2.390266131579796e-05,
      "loss": 0.5139,
      "num_input_tokens_seen": 7004704,
      "step": 10735
    },
    {
      "epoch": 5.628930817610063,
      "grad_norm": 6.78125,
      "learning_rate": 2.3879815248633768e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 7007424,
      "step": 10740
    },
    {
      "epoch": 5.631551362683438,
      "grad_norm": 4.25,
      "learning_rate": 2.385697011878902e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 7010688,
      "step": 10745
    },
    {
      "epoch": 5.634171907756813,
      "grad_norm": 4.1875,
      "learning_rate": 2.383412594537945e-05,
      "loss": 0.3695,
      "num_input_tokens_seen": 7014048,
      "step": 10750
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 4.9375,
      "learning_rate": 2.3811282747520038e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 7017856,
      "step": 10755
    },
    {
      "epoch": 5.639412997903564,
      "grad_norm": 5.25,
      "learning_rate": 2.378844054432493e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 7021248,
      "step": 10760
    },
    {
      "epoch": 5.6420335429769395,
      "grad_norm": 7.34375,
      "learning_rate": 2.3765599354907427e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 7024320,
      "step": 10765
    },
    {
      "epoch": 5.644654088050315,
      "grad_norm": 11.5,
      "learning_rate": 2.374275919838e-05,
      "loss": 0.5278,
      "num_input_tokens_seen": 7027296,
      "step": 10770
    },
    {
      "epoch": 5.64727463312369,
      "grad_norm": 2.734375,
      "learning_rate": 2.371992009385425e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 7030784,
      "step": 10775
    },
    {
      "epoch": 5.649895178197065,
      "grad_norm": 4.46875,
      "learning_rate": 2.369708206044089e-05,
      "loss": 0.7007,
      "num_input_tokens_seen": 7034112,
      "step": 10780
    },
    {
      "epoch": 5.65251572327044,
      "grad_norm": 5.40625,
      "learning_rate": 2.3674245117249747e-05,
      "loss": 0.4102,
      "num_input_tokens_seen": 7039072,
      "step": 10785
    },
    {
      "epoch": 5.655136268343815,
      "grad_norm": 4.9375,
      "learning_rate": 2.3651409283389743e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 7042304,
      "step": 10790
    },
    {
      "epoch": 5.6577568134171905,
      "grad_norm": 3.8125,
      "learning_rate": 2.3628574577968834e-05,
      "loss": 0.5359,
      "num_input_tokens_seen": 7046400,
      "step": 10795
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 3.6875,
      "learning_rate": 2.360574102009408e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 7050656,
      "step": 10800
    },
    {
      "epoch": 5.662997903563941,
      "grad_norm": 6.0625,
      "learning_rate": 2.3582908628871554e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 7053824,
      "step": 10805
    },
    {
      "epoch": 5.665618448637317,
      "grad_norm": 6.15625,
      "learning_rate": 2.3560077423406355e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 7057760,
      "step": 10810
    },
    {
      "epoch": 5.668238993710692,
      "grad_norm": 3.9375,
      "learning_rate": 2.3537247422802595e-05,
      "loss": 0.419,
      "num_input_tokens_seen": 7061888,
      "step": 10815
    },
    {
      "epoch": 5.670859538784067,
      "grad_norm": 4.09375,
      "learning_rate": 2.351441864616338e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 7064960,
      "step": 10820
    },
    {
      "epoch": 5.673480083857442,
      "grad_norm": 5.03125,
      "learning_rate": 2.3491591112590776e-05,
      "loss": 0.458,
      "num_input_tokens_seen": 7067968,
      "step": 10825
    },
    {
      "epoch": 5.676100628930818,
      "grad_norm": 3.328125,
      "learning_rate": 2.346876484118584e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 7071424,
      "step": 10830
    },
    {
      "epoch": 5.678721174004193,
      "grad_norm": 9.0,
      "learning_rate": 2.3445939851048533e-05,
      "loss": 0.5554,
      "num_input_tokens_seen": 7074656,
      "step": 10835
    },
    {
      "epoch": 5.681341719077568,
      "grad_norm": 6.53125,
      "learning_rate": 2.342311616127777e-05,
      "loss": 0.5226,
      "num_input_tokens_seen": 7077952,
      "step": 10840
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 6.75,
      "learning_rate": 2.3400293790971378e-05,
      "loss": 0.3602,
      "num_input_tokens_seen": 7080512,
      "step": 10845
    },
    {
      "epoch": 5.686582809224319,
      "grad_norm": 9.5625,
      "learning_rate": 2.3377472759226064e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 7083904,
      "step": 10850
    },
    {
      "epoch": 5.689203354297694,
      "grad_norm": 5.03125,
      "learning_rate": 2.3354653085137433e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 7086144,
      "step": 10855
    },
    {
      "epoch": 5.6918238993710695,
      "grad_norm": 5.5625,
      "learning_rate": 2.333183478779995e-05,
      "loss": 0.4803,
      "num_input_tokens_seen": 7088608,
      "step": 10860
    },
    {
      "epoch": 5.694444444444445,
      "grad_norm": 3.71875,
      "learning_rate": 2.330901788630691e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 7091296,
      "step": 10865
    },
    {
      "epoch": 5.69706498951782,
      "grad_norm": 6.5625,
      "learning_rate": 2.3286202399750463e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 7095008,
      "step": 10870
    },
    {
      "epoch": 5.699685534591195,
      "grad_norm": 6.875,
      "learning_rate": 2.3263388347221575e-05,
      "loss": 0.458,
      "num_input_tokens_seen": 7098016,
      "step": 10875
    },
    {
      "epoch": 5.70230607966457,
      "grad_norm": 2.59375,
      "learning_rate": 2.3240575747809984e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 7101056,
      "step": 10880
    },
    {
      "epoch": 5.704926624737945,
      "grad_norm": 7.6875,
      "learning_rate": 2.3217764620604233e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 7103488,
      "step": 10885
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 6.125,
      "learning_rate": 2.3194954984691656e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 7107072,
      "step": 10890
    },
    {
      "epoch": 5.710167714884696,
      "grad_norm": 8.75,
      "learning_rate": 2.3172146859158282e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 7110336,
      "step": 10895
    },
    {
      "epoch": 5.712788259958071,
      "grad_norm": 6.1875,
      "learning_rate": 2.314934026308893e-05,
      "loss": 0.5857,
      "num_input_tokens_seen": 7114240,
      "step": 10900
    },
    {
      "epoch": 5.715408805031447,
      "grad_norm": 4.46875,
      "learning_rate": 2.3126535215567112e-05,
      "loss": 0.5657,
      "num_input_tokens_seen": 7121216,
      "step": 10905
    },
    {
      "epoch": 5.718029350104822,
      "grad_norm": 11.4375,
      "learning_rate": 2.3103731735675045e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 7124800,
      "step": 10910
    },
    {
      "epoch": 5.720649895178197,
      "grad_norm": 7.40625,
      "learning_rate": 2.308092984249365e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 7128224,
      "step": 10915
    },
    {
      "epoch": 5.723270440251572,
      "grad_norm": 12.625,
      "learning_rate": 2.3058129555102498e-05,
      "loss": 0.4022,
      "num_input_tokens_seen": 7131456,
      "step": 10920
    },
    {
      "epoch": 5.725890985324948,
      "grad_norm": 6.96875,
      "learning_rate": 2.3035330892579825e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 7134944,
      "step": 10925
    },
    {
      "epoch": 5.728511530398323,
      "grad_norm": 2.703125,
      "learning_rate": 2.3012533874002534e-05,
      "loss": 0.6,
      "num_input_tokens_seen": 7139200,
      "step": 10930
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 8.375,
      "learning_rate": 2.2989738518446104e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 7141792,
      "step": 10935
    },
    {
      "epoch": 5.733752620545073,
      "grad_norm": 4.25,
      "learning_rate": 2.2966944844984658e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 7145152,
      "step": 10940
    },
    {
      "epoch": 5.736373165618449,
      "grad_norm": 4.65625,
      "learning_rate": 2.29441528726909e-05,
      "loss": 0.5972,
      "num_input_tokens_seen": 7148192,
      "step": 10945
    },
    {
      "epoch": 5.738993710691824,
      "grad_norm": 4.28125,
      "learning_rate": 2.292136262063611e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 7151520,
      "step": 10950
    },
    {
      "epoch": 5.7416142557651995,
      "grad_norm": 7.90625,
      "learning_rate": 2.289857410789013e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 7154496,
      "step": 10955
    },
    {
      "epoch": 5.744234800838575,
      "grad_norm": 5.03125,
      "learning_rate": 2.287578735352136e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 7157600,
      "step": 10960
    },
    {
      "epoch": 5.74685534591195,
      "grad_norm": 6.65625,
      "learning_rate": 2.285300237659668e-05,
      "loss": 0.5708,
      "num_input_tokens_seen": 7160224,
      "step": 10965
    },
    {
      "epoch": 5.749475890985325,
      "grad_norm": 5.78125,
      "learning_rate": 2.283021919618155e-05,
      "loss": 0.3946,
      "num_input_tokens_seen": 7162720,
      "step": 10970
    },
    {
      "epoch": 5.7520964360587,
      "grad_norm": 4.78125,
      "learning_rate": 2.28074378313399e-05,
      "loss": 0.594,
      "num_input_tokens_seen": 7166240,
      "step": 10975
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 4.71875,
      "learning_rate": 2.2784658301134105e-05,
      "loss": 0.5653,
      "num_input_tokens_seen": 7172480,
      "step": 10980
    },
    {
      "epoch": 5.7573375262054505,
      "grad_norm": 4.71875,
      "learning_rate": 2.2761880624625048e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 7176416,
      "step": 10985
    },
    {
      "epoch": 5.759958071278826,
      "grad_norm": 3.140625,
      "learning_rate": 2.2739104820872062e-05,
      "loss": 0.6107,
      "num_input_tokens_seen": 7180832,
      "step": 10990
    },
    {
      "epoch": 5.762578616352201,
      "grad_norm": 4.875,
      "learning_rate": 2.271633090893288e-05,
      "loss": 0.5238,
      "num_input_tokens_seen": 7184416,
      "step": 10995
    },
    {
      "epoch": 5.765199161425577,
      "grad_norm": 7.34375,
      "learning_rate": 2.269355890786368e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 7186816,
      "step": 11000
    },
    {
      "epoch": 5.767819706498952,
      "grad_norm": 4.53125,
      "learning_rate": 2.2670788836719037e-05,
      "loss": 0.5879,
      "num_input_tokens_seen": 7189824,
      "step": 11005
    },
    {
      "epoch": 5.770440251572327,
      "grad_norm": 8.6875,
      "learning_rate": 2.2648020714551897e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 7192640,
      "step": 11010
    },
    {
      "epoch": 5.773060796645702,
      "grad_norm": 4.0625,
      "learning_rate": 2.26252545604136e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 7195328,
      "step": 11015
    },
    {
      "epoch": 5.7756813417190775,
      "grad_norm": 11.5,
      "learning_rate": 2.2602490393353798e-05,
      "loss": 0.7071,
      "num_input_tokens_seen": 7198464,
      "step": 11020
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 4.1875,
      "learning_rate": 2.2579728232420525e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 7202080,
      "step": 11025
    },
    {
      "epoch": 5.780922431865828,
      "grad_norm": 6.125,
      "learning_rate": 2.255696809666012e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 7204832,
      "step": 11030
    },
    {
      "epoch": 5.783542976939203,
      "grad_norm": 8.3125,
      "learning_rate": 2.253421000511721e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 7208128,
      "step": 11035
    },
    {
      "epoch": 5.786163522012579,
      "grad_norm": 4.5625,
      "learning_rate": 2.2511453976834733e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 7211232,
      "step": 11040
    },
    {
      "epoch": 5.788784067085954,
      "grad_norm": 3.078125,
      "learning_rate": 2.2488700030853907e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 7214720,
      "step": 11045
    },
    {
      "epoch": 5.7914046121593294,
      "grad_norm": 7.34375,
      "learning_rate": 2.2465948186214175e-05,
      "loss": 0.5959,
      "num_input_tokens_seen": 7219200,
      "step": 11050
    },
    {
      "epoch": 5.794025157232705,
      "grad_norm": 4.625,
      "learning_rate": 2.244319846195325e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 7223808,
      "step": 11055
    },
    {
      "epoch": 5.79664570230608,
      "grad_norm": 10.625,
      "learning_rate": 2.2420450877107075e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 7227264,
      "step": 11060
    },
    {
      "epoch": 5.799266247379455,
      "grad_norm": 5.25,
      "learning_rate": 2.2397705450709763e-05,
      "loss": 0.5715,
      "num_input_tokens_seen": 7232672,
      "step": 11065
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 9.0,
      "learning_rate": 2.237496220179366e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 7235584,
      "step": 11070
    },
    {
      "epoch": 5.804507337526205,
      "grad_norm": 5.375,
      "learning_rate": 2.235222114938929e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 7238592,
      "step": 11075
    },
    {
      "epoch": 5.8071278825995805,
      "grad_norm": 7.0,
      "learning_rate": 2.232948231252531e-05,
      "loss": 0.5162,
      "num_input_tokens_seen": 7241952,
      "step": 11080
    },
    {
      "epoch": 5.809748427672956,
      "grad_norm": 6.71875,
      "learning_rate": 2.2306745710228545e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 7245056,
      "step": 11085
    },
    {
      "epoch": 5.812368972746331,
      "grad_norm": 3.375,
      "learning_rate": 2.2284011361523954e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 7248896,
      "step": 11090
    },
    {
      "epoch": 5.814989517819707,
      "grad_norm": 6.15625,
      "learning_rate": 2.2261279285434588e-05,
      "loss": 0.5331,
      "num_input_tokens_seen": 7251680,
      "step": 11095
    },
    {
      "epoch": 5.817610062893082,
      "grad_norm": 4.65625,
      "learning_rate": 2.2238549500981626e-05,
      "loss": 0.5917,
      "num_input_tokens_seen": 7254432,
      "step": 11100
    },
    {
      "epoch": 5.820230607966457,
      "grad_norm": 21.125,
      "learning_rate": 2.2215822027184294e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 7257280,
      "step": 11105
    },
    {
      "epoch": 5.822851153039832,
      "grad_norm": 3.59375,
      "learning_rate": 2.2193096883059913e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 7259840,
      "step": 11110
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 6.84375,
      "learning_rate": 2.2170374087623853e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 7262624,
      "step": 11115
    },
    {
      "epoch": 5.828092243186583,
      "grad_norm": 3.453125,
      "learning_rate": 2.2147653659889494e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 7265568,
      "step": 11120
    },
    {
      "epoch": 5.830712788259958,
      "grad_norm": 7.5625,
      "learning_rate": 2.2124935618868266e-05,
      "loss": 0.3902,
      "num_input_tokens_seen": 7268896,
      "step": 11125
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 9.4375,
      "learning_rate": 2.210221998356959e-05,
      "loss": 0.3798,
      "num_input_tokens_seen": 7271328,
      "step": 11130
    },
    {
      "epoch": 5.835953878406709,
      "grad_norm": 7.78125,
      "learning_rate": 2.2079506773000862e-05,
      "loss": 0.5759,
      "num_input_tokens_seen": 7274752,
      "step": 11135
    },
    {
      "epoch": 5.838574423480084,
      "grad_norm": 3.765625,
      "learning_rate": 2.205679600616746e-05,
      "loss": 0.5695,
      "num_input_tokens_seen": 7279168,
      "step": 11140
    },
    {
      "epoch": 5.841194968553459,
      "grad_norm": 8.0,
      "learning_rate": 2.2034087702072736e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 7282944,
      "step": 11145
    },
    {
      "epoch": 5.843815513626835,
      "grad_norm": 6.15625,
      "learning_rate": 2.2011381879717928e-05,
      "loss": 0.5353,
      "num_input_tokens_seen": 7285824,
      "step": 11150
    },
    {
      "epoch": 5.84643605870021,
      "grad_norm": 7.71875,
      "learning_rate": 2.1988678558102255e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 7288384,
      "step": 11155
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 8.5,
      "learning_rate": 2.1965977756222816e-05,
      "loss": 0.5638,
      "num_input_tokens_seen": 7291840,
      "step": 11160
    },
    {
      "epoch": 5.85167714884696,
      "grad_norm": 3.546875,
      "learning_rate": 2.1943279493074595e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 7296032,
      "step": 11165
    },
    {
      "epoch": 5.854297693920335,
      "grad_norm": 7.96875,
      "learning_rate": 2.192058378765047e-05,
      "loss": 0.6191,
      "num_input_tokens_seen": 7298560,
      "step": 11170
    },
    {
      "epoch": 5.8569182389937104,
      "grad_norm": 5.46875,
      "learning_rate": 2.1897890658941175e-05,
      "loss": 0.3978,
      "num_input_tokens_seen": 7301376,
      "step": 11175
    },
    {
      "epoch": 5.859538784067086,
      "grad_norm": 8.75,
      "learning_rate": 2.1875200125935273e-05,
      "loss": 0.5247,
      "num_input_tokens_seen": 7304896,
      "step": 11180
    },
    {
      "epoch": 5.862159329140461,
      "grad_norm": 5.625,
      "learning_rate": 2.185251220761917e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 7307968,
      "step": 11185
    },
    {
      "epoch": 5.864779874213837,
      "grad_norm": 9.3125,
      "learning_rate": 2.182982692297709e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 7311040,
      "step": 11190
    },
    {
      "epoch": 5.867400419287212,
      "grad_norm": 5.5,
      "learning_rate": 2.180714429099102e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 7314848,
      "step": 11195
    },
    {
      "epoch": 5.870020964360587,
      "grad_norm": 6.25,
      "learning_rate": 2.1784464330640774e-05,
      "loss": 0.5458,
      "num_input_tokens_seen": 7317440,
      "step": 11200
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 6.25,
      "learning_rate": 2.1761787060903888e-05,
      "loss": 0.4703,
      "num_input_tokens_seen": 7320288,
      "step": 11205
    },
    {
      "epoch": 5.8752620545073375,
      "grad_norm": 7.03125,
      "learning_rate": 2.1739112500755673e-05,
      "loss": 0.5414,
      "num_input_tokens_seen": 7323168,
      "step": 11210
    },
    {
      "epoch": 5.877882599580713,
      "grad_norm": 6.5625,
      "learning_rate": 2.1716440669169175e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 7326048,
      "step": 11215
    },
    {
      "epoch": 5.880503144654088,
      "grad_norm": 6.3125,
      "learning_rate": 2.169377158511513e-05,
      "loss": 0.529,
      "num_input_tokens_seen": 7328288,
      "step": 11220
    },
    {
      "epoch": 5.883123689727463,
      "grad_norm": 12.375,
      "learning_rate": 2.1671105267562e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 7331328,
      "step": 11225
    },
    {
      "epoch": 5.885744234800838,
      "grad_norm": 6.03125,
      "learning_rate": 2.1648441735475936e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 7334720,
      "step": 11230
    },
    {
      "epoch": 5.888364779874214,
      "grad_norm": 5.8125,
      "learning_rate": 2.1625781007820723e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 7337312,
      "step": 11235
    },
    {
      "epoch": 5.890985324947589,
      "grad_norm": 6.3125,
      "learning_rate": 2.160312310355783e-05,
      "loss": 0.5241,
      "num_input_tokens_seen": 7340384,
      "step": 11240
    },
    {
      "epoch": 5.893605870020965,
      "grad_norm": 3.46875,
      "learning_rate": 2.1580468041646378e-05,
      "loss": 0.5174,
      "num_input_tokens_seen": 7343936,
      "step": 11245
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 5.53125,
      "learning_rate": 2.155781584104306e-05,
      "loss": 0.4911,
      "num_input_tokens_seen": 7346848,
      "step": 11250
    },
    {
      "epoch": 5.898846960167715,
      "grad_norm": 6.875,
      "learning_rate": 2.153516652070221e-05,
      "loss": 0.5295,
      "num_input_tokens_seen": 7350112,
      "step": 11255
    },
    {
      "epoch": 5.90146750524109,
      "grad_norm": 3.609375,
      "learning_rate": 2.1512520099575756e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 7353952,
      "step": 11260
    },
    {
      "epoch": 5.904088050314465,
      "grad_norm": 8.625,
      "learning_rate": 2.1489876596613176e-05,
      "loss": 0.6205,
      "num_input_tokens_seen": 7357440,
      "step": 11265
    },
    {
      "epoch": 5.90670859538784,
      "grad_norm": 5.5,
      "learning_rate": 2.146723603076152e-05,
      "loss": 0.6477,
      "num_input_tokens_seen": 7360800,
      "step": 11270
    },
    {
      "epoch": 5.909329140461216,
      "grad_norm": 6.6875,
      "learning_rate": 2.14445984209654e-05,
      "loss": 0.5529,
      "num_input_tokens_seen": 7364032,
      "step": 11275
    },
    {
      "epoch": 5.911949685534591,
      "grad_norm": 8.1875,
      "learning_rate": 2.14219637861669e-05,
      "loss": 0.5655,
      "num_input_tokens_seen": 7366880,
      "step": 11280
    },
    {
      "epoch": 5.914570230607967,
      "grad_norm": 4.6875,
      "learning_rate": 2.1399332145305678e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 7369664,
      "step": 11285
    },
    {
      "epoch": 5.917190775681342,
      "grad_norm": 5.25,
      "learning_rate": 2.1376703517318837e-05,
      "loss": 0.5886,
      "num_input_tokens_seen": 7372192,
      "step": 11290
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 12.0625,
      "learning_rate": 2.1354077921140984e-05,
      "loss": 0.5125,
      "num_input_tokens_seen": 7374720,
      "step": 11295
    },
    {
      "epoch": 5.922431865828092,
      "grad_norm": 6.1875,
      "learning_rate": 2.1331455375704195e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 7378112,
      "step": 11300
    },
    {
      "epoch": 5.9250524109014675,
      "grad_norm": 7.40625,
      "learning_rate": 2.1308835899937972e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 7381728,
      "step": 11305
    },
    {
      "epoch": 5.927672955974843,
      "grad_norm": 10.0625,
      "learning_rate": 2.128621951276926e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 7385504,
      "step": 11310
    },
    {
      "epoch": 5.930293501048218,
      "grad_norm": 5.15625,
      "learning_rate": 2.126360623312243e-05,
      "loss": 0.4101,
      "num_input_tokens_seen": 7389600,
      "step": 11315
    },
    {
      "epoch": 5.932914046121593,
      "grad_norm": 5.53125,
      "learning_rate": 2.124099607991922e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 7394240,
      "step": 11320
    },
    {
      "epoch": 5.935534591194968,
      "grad_norm": 3.984375,
      "learning_rate": 2.121838907207879e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 7396704,
      "step": 11325
    },
    {
      "epoch": 5.938155136268344,
      "grad_norm": 3.84375,
      "learning_rate": 2.1195785228517658e-05,
      "loss": 0.4597,
      "num_input_tokens_seen": 7400352,
      "step": 11330
    },
    {
      "epoch": 5.940775681341719,
      "grad_norm": 5.21875,
      "learning_rate": 2.117318456814967e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 7402816,
      "step": 11335
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 6.46875,
      "learning_rate": 2.1150587109886026e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 7405568,
      "step": 11340
    },
    {
      "epoch": 5.94601677148847,
      "grad_norm": 5.59375,
      "learning_rate": 2.1127992872635263e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 7408608,
      "step": 11345
    },
    {
      "epoch": 5.948637316561845,
      "grad_norm": 9.9375,
      "learning_rate": 2.1105401875303193e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 7412448,
      "step": 11350
    },
    {
      "epoch": 5.95125786163522,
      "grad_norm": 7.96875,
      "learning_rate": 2.1082814136792937e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 7415104,
      "step": 11355
    },
    {
      "epoch": 5.953878406708595,
      "grad_norm": 6.46875,
      "learning_rate": 2.1060229676004887e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 7418944,
      "step": 11360
    },
    {
      "epoch": 5.95649895178197,
      "grad_norm": 3.78125,
      "learning_rate": 2.1037648511836675e-05,
      "loss": 0.4168,
      "num_input_tokens_seen": 7422816,
      "step": 11365
    },
    {
      "epoch": 5.959119496855346,
      "grad_norm": 8.4375,
      "learning_rate": 2.1015070663183195e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 7425824,
      "step": 11370
    },
    {
      "epoch": 5.961740041928721,
      "grad_norm": 5.4375,
      "learning_rate": 2.0992496148936573e-05,
      "loss": 0.557,
      "num_input_tokens_seen": 7429088,
      "step": 11375
    },
    {
      "epoch": 5.964360587002097,
      "grad_norm": 5.4375,
      "learning_rate": 2.0969924987986107e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 7432736,
      "step": 11380
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 5.59375,
      "learning_rate": 2.0947357199218325e-05,
      "loss": 0.5555,
      "num_input_tokens_seen": 7435840,
      "step": 11385
    },
    {
      "epoch": 5.969601677148847,
      "grad_norm": 7.78125,
      "learning_rate": 2.0924792801516922e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 7441056,
      "step": 11390
    },
    {
      "epoch": 5.972222222222222,
      "grad_norm": 3.734375,
      "learning_rate": 2.0902231813762753e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 7444640,
      "step": 11395
    },
    {
      "epoch": 5.9748427672955975,
      "grad_norm": 6.5625,
      "learning_rate": 2.0879674254833828e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 7448096,
      "step": 11400
    },
    {
      "epoch": 5.977463312368973,
      "grad_norm": 3.96875,
      "learning_rate": 2.085712014360527e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 7451520,
      "step": 11405
    },
    {
      "epoch": 5.980083857442348,
      "grad_norm": 11.125,
      "learning_rate": 2.0834569498949342e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 7454048,
      "step": 11410
    },
    {
      "epoch": 5.982704402515723,
      "grad_norm": 24.125,
      "learning_rate": 2.0812022339735395e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 7457216,
      "step": 11415
    },
    {
      "epoch": 5.985324947589098,
      "grad_norm": 3.84375,
      "learning_rate": 2.0789478684829846e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 7460992,
      "step": 11420
    },
    {
      "epoch": 5.987945492662474,
      "grad_norm": 7.53125,
      "learning_rate": 2.0766938553096204e-05,
      "loss": 0.4838,
      "num_input_tokens_seen": 7463488,
      "step": 11425
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 4.375,
      "learning_rate": 2.0744401963395027e-05,
      "loss": 0.4728,
      "num_input_tokens_seen": 7466304,
      "step": 11430
    },
    {
      "epoch": 5.993186582809225,
      "grad_norm": 4.3125,
      "learning_rate": 2.0721868934583897e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 7470176,
      "step": 11435
    },
    {
      "epoch": 5.9958071278826,
      "grad_norm": 7.0,
      "learning_rate": 2.0699339485517422e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 7473376,
      "step": 11440
    },
    {
      "epoch": 5.998427672955975,
      "grad_norm": 6.625,
      "learning_rate": 2.0676813635047225e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 7476864,
      "step": 11445
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.49464187026023865,
      "eval_runtime": 24.2044,
      "eval_samples_per_second": 35.035,
      "eval_steps_per_second": 8.759,
      "num_input_tokens_seen": 7478504,
      "step": 11448
    },
    {
      "epoch": 6.00104821802935,
      "grad_norm": 5.375,
      "learning_rate": 2.0654291402021896e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 7479688,
      "step": 11450
    },
    {
      "epoch": 6.003668763102725,
      "grad_norm": 3.65625,
      "learning_rate": 2.063177280528702e-05,
      "loss": 0.4328,
      "num_input_tokens_seen": 7483176,
      "step": 11455
    },
    {
      "epoch": 6.0062893081761,
      "grad_norm": 4.6875,
      "learning_rate": 2.0609257863685142e-05,
      "loss": 0.5007,
      "num_input_tokens_seen": 7486120,
      "step": 11460
    },
    {
      "epoch": 6.008909853249476,
      "grad_norm": 9.125,
      "learning_rate": 2.0586746596055706e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 7488872,
      "step": 11465
    },
    {
      "epoch": 6.011530398322851,
      "grad_norm": 13.75,
      "learning_rate": 2.0564239021235128e-05,
      "loss": 0.5152,
      "num_input_tokens_seen": 7491880,
      "step": 11470
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 7.15625,
      "learning_rate": 2.0541735158056733e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 7494952,
      "step": 11475
    },
    {
      "epoch": 6.016771488469602,
      "grad_norm": 4.03125,
      "learning_rate": 2.0519235025350704e-05,
      "loss": 0.5352,
      "num_input_tokens_seen": 7498600,
      "step": 11480
    },
    {
      "epoch": 6.019392033542977,
      "grad_norm": 4.28125,
      "learning_rate": 2.0496738641944133e-05,
      "loss": 0.5115,
      "num_input_tokens_seen": 7502696,
      "step": 11485
    },
    {
      "epoch": 6.022012578616352,
      "grad_norm": 5.125,
      "learning_rate": 2.0474246026660966e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 7505800,
      "step": 11490
    },
    {
      "epoch": 6.0246331236897275,
      "grad_norm": 10.4375,
      "learning_rate": 2.0451757198321992e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 7508712,
      "step": 11495
    },
    {
      "epoch": 6.027253668763103,
      "grad_norm": 4.375,
      "learning_rate": 2.042927217574485e-05,
      "loss": 0.4058,
      "num_input_tokens_seen": 7511400,
      "step": 11500
    },
    {
      "epoch": 6.029874213836478,
      "grad_norm": 3.8125,
      "learning_rate": 2.040679097774396e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 7515208,
      "step": 11505
    },
    {
      "epoch": 6.032494758909853,
      "grad_norm": 10.375,
      "learning_rate": 2.0384313623130565e-05,
      "loss": 0.3077,
      "num_input_tokens_seen": 7520712,
      "step": 11510
    },
    {
      "epoch": 6.035115303983228,
      "grad_norm": 3.421875,
      "learning_rate": 2.0361840130712706e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 7522984,
      "step": 11515
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 4.4375,
      "learning_rate": 2.033937051929516e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 7526024,
      "step": 11520
    },
    {
      "epoch": 6.040356394129979,
      "grad_norm": 12.9375,
      "learning_rate": 2.0316904807679464e-05,
      "loss": 0.534,
      "num_input_tokens_seen": 7529288,
      "step": 11525
    },
    {
      "epoch": 6.0429769392033545,
      "grad_norm": 5.3125,
      "learning_rate": 2.0294443014663923e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 7531752,
      "step": 11530
    },
    {
      "epoch": 6.04559748427673,
      "grad_norm": 8.6875,
      "learning_rate": 2.0271985159043518e-05,
      "loss": 0.4784,
      "num_input_tokens_seen": 7535112,
      "step": 11535
    },
    {
      "epoch": 6.048218029350105,
      "grad_norm": 8.5625,
      "learning_rate": 2.0249531259609965e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 7538344,
      "step": 11540
    },
    {
      "epoch": 6.05083857442348,
      "grad_norm": 14.125,
      "learning_rate": 2.0227081335151675e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 7541064,
      "step": 11545
    },
    {
      "epoch": 6.053459119496855,
      "grad_norm": 12.4375,
      "learning_rate": 2.0204635404453688e-05,
      "loss": 0.4579,
      "num_input_tokens_seen": 7544808,
      "step": 11550
    },
    {
      "epoch": 6.05607966457023,
      "grad_norm": 6.625,
      "learning_rate": 2.0182193486297755e-05,
      "loss": 0.3646,
      "num_input_tokens_seen": 7548872,
      "step": 11555
    },
    {
      "epoch": 6.058700209643606,
      "grad_norm": 4.84375,
      "learning_rate": 2.0159755599462256e-05,
      "loss": 0.545,
      "num_input_tokens_seen": 7551912,
      "step": 11560
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 5.875,
      "learning_rate": 2.0137321762722166e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 7554856,
      "step": 11565
    },
    {
      "epoch": 6.063941299790357,
      "grad_norm": 18.625,
      "learning_rate": 2.0114891994849112e-05,
      "loss": 0.5684,
      "num_input_tokens_seen": 7558344,
      "step": 11570
    },
    {
      "epoch": 6.066561844863732,
      "grad_norm": 5.0,
      "learning_rate": 2.0092466314611287e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 7561704,
      "step": 11575
    },
    {
      "epoch": 6.069182389937107,
      "grad_norm": 7.75,
      "learning_rate": 2.0070044740773487e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 7563816,
      "step": 11580
    },
    {
      "epoch": 6.071802935010482,
      "grad_norm": 5.53125,
      "learning_rate": 2.0047627292097067e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 7566408,
      "step": 11585
    },
    {
      "epoch": 6.0744234800838575,
      "grad_norm": 6.3125,
      "learning_rate": 2.002521398733989e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 7569960,
      "step": 11590
    },
    {
      "epoch": 6.077044025157233,
      "grad_norm": 10.5,
      "learning_rate": 2.0002804845256423e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 7572872,
      "step": 11595
    },
    {
      "epoch": 6.079664570230608,
      "grad_norm": 7.40625,
      "learning_rate": 1.9980399884597605e-05,
      "loss": 0.6097,
      "num_input_tokens_seen": 7576104,
      "step": 11600
    },
    {
      "epoch": 6.082285115303983,
      "grad_norm": 4.71875,
      "learning_rate": 1.995799912411087e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 7578856,
      "step": 11605
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 5.53125,
      "learning_rate": 1.993560258254016e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 7582408,
      "step": 11610
    },
    {
      "epoch": 6.087526205450734,
      "grad_norm": 4.46875,
      "learning_rate": 1.9913210278625876e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 7585224,
      "step": 11615
    },
    {
      "epoch": 6.090146750524109,
      "grad_norm": 10.1875,
      "learning_rate": 1.9890822231104872e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 7587880,
      "step": 11620
    },
    {
      "epoch": 6.0927672955974845,
      "grad_norm": 5.09375,
      "learning_rate": 1.9868438458710447e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 7591784,
      "step": 11625
    },
    {
      "epoch": 6.09538784067086,
      "grad_norm": 7.6875,
      "learning_rate": 1.984605898017233e-05,
      "loss": 0.4998,
      "num_input_tokens_seen": 7594920,
      "step": 11630
    },
    {
      "epoch": 6.098008385744235,
      "grad_norm": 9.25,
      "learning_rate": 1.9823683814216622e-05,
      "loss": 0.6123,
      "num_input_tokens_seen": 7598504,
      "step": 11635
    },
    {
      "epoch": 6.10062893081761,
      "grad_norm": 4.4375,
      "learning_rate": 1.980131297956586e-05,
      "loss": 0.5611,
      "num_input_tokens_seen": 7601640,
      "step": 11640
    },
    {
      "epoch": 6.103249475890985,
      "grad_norm": 7.09375,
      "learning_rate": 1.977894649493894e-05,
      "loss": 0.5093,
      "num_input_tokens_seen": 7604520,
      "step": 11645
    },
    {
      "epoch": 6.10587002096436,
      "grad_norm": 4.96875,
      "learning_rate": 1.9756584379051092e-05,
      "loss": 0.3638,
      "num_input_tokens_seen": 7608136,
      "step": 11650
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 9.125,
      "learning_rate": 1.9734226650613928e-05,
      "loss": 0.5469,
      "num_input_tokens_seen": 7611592,
      "step": 11655
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 10.375,
      "learning_rate": 1.9711873328335374e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 7614184,
      "step": 11660
    },
    {
      "epoch": 6.113731656184487,
      "grad_norm": 6.6875,
      "learning_rate": 1.9689524430919664e-05,
      "loss": 0.4102,
      "num_input_tokens_seen": 7617064,
      "step": 11665
    },
    {
      "epoch": 6.116352201257862,
      "grad_norm": 4.90625,
      "learning_rate": 1.9667179977067344e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 7620680,
      "step": 11670
    },
    {
      "epoch": 6.118972746331237,
      "grad_norm": 7.8125,
      "learning_rate": 1.9644839985475216e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 7623944,
      "step": 11675
    },
    {
      "epoch": 6.121593291404612,
      "grad_norm": 10.375,
      "learning_rate": 1.962250447483638e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 7627144,
      "step": 11680
    },
    {
      "epoch": 6.1242138364779874,
      "grad_norm": 6.4375,
      "learning_rate": 1.960017346384017e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 7630216,
      "step": 11685
    },
    {
      "epoch": 6.126834381551363,
      "grad_norm": 8.1875,
      "learning_rate": 1.9577846971172144e-05,
      "loss": 0.4932,
      "num_input_tokens_seen": 7633000,
      "step": 11690
    },
    {
      "epoch": 6.129454926624738,
      "grad_norm": 6.65625,
      "learning_rate": 1.9555525015514097e-05,
      "loss": 0.3974,
      "num_input_tokens_seen": 7636104,
      "step": 11695
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 10.6875,
      "learning_rate": 1.9533207615544034e-05,
      "loss": 0.6358,
      "num_input_tokens_seen": 7639208,
      "step": 11700
    },
    {
      "epoch": 6.134696016771488,
      "grad_norm": 3.703125,
      "learning_rate": 1.9510894789936113e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 7642184,
      "step": 11705
    },
    {
      "epoch": 6.137316561844864,
      "grad_norm": 12.1875,
      "learning_rate": 1.9488586557360703e-05,
      "loss": 0.5101,
      "num_input_tokens_seen": 7645480,
      "step": 11710
    },
    {
      "epoch": 6.139937106918239,
      "grad_norm": 5.21875,
      "learning_rate": 1.9466282936484313e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 7648936,
      "step": 11715
    },
    {
      "epoch": 6.1425576519916145,
      "grad_norm": 6.75,
      "learning_rate": 1.944398394596959e-05,
      "loss": 0.3906,
      "num_input_tokens_seen": 7651816,
      "step": 11720
    },
    {
      "epoch": 6.14517819706499,
      "grad_norm": 3.640625,
      "learning_rate": 1.942168960447531e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 7654568,
      "step": 11725
    },
    {
      "epoch": 6.147798742138365,
      "grad_norm": 5.96875,
      "learning_rate": 1.9399399930656377e-05,
      "loss": 0.3169,
      "num_input_tokens_seen": 7657544,
      "step": 11730
    },
    {
      "epoch": 6.15041928721174,
      "grad_norm": 8.0625,
      "learning_rate": 1.937711494316374e-05,
      "loss": 0.53,
      "num_input_tokens_seen": 7660936,
      "step": 11735
    },
    {
      "epoch": 6.153039832285115,
      "grad_norm": 6.0625,
      "learning_rate": 1.9354834660644478e-05,
      "loss": 0.5217,
      "num_input_tokens_seen": 7663336,
      "step": 11740
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 3.96875,
      "learning_rate": 1.9332559101741715e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 7666312,
      "step": 11745
    },
    {
      "epoch": 6.1582809224318655,
      "grad_norm": 5.59375,
      "learning_rate": 1.9310288285094615e-05,
      "loss": 0.5441,
      "num_input_tokens_seen": 7669224,
      "step": 11750
    },
    {
      "epoch": 6.160901467505241,
      "grad_norm": 5.15625,
      "learning_rate": 1.9288022229338384e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 7672264,
      "step": 11755
    },
    {
      "epoch": 6.163522012578617,
      "grad_norm": 7.34375,
      "learning_rate": 1.9265760953104235e-05,
      "loss": 0.5383,
      "num_input_tokens_seen": 7674824,
      "step": 11760
    },
    {
      "epoch": 6.166142557651992,
      "grad_norm": 7.8125,
      "learning_rate": 1.924350447501939e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 7677608,
      "step": 11765
    },
    {
      "epoch": 6.168763102725367,
      "grad_norm": 6.65625,
      "learning_rate": 1.922125281370707e-05,
      "loss": 0.4934,
      "num_input_tokens_seen": 7680680,
      "step": 11770
    },
    {
      "epoch": 6.171383647798742,
      "grad_norm": 4.59375,
      "learning_rate": 1.919900598778642e-05,
      "loss": 0.514,
      "num_input_tokens_seen": 7684072,
      "step": 11775
    },
    {
      "epoch": 6.174004192872117,
      "grad_norm": 3.640625,
      "learning_rate": 1.9176764015872578e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 7687624,
      "step": 11780
    },
    {
      "epoch": 6.176624737945493,
      "grad_norm": 4.40625,
      "learning_rate": 1.9154526916576618e-05,
      "loss": 0.5793,
      "num_input_tokens_seen": 7691656,
      "step": 11785
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 4.875,
      "learning_rate": 1.913229470850552e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 7694568,
      "step": 11790
    },
    {
      "epoch": 6.181865828092243,
      "grad_norm": 3.734375,
      "learning_rate": 1.9110067410262185e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 7697544,
      "step": 11795
    },
    {
      "epoch": 6.184486373165618,
      "grad_norm": 5.125,
      "learning_rate": 1.90878450404454e-05,
      "loss": 0.7579,
      "num_input_tokens_seen": 7700456,
      "step": 11800
    },
    {
      "epoch": 6.187106918238993,
      "grad_norm": 5.5,
      "learning_rate": 1.9065627617649828e-05,
      "loss": 0.4568,
      "num_input_tokens_seen": 7703112,
      "step": 11805
    },
    {
      "epoch": 6.189727463312369,
      "grad_norm": 5.65625,
      "learning_rate": 1.9043415160465993e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 7705992,
      "step": 11810
    },
    {
      "epoch": 6.1923480083857445,
      "grad_norm": 6.90625,
      "learning_rate": 1.9021207687480278e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 7709576,
      "step": 11815
    },
    {
      "epoch": 6.19496855345912,
      "grad_norm": 7.9375,
      "learning_rate": 1.8999005217274857e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 7712712,
      "step": 11820
    },
    {
      "epoch": 6.197589098532495,
      "grad_norm": 7.5625,
      "learning_rate": 1.897680776842775e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 7715432,
      "step": 11825
    },
    {
      "epoch": 6.20020964360587,
      "grad_norm": 4.03125,
      "learning_rate": 1.895461535951279e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 7718216,
      "step": 11830
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 6.28125,
      "learning_rate": 1.8932428009099545e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 7721096,
      "step": 11835
    },
    {
      "epoch": 6.20545073375262,
      "grad_norm": 7.5,
      "learning_rate": 1.891024573575339e-05,
      "loss": 0.5903,
      "num_input_tokens_seen": 7724264,
      "step": 11840
    },
    {
      "epoch": 6.2080712788259955,
      "grad_norm": 8.125,
      "learning_rate": 1.8888068558035435e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 7727176,
      "step": 11845
    },
    {
      "epoch": 6.210691823899371,
      "grad_norm": 6.46875,
      "learning_rate": 1.8865896494502525e-05,
      "loss": 0.4846,
      "num_input_tokens_seen": 7729544,
      "step": 11850
    },
    {
      "epoch": 6.213312368972747,
      "grad_norm": 3.8125,
      "learning_rate": 1.8843729563707247e-05,
      "loss": 0.5657,
      "num_input_tokens_seen": 7733288,
      "step": 11855
    },
    {
      "epoch": 6.215932914046122,
      "grad_norm": 9.125,
      "learning_rate": 1.8821567784197847e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 7737096,
      "step": 11860
    },
    {
      "epoch": 6.218553459119497,
      "grad_norm": 11.0,
      "learning_rate": 1.8799411174518306e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 7741512,
      "step": 11865
    },
    {
      "epoch": 6.221174004192872,
      "grad_norm": 5.28125,
      "learning_rate": 1.8777259753208275e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 7744648,
      "step": 11870
    },
    {
      "epoch": 6.223794549266247,
      "grad_norm": 14.125,
      "learning_rate": 1.8755113538803026e-05,
      "loss": 0.3801,
      "num_input_tokens_seen": 7748392,
      "step": 11875
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 15.3125,
      "learning_rate": 1.8732972549833516e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 7752360,
      "step": 11880
    },
    {
      "epoch": 6.229035639412998,
      "grad_norm": 8.4375,
      "learning_rate": 1.8710836804826314e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 7755016,
      "step": 11885
    },
    {
      "epoch": 6.231656184486373,
      "grad_norm": 6.34375,
      "learning_rate": 1.8688706322303595e-05,
      "loss": 0.5209,
      "num_input_tokens_seen": 7758120,
      "step": 11890
    },
    {
      "epoch": 6.234276729559748,
      "grad_norm": 13.0625,
      "learning_rate": 1.8666581120783134e-05,
      "loss": 0.6579,
      "num_input_tokens_seen": 7761032,
      "step": 11895
    },
    {
      "epoch": 6.236897274633123,
      "grad_norm": 10.1875,
      "learning_rate": 1.8644461218778304e-05,
      "loss": 0.4384,
      "num_input_tokens_seen": 7763240,
      "step": 11900
    },
    {
      "epoch": 6.239517819706499,
      "grad_norm": 4.09375,
      "learning_rate": 1.8622346634798e-05,
      "loss": 0.4701,
      "num_input_tokens_seen": 7766664,
      "step": 11905
    },
    {
      "epoch": 6.2421383647798745,
      "grad_norm": 7.0,
      "learning_rate": 1.8600237387346716e-05,
      "loss": 0.578,
      "num_input_tokens_seen": 7769448,
      "step": 11910
    },
    {
      "epoch": 6.24475890985325,
      "grad_norm": 7.21875,
      "learning_rate": 1.8578133494924473e-05,
      "loss": 0.6017,
      "num_input_tokens_seen": 7772296,
      "step": 11915
    },
    {
      "epoch": 6.247379454926625,
      "grad_norm": 5.15625,
      "learning_rate": 1.8556034976026764e-05,
      "loss": 0.4085,
      "num_input_tokens_seen": 7775912,
      "step": 11920
    },
    {
      "epoch": 6.25,
      "grad_norm": 3.546875,
      "learning_rate": 1.8533941849144642e-05,
      "loss": 0.5342,
      "num_input_tokens_seen": 7779784,
      "step": 11925
    },
    {
      "epoch": 6.252620545073375,
      "grad_norm": 5.71875,
      "learning_rate": 1.8511854132764627e-05,
      "loss": 0.3606,
      "num_input_tokens_seen": 7782504,
      "step": 11930
    },
    {
      "epoch": 6.25524109014675,
      "grad_norm": 7.9375,
      "learning_rate": 1.84897718453687e-05,
      "loss": 0.3539,
      "num_input_tokens_seen": 7785864,
      "step": 11935
    },
    {
      "epoch": 6.2578616352201255,
      "grad_norm": 10.375,
      "learning_rate": 1.846769500543434e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 7788808,
      "step": 11940
    },
    {
      "epoch": 6.260482180293501,
      "grad_norm": 7.25,
      "learning_rate": 1.844562363143439e-05,
      "loss": 0.4453,
      "num_input_tokens_seen": 7792232,
      "step": 11945
    },
    {
      "epoch": 6.263102725366876,
      "grad_norm": 3.703125,
      "learning_rate": 1.8423557741837198e-05,
      "loss": 0.443,
      "num_input_tokens_seen": 7794856,
      "step": 11950
    },
    {
      "epoch": 6.265723270440252,
      "grad_norm": 5.34375,
      "learning_rate": 1.84014973551065e-05,
      "loss": 0.6001,
      "num_input_tokens_seen": 7797800,
      "step": 11955
    },
    {
      "epoch": 6.268343815513627,
      "grad_norm": 3.328125,
      "learning_rate": 1.8379442489701396e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 7801256,
      "step": 11960
    },
    {
      "epoch": 6.270964360587002,
      "grad_norm": 5.59375,
      "learning_rate": 1.8357393164076403e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 7804296,
      "step": 11965
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 5.84375,
      "learning_rate": 1.8335349396681394e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 7807464,
      "step": 11970
    },
    {
      "epoch": 6.276205450733753,
      "grad_norm": 2.234375,
      "learning_rate": 1.8313311205961577e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 7812008,
      "step": 11975
    },
    {
      "epoch": 6.278825995807128,
      "grad_norm": 10.3125,
      "learning_rate": 1.82912786103575e-05,
      "loss": 0.4739,
      "num_input_tokens_seen": 7815336,
      "step": 11980
    },
    {
      "epoch": 6.281446540880503,
      "grad_norm": 2.734375,
      "learning_rate": 1.826925162830505e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 7818856,
      "step": 11985
    },
    {
      "epoch": 6.284067085953878,
      "grad_norm": 6.125,
      "learning_rate": 1.8247230278235384e-05,
      "loss": 0.5816,
      "num_input_tokens_seen": 7822024,
      "step": 11990
    },
    {
      "epoch": 6.286687631027253,
      "grad_norm": 7.15625,
      "learning_rate": 1.8225214578574967e-05,
      "loss": 0.6469,
      "num_input_tokens_seen": 7825640,
      "step": 11995
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 4.8125,
      "learning_rate": 1.820320454774554e-05,
      "loss": 0.4734,
      "num_input_tokens_seen": 7827944,
      "step": 12000
    },
    {
      "epoch": 6.2919287211740045,
      "grad_norm": 5.9375,
      "learning_rate": 1.8181200204164073e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 7830600,
      "step": 12005
    },
    {
      "epoch": 6.29454926624738,
      "grad_norm": 3.53125,
      "learning_rate": 1.8159201566242806e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 7834056,
      "step": 12010
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 10.875,
      "learning_rate": 1.81372086523892e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 7836904,
      "step": 12015
    },
    {
      "epoch": 6.29979035639413,
      "grad_norm": 10.5,
      "learning_rate": 1.8115221481005904e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 7839880,
      "step": 12020
    },
    {
      "epoch": 6.302410901467505,
      "grad_norm": 4.5,
      "learning_rate": 1.809324007049079e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 7843432,
      "step": 12025
    },
    {
      "epoch": 6.30503144654088,
      "grad_norm": 5.0,
      "learning_rate": 1.8071264439236903e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 7846216,
      "step": 12030
    },
    {
      "epoch": 6.3076519916142555,
      "grad_norm": 8.375,
      "learning_rate": 1.8049294605632434e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 7849544,
      "step": 12035
    },
    {
      "epoch": 6.310272536687631,
      "grad_norm": 5.6875,
      "learning_rate": 1.8027330588060757e-05,
      "loss": 0.5142,
      "num_input_tokens_seen": 7852552,
      "step": 12040
    },
    {
      "epoch": 6.312893081761006,
      "grad_norm": 8.3125,
      "learning_rate": 1.8005372404900335e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 7855336,
      "step": 12045
    },
    {
      "epoch": 6.315513626834382,
      "grad_norm": 6.4375,
      "learning_rate": 1.7983420074524777e-05,
      "loss": 0.5186,
      "num_input_tokens_seen": 7858184,
      "step": 12050
    },
    {
      "epoch": 6.318134171907757,
      "grad_norm": 19.75,
      "learning_rate": 1.7961473615302805e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 7861192,
      "step": 12055
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 7.46875,
      "learning_rate": 1.79395330455982e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 7864520,
      "step": 12060
    },
    {
      "epoch": 6.323375262054507,
      "grad_norm": 9.75,
      "learning_rate": 1.7917598383769836e-05,
      "loss": 0.5723,
      "num_input_tokens_seen": 7867240,
      "step": 12065
    },
    {
      "epoch": 6.325995807127883,
      "grad_norm": 9.0625,
      "learning_rate": 1.789566964817163e-05,
      "loss": 0.5357,
      "num_input_tokens_seen": 7870120,
      "step": 12070
    },
    {
      "epoch": 6.328616352201258,
      "grad_norm": 8.9375,
      "learning_rate": 1.7873746857152552e-05,
      "loss": 0.3901,
      "num_input_tokens_seen": 7873736,
      "step": 12075
    },
    {
      "epoch": 6.331236897274633,
      "grad_norm": 7.5625,
      "learning_rate": 1.7851830029056587e-05,
      "loss": 0.4422,
      "num_input_tokens_seen": 7877000,
      "step": 12080
    },
    {
      "epoch": 6.333857442348008,
      "grad_norm": 10.6875,
      "learning_rate": 1.7829919182222752e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 7879784,
      "step": 12085
    },
    {
      "epoch": 6.336477987421383,
      "grad_norm": 6.0,
      "learning_rate": 1.780801433498501e-05,
      "loss": 0.5542,
      "num_input_tokens_seen": 7882792,
      "step": 12090
    },
    {
      "epoch": 6.339098532494759,
      "grad_norm": 9.0,
      "learning_rate": 1.7786115505672364e-05,
      "loss": 0.5151,
      "num_input_tokens_seen": 7885416,
      "step": 12095
    },
    {
      "epoch": 6.3417190775681345,
      "grad_norm": 5.5625,
      "learning_rate": 1.7764222712608753e-05,
      "loss": 0.5096,
      "num_input_tokens_seen": 7888936,
      "step": 12100
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 3.203125,
      "learning_rate": 1.7742335974113046e-05,
      "loss": 0.5409,
      "num_input_tokens_seen": 7891976,
      "step": 12105
    },
    {
      "epoch": 6.346960167714885,
      "grad_norm": 10.375,
      "learning_rate": 1.7720455308499084e-05,
      "loss": 0.5748,
      "num_input_tokens_seen": 7895848,
      "step": 12110
    },
    {
      "epoch": 6.34958071278826,
      "grad_norm": 8.1875,
      "learning_rate": 1.769858073407561e-05,
      "loss": 0.726,
      "num_input_tokens_seen": 7898760,
      "step": 12115
    },
    {
      "epoch": 6.352201257861635,
      "grad_norm": 8.875,
      "learning_rate": 1.767671226914625e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 7902120,
      "step": 12120
    },
    {
      "epoch": 6.35482180293501,
      "grad_norm": 7.28125,
      "learning_rate": 1.7654849932009566e-05,
      "loss": 0.6715,
      "num_input_tokens_seen": 7905224,
      "step": 12125
    },
    {
      "epoch": 6.3574423480083855,
      "grad_norm": 6.125,
      "learning_rate": 1.763299374095893e-05,
      "loss": 0.5842,
      "num_input_tokens_seen": 7908424,
      "step": 12130
    },
    {
      "epoch": 6.360062893081761,
      "grad_norm": 5.8125,
      "learning_rate": 1.761114371428262e-05,
      "loss": 0.4707,
      "num_input_tokens_seen": 7911336,
      "step": 12135
    },
    {
      "epoch": 6.362683438155136,
      "grad_norm": 4.96875,
      "learning_rate": 1.7589299870263753e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 7913704,
      "step": 12140
    },
    {
      "epoch": 6.365303983228512,
      "grad_norm": 9.0,
      "learning_rate": 1.756746222718024e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 7916584,
      "step": 12145
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 3.78125,
      "learning_rate": 1.7545630803304826e-05,
      "loss": 0.5927,
      "num_input_tokens_seen": 7919880,
      "step": 12150
    },
    {
      "epoch": 6.370545073375262,
      "grad_norm": 6.9375,
      "learning_rate": 1.7523805616905063e-05,
      "loss": 0.5451,
      "num_input_tokens_seen": 7923752,
      "step": 12155
    },
    {
      "epoch": 6.373165618448637,
      "grad_norm": 4.0,
      "learning_rate": 1.7501986686243256e-05,
      "loss": 0.5704,
      "num_input_tokens_seen": 7927464,
      "step": 12160
    },
    {
      "epoch": 6.3757861635220126,
      "grad_norm": 6.125,
      "learning_rate": 1.748017402957649e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 7930792,
      "step": 12165
    },
    {
      "epoch": 6.378406708595388,
      "grad_norm": 3.25,
      "learning_rate": 1.7458367665156615e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 7935112,
      "step": 12170
    },
    {
      "epoch": 6.381027253668763,
      "grad_norm": 5.125,
      "learning_rate": 1.743656761123018e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 7938152,
      "step": 12175
    },
    {
      "epoch": 6.383647798742138,
      "grad_norm": 4.875,
      "learning_rate": 1.7414773886038487e-05,
      "loss": 0.5122,
      "num_input_tokens_seen": 7942536,
      "step": 12180
    },
    {
      "epoch": 6.386268343815513,
      "grad_norm": 7.875,
      "learning_rate": 1.7392986507817532e-05,
      "loss": 0.4233,
      "num_input_tokens_seen": 7945224,
      "step": 12185
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 5.875,
      "learning_rate": 1.7371205494797987e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 7948520,
      "step": 12190
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 8.125,
      "learning_rate": 1.7349430865205215e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 7951368,
      "step": 12195
    },
    {
      "epoch": 6.39412997903564,
      "grad_norm": 6.15625,
      "learning_rate": 1.7327662637259234e-05,
      "loss": 0.6276,
      "num_input_tokens_seen": 7954472,
      "step": 12200
    },
    {
      "epoch": 6.396750524109015,
      "grad_norm": 6.4375,
      "learning_rate": 1.7305900829174697e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 7957000,
      "step": 12205
    },
    {
      "epoch": 6.39937106918239,
      "grad_norm": 5.65625,
      "learning_rate": 1.7284145459160893e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 7959400,
      "step": 12210
    },
    {
      "epoch": 6.401991614255765,
      "grad_norm": 16.25,
      "learning_rate": 1.7262396545421728e-05,
      "loss": 0.5196,
      "num_input_tokens_seen": 7962312,
      "step": 12215
    },
    {
      "epoch": 6.40461215932914,
      "grad_norm": 9.0,
      "learning_rate": 1.7240654106155688e-05,
      "loss": 0.5779,
      "num_input_tokens_seen": 7965768,
      "step": 12220
    },
    {
      "epoch": 6.4072327044025155,
      "grad_norm": 4.46875,
      "learning_rate": 1.721891815955587e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 7972840,
      "step": 12225
    },
    {
      "epoch": 6.409853249475891,
      "grad_norm": 8.875,
      "learning_rate": 1.71971887238099e-05,
      "loss": 0.5445,
      "num_input_tokens_seen": 7975304,
      "step": 12230
    },
    {
      "epoch": 6.412473794549266,
      "grad_norm": 7.3125,
      "learning_rate": 1.7175465817099988e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 7977928,
      "step": 12235
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 5.4375,
      "learning_rate": 1.7153749457602874e-05,
      "loss": 0.4702,
      "num_input_tokens_seen": 7982120,
      "step": 12240
    },
    {
      "epoch": 6.417714884696017,
      "grad_norm": 8.4375,
      "learning_rate": 1.7132039663489806e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 7985032,
      "step": 12245
    },
    {
      "epoch": 6.420335429769392,
      "grad_norm": 6.21875,
      "learning_rate": 1.7110336452926555e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 7987272,
      "step": 12250
    },
    {
      "epoch": 6.422955974842767,
      "grad_norm": 3.84375,
      "learning_rate": 1.708863984407338e-05,
      "loss": 0.5181,
      "num_input_tokens_seen": 7990152,
      "step": 12255
    },
    {
      "epoch": 6.4255765199161425,
      "grad_norm": 5.65625,
      "learning_rate": 1.7066949855085e-05,
      "loss": 0.5837,
      "num_input_tokens_seen": 7992648,
      "step": 12260
    },
    {
      "epoch": 6.428197064989518,
      "grad_norm": 5.03125,
      "learning_rate": 1.704526650411062e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 7995880,
      "step": 12265
    },
    {
      "epoch": 6.430817610062893,
      "grad_norm": 10.25,
      "learning_rate": 1.7023589809293876e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 7998312,
      "step": 12270
    },
    {
      "epoch": 6.433438155136268,
      "grad_norm": 4.625,
      "learning_rate": 1.7001919788772824e-05,
      "loss": 0.4429,
      "num_input_tokens_seen": 8001768,
      "step": 12275
    },
    {
      "epoch": 6.436058700209643,
      "grad_norm": 9.1875,
      "learning_rate": 1.6980256460679953e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 8004392,
      "step": 12280
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 8.4375,
      "learning_rate": 1.6958599843142153e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 8006984,
      "step": 12285
    },
    {
      "epoch": 6.441299790356394,
      "grad_norm": 3.65625,
      "learning_rate": 1.6936949954280686e-05,
      "loss": 0.6338,
      "num_input_tokens_seen": 8010920,
      "step": 12290
    },
    {
      "epoch": 6.44392033542977,
      "grad_norm": 7.34375,
      "learning_rate": 1.691530681221119e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 8013800,
      "step": 12295
    },
    {
      "epoch": 6.446540880503145,
      "grad_norm": 5.875,
      "learning_rate": 1.6893670435043666e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 8016616,
      "step": 12300
    },
    {
      "epoch": 6.44916142557652,
      "grad_norm": 3.96875,
      "learning_rate": 1.6872040840882434e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 8020040,
      "step": 12305
    },
    {
      "epoch": 6.451781970649895,
      "grad_norm": 5.3125,
      "learning_rate": 1.6850418047826167e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 8024008,
      "step": 12310
    },
    {
      "epoch": 6.45440251572327,
      "grad_norm": 7.1875,
      "learning_rate": 1.6828802073967805e-05,
      "loss": 0.5017,
      "num_input_tokens_seen": 8027528,
      "step": 12315
    },
    {
      "epoch": 6.4570230607966455,
      "grad_norm": 10.6875,
      "learning_rate": 1.6807192937394624e-05,
      "loss": 0.5594,
      "num_input_tokens_seen": 8030312,
      "step": 12320
    },
    {
      "epoch": 6.459643605870021,
      "grad_norm": 5.90625,
      "learning_rate": 1.6785590656188167e-05,
      "loss": 0.3733,
      "num_input_tokens_seen": 8034280,
      "step": 12325
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 4.375,
      "learning_rate": 1.6763995248424223e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 8038408,
      "step": 12330
    },
    {
      "epoch": 6.464884696016772,
      "grad_norm": 6.375,
      "learning_rate": 1.6742406732172854e-05,
      "loss": 0.4726,
      "num_input_tokens_seen": 8041544,
      "step": 12335
    },
    {
      "epoch": 6.467505241090147,
      "grad_norm": 3.21875,
      "learning_rate": 1.6720825125498342e-05,
      "loss": 0.3462,
      "num_input_tokens_seen": 8045096,
      "step": 12340
    },
    {
      "epoch": 6.470125786163522,
      "grad_norm": 4.15625,
      "learning_rate": 1.6699250446459182e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 8049064,
      "step": 12345
    },
    {
      "epoch": 6.472746331236897,
      "grad_norm": 5.3125,
      "learning_rate": 1.6677682713108082e-05,
      "loss": 0.5459,
      "num_input_tokens_seen": 8051816,
      "step": 12350
    },
    {
      "epoch": 6.4753668763102725,
      "grad_norm": 8.5,
      "learning_rate": 1.6656121943491954e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 8055528,
      "step": 12355
    },
    {
      "epoch": 6.477987421383648,
      "grad_norm": 6.03125,
      "learning_rate": 1.6634568155651842e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 8058472,
      "step": 12360
    },
    {
      "epoch": 6.480607966457023,
      "grad_norm": 4.875,
      "learning_rate": 1.6613021367622978e-05,
      "loss": 0.707,
      "num_input_tokens_seen": 8061608,
      "step": 12365
    },
    {
      "epoch": 6.483228511530398,
      "grad_norm": 4.1875,
      "learning_rate": 1.6591481597434733e-05,
      "loss": 0.4944,
      "num_input_tokens_seen": 8064648,
      "step": 12370
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 7.28125,
      "learning_rate": 1.65699488631106e-05,
      "loss": 0.6703,
      "num_input_tokens_seen": 8067496,
      "step": 12375
    },
    {
      "epoch": 6.488469601677149,
      "grad_norm": 6.28125,
      "learning_rate": 1.6548423182668186e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 8069864,
      "step": 12380
    },
    {
      "epoch": 6.491090146750524,
      "grad_norm": 5.40625,
      "learning_rate": 1.6526904574119213e-05,
      "loss": 0.5804,
      "num_input_tokens_seen": 8072936,
      "step": 12385
    },
    {
      "epoch": 6.4937106918239,
      "grad_norm": 5.65625,
      "learning_rate": 1.6505393055469444e-05,
      "loss": 0.5644,
      "num_input_tokens_seen": 8075784,
      "step": 12390
    },
    {
      "epoch": 6.496331236897275,
      "grad_norm": 11.125,
      "learning_rate": 1.648388864471875e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 8078280,
      "step": 12395
    },
    {
      "epoch": 6.49895178197065,
      "grad_norm": 5.8125,
      "learning_rate": 1.646239135986105e-05,
      "loss": 0.3888,
      "num_input_tokens_seen": 8081992,
      "step": 12400
    },
    {
      "epoch": 6.5,
      "eval_loss": 0.4947109520435333,
      "eval_runtime": 24.1731,
      "eval_samples_per_second": 35.08,
      "eval_steps_per_second": 8.77,
      "num_input_tokens_seen": 8083560,
      "step": 12402
    },
    {
      "epoch": 6.501572327044025,
      "grad_norm": 7.40625,
      "learning_rate": 1.6440901218884264e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 8085672,
      "step": 12405
    },
    {
      "epoch": 6.5041928721174,
      "grad_norm": 3.640625,
      "learning_rate": 1.641941823977038e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 8089000,
      "step": 12410
    },
    {
      "epoch": 6.506813417190775,
      "grad_norm": 4.34375,
      "learning_rate": 1.6397942440495363e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 8092168,
      "step": 12415
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 6.09375,
      "learning_rate": 1.6376473839029188e-05,
      "loss": 0.5229,
      "num_input_tokens_seen": 8094728,
      "step": 12420
    },
    {
      "epoch": 6.512054507337526,
      "grad_norm": 13.375,
      "learning_rate": 1.63550124533358e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 8098088,
      "step": 12425
    },
    {
      "epoch": 6.514675052410902,
      "grad_norm": 6.5625,
      "learning_rate": 1.63335583013731e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 8100904,
      "step": 12430
    },
    {
      "epoch": 6.517295597484277,
      "grad_norm": 6.46875,
      "learning_rate": 1.6312111401092946e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 8104264,
      "step": 12435
    },
    {
      "epoch": 6.519916142557652,
      "grad_norm": 5.9375,
      "learning_rate": 1.6290671770441135e-05,
      "loss": 0.3946,
      "num_input_tokens_seen": 8107144,
      "step": 12440
    },
    {
      "epoch": 6.522536687631027,
      "grad_norm": 5.6875,
      "learning_rate": 1.6269239427357348e-05,
      "loss": 0.5093,
      "num_input_tokens_seen": 8111304,
      "step": 12445
    },
    {
      "epoch": 6.5251572327044025,
      "grad_norm": 7.28125,
      "learning_rate": 1.62478143897752e-05,
      "loss": 0.5235,
      "num_input_tokens_seen": 8114056,
      "step": 12450
    },
    {
      "epoch": 6.527777777777778,
      "grad_norm": 10.75,
      "learning_rate": 1.6226396675622203e-05,
      "loss": 0.588,
      "num_input_tokens_seen": 8116808,
      "step": 12455
    },
    {
      "epoch": 6.530398322851153,
      "grad_norm": 7.90625,
      "learning_rate": 1.6204986302819693e-05,
      "loss": 0.5274,
      "num_input_tokens_seen": 8119496,
      "step": 12460
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 8.3125,
      "learning_rate": 1.6183583289282906e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 8122728,
      "step": 12465
    },
    {
      "epoch": 6.535639412997903,
      "grad_norm": 7.40625,
      "learning_rate": 1.616218765292091e-05,
      "loss": 0.61,
      "num_input_tokens_seen": 8126504,
      "step": 12470
    },
    {
      "epoch": 6.538259958071279,
      "grad_norm": 7.65625,
      "learning_rate": 1.6140799411636586e-05,
      "loss": 0.6672,
      "num_input_tokens_seen": 8128776,
      "step": 12475
    },
    {
      "epoch": 6.540880503144654,
      "grad_norm": 5.96875,
      "learning_rate": 1.611941858332664e-05,
      "loss": 0.564,
      "num_input_tokens_seen": 8131720,
      "step": 12480
    },
    {
      "epoch": 6.54350104821803,
      "grad_norm": 3.578125,
      "learning_rate": 1.6098045185881587e-05,
      "loss": 0.5799,
      "num_input_tokens_seen": 8135624,
      "step": 12485
    },
    {
      "epoch": 6.546121593291405,
      "grad_norm": 6.625,
      "learning_rate": 1.6076679237185682e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 8138632,
      "step": 12490
    },
    {
      "epoch": 6.54874213836478,
      "grad_norm": 5.90625,
      "learning_rate": 1.6055320755117004e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 8142184,
      "step": 12495
    },
    {
      "epoch": 6.551362683438155,
      "grad_norm": 3.375,
      "learning_rate": 1.6033969757547336e-05,
      "loss": 0.5607,
      "num_input_tokens_seen": 8146792,
      "step": 12500
    },
    {
      "epoch": 6.55398322851153,
      "grad_norm": 6.9375,
      "learning_rate": 1.601262626234222e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 8150504,
      "step": 12505
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 4.75,
      "learning_rate": 1.5991290287360925e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 8153000,
      "step": 12510
    },
    {
      "epoch": 6.559224318658281,
      "grad_norm": 11.5,
      "learning_rate": 1.5969961850456412e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 8155720,
      "step": 12515
    },
    {
      "epoch": 6.561844863731656,
      "grad_norm": 3.859375,
      "learning_rate": 1.5948640969475346e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 8159400,
      "step": 12520
    },
    {
      "epoch": 6.564465408805032,
      "grad_norm": 3.984375,
      "learning_rate": 1.592732766225808e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 8162952,
      "step": 12525
    },
    {
      "epoch": 6.567085953878407,
      "grad_norm": 9.75,
      "learning_rate": 1.5906021946638585e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 8166376,
      "step": 12530
    },
    {
      "epoch": 6.569706498951782,
      "grad_norm": 9.5625,
      "learning_rate": 1.5884723840444532e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 8169064,
      "step": 12535
    },
    {
      "epoch": 6.572327044025157,
      "grad_norm": 4.875,
      "learning_rate": 1.5863433361497214e-05,
      "loss": 0.6039,
      "num_input_tokens_seen": 8171976,
      "step": 12540
    },
    {
      "epoch": 6.5749475890985325,
      "grad_norm": 6.125,
      "learning_rate": 1.5842150527611506e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 8176200,
      "step": 12545
    },
    {
      "epoch": 6.577568134171908,
      "grad_norm": 4.65625,
      "learning_rate": 1.5820875356595925e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 8179368,
      "step": 12550
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 7.5625,
      "learning_rate": 1.579960786625256e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 8182248,
      "step": 12555
    },
    {
      "epoch": 6.582809224318658,
      "grad_norm": 5.71875,
      "learning_rate": 1.5778348074377074e-05,
      "loss": 0.5931,
      "num_input_tokens_seen": 8184776,
      "step": 12560
    },
    {
      "epoch": 6.585429769392033,
      "grad_norm": 4.53125,
      "learning_rate": 1.575709599875869e-05,
      "loss": 0.4146,
      "num_input_tokens_seen": 8187976,
      "step": 12565
    },
    {
      "epoch": 6.588050314465409,
      "grad_norm": 6.71875,
      "learning_rate": 1.5735851657180184e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 8190984,
      "step": 12570
    },
    {
      "epoch": 6.590670859538784,
      "grad_norm": 6.96875,
      "learning_rate": 1.571461506741783e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 8194568,
      "step": 12575
    },
    {
      "epoch": 6.59329140461216,
      "grad_norm": 7.84375,
      "learning_rate": 1.5693386247241453e-05,
      "loss": 0.6169,
      "num_input_tokens_seen": 8197512,
      "step": 12580
    },
    {
      "epoch": 6.595911949685535,
      "grad_norm": 6.5625,
      "learning_rate": 1.5672165214414362e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 8201128,
      "step": 12585
    },
    {
      "epoch": 6.59853249475891,
      "grad_norm": 4.53125,
      "learning_rate": 1.5650951986693334e-05,
      "loss": 0.3319,
      "num_input_tokens_seen": 8204392,
      "step": 12590
    },
    {
      "epoch": 6.601153039832285,
      "grad_norm": 5.53125,
      "learning_rate": 1.5629746581828642e-05,
      "loss": 0.5252,
      "num_input_tokens_seen": 8206952,
      "step": 12595
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 16.625,
      "learning_rate": 1.560854901756399e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 8209768,
      "step": 12600
    },
    {
      "epoch": 6.606394129979035,
      "grad_norm": 2.1875,
      "learning_rate": 1.558735931163653e-05,
      "loss": 0.5677,
      "num_input_tokens_seen": 8216680,
      "step": 12605
    },
    {
      "epoch": 6.609014675052411,
      "grad_norm": 5.0625,
      "learning_rate": 1.5566177481776857e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 8220552,
      "step": 12610
    },
    {
      "epoch": 6.611635220125786,
      "grad_norm": 4.3125,
      "learning_rate": 1.554500354570894e-05,
      "loss": 0.5119,
      "num_input_tokens_seen": 8225064,
      "step": 12615
    },
    {
      "epoch": 6.614255765199162,
      "grad_norm": 8.0,
      "learning_rate": 1.552383752115017e-05,
      "loss": 0.5461,
      "num_input_tokens_seen": 8228168,
      "step": 12620
    },
    {
      "epoch": 6.616876310272537,
      "grad_norm": 14.6875,
      "learning_rate": 1.550267942581132e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 8231944,
      "step": 12625
    },
    {
      "epoch": 6.619496855345912,
      "grad_norm": 8.125,
      "learning_rate": 1.548152927739649e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 8235848,
      "step": 12630
    },
    {
      "epoch": 6.622117400419287,
      "grad_norm": 4.59375,
      "learning_rate": 1.5460387093603178e-05,
      "loss": 0.5556,
      "num_input_tokens_seen": 8238792,
      "step": 12635
    },
    {
      "epoch": 6.6247379454926625,
      "grad_norm": 7.84375,
      "learning_rate": 1.5439252892122197e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 8242088,
      "step": 12640
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 6.34375,
      "learning_rate": 1.5418126690637673e-05,
      "loss": 0.616,
      "num_input_tokens_seen": 8244616,
      "step": 12645
    },
    {
      "epoch": 6.629979035639413,
      "grad_norm": 5.34375,
      "learning_rate": 1.5397008506827057e-05,
      "loss": 0.4303,
      "num_input_tokens_seen": 8247912,
      "step": 12650
    },
    {
      "epoch": 6.632599580712788,
      "grad_norm": 4.59375,
      "learning_rate": 1.537589835836108e-05,
      "loss": 0.4869,
      "num_input_tokens_seen": 8251816,
      "step": 12655
    },
    {
      "epoch": 6.635220125786163,
      "grad_norm": 6.84375,
      "learning_rate": 1.5354796262903736e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 8255432,
      "step": 12660
    },
    {
      "epoch": 6.637840670859539,
      "grad_norm": 3.125,
      "learning_rate": 1.5333702238112306e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 8258824,
      "step": 12665
    },
    {
      "epoch": 6.640461215932914,
      "grad_norm": 6.6875,
      "learning_rate": 1.5312616301637313e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 8262888,
      "step": 12670
    },
    {
      "epoch": 6.6430817610062896,
      "grad_norm": 4.5625,
      "learning_rate": 1.5291538471122488e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 8266568,
      "step": 12675
    },
    {
      "epoch": 6.645702306079665,
      "grad_norm": 6.75,
      "learning_rate": 1.527046876420481e-05,
      "loss": 0.5451,
      "num_input_tokens_seen": 8269800,
      "step": 12680
    },
    {
      "epoch": 6.64832285115304,
      "grad_norm": 4.09375,
      "learning_rate": 1.524940719851444e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 8272936,
      "step": 12685
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 4.34375,
      "learning_rate": 1.5228353791674734e-05,
      "loss": 0.501,
      "num_input_tokens_seen": 8276104,
      "step": 12690
    },
    {
      "epoch": 6.65356394129979,
      "grad_norm": 4.09375,
      "learning_rate": 1.520730856130223e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 8278728,
      "step": 12695
    },
    {
      "epoch": 6.656184486373165,
      "grad_norm": 5.09375,
      "learning_rate": 1.5186271525006607e-05,
      "loss": 0.4528,
      "num_input_tokens_seen": 8282312,
      "step": 12700
    },
    {
      "epoch": 6.658805031446541,
      "grad_norm": 5.46875,
      "learning_rate": 1.5165242700390697e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 8285800,
      "step": 12705
    },
    {
      "epoch": 6.661425576519916,
      "grad_norm": 6.5625,
      "learning_rate": 1.5144222105050471e-05,
      "loss": 0.6055,
      "num_input_tokens_seen": 8288648,
      "step": 12710
    },
    {
      "epoch": 6.664046121593291,
      "grad_norm": 6.0625,
      "learning_rate": 1.5123209756574986e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 8291688,
      "step": 12715
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 9.9375,
      "learning_rate": 1.5102205672546416e-05,
      "loss": 0.5687,
      "num_input_tokens_seen": 8295400,
      "step": 12720
    },
    {
      "epoch": 6.669287211740042,
      "grad_norm": 4.75,
      "learning_rate": 1.508120987054004e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 8297704,
      "step": 12725
    },
    {
      "epoch": 6.671907756813417,
      "grad_norm": 9.375,
      "learning_rate": 1.5060222368124163e-05,
      "loss": 0.6368,
      "num_input_tokens_seen": 8300808,
      "step": 12730
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 6.1875,
      "learning_rate": 1.5039243182860177e-05,
      "loss": 0.6931,
      "num_input_tokens_seen": 8304104,
      "step": 12735
    },
    {
      "epoch": 6.677148846960168,
      "grad_norm": 7.3125,
      "learning_rate": 1.5018272332302513e-05,
      "loss": 0.5791,
      "num_input_tokens_seen": 8307784,
      "step": 12740
    },
    {
      "epoch": 6.679769392033543,
      "grad_norm": 4.375,
      "learning_rate": 1.4997309833998607e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 8311848,
      "step": 12745
    },
    {
      "epoch": 6.682389937106918,
      "grad_norm": 4.03125,
      "learning_rate": 1.4976355705488932e-05,
      "loss": 0.5098,
      "num_input_tokens_seen": 8315272,
      "step": 12750
    },
    {
      "epoch": 6.685010482180293,
      "grad_norm": 6.8125,
      "learning_rate": 1.4955409964306946e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 8318376,
      "step": 12755
    },
    {
      "epoch": 6.687631027253669,
      "grad_norm": 3.921875,
      "learning_rate": 1.4934472627979067e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 8322248,
      "step": 12760
    },
    {
      "epoch": 6.690251572327044,
      "grad_norm": 3.671875,
      "learning_rate": 1.491354371402473e-05,
      "loss": 0.3404,
      "num_input_tokens_seen": 8326056,
      "step": 12765
    },
    {
      "epoch": 6.6928721174004195,
      "grad_norm": 4.4375,
      "learning_rate": 1.4892623239956289e-05,
      "loss": 0.5664,
      "num_input_tokens_seen": 8329608,
      "step": 12770
    },
    {
      "epoch": 6.695492662473795,
      "grad_norm": 7.21875,
      "learning_rate": 1.4871711223279022e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 8332072,
      "step": 12775
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 5.875,
      "learning_rate": 1.4850807681491169e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 8334760,
      "step": 12780
    },
    {
      "epoch": 6.700733752620545,
      "grad_norm": 5.59375,
      "learning_rate": 1.4829912632083845e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 8337736,
      "step": 12785
    },
    {
      "epoch": 6.70335429769392,
      "grad_norm": 6.0625,
      "learning_rate": 1.4809026092541078e-05,
      "loss": 0.3939,
      "num_input_tokens_seen": 8340488,
      "step": 12790
    },
    {
      "epoch": 6.705974842767295,
      "grad_norm": 11.4375,
      "learning_rate": 1.4788148080339787e-05,
      "loss": 0.5238,
      "num_input_tokens_seen": 8343176,
      "step": 12795
    },
    {
      "epoch": 6.7085953878406706,
      "grad_norm": 5.96875,
      "learning_rate": 1.4767278612949703e-05,
      "loss": 0.587,
      "num_input_tokens_seen": 8346376,
      "step": 12800
    },
    {
      "epoch": 6.711215932914046,
      "grad_norm": 5.625,
      "learning_rate": 1.474641770783347e-05,
      "loss": 0.4517,
      "num_input_tokens_seen": 8349672,
      "step": 12805
    },
    {
      "epoch": 6.713836477987421,
      "grad_norm": 5.90625,
      "learning_rate": 1.4725565382446549e-05,
      "loss": 0.5274,
      "num_input_tokens_seen": 8352232,
      "step": 12810
    },
    {
      "epoch": 6.716457023060797,
      "grad_norm": 4.6875,
      "learning_rate": 1.4704721654237185e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 8355592,
      "step": 12815
    },
    {
      "epoch": 6.719077568134172,
      "grad_norm": 4.0625,
      "learning_rate": 1.4683886540646468e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 8357736,
      "step": 12820
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 5.1875,
      "learning_rate": 1.4663060059108282e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 8362568,
      "step": 12825
    },
    {
      "epoch": 6.7243186582809225,
      "grad_norm": 5.3125,
      "learning_rate": 1.464224222704926e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 8366152,
      "step": 12830
    },
    {
      "epoch": 6.726939203354298,
      "grad_norm": 15.8125,
      "learning_rate": 1.462143306188882e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 8368456,
      "step": 12835
    },
    {
      "epoch": 6.729559748427673,
      "grad_norm": 6.6875,
      "learning_rate": 1.4600632581039123e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 8371560,
      "step": 12840
    },
    {
      "epoch": 6.732180293501048,
      "grad_norm": 9.375,
      "learning_rate": 1.457984080190506e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 8374792,
      "step": 12845
    },
    {
      "epoch": 6.734800838574423,
      "grad_norm": 7.34375,
      "learning_rate": 1.4559057741884227e-05,
      "loss": 0.4722,
      "num_input_tokens_seen": 8377832,
      "step": 12850
    },
    {
      "epoch": 6.737421383647799,
      "grad_norm": 4.78125,
      "learning_rate": 1.4538283418366965e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 8380872,
      "step": 12855
    },
    {
      "epoch": 6.740041928721174,
      "grad_norm": 3.921875,
      "learning_rate": 1.4517517848736267e-05,
      "loss": 0.4755,
      "num_input_tokens_seen": 8384264,
      "step": 12860
    },
    {
      "epoch": 6.7426624737945495,
      "grad_norm": 3.375,
      "learning_rate": 1.449676105036781e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 8389064,
      "step": 12865
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 5.71875,
      "learning_rate": 1.4476013040629938e-05,
      "loss": 0.5602,
      "num_input_tokens_seen": 8391976,
      "step": 12870
    },
    {
      "epoch": 6.7479035639413,
      "grad_norm": 4.5625,
      "learning_rate": 1.4455273836883629e-05,
      "loss": 0.3838,
      "num_input_tokens_seen": 8395176,
      "step": 12875
    },
    {
      "epoch": 6.750524109014675,
      "grad_norm": 7.375,
      "learning_rate": 1.443454345648252e-05,
      "loss": 0.5768,
      "num_input_tokens_seen": 8398216,
      "step": 12880
    },
    {
      "epoch": 6.75314465408805,
      "grad_norm": 3.765625,
      "learning_rate": 1.4413821916772832e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 8402120,
      "step": 12885
    },
    {
      "epoch": 6.755765199161425,
      "grad_norm": 5.875,
      "learning_rate": 1.4393109235093399e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 8405736,
      "step": 12890
    },
    {
      "epoch": 6.7583857442348005,
      "grad_norm": 20.25,
      "learning_rate": 1.4372405428775664e-05,
      "loss": 0.6308,
      "num_input_tokens_seen": 8408808,
      "step": 12895
    },
    {
      "epoch": 6.761006289308176,
      "grad_norm": 5.4375,
      "learning_rate": 1.4351710515143618e-05,
      "loss": 0.3984,
      "num_input_tokens_seen": 8411400,
      "step": 12900
    },
    {
      "epoch": 6.763626834381551,
      "grad_norm": 5.15625,
      "learning_rate": 1.4331024511513808e-05,
      "loss": 0.365,
      "num_input_tokens_seen": 8414632,
      "step": 12905
    },
    {
      "epoch": 6.766247379454927,
      "grad_norm": 10.375,
      "learning_rate": 1.4310347435195368e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 8417256,
      "step": 12910
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 4.1875,
      "learning_rate": 1.428967930348989e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 8420584,
      "step": 12915
    },
    {
      "epoch": 6.771488469601677,
      "grad_norm": 5.28125,
      "learning_rate": 1.4269020133691542e-05,
      "loss": 0.5902,
      "num_input_tokens_seen": 8423912,
      "step": 12920
    },
    {
      "epoch": 6.774109014675052,
      "grad_norm": 6.75,
      "learning_rate": 1.4248369943086998e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 8427368,
      "step": 12925
    },
    {
      "epoch": 6.776729559748428,
      "grad_norm": 3.8125,
      "learning_rate": 1.4227728748955345e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 8431176,
      "step": 12930
    },
    {
      "epoch": 6.779350104821803,
      "grad_norm": 5.09375,
      "learning_rate": 1.4207096568568232e-05,
      "loss": 0.4698,
      "num_input_tokens_seen": 8433768,
      "step": 12935
    },
    {
      "epoch": 6.781970649895178,
      "grad_norm": 6.9375,
      "learning_rate": 1.418647341918971e-05,
      "loss": 0.4002,
      "num_input_tokens_seen": 8437384,
      "step": 12940
    },
    {
      "epoch": 6.784591194968553,
      "grad_norm": 5.9375,
      "learning_rate": 1.4165859318076276e-05,
      "loss": 0.4003,
      "num_input_tokens_seen": 8441192,
      "step": 12945
    },
    {
      "epoch": 6.787211740041929,
      "grad_norm": 5.21875,
      "learning_rate": 1.4145254282476895e-05,
      "loss": 0.627,
      "num_input_tokens_seen": 8446504,
      "step": 12950
    },
    {
      "epoch": 6.789832285115304,
      "grad_norm": 4.34375,
      "learning_rate": 1.4124658329632901e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 8449384,
      "step": 12955
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 12.4375,
      "learning_rate": 1.4104071476778044e-05,
      "loss": 0.4389,
      "num_input_tokens_seen": 8452456,
      "step": 12960
    },
    {
      "epoch": 6.795073375262055,
      "grad_norm": 6.28125,
      "learning_rate": 1.4083493741138486e-05,
      "loss": 0.5259,
      "num_input_tokens_seen": 8457000,
      "step": 12965
    },
    {
      "epoch": 6.79769392033543,
      "grad_norm": 8.125,
      "learning_rate": 1.4062925139932703e-05,
      "loss": 0.4612,
      "num_input_tokens_seen": 8461576,
      "step": 12970
    },
    {
      "epoch": 6.800314465408805,
      "grad_norm": 8.375,
      "learning_rate": 1.4042365690371587e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 8464712,
      "step": 12975
    },
    {
      "epoch": 6.80293501048218,
      "grad_norm": 5.9375,
      "learning_rate": 1.4021815409658335e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 8467880,
      "step": 12980
    },
    {
      "epoch": 6.805555555555555,
      "grad_norm": 5.0,
      "learning_rate": 1.4001274314988475e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 8471208,
      "step": 12985
    },
    {
      "epoch": 6.8081761006289305,
      "grad_norm": 4.84375,
      "learning_rate": 1.3980742423549875e-05,
      "loss": 0.5132,
      "num_input_tokens_seen": 8474248,
      "step": 12990
    },
    {
      "epoch": 6.810796645702306,
      "grad_norm": 3.90625,
      "learning_rate": 1.3960219752522679e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 8477064,
      "step": 12995
    },
    {
      "epoch": 6.813417190775681,
      "grad_norm": 6.03125,
      "learning_rate": 1.3939706319079305e-05,
      "loss": 0.5512,
      "num_input_tokens_seen": 8479528,
      "step": 13000
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 7.125,
      "learning_rate": 1.391920214038448e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 8482856,
      "step": 13005
    },
    {
      "epoch": 6.818658280922432,
      "grad_norm": 3.65625,
      "learning_rate": 1.3898707233595153e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 8486312,
      "step": 13010
    },
    {
      "epoch": 6.821278825995807,
      "grad_norm": 6.84375,
      "learning_rate": 1.3878221615860527e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 8489096,
      "step": 13015
    },
    {
      "epoch": 6.823899371069182,
      "grad_norm": 3.0625,
      "learning_rate": 1.3857745304322017e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 8493448,
      "step": 13020
    },
    {
      "epoch": 6.826519916142558,
      "grad_norm": 4.9375,
      "learning_rate": 1.3837278316113293e-05,
      "loss": 0.3712,
      "num_input_tokens_seen": 8496968,
      "step": 13025
    },
    {
      "epoch": 6.829140461215933,
      "grad_norm": 13.5625,
      "learning_rate": 1.3816820668360177e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 8500616,
      "step": 13030
    },
    {
      "epoch": 6.831761006289308,
      "grad_norm": 3.796875,
      "learning_rate": 1.3796372378180691e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 8504136,
      "step": 13035
    },
    {
      "epoch": 6.834381551362683,
      "grad_norm": 6.59375,
      "learning_rate": 1.3775933462685047e-05,
      "loss": 0.5659,
      "num_input_tokens_seen": 8507464,
      "step": 13040
    },
    {
      "epoch": 6.837002096436059,
      "grad_norm": 8.25,
      "learning_rate": 1.375550393897559e-05,
      "loss": 0.5668,
      "num_input_tokens_seen": 8510216,
      "step": 13045
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 8.0625,
      "learning_rate": 1.3735083824146793e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 8513960,
      "step": 13050
    },
    {
      "epoch": 6.8422431865828095,
      "grad_norm": 4.78125,
      "learning_rate": 1.3714673135285316e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 8517768,
      "step": 13055
    },
    {
      "epoch": 6.844863731656185,
      "grad_norm": 3.296875,
      "learning_rate": 1.3694271889469844e-05,
      "loss": 0.5667,
      "num_input_tokens_seen": 8521640,
      "step": 13060
    },
    {
      "epoch": 6.84748427672956,
      "grad_norm": 4.0625,
      "learning_rate": 1.3673880103771241e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 8524968,
      "step": 13065
    },
    {
      "epoch": 6.850104821802935,
      "grad_norm": 6.34375,
      "learning_rate": 1.365349779525241e-05,
      "loss": 0.6819,
      "num_input_tokens_seen": 8528200,
      "step": 13070
    },
    {
      "epoch": 6.85272536687631,
      "grad_norm": 4.875,
      "learning_rate": 1.3633124980968327e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 8531144,
      "step": 13075
    },
    {
      "epoch": 6.855345911949685,
      "grad_norm": 9.875,
      "learning_rate": 1.3612761677966051e-05,
      "loss": 0.5096,
      "num_input_tokens_seen": 8535240,
      "step": 13080
    },
    {
      "epoch": 6.8579664570230605,
      "grad_norm": 5.21875,
      "learning_rate": 1.3592407903284654e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 8537352,
      "step": 13085
    },
    {
      "epoch": 6.860587002096436,
      "grad_norm": 7.34375,
      "learning_rate": 1.3572063673955238e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 8540776,
      "step": 13090
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 7.8125,
      "learning_rate": 1.355172900700095e-05,
      "loss": 0.5593,
      "num_input_tokens_seen": 8543944,
      "step": 13095
    },
    {
      "epoch": 6.865828092243187,
      "grad_norm": 9.0625,
      "learning_rate": 1.3531403919436875e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 8546984,
      "step": 13100
    },
    {
      "epoch": 6.868448637316562,
      "grad_norm": 8.8125,
      "learning_rate": 1.3511088428270142e-05,
      "loss": 0.8197,
      "num_input_tokens_seen": 8550376,
      "step": 13105
    },
    {
      "epoch": 6.871069182389937,
      "grad_norm": 7.03125,
      "learning_rate": 1.3490782550499823e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 8553512,
      "step": 13110
    },
    {
      "epoch": 6.873689727463312,
      "grad_norm": 4.03125,
      "learning_rate": 1.3470486303116936e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 8557320,
      "step": 13115
    },
    {
      "epoch": 6.876310272536688,
      "grad_norm": 7.75,
      "learning_rate": 1.3450199703104471e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 8559848,
      "step": 13120
    },
    {
      "epoch": 6.878930817610063,
      "grad_norm": 5.1875,
      "learning_rate": 1.3429922767437319e-05,
      "loss": 0.3279,
      "num_input_tokens_seen": 8563400,
      "step": 13125
    },
    {
      "epoch": 6.881551362683438,
      "grad_norm": 3.75,
      "learning_rate": 1.3409655513082291e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 8566728,
      "step": 13130
    },
    {
      "epoch": 6.884171907756813,
      "grad_norm": 8.4375,
      "learning_rate": 1.3389397956998111e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 8570056,
      "step": 13135
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 6.65625,
      "learning_rate": 1.336915011613537e-05,
      "loss": 0.4539,
      "num_input_tokens_seen": 8573064,
      "step": 13140
    },
    {
      "epoch": 6.889412997903564,
      "grad_norm": 6.09375,
      "learning_rate": 1.3348912007436537e-05,
      "loss": 0.5976,
      "num_input_tokens_seen": 8577640,
      "step": 13145
    },
    {
      "epoch": 6.8920335429769395,
      "grad_norm": 5.59375,
      "learning_rate": 1.3328683647835933e-05,
      "loss": 0.6055,
      "num_input_tokens_seen": 8581256,
      "step": 13150
    },
    {
      "epoch": 6.894654088050315,
      "grad_norm": 4.6875,
      "learning_rate": 1.330846505425972e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 8584168,
      "step": 13155
    },
    {
      "epoch": 6.89727463312369,
      "grad_norm": 4.375,
      "learning_rate": 1.3288256243625911e-05,
      "loss": 0.5018,
      "num_input_tokens_seen": 8587432,
      "step": 13160
    },
    {
      "epoch": 6.899895178197065,
      "grad_norm": 4.75,
      "learning_rate": 1.3268057232844305e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 8590728,
      "step": 13165
    },
    {
      "epoch": 6.90251572327044,
      "grad_norm": 7.125,
      "learning_rate": 1.3247868038816504e-05,
      "loss": 0.4051,
      "num_input_tokens_seen": 8594120,
      "step": 13170
    },
    {
      "epoch": 6.905136268343815,
      "grad_norm": 4.84375,
      "learning_rate": 1.3227688678435924e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 8596904,
      "step": 13175
    },
    {
      "epoch": 6.9077568134171905,
      "grad_norm": 4.65625,
      "learning_rate": 1.3207519168587717e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 8599848,
      "step": 13180
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 4.15625,
      "learning_rate": 1.3187359526148813e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 8603272,
      "step": 13185
    },
    {
      "epoch": 6.912997903563941,
      "grad_norm": 5.59375,
      "learning_rate": 1.3167209767987868e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 8607240,
      "step": 13190
    },
    {
      "epoch": 6.915618448637317,
      "grad_norm": 18.625,
      "learning_rate": 1.3147069910965298e-05,
      "loss": 0.638,
      "num_input_tokens_seen": 8613160,
      "step": 13195
    },
    {
      "epoch": 6.918238993710692,
      "grad_norm": 4.1875,
      "learning_rate": 1.3126939971933205e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 8617480,
      "step": 13200
    },
    {
      "epoch": 6.920859538784067,
      "grad_norm": 4.59375,
      "learning_rate": 1.3106819967735395e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 8620520,
      "step": 13205
    },
    {
      "epoch": 6.923480083857442,
      "grad_norm": 4.9375,
      "learning_rate": 1.3086709915207388e-05,
      "loss": 0.3326,
      "num_input_tokens_seen": 8626568,
      "step": 13210
    },
    {
      "epoch": 6.926100628930818,
      "grad_norm": 4.625,
      "learning_rate": 1.3066609831176346e-05,
      "loss": 0.5117,
      "num_input_tokens_seen": 8630184,
      "step": 13215
    },
    {
      "epoch": 6.928721174004193,
      "grad_norm": 5.6875,
      "learning_rate": 1.3046519732461094e-05,
      "loss": 0.4645,
      "num_input_tokens_seen": 8633576,
      "step": 13220
    },
    {
      "epoch": 6.931341719077568,
      "grad_norm": 4.9375,
      "learning_rate": 1.302643963587213e-05,
      "loss": 0.4761,
      "num_input_tokens_seen": 8636520,
      "step": 13225
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 5.75,
      "learning_rate": 1.3006369558211534e-05,
      "loss": 0.3682,
      "num_input_tokens_seen": 8639816,
      "step": 13230
    },
    {
      "epoch": 6.936582809224319,
      "grad_norm": 6.8125,
      "learning_rate": 1.2986309516273043e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 8642888,
      "step": 13235
    },
    {
      "epoch": 6.939203354297694,
      "grad_norm": 4.1875,
      "learning_rate": 1.2966259526842006e-05,
      "loss": 0.5245,
      "num_input_tokens_seen": 8645704,
      "step": 13240
    },
    {
      "epoch": 6.9418238993710695,
      "grad_norm": 5.65625,
      "learning_rate": 1.2946219606695297e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 8648648,
      "step": 13245
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 5.40625,
      "learning_rate": 1.2926189772601438e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 8651560,
      "step": 13250
    },
    {
      "epoch": 6.94706498951782,
      "grad_norm": 5.5,
      "learning_rate": 1.2906170041320468e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 8655208,
      "step": 13255
    },
    {
      "epoch": 6.949685534591195,
      "grad_norm": 6.0625,
      "learning_rate": 1.2886160429603972e-05,
      "loss": 0.3521,
      "num_input_tokens_seen": 8657576,
      "step": 13260
    },
    {
      "epoch": 6.95230607966457,
      "grad_norm": 10.0,
      "learning_rate": 1.2866160954195112e-05,
      "loss": 0.7338,
      "num_input_tokens_seen": 8660872,
      "step": 13265
    },
    {
      "epoch": 6.954926624737945,
      "grad_norm": 5.34375,
      "learning_rate": 1.284617163182849e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 8665192,
      "step": 13270
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 2.3125,
      "learning_rate": 1.2826192479230287e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 8670088,
      "step": 13275
    },
    {
      "epoch": 6.960167714884696,
      "grad_norm": 12.375,
      "learning_rate": 1.2806223513118154e-05,
      "loss": 0.5037,
      "num_input_tokens_seen": 8673576,
      "step": 13280
    },
    {
      "epoch": 6.962788259958071,
      "grad_norm": 5.40625,
      "learning_rate": 1.2786264750201182e-05,
      "loss": 0.6897,
      "num_input_tokens_seen": 8676424,
      "step": 13285
    },
    {
      "epoch": 6.965408805031447,
      "grad_norm": 6.71875,
      "learning_rate": 1.2766316207179973e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 8679912,
      "step": 13290
    },
    {
      "epoch": 6.968029350104822,
      "grad_norm": 3.75,
      "learning_rate": 1.2746377900746548e-05,
      "loss": 0.7278,
      "num_input_tokens_seen": 8683880,
      "step": 13295
    },
    {
      "epoch": 6.970649895178197,
      "grad_norm": 3.984375,
      "learning_rate": 1.2726449847584365e-05,
      "loss": 0.5597,
      "num_input_tokens_seen": 8687784,
      "step": 13300
    },
    {
      "epoch": 6.973270440251572,
      "grad_norm": 4.75,
      "learning_rate": 1.2706532064368326e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 8690920,
      "step": 13305
    },
    {
      "epoch": 6.975890985324948,
      "grad_norm": 8.9375,
      "learning_rate": 1.268662456776471e-05,
      "loss": 0.3854,
      "num_input_tokens_seen": 8693960,
      "step": 13310
    },
    {
      "epoch": 6.978511530398323,
      "grad_norm": 2.5,
      "learning_rate": 1.2666727374431198e-05,
      "loss": 0.5485,
      "num_input_tokens_seen": 8697640,
      "step": 13315
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 5.15625,
      "learning_rate": 1.2646840501016863e-05,
      "loss": 0.7002,
      "num_input_tokens_seen": 8699912,
      "step": 13320
    },
    {
      "epoch": 6.983752620545073,
      "grad_norm": 5.625,
      "learning_rate": 1.262696396416213e-05,
      "loss": 0.6436,
      "num_input_tokens_seen": 8703240,
      "step": 13325
    },
    {
      "epoch": 6.986373165618449,
      "grad_norm": 6.5625,
      "learning_rate": 1.2607097780498772e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 8706376,
      "step": 13330
    },
    {
      "epoch": 6.988993710691824,
      "grad_norm": 4.625,
      "learning_rate": 1.2587241966649908e-05,
      "loss": 0.6149,
      "num_input_tokens_seen": 8709224,
      "step": 13335
    },
    {
      "epoch": 6.9916142557651995,
      "grad_norm": 11.5625,
      "learning_rate": 1.2567396539229965e-05,
      "loss": 0.5727,
      "num_input_tokens_seen": 8711816,
      "step": 13340
    },
    {
      "epoch": 6.994234800838575,
      "grad_norm": 9.4375,
      "learning_rate": 1.2547561514844704e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 8715496,
      "step": 13345
    },
    {
      "epoch": 6.99685534591195,
      "grad_norm": 6.21875,
      "learning_rate": 1.2527736910091168e-05,
      "loss": 0.4012,
      "num_input_tokens_seen": 8718312,
      "step": 13350
    },
    {
      "epoch": 6.999475890985325,
      "grad_norm": 10.5625,
      "learning_rate": 1.2507922741557665e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 8722632,
      "step": 13355
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.49383020401000977,
      "eval_runtime": 24.2161,
      "eval_samples_per_second": 35.018,
      "eval_steps_per_second": 8.755,
      "num_input_tokens_seen": 8722744,
      "step": 13356
    },
    {
      "epoch": 7.0020964360587,
      "grad_norm": 7.03125,
      "learning_rate": 1.2488119025823802e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 8725944,
      "step": 13360
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 5.59375,
      "learning_rate": 1.2468325779460424e-05,
      "loss": 0.4888,
      "num_input_tokens_seen": 8729592,
      "step": 13365
    },
    {
      "epoch": 7.0073375262054505,
      "grad_norm": 5.125,
      "learning_rate": 1.2448543019029607e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 8734264,
      "step": 13370
    },
    {
      "epoch": 7.009958071278826,
      "grad_norm": 4.84375,
      "learning_rate": 1.2428770761084655e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 8738744,
      "step": 13375
    },
    {
      "epoch": 7.012578616352202,
      "grad_norm": 6.125,
      "learning_rate": 1.2409009022170109e-05,
      "loss": 0.412,
      "num_input_tokens_seen": 8741720,
      "step": 13380
    },
    {
      "epoch": 7.015199161425577,
      "grad_norm": 4.9375,
      "learning_rate": 1.2389257818821679e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 8745112,
      "step": 13385
    },
    {
      "epoch": 7.017819706498952,
      "grad_norm": 17.25,
      "learning_rate": 1.236951716756626e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 8747960,
      "step": 13390
    },
    {
      "epoch": 7.020440251572327,
      "grad_norm": 9.3125,
      "learning_rate": 1.2349787084921952e-05,
      "loss": 0.5208,
      "num_input_tokens_seen": 8750808,
      "step": 13395
    },
    {
      "epoch": 7.023060796645702,
      "grad_norm": 6.71875,
      "learning_rate": 1.233006758739797e-05,
      "loss": 0.4881,
      "num_input_tokens_seen": 8753368,
      "step": 13400
    },
    {
      "epoch": 7.0256813417190775,
      "grad_norm": 10.625,
      "learning_rate": 1.2310358691494681e-05,
      "loss": 0.5373,
      "num_input_tokens_seen": 8756056,
      "step": 13405
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 10.25,
      "learning_rate": 1.229066041370362e-05,
      "loss": 0.4852,
      "num_input_tokens_seen": 8759256,
      "step": 13410
    },
    {
      "epoch": 7.030922431865828,
      "grad_norm": 8.9375,
      "learning_rate": 1.2270972770507364e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 8763032,
      "step": 13415
    },
    {
      "epoch": 7.033542976939203,
      "grad_norm": 8.8125,
      "learning_rate": 1.2251295778379657e-05,
      "loss": 0.3657,
      "num_input_tokens_seen": 8766072,
      "step": 13420
    },
    {
      "epoch": 7.036163522012578,
      "grad_norm": 7.6875,
      "learning_rate": 1.2231629453785324e-05,
      "loss": 0.503,
      "num_input_tokens_seen": 8769976,
      "step": 13425
    },
    {
      "epoch": 7.038784067085954,
      "grad_norm": 5.5,
      "learning_rate": 1.2211973813180209e-05,
      "loss": 0.5469,
      "num_input_tokens_seen": 8773528,
      "step": 13430
    },
    {
      "epoch": 7.0414046121593294,
      "grad_norm": 13.0625,
      "learning_rate": 1.2192328873011283e-05,
      "loss": 0.5117,
      "num_input_tokens_seen": 8776440,
      "step": 13435
    },
    {
      "epoch": 7.044025157232705,
      "grad_norm": 5.5,
      "learning_rate": 1.2172694649716524e-05,
      "loss": 0.3642,
      "num_input_tokens_seen": 8780408,
      "step": 13440
    },
    {
      "epoch": 7.04664570230608,
      "grad_norm": 5.40625,
      "learning_rate": 1.2153071159724947e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 8782872,
      "step": 13445
    },
    {
      "epoch": 7.049266247379455,
      "grad_norm": 3.703125,
      "learning_rate": 1.2133458419456614e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 8786776,
      "step": 13450
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 4.84375,
      "learning_rate": 1.2113856445322541e-05,
      "loss": 0.5565,
      "num_input_tokens_seen": 8791576,
      "step": 13455
    },
    {
      "epoch": 7.054507337526205,
      "grad_norm": 7.59375,
      "learning_rate": 1.2094265253724777e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 8794072,
      "step": 13460
    },
    {
      "epoch": 7.0571278825995805,
      "grad_norm": 4.75,
      "learning_rate": 1.207468486105636e-05,
      "loss": 0.5007,
      "num_input_tokens_seen": 8798424,
      "step": 13465
    },
    {
      "epoch": 7.059748427672956,
      "grad_norm": 7.90625,
      "learning_rate": 1.2055115283701224e-05,
      "loss": 0.3895,
      "num_input_tokens_seen": 8801208,
      "step": 13470
    },
    {
      "epoch": 7.062368972746331,
      "grad_norm": 6.8125,
      "learning_rate": 1.2035556538034332e-05,
      "loss": 0.514,
      "num_input_tokens_seen": 8804472,
      "step": 13475
    },
    {
      "epoch": 7.064989517819707,
      "grad_norm": 4.5625,
      "learning_rate": 1.2016008640421533e-05,
      "loss": 0.5287,
      "num_input_tokens_seen": 8808152,
      "step": 13480
    },
    {
      "epoch": 7.067610062893082,
      "grad_norm": 4.15625,
      "learning_rate": 1.1996471607219612e-05,
      "loss": 0.4592,
      "num_input_tokens_seen": 8811352,
      "step": 13485
    },
    {
      "epoch": 7.070230607966457,
      "grad_norm": 14.5625,
      "learning_rate": 1.1976945454776284e-05,
      "loss": 0.5112,
      "num_input_tokens_seen": 8814328,
      "step": 13490
    },
    {
      "epoch": 7.072851153039832,
      "grad_norm": 3.71875,
      "learning_rate": 1.1957430199430128e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 8817976,
      "step": 13495
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 3.484375,
      "learning_rate": 1.1937925857510609e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 8820856,
      "step": 13500
    },
    {
      "epoch": 7.078092243186583,
      "grad_norm": 5.9375,
      "learning_rate": 1.1918432445338092e-05,
      "loss": 0.5858,
      "num_input_tokens_seen": 8824120,
      "step": 13505
    },
    {
      "epoch": 7.080712788259958,
      "grad_norm": 5.40625,
      "learning_rate": 1.1898949979223765e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 8827192,
      "step": 13510
    },
    {
      "epoch": 7.083333333333333,
      "grad_norm": 7.0,
      "learning_rate": 1.187947847546966e-05,
      "loss": 0.4068,
      "num_input_tokens_seen": 8831064,
      "step": 13515
    },
    {
      "epoch": 7.085953878406708,
      "grad_norm": 5.28125,
      "learning_rate": 1.1860017950368646e-05,
      "loss": 0.5727,
      "num_input_tokens_seen": 8834200,
      "step": 13520
    },
    {
      "epoch": 7.088574423480084,
      "grad_norm": 8.6875,
      "learning_rate": 1.1840568420204392e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 8837240,
      "step": 13525
    },
    {
      "epoch": 7.091194968553459,
      "grad_norm": 3.3125,
      "learning_rate": 1.1821129901251396e-05,
      "loss": 0.6377,
      "num_input_tokens_seen": 8841976,
      "step": 13530
    },
    {
      "epoch": 7.093815513626835,
      "grad_norm": 8.0,
      "learning_rate": 1.1801702409774909e-05,
      "loss": 0.4213,
      "num_input_tokens_seen": 8845336,
      "step": 13535
    },
    {
      "epoch": 7.09643605870021,
      "grad_norm": 6.53125,
      "learning_rate": 1.1782285962030965e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 8848184,
      "step": 13540
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 11.8125,
      "learning_rate": 1.1762880574266374e-05,
      "loss": 0.5226,
      "num_input_tokens_seen": 8850840,
      "step": 13545
    },
    {
      "epoch": 7.10167714884696,
      "grad_norm": 13.875,
      "learning_rate": 1.1743486262718673e-05,
      "loss": 0.6105,
      "num_input_tokens_seen": 8854296,
      "step": 13550
    },
    {
      "epoch": 7.104297693920335,
      "grad_norm": 7.78125,
      "learning_rate": 1.1724103043616134e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 8857560,
      "step": 13555
    },
    {
      "epoch": 7.1069182389937104,
      "grad_norm": 2.671875,
      "learning_rate": 1.1704730933177738e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 8861496,
      "step": 13560
    },
    {
      "epoch": 7.109538784067086,
      "grad_norm": 7.1875,
      "learning_rate": 1.1685369947613204e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 8864344,
      "step": 13565
    },
    {
      "epoch": 7.112159329140461,
      "grad_norm": 3.109375,
      "learning_rate": 1.1666020103122907e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 8868152,
      "step": 13570
    },
    {
      "epoch": 7.114779874213837,
      "grad_norm": 7.25,
      "learning_rate": 1.1646681415897912e-05,
      "loss": 0.3424,
      "num_input_tokens_seen": 8870776,
      "step": 13575
    },
    {
      "epoch": 7.117400419287212,
      "grad_norm": 4.75,
      "learning_rate": 1.1627353902119958e-05,
      "loss": 0.5892,
      "num_input_tokens_seen": 8873944,
      "step": 13580
    },
    {
      "epoch": 7.120020964360587,
      "grad_norm": 14.0625,
      "learning_rate": 1.1608037577961423e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 8876792,
      "step": 13585
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 11.4375,
      "learning_rate": 1.158873245958531e-05,
      "loss": 0.6035,
      "num_input_tokens_seen": 8879672,
      "step": 13590
    },
    {
      "epoch": 7.1252620545073375,
      "grad_norm": 8.3125,
      "learning_rate": 1.1569438563145297e-05,
      "loss": 0.3822,
      "num_input_tokens_seen": 8882904,
      "step": 13595
    },
    {
      "epoch": 7.127882599580713,
      "grad_norm": 8.75,
      "learning_rate": 1.1550155904785587e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 8886328,
      "step": 13600
    },
    {
      "epoch": 7.130503144654088,
      "grad_norm": 10.8125,
      "learning_rate": 1.1530884500641063e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 8888888,
      "step": 13605
    },
    {
      "epoch": 7.133123689727463,
      "grad_norm": 5.96875,
      "learning_rate": 1.1511624366837143e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 8892536,
      "step": 13610
    },
    {
      "epoch": 7.135744234800838,
      "grad_norm": 4.25,
      "learning_rate": 1.149237551948982e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 8895768,
      "step": 13615
    },
    {
      "epoch": 7.138364779874214,
      "grad_norm": 6.09375,
      "learning_rate": 1.147313797470567e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 8899320,
      "step": 13620
    },
    {
      "epoch": 7.140985324947589,
      "grad_norm": 8.125,
      "learning_rate": 1.1453911748581778e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 8901816,
      "step": 13625
    },
    {
      "epoch": 7.143605870020965,
      "grad_norm": 4.40625,
      "learning_rate": 1.1434696857205765e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 8904984,
      "step": 13630
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 4.28125,
      "learning_rate": 1.1415493316655804e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 8909272,
      "step": 13635
    },
    {
      "epoch": 7.148846960167715,
      "grad_norm": 4.96875,
      "learning_rate": 1.1396301143000499e-05,
      "loss": 0.5336,
      "num_input_tokens_seen": 8912088,
      "step": 13640
    },
    {
      "epoch": 7.15146750524109,
      "grad_norm": 5.1875,
      "learning_rate": 1.1377120352299014e-05,
      "loss": 0.5546,
      "num_input_tokens_seen": 8915160,
      "step": 13645
    },
    {
      "epoch": 7.154088050314465,
      "grad_norm": 4.28125,
      "learning_rate": 1.1357950960600955e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 8918200,
      "step": 13650
    },
    {
      "epoch": 7.15670859538784,
      "grad_norm": 4.28125,
      "learning_rate": 1.1338792983946376e-05,
      "loss": 0.38,
      "num_input_tokens_seen": 8921464,
      "step": 13655
    },
    {
      "epoch": 7.159329140461216,
      "grad_norm": 8.9375,
      "learning_rate": 1.1319646438365817e-05,
      "loss": 0.4971,
      "num_input_tokens_seen": 8924344,
      "step": 13660
    },
    {
      "epoch": 7.161949685534591,
      "grad_norm": 8.5625,
      "learning_rate": 1.1300511339880227e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 8927352,
      "step": 13665
    },
    {
      "epoch": 7.164570230607967,
      "grad_norm": 5.65625,
      "learning_rate": 1.128138770450097e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 8930936,
      "step": 13670
    },
    {
      "epoch": 7.167190775681342,
      "grad_norm": 9.5625,
      "learning_rate": 1.126227554822985e-05,
      "loss": 0.568,
      "num_input_tokens_seen": 8934296,
      "step": 13675
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 3.421875,
      "learning_rate": 1.1243174887059038e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 8937016,
      "step": 13680
    },
    {
      "epoch": 7.172431865828092,
      "grad_norm": 7.53125,
      "learning_rate": 1.1224085736971093e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 8940792,
      "step": 13685
    },
    {
      "epoch": 7.1750524109014675,
      "grad_norm": 5.625,
      "learning_rate": 1.1205008113938934e-05,
      "loss": 0.5424,
      "num_input_tokens_seen": 8944024,
      "step": 13690
    },
    {
      "epoch": 7.177672955974843,
      "grad_norm": 7.21875,
      "learning_rate": 1.1185942033925867e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 8946872,
      "step": 13695
    },
    {
      "epoch": 7.180293501048218,
      "grad_norm": 5.28125,
      "learning_rate": 1.1166887512885505e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 8950168,
      "step": 13700
    },
    {
      "epoch": 7.182914046121593,
      "grad_norm": 9.75,
      "learning_rate": 1.11478445667618e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 8953112,
      "step": 13705
    },
    {
      "epoch": 7.185534591194968,
      "grad_norm": 9.0625,
      "learning_rate": 1.1128813211489012e-05,
      "loss": 0.4017,
      "num_input_tokens_seen": 8956312,
      "step": 13710
    },
    {
      "epoch": 7.188155136268344,
      "grad_norm": 7.4375,
      "learning_rate": 1.1109793462991725e-05,
      "loss": 0.4203,
      "num_input_tokens_seen": 8959032,
      "step": 13715
    },
    {
      "epoch": 7.190775681341719,
      "grad_norm": 8.3125,
      "learning_rate": 1.109078533718479e-05,
      "loss": 0.5876,
      "num_input_tokens_seen": 8962744,
      "step": 13720
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 3.6875,
      "learning_rate": 1.107178884997334e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 8966520,
      "step": 13725
    },
    {
      "epoch": 7.19601677148847,
      "grad_norm": 6.96875,
      "learning_rate": 1.1052804017252751e-05,
      "loss": 0.5575,
      "num_input_tokens_seen": 8969304,
      "step": 13730
    },
    {
      "epoch": 7.198637316561845,
      "grad_norm": 3.375,
      "learning_rate": 1.1033830854908691e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 8973432,
      "step": 13735
    },
    {
      "epoch": 7.20125786163522,
      "grad_norm": 7.0,
      "learning_rate": 1.1014869378817022e-05,
      "loss": 0.5403,
      "num_input_tokens_seen": 8976248,
      "step": 13740
    },
    {
      "epoch": 7.203878406708595,
      "grad_norm": 5.15625,
      "learning_rate": 1.0995919604843832e-05,
      "loss": 0.4678,
      "num_input_tokens_seen": 8980792,
      "step": 13745
    },
    {
      "epoch": 7.20649895178197,
      "grad_norm": 6.03125,
      "learning_rate": 1.0976981548845444e-05,
      "loss": 0.5889,
      "num_input_tokens_seen": 8983480,
      "step": 13750
    },
    {
      "epoch": 7.209119496855346,
      "grad_norm": 9.625,
      "learning_rate": 1.095805522666835e-05,
      "loss": 0.5444,
      "num_input_tokens_seen": 8986200,
      "step": 13755
    },
    {
      "epoch": 7.211740041928721,
      "grad_norm": 6.40625,
      "learning_rate": 1.0939140654149225e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 8989816,
      "step": 13760
    },
    {
      "epoch": 7.214360587002097,
      "grad_norm": 5.125,
      "learning_rate": 1.0920237847114944e-05,
      "loss": 0.3147,
      "num_input_tokens_seen": 8992600,
      "step": 13765
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 6.0,
      "learning_rate": 1.0901346821382476e-05,
      "loss": 0.501,
      "num_input_tokens_seen": 8997432,
      "step": 13770
    },
    {
      "epoch": 7.219601677148847,
      "grad_norm": 6.875,
      "learning_rate": 1.0882467592758989e-05,
      "loss": 0.5053,
      "num_input_tokens_seen": 8999864,
      "step": 13775
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 4.375,
      "learning_rate": 1.0863600177041772e-05,
      "loss": 0.3862,
      "num_input_tokens_seen": 9003160,
      "step": 13780
    },
    {
      "epoch": 7.2248427672955975,
      "grad_norm": 12.5625,
      "learning_rate": 1.0844744590018186e-05,
      "loss": 0.4619,
      "num_input_tokens_seen": 9006040,
      "step": 13785
    },
    {
      "epoch": 7.227463312368973,
      "grad_norm": 7.125,
      "learning_rate": 1.0825900847465748e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 9009624,
      "step": 13790
    },
    {
      "epoch": 7.230083857442348,
      "grad_norm": 4.28125,
      "learning_rate": 1.0807068965152033e-05,
      "loss": 0.4737,
      "num_input_tokens_seen": 9013912,
      "step": 13795
    },
    {
      "epoch": 7.232704402515723,
      "grad_norm": 6.625,
      "learning_rate": 1.0788248958834695e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 9017016,
      "step": 13800
    },
    {
      "epoch": 7.235324947589098,
      "grad_norm": 11.0,
      "learning_rate": 1.0769440844261481e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 9019000,
      "step": 13805
    },
    {
      "epoch": 7.237945492662474,
      "grad_norm": 7.28125,
      "learning_rate": 1.0750644637170122e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 9022520,
      "step": 13810
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 5.53125,
      "learning_rate": 1.0731860353288445e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 9025496,
      "step": 13815
    },
    {
      "epoch": 7.243186582809225,
      "grad_norm": 4.21875,
      "learning_rate": 1.0713088008334302e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 9029464,
      "step": 13820
    },
    {
      "epoch": 7.2458071278826,
      "grad_norm": 7.21875,
      "learning_rate": 1.0694327618015493e-05,
      "loss": 0.5512,
      "num_input_tokens_seen": 9032088,
      "step": 13825
    },
    {
      "epoch": 7.248427672955975,
      "grad_norm": 8.125,
      "learning_rate": 1.0675579198029887e-05,
      "loss": 0.4245,
      "num_input_tokens_seen": 9035768,
      "step": 13830
    },
    {
      "epoch": 7.25104821802935,
      "grad_norm": 4.25,
      "learning_rate": 1.0656842764065295e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 9038840,
      "step": 13835
    },
    {
      "epoch": 7.253668763102725,
      "grad_norm": 7.9375,
      "learning_rate": 1.0638118331799499e-05,
      "loss": 0.3548,
      "num_input_tokens_seen": 9042328,
      "step": 13840
    },
    {
      "epoch": 7.2562893081761,
      "grad_norm": 7.4375,
      "learning_rate": 1.061940591690027e-05,
      "loss": 0.492,
      "num_input_tokens_seen": 9045336,
      "step": 13845
    },
    {
      "epoch": 7.258909853249476,
      "grad_norm": 5.78125,
      "learning_rate": 1.0600705535025285e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 9048312,
      "step": 13850
    },
    {
      "epoch": 7.261530398322851,
      "grad_norm": 3.65625,
      "learning_rate": 1.058201720182217e-05,
      "loss": 0.5964,
      "num_input_tokens_seen": 9052760,
      "step": 13855
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 12.9375,
      "learning_rate": 1.056334093292848e-05,
      "loss": 0.7057,
      "num_input_tokens_seen": 9055864,
      "step": 13860
    },
    {
      "epoch": 7.266771488469602,
      "grad_norm": 9.25,
      "learning_rate": 1.054467674397166e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 9058776,
      "step": 13865
    },
    {
      "epoch": 7.269392033542977,
      "grad_norm": 9.5,
      "learning_rate": 1.0526024650569047e-05,
      "loss": 0.6474,
      "num_input_tokens_seen": 9062360,
      "step": 13870
    },
    {
      "epoch": 7.272012578616352,
      "grad_norm": 3.234375,
      "learning_rate": 1.0507384668327852e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 9065016,
      "step": 13875
    },
    {
      "epoch": 7.2746331236897275,
      "grad_norm": 6.4375,
      "learning_rate": 1.048875681284518e-05,
      "loss": 0.503,
      "num_input_tokens_seen": 9071992,
      "step": 13880
    },
    {
      "epoch": 7.277253668763103,
      "grad_norm": 5.96875,
      "learning_rate": 1.0470141099707959e-05,
      "loss": 0.5254,
      "num_input_tokens_seen": 9075672,
      "step": 13885
    },
    {
      "epoch": 7.279874213836478,
      "grad_norm": 6.53125,
      "learning_rate": 1.0451537544492968e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 9078936,
      "step": 13890
    },
    {
      "epoch": 7.282494758909853,
      "grad_norm": 9.0625,
      "learning_rate": 1.0432946162766805e-05,
      "loss": 0.6699,
      "num_input_tokens_seen": 9082168,
      "step": 13895
    },
    {
      "epoch": 7.285115303983228,
      "grad_norm": 13.375,
      "learning_rate": 1.0414366970085906e-05,
      "loss": 0.6301,
      "num_input_tokens_seen": 9084664,
      "step": 13900
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 3.75,
      "learning_rate": 1.0395799981996479e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 9087864,
      "step": 13905
    },
    {
      "epoch": 7.290356394129979,
      "grad_norm": 6.15625,
      "learning_rate": 1.0377245214034537e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 9091416,
      "step": 13910
    },
    {
      "epoch": 7.2929769392033545,
      "grad_norm": 7.96875,
      "learning_rate": 1.0358702681725848e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 9094296,
      "step": 13915
    },
    {
      "epoch": 7.29559748427673,
      "grad_norm": 7.5,
      "learning_rate": 1.0340172400585977e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 9097112,
      "step": 13920
    },
    {
      "epoch": 7.298218029350105,
      "grad_norm": 5.625,
      "learning_rate": 1.0321654386120205e-05,
      "loss": 0.4595,
      "num_input_tokens_seen": 9100632,
      "step": 13925
    },
    {
      "epoch": 7.30083857442348,
      "grad_norm": 6.125,
      "learning_rate": 1.0303148653823557e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 9103864,
      "step": 13930
    },
    {
      "epoch": 7.303459119496855,
      "grad_norm": 6.5625,
      "learning_rate": 1.0284655219180797e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 9107928,
      "step": 13935
    },
    {
      "epoch": 7.30607966457023,
      "grad_norm": 5.25,
      "learning_rate": 1.026617409766638e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 9111384,
      "step": 13940
    },
    {
      "epoch": 7.308700209643606,
      "grad_norm": 5.21875,
      "learning_rate": 1.0247705304744457e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 9114488,
      "step": 13945
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 4.53125,
      "learning_rate": 1.0229248855868892e-05,
      "loss": 0.5971,
      "num_input_tokens_seen": 9117208,
      "step": 13950
    },
    {
      "epoch": 7.313941299790357,
      "grad_norm": 4.21875,
      "learning_rate": 1.0210804766483168e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 9121176,
      "step": 13955
    },
    {
      "epoch": 7.316561844863732,
      "grad_norm": 4.96875,
      "learning_rate": 1.019237305202048e-05,
      "loss": 0.57,
      "num_input_tokens_seen": 9124216,
      "step": 13960
    },
    {
      "epoch": 7.319182389937107,
      "grad_norm": 5.34375,
      "learning_rate": 1.0173953727903634e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 9126936,
      "step": 13965
    },
    {
      "epoch": 7.321802935010482,
      "grad_norm": 5.9375,
      "learning_rate": 1.0155546809545077e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 9129688,
      "step": 13970
    },
    {
      "epoch": 7.3244234800838575,
      "grad_norm": 7.875,
      "learning_rate": 1.013715231234689e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 9132152,
      "step": 13975
    },
    {
      "epoch": 7.327044025157233,
      "grad_norm": 8.5,
      "learning_rate": 1.0118770251700741e-05,
      "loss": 0.567,
      "num_input_tokens_seen": 9135352,
      "step": 13980
    },
    {
      "epoch": 7.329664570230608,
      "grad_norm": 3.828125,
      "learning_rate": 1.0100400642987886e-05,
      "loss": 0.6289,
      "num_input_tokens_seen": 9139064,
      "step": 13985
    },
    {
      "epoch": 7.332285115303983,
      "grad_norm": 6.9375,
      "learning_rate": 1.0082043501579205e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 9141624,
      "step": 13990
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 6.15625,
      "learning_rate": 1.0063698842835082e-05,
      "loss": 0.407,
      "num_input_tokens_seen": 9144120,
      "step": 13995
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 8.4375,
      "learning_rate": 1.0045366682105511e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 9147384,
      "step": 14000
    },
    {
      "epoch": 7.340146750524109,
      "grad_norm": 4.4375,
      "learning_rate": 1.002704703473e-05,
      "loss": 0.4585,
      "num_input_tokens_seen": 9150744,
      "step": 14005
    },
    {
      "epoch": 7.3427672955974845,
      "grad_norm": 4.0625,
      "learning_rate": 1.0008739916037585e-05,
      "loss": 0.4879,
      "num_input_tokens_seen": 9153560,
      "step": 14010
    },
    {
      "epoch": 7.34538784067086,
      "grad_norm": 6.28125,
      "learning_rate": 9.990445341346846e-06,
      "loss": 0.3501,
      "num_input_tokens_seen": 9156696,
      "step": 14015
    },
    {
      "epoch": 7.348008385744235,
      "grad_norm": 4.9375,
      "learning_rate": 9.972163325965833e-06,
      "loss": 0.4117,
      "num_input_tokens_seen": 9159416,
      "step": 14020
    },
    {
      "epoch": 7.35062893081761,
      "grad_norm": 5.8125,
      "learning_rate": 9.953893885192097e-06,
      "loss": 0.3824,
      "num_input_tokens_seen": 9162584,
      "step": 14025
    },
    {
      "epoch": 7.353249475890985,
      "grad_norm": 6.78125,
      "learning_rate": 9.93563703431269e-06,
      "loss": 0.4515,
      "num_input_tokens_seen": 9165848,
      "step": 14030
    },
    {
      "epoch": 7.35587002096436,
      "grad_norm": 7.59375,
      "learning_rate": 9.917392788604097e-06,
      "loss": 0.4324,
      "num_input_tokens_seen": 9168600,
      "step": 14035
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 3.21875,
      "learning_rate": 9.899161163332274e-06,
      "loss": 0.3929,
      "num_input_tokens_seen": 9173336,
      "step": 14040
    },
    {
      "epoch": 7.361111111111111,
      "grad_norm": 8.0625,
      "learning_rate": 9.880942173752602e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 9176376,
      "step": 14045
    },
    {
      "epoch": 7.363731656184487,
      "grad_norm": 5.46875,
      "learning_rate": 9.862735835109915e-06,
      "loss": 0.5112,
      "num_input_tokens_seen": 9180088,
      "step": 14050
    },
    {
      "epoch": 7.366352201257862,
      "grad_norm": 12.0,
      "learning_rate": 9.844542162638442e-06,
      "loss": 0.4405,
      "num_input_tokens_seen": 9183064,
      "step": 14055
    },
    {
      "epoch": 7.368972746331237,
      "grad_norm": 9.9375,
      "learning_rate": 9.826361171561802e-06,
      "loss": 0.4667,
      "num_input_tokens_seen": 9186040,
      "step": 14060
    },
    {
      "epoch": 7.371593291404612,
      "grad_norm": 5.5625,
      "learning_rate": 9.808192877093039e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 9189080,
      "step": 14065
    },
    {
      "epoch": 7.3742138364779874,
      "grad_norm": 3.625,
      "learning_rate": 9.790037294434545e-06,
      "loss": 0.5924,
      "num_input_tokens_seen": 9191992,
      "step": 14070
    },
    {
      "epoch": 7.376834381551363,
      "grad_norm": 8.875,
      "learning_rate": 9.771894438778075e-06,
      "loss": 0.4965,
      "num_input_tokens_seen": 9194520,
      "step": 14075
    },
    {
      "epoch": 7.379454926624738,
      "grad_norm": 9.3125,
      "learning_rate": 9.753764325304751e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 9196632,
      "step": 14080
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 10.375,
      "learning_rate": 9.735646969185008e-06,
      "loss": 0.5363,
      "num_input_tokens_seen": 9199832,
      "step": 14085
    },
    {
      "epoch": 7.384696016771488,
      "grad_norm": 6.46875,
      "learning_rate": 9.717542385578645e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 9203384,
      "step": 14090
    },
    {
      "epoch": 7.387316561844864,
      "grad_norm": 12.4375,
      "learning_rate": 9.699450589634736e-06,
      "loss": 0.5451,
      "num_input_tokens_seen": 9206840,
      "step": 14095
    },
    {
      "epoch": 7.389937106918239,
      "grad_norm": 12.125,
      "learning_rate": 9.681371596491665e-06,
      "loss": 0.6486,
      "num_input_tokens_seen": 9209432,
      "step": 14100
    },
    {
      "epoch": 7.3925576519916145,
      "grad_norm": 5.9375,
      "learning_rate": 9.663305421277125e-06,
      "loss": 0.4557,
      "num_input_tokens_seen": 9212056,
      "step": 14105
    },
    {
      "epoch": 7.39517819706499,
      "grad_norm": 4.96875,
      "learning_rate": 9.645252079108055e-06,
      "loss": 0.3875,
      "num_input_tokens_seen": 9215224,
      "step": 14110
    },
    {
      "epoch": 7.397798742138365,
      "grad_norm": 3.125,
      "learning_rate": 9.62721158509066e-06,
      "loss": 0.3925,
      "num_input_tokens_seen": 9219672,
      "step": 14115
    },
    {
      "epoch": 7.40041928721174,
      "grad_norm": 11.9375,
      "learning_rate": 9.609183954320425e-06,
      "loss": 0.4804,
      "num_input_tokens_seen": 9222616,
      "step": 14120
    },
    {
      "epoch": 7.403039832285115,
      "grad_norm": 4.375,
      "learning_rate": 9.59116920188202e-06,
      "loss": 0.728,
      "num_input_tokens_seen": 9227032,
      "step": 14125
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 13.6875,
      "learning_rate": 9.573167342849375e-06,
      "loss": 0.5536,
      "num_input_tokens_seen": 9229688,
      "step": 14130
    },
    {
      "epoch": 7.4082809224318655,
      "grad_norm": 6.875,
      "learning_rate": 9.555178392285647e-06,
      "loss": 0.5218,
      "num_input_tokens_seen": 9233016,
      "step": 14135
    },
    {
      "epoch": 7.410901467505241,
      "grad_norm": 7.25,
      "learning_rate": 9.53720236524313e-06,
      "loss": 0.4596,
      "num_input_tokens_seen": 9236504,
      "step": 14140
    },
    {
      "epoch": 7.413522012578617,
      "grad_norm": 5.96875,
      "learning_rate": 9.519239276763376e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 9240344,
      "step": 14145
    },
    {
      "epoch": 7.416142557651992,
      "grad_norm": 7.09375,
      "learning_rate": 9.501289141877056e-06,
      "loss": 0.466,
      "num_input_tokens_seen": 9242904,
      "step": 14150
    },
    {
      "epoch": 7.418763102725367,
      "grad_norm": 10.625,
      "learning_rate": 9.483351975604025e-06,
      "loss": 0.5978,
      "num_input_tokens_seen": 9246072,
      "step": 14155
    },
    {
      "epoch": 7.421383647798742,
      "grad_norm": 9.75,
      "learning_rate": 9.465427792953293e-06,
      "loss": 0.5193,
      "num_input_tokens_seen": 9248696,
      "step": 14160
    },
    {
      "epoch": 7.424004192872117,
      "grad_norm": 7.375,
      "learning_rate": 9.447516608922996e-06,
      "loss": 0.672,
      "num_input_tokens_seen": 9251160,
      "step": 14165
    },
    {
      "epoch": 7.426624737945493,
      "grad_norm": 6.0625,
      "learning_rate": 9.429618438500381e-06,
      "loss": 0.5251,
      "num_input_tokens_seen": 9254040,
      "step": 14170
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 4.34375,
      "learning_rate": 9.411733296661852e-06,
      "loss": 0.4926,
      "num_input_tokens_seen": 9257816,
      "step": 14175
    },
    {
      "epoch": 7.431865828092243,
      "grad_norm": 9.0625,
      "learning_rate": 9.393861198372836e-06,
      "loss": 0.4178,
      "num_input_tokens_seen": 9260184,
      "step": 14180
    },
    {
      "epoch": 7.434486373165618,
      "grad_norm": 6.5625,
      "learning_rate": 9.376002158587915e-06,
      "loss": 0.5949,
      "num_input_tokens_seen": 9263032,
      "step": 14185
    },
    {
      "epoch": 7.437106918238994,
      "grad_norm": 6.15625,
      "learning_rate": 9.358156192250717e-06,
      "loss": 0.5671,
      "num_input_tokens_seen": 9267448,
      "step": 14190
    },
    {
      "epoch": 7.439727463312369,
      "grad_norm": 7.28125,
      "learning_rate": 9.340323314293917e-06,
      "loss": 0.435,
      "num_input_tokens_seen": 9270200,
      "step": 14195
    },
    {
      "epoch": 7.4423480083857445,
      "grad_norm": 8.375,
      "learning_rate": 9.322503539639269e-06,
      "loss": 0.4541,
      "num_input_tokens_seen": 9272856,
      "step": 14200
    },
    {
      "epoch": 7.44496855345912,
      "grad_norm": 3.859375,
      "learning_rate": 9.304696883197542e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 9276504,
      "step": 14205
    },
    {
      "epoch": 7.447589098532495,
      "grad_norm": 4.6875,
      "learning_rate": 9.286903359868518e-06,
      "loss": 0.5387,
      "num_input_tokens_seen": 9279736,
      "step": 14210
    },
    {
      "epoch": 7.45020964360587,
      "grad_norm": 7.40625,
      "learning_rate": 9.269122984541029e-06,
      "loss": 0.442,
      "num_input_tokens_seen": 9282680,
      "step": 14215
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 8.125,
      "learning_rate": 9.251355772092867e-06,
      "loss": 0.475,
      "num_input_tokens_seen": 9285912,
      "step": 14220
    },
    {
      "epoch": 7.45545073375262,
      "grad_norm": 6.25,
      "learning_rate": 9.233601737390826e-06,
      "loss": 0.3875,
      "num_input_tokens_seen": 9289176,
      "step": 14225
    },
    {
      "epoch": 7.4580712788259955,
      "grad_norm": 6.5625,
      "learning_rate": 9.215860895290662e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 9291736,
      "step": 14230
    },
    {
      "epoch": 7.460691823899371,
      "grad_norm": 4.46875,
      "learning_rate": 9.198133260637121e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 9296312,
      "step": 14235
    },
    {
      "epoch": 7.463312368972747,
      "grad_norm": 9.125,
      "learning_rate": 9.180418848263866e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 9298904,
      "step": 14240
    },
    {
      "epoch": 7.465932914046122,
      "grad_norm": 5.1875,
      "learning_rate": 9.162717672993499e-06,
      "loss": 0.4906,
      "num_input_tokens_seen": 9301432,
      "step": 14245
    },
    {
      "epoch": 7.468553459119497,
      "grad_norm": 6.375,
      "learning_rate": 9.145029749637576e-06,
      "loss": 0.5056,
      "num_input_tokens_seen": 9304376,
      "step": 14250
    },
    {
      "epoch": 7.471174004192872,
      "grad_norm": 8.875,
      "learning_rate": 9.127355092996532e-06,
      "loss": 0.4814,
      "num_input_tokens_seen": 9307864,
      "step": 14255
    },
    {
      "epoch": 7.473794549266247,
      "grad_norm": 5.8125,
      "learning_rate": 9.10969371785971e-06,
      "loss": 0.5368,
      "num_input_tokens_seen": 9311032,
      "step": 14260
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 4.8125,
      "learning_rate": 9.092045639005347e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 9314264,
      "step": 14265
    },
    {
      "epoch": 7.479035639412998,
      "grad_norm": 4.53125,
      "learning_rate": 9.07441087120054e-06,
      "loss": 0.404,
      "num_input_tokens_seen": 9317272,
      "step": 14270
    },
    {
      "epoch": 7.481656184486373,
      "grad_norm": 9.8125,
      "learning_rate": 9.05678942920127e-06,
      "loss": 0.4884,
      "num_input_tokens_seen": 9320504,
      "step": 14275
    },
    {
      "epoch": 7.484276729559748,
      "grad_norm": 8.125,
      "learning_rate": 9.03918132775235e-06,
      "loss": 0.46,
      "num_input_tokens_seen": 9324792,
      "step": 14280
    },
    {
      "epoch": 7.486897274633124,
      "grad_norm": 5.3125,
      "learning_rate": 9.021586581587425e-06,
      "loss": 0.4002,
      "num_input_tokens_seen": 9327544,
      "step": 14285
    },
    {
      "epoch": 7.489517819706499,
      "grad_norm": 6.34375,
      "learning_rate": 9.004005205428992e-06,
      "loss": 0.5383,
      "num_input_tokens_seen": 9330488,
      "step": 14290
    },
    {
      "epoch": 7.4921383647798745,
      "grad_norm": 4.5625,
      "learning_rate": 8.986437213988336e-06,
      "loss": 0.5497,
      "num_input_tokens_seen": 9335288,
      "step": 14295
    },
    {
      "epoch": 7.49475890985325,
      "grad_norm": 3.8125,
      "learning_rate": 8.968882621965542e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 9339384,
      "step": 14300
    },
    {
      "epoch": 7.497379454926625,
      "grad_norm": 5.125,
      "learning_rate": 8.951341444049513e-06,
      "loss": 0.6062,
      "num_input_tokens_seen": 9342872,
      "step": 14305
    },
    {
      "epoch": 7.5,
      "grad_norm": 5.625,
      "learning_rate": 8.933813694917873e-06,
      "loss": 0.4332,
      "num_input_tokens_seen": 9345976,
      "step": 14310
    },
    {
      "epoch": 7.5,
      "eval_loss": 0.49340954422950745,
      "eval_runtime": 24.211,
      "eval_samples_per_second": 35.025,
      "eval_steps_per_second": 8.756,
      "num_input_tokens_seen": 9345976,
      "step": 14310
    },
    {
      "epoch": 7.502620545073375,
      "grad_norm": 7.0625,
      "learning_rate": 8.916299389237067e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 9349208,
      "step": 14315
    },
    {
      "epoch": 7.50524109014675,
      "grad_norm": 5.65625,
      "learning_rate": 8.898798541662259e-06,
      "loss": 0.4597,
      "num_input_tokens_seen": 9352184,
      "step": 14320
    },
    {
      "epoch": 7.5078616352201255,
      "grad_norm": 7.375,
      "learning_rate": 8.88131116683735e-06,
      "loss": 0.5335,
      "num_input_tokens_seen": 9356216,
      "step": 14325
    },
    {
      "epoch": 7.510482180293501,
      "grad_norm": 12.3125,
      "learning_rate": 8.863837279394993e-06,
      "loss": 0.4811,
      "num_input_tokens_seen": 9359544,
      "step": 14330
    },
    {
      "epoch": 7.513102725366876,
      "grad_norm": 4.375,
      "learning_rate": 8.84637689395653e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 9362200,
      "step": 14335
    },
    {
      "epoch": 7.515723270440252,
      "grad_norm": 5.84375,
      "learning_rate": 8.828930025132006e-06,
      "loss": 0.4246,
      "num_input_tokens_seen": 9366520,
      "step": 14340
    },
    {
      "epoch": 7.518343815513627,
      "grad_norm": 4.4375,
      "learning_rate": 8.81149668752018e-06,
      "loss": 0.4085,
      "num_input_tokens_seen": 9372440,
      "step": 14345
    },
    {
      "epoch": 7.520964360587002,
      "grad_norm": 9.125,
      "learning_rate": 8.794076895708463e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 9375064,
      "step": 14350
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 6.40625,
      "learning_rate": 8.776670664272946e-06,
      "loss": 0.3988,
      "num_input_tokens_seen": 9377720,
      "step": 14355
    },
    {
      "epoch": 7.526205450733753,
      "grad_norm": 5.28125,
      "learning_rate": 8.759278007778362e-06,
      "loss": 0.4933,
      "num_input_tokens_seen": 9380984,
      "step": 14360
    },
    {
      "epoch": 7.528825995807128,
      "grad_norm": 7.78125,
      "learning_rate": 8.741898940778088e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 9383992,
      "step": 14365
    },
    {
      "epoch": 7.531446540880503,
      "grad_norm": 10.9375,
      "learning_rate": 8.724533477814148e-06,
      "loss": 0.4852,
      "num_input_tokens_seen": 9386360,
      "step": 14370
    },
    {
      "epoch": 7.534067085953878,
      "grad_norm": 6.9375,
      "learning_rate": 8.707181633417159e-06,
      "loss": 0.5446,
      "num_input_tokens_seen": 9389144,
      "step": 14375
    },
    {
      "epoch": 7.536687631027254,
      "grad_norm": 5.71875,
      "learning_rate": 8.689843422106345e-06,
      "loss": 0.3592,
      "num_input_tokens_seen": 9392024,
      "step": 14380
    },
    {
      "epoch": 7.539308176100629,
      "grad_norm": 9.5,
      "learning_rate": 8.672518858389548e-06,
      "loss": 0.4587,
      "num_input_tokens_seen": 9394872,
      "step": 14385
    },
    {
      "epoch": 7.5419287211740045,
      "grad_norm": 6.71875,
      "learning_rate": 8.655207956763159e-06,
      "loss": 0.6501,
      "num_input_tokens_seen": 9397304,
      "step": 14390
    },
    {
      "epoch": 7.54454926624738,
      "grad_norm": 8.375,
      "learning_rate": 8.63791073171215e-06,
      "loss": 0.5291,
      "num_input_tokens_seen": 9400664,
      "step": 14395
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 3.59375,
      "learning_rate": 8.620627197710044e-06,
      "loss": 0.5289,
      "num_input_tokens_seen": 9403864,
      "step": 14400
    },
    {
      "epoch": 7.54979035639413,
      "grad_norm": 9.0625,
      "learning_rate": 8.603357369218928e-06,
      "loss": 0.4769,
      "num_input_tokens_seen": 9407544,
      "step": 14405
    },
    {
      "epoch": 7.552410901467505,
      "grad_norm": 4.125,
      "learning_rate": 8.586101260689397e-06,
      "loss": 0.3851,
      "num_input_tokens_seen": 9410968,
      "step": 14410
    },
    {
      "epoch": 7.55503144654088,
      "grad_norm": 13.875,
      "learning_rate": 8.568858886560563e-06,
      "loss": 0.5452,
      "num_input_tokens_seen": 9413400,
      "step": 14415
    },
    {
      "epoch": 7.5576519916142555,
      "grad_norm": 13.4375,
      "learning_rate": 8.551630261260079e-06,
      "loss": 0.3896,
      "num_input_tokens_seen": 9415832,
      "step": 14420
    },
    {
      "epoch": 7.560272536687631,
      "grad_norm": 7.78125,
      "learning_rate": 8.53441539920406e-06,
      "loss": 0.5747,
      "num_input_tokens_seen": 9419064,
      "step": 14425
    },
    {
      "epoch": 7.562893081761006,
      "grad_norm": 4.96875,
      "learning_rate": 8.517214314797108e-06,
      "loss": 0.5334,
      "num_input_tokens_seen": 9421752,
      "step": 14430
    },
    {
      "epoch": 7.565513626834382,
      "grad_norm": 7.03125,
      "learning_rate": 8.500027022432333e-06,
      "loss": 0.5539,
      "num_input_tokens_seen": 9424984,
      "step": 14435
    },
    {
      "epoch": 7.568134171907757,
      "grad_norm": 5.03125,
      "learning_rate": 8.482853536491239e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 9428216,
      "step": 14440
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 7.5,
      "learning_rate": 8.465693871343842e-06,
      "loss": 0.4875,
      "num_input_tokens_seen": 9431576,
      "step": 14445
    },
    {
      "epoch": 7.573375262054507,
      "grad_norm": 11.125,
      "learning_rate": 8.448548041348552e-06,
      "loss": 0.621,
      "num_input_tokens_seen": 9434680,
      "step": 14450
    },
    {
      "epoch": 7.575995807127883,
      "grad_norm": 14.125,
      "learning_rate": 8.431416060852218e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 9437432,
      "step": 14455
    },
    {
      "epoch": 7.578616352201258,
      "grad_norm": 5.46875,
      "learning_rate": 8.414297944190108e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 9441208,
      "step": 14460
    },
    {
      "epoch": 7.581236897274633,
      "grad_norm": 5.625,
      "learning_rate": 8.397193705685873e-06,
      "loss": 0.5009,
      "num_input_tokens_seen": 9444056,
      "step": 14465
    },
    {
      "epoch": 7.583857442348008,
      "grad_norm": 14.1875,
      "learning_rate": 8.380103359651553e-06,
      "loss": 0.6297,
      "num_input_tokens_seen": 9446968,
      "step": 14470
    },
    {
      "epoch": 7.586477987421384,
      "grad_norm": 3.90625,
      "learning_rate": 8.36302692038759e-06,
      "loss": 0.4812,
      "num_input_tokens_seen": 9450712,
      "step": 14475
    },
    {
      "epoch": 7.589098532494759,
      "grad_norm": 5.75,
      "learning_rate": 8.345964402182739e-06,
      "loss": 0.4474,
      "num_input_tokens_seen": 9453240,
      "step": 14480
    },
    {
      "epoch": 7.5917190775681345,
      "grad_norm": 4.65625,
      "learning_rate": 8.328915819314148e-06,
      "loss": 0.4142,
      "num_input_tokens_seen": 9456728,
      "step": 14485
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 6.96875,
      "learning_rate": 8.31188118604731e-06,
      "loss": 0.473,
      "num_input_tokens_seen": 9459800,
      "step": 14490
    },
    {
      "epoch": 7.596960167714885,
      "grad_norm": 5.03125,
      "learning_rate": 8.294860516636e-06,
      "loss": 0.4675,
      "num_input_tokens_seen": 9463384,
      "step": 14495
    },
    {
      "epoch": 7.59958071278826,
      "grad_norm": 6.21875,
      "learning_rate": 8.277853825322355e-06,
      "loss": 0.4881,
      "num_input_tokens_seen": 9466520,
      "step": 14500
    },
    {
      "epoch": 7.602201257861635,
      "grad_norm": 5.9375,
      "learning_rate": 8.260861126336794e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 9469944,
      "step": 14505
    },
    {
      "epoch": 7.60482180293501,
      "grad_norm": 43.0,
      "learning_rate": 8.243882433898018e-06,
      "loss": 0.5544,
      "num_input_tokens_seen": 9472280,
      "step": 14510
    },
    {
      "epoch": 7.6074423480083855,
      "grad_norm": 9.25,
      "learning_rate": 8.226917762213044e-06,
      "loss": 0.488,
      "num_input_tokens_seen": 9474808,
      "step": 14515
    },
    {
      "epoch": 7.610062893081761,
      "grad_norm": 3.84375,
      "learning_rate": 8.209967125477119e-06,
      "loss": 0.4865,
      "num_input_tokens_seen": 9479064,
      "step": 14520
    },
    {
      "epoch": 7.612683438155136,
      "grad_norm": 7.03125,
      "learning_rate": 8.193030537873761e-06,
      "loss": 0.5111,
      "num_input_tokens_seen": 9483128,
      "step": 14525
    },
    {
      "epoch": 7.615303983228512,
      "grad_norm": 9.375,
      "learning_rate": 8.176108013574743e-06,
      "loss": 0.4681,
      "num_input_tokens_seen": 9486136,
      "step": 14530
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 14.875,
      "learning_rate": 8.159199566740055e-06,
      "loss": 0.8264,
      "num_input_tokens_seen": 9488568,
      "step": 14535
    },
    {
      "epoch": 7.620545073375262,
      "grad_norm": 9.625,
      "learning_rate": 8.142305211517914e-06,
      "loss": 0.4946,
      "num_input_tokens_seen": 9492440,
      "step": 14540
    },
    {
      "epoch": 7.623165618448637,
      "grad_norm": 5.9375,
      "learning_rate": 8.125424962044742e-06,
      "loss": 0.437,
      "num_input_tokens_seen": 9495192,
      "step": 14545
    },
    {
      "epoch": 7.6257861635220126,
      "grad_norm": 14.0625,
      "learning_rate": 8.108558832445157e-06,
      "loss": 0.5105,
      "num_input_tokens_seen": 9497688,
      "step": 14550
    },
    {
      "epoch": 7.628406708595388,
      "grad_norm": 4.625,
      "learning_rate": 8.091706836831974e-06,
      "loss": 0.4795,
      "num_input_tokens_seen": 9500984,
      "step": 14555
    },
    {
      "epoch": 7.631027253668763,
      "grad_norm": 5.34375,
      "learning_rate": 8.074868989306173e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 9504152,
      "step": 14560
    },
    {
      "epoch": 7.633647798742138,
      "grad_norm": 9.0625,
      "learning_rate": 8.058045303956885e-06,
      "loss": 0.4915,
      "num_input_tokens_seen": 9506840,
      "step": 14565
    },
    {
      "epoch": 7.636268343815514,
      "grad_norm": 3.09375,
      "learning_rate": 8.041235794861416e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 9509880,
      "step": 14570
    },
    {
      "epoch": 7.638888888888889,
      "grad_norm": 7.0,
      "learning_rate": 8.024440476085188e-06,
      "loss": 0.4663,
      "num_input_tokens_seen": 9513464,
      "step": 14575
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 14.1875,
      "learning_rate": 8.007659361681758e-06,
      "loss": 0.5352,
      "num_input_tokens_seen": 9516216,
      "step": 14580
    },
    {
      "epoch": 7.64412997903564,
      "grad_norm": 4.625,
      "learning_rate": 7.990892465692787e-06,
      "loss": 0.3943,
      "num_input_tokens_seen": 9519032,
      "step": 14585
    },
    {
      "epoch": 7.646750524109015,
      "grad_norm": 4.375,
      "learning_rate": 7.974139802148065e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 9523640,
      "step": 14590
    },
    {
      "epoch": 7.64937106918239,
      "grad_norm": 6.5,
      "learning_rate": 7.957401385065444e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 9526616,
      "step": 14595
    },
    {
      "epoch": 7.651991614255765,
      "grad_norm": 5.28125,
      "learning_rate": 7.94067722845086e-06,
      "loss": 0.5146,
      "num_input_tokens_seen": 9529656,
      "step": 14600
    },
    {
      "epoch": 7.65461215932914,
      "grad_norm": 3.84375,
      "learning_rate": 7.923967346298345e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 9532984,
      "step": 14605
    },
    {
      "epoch": 7.6572327044025155,
      "grad_norm": 12.5,
      "learning_rate": 7.907271752589951e-06,
      "loss": 0.4537,
      "num_input_tokens_seen": 9535928,
      "step": 14610
    },
    {
      "epoch": 7.659853249475891,
      "grad_norm": 5.875,
      "learning_rate": 7.89059046129578e-06,
      "loss": 0.4959,
      "num_input_tokens_seen": 9539256,
      "step": 14615
    },
    {
      "epoch": 7.662473794549266,
      "grad_norm": 8.25,
      "learning_rate": 7.873923486374001e-06,
      "loss": 0.47,
      "num_input_tokens_seen": 9542296,
      "step": 14620
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 6.5625,
      "learning_rate": 7.857270841770745e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 9545368,
      "step": 14625
    },
    {
      "epoch": 7.667714884696017,
      "grad_norm": 4.96875,
      "learning_rate": 7.8406325414202e-06,
      "loss": 0.4533,
      "num_input_tokens_seen": 9548472,
      "step": 14630
    },
    {
      "epoch": 7.670335429769392,
      "grad_norm": 3.359375,
      "learning_rate": 7.824008599244553e-06,
      "loss": 0.4059,
      "num_input_tokens_seen": 9552248,
      "step": 14635
    },
    {
      "epoch": 7.672955974842767,
      "grad_norm": 6.21875,
      "learning_rate": 7.807399029153925e-06,
      "loss": 0.727,
      "num_input_tokens_seen": 9555320,
      "step": 14640
    },
    {
      "epoch": 7.6755765199161425,
      "grad_norm": 7.125,
      "learning_rate": 7.790803845046474e-06,
      "loss": 0.4863,
      "num_input_tokens_seen": 9558904,
      "step": 14645
    },
    {
      "epoch": 7.678197064989518,
      "grad_norm": 12.5625,
      "learning_rate": 7.774223060808277e-06,
      "loss": 0.5635,
      "num_input_tokens_seen": 9561624,
      "step": 14650
    },
    {
      "epoch": 7.680817610062893,
      "grad_norm": 8.9375,
      "learning_rate": 7.757656690313375e-06,
      "loss": 0.6298,
      "num_input_tokens_seen": 9564568,
      "step": 14655
    },
    {
      "epoch": 7.683438155136268,
      "grad_norm": 5.25,
      "learning_rate": 7.741104747423769e-06,
      "loss": 0.5601,
      "num_input_tokens_seen": 9567704,
      "step": 14660
    },
    {
      "epoch": 7.686058700209644,
      "grad_norm": 5.28125,
      "learning_rate": 7.724567245989342e-06,
      "loss": 0.4824,
      "num_input_tokens_seen": 9571512,
      "step": 14665
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 7.4375,
      "learning_rate": 7.708044199847934e-06,
      "loss": 0.561,
      "num_input_tokens_seen": 9575448,
      "step": 14670
    },
    {
      "epoch": 7.691299790356394,
      "grad_norm": 4.34375,
      "learning_rate": 7.691535622825288e-06,
      "loss": 0.2904,
      "num_input_tokens_seen": 9583704,
      "step": 14675
    },
    {
      "epoch": 7.69392033542977,
      "grad_norm": 6.6875,
      "learning_rate": 7.675041528735e-06,
      "loss": 0.4628,
      "num_input_tokens_seen": 9586392,
      "step": 14680
    },
    {
      "epoch": 7.696540880503145,
      "grad_norm": 6.0,
      "learning_rate": 7.658561931378594e-06,
      "loss": 0.6237,
      "num_input_tokens_seen": 9589336,
      "step": 14685
    },
    {
      "epoch": 7.69916142557652,
      "grad_norm": 4.625,
      "learning_rate": 7.64209684454544e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 9593016,
      "step": 14690
    },
    {
      "epoch": 7.701781970649895,
      "grad_norm": 14.5625,
      "learning_rate": 7.625646282012763e-06,
      "loss": 0.4494,
      "num_input_tokens_seen": 9595512,
      "step": 14695
    },
    {
      "epoch": 7.70440251572327,
      "grad_norm": 7.40625,
      "learning_rate": 7.6092102575456546e-06,
      "loss": 0.475,
      "num_input_tokens_seen": 9598904,
      "step": 14700
    },
    {
      "epoch": 7.7070230607966455,
      "grad_norm": 6.0625,
      "learning_rate": 7.592788784897023e-06,
      "loss": 0.3123,
      "num_input_tokens_seen": 9602712,
      "step": 14705
    },
    {
      "epoch": 7.709643605870021,
      "grad_norm": 4.875,
      "learning_rate": 7.576381877807598e-06,
      "loss": 0.4979,
      "num_input_tokens_seen": 9605560,
      "step": 14710
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 5.4375,
      "learning_rate": 7.559989550005947e-06,
      "loss": 0.6134,
      "num_input_tokens_seen": 9609048,
      "step": 14715
    },
    {
      "epoch": 7.714884696016772,
      "grad_norm": 7.0,
      "learning_rate": 7.543611815208415e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 9612824,
      "step": 14720
    },
    {
      "epoch": 7.717505241090147,
      "grad_norm": 3.03125,
      "learning_rate": 7.5272486871191375e-06,
      "loss": 0.4229,
      "num_input_tokens_seen": 9615704,
      "step": 14725
    },
    {
      "epoch": 7.720125786163522,
      "grad_norm": 6.40625,
      "learning_rate": 7.510900179430036e-06,
      "loss": 0.5519,
      "num_input_tokens_seen": 9618840,
      "step": 14730
    },
    {
      "epoch": 7.722746331236897,
      "grad_norm": 4.53125,
      "learning_rate": 7.494566305820788e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 9622552,
      "step": 14735
    },
    {
      "epoch": 7.7253668763102725,
      "grad_norm": 3.734375,
      "learning_rate": 7.478247079958845e-06,
      "loss": 0.6236,
      "num_input_tokens_seen": 9625720,
      "step": 14740
    },
    {
      "epoch": 7.727987421383648,
      "grad_norm": 6.8125,
      "learning_rate": 7.461942515499384e-06,
      "loss": 0.4291,
      "num_input_tokens_seen": 9628664,
      "step": 14745
    },
    {
      "epoch": 7.730607966457023,
      "grad_norm": 4.1875,
      "learning_rate": 7.445652626085312e-06,
      "loss": 0.4285,
      "num_input_tokens_seen": 9632088,
      "step": 14750
    },
    {
      "epoch": 7.733228511530398,
      "grad_norm": 5.59375,
      "learning_rate": 7.429377425347281e-06,
      "loss": 0.5203,
      "num_input_tokens_seen": 9635160,
      "step": 14755
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 4.84375,
      "learning_rate": 7.413116926903624e-06,
      "loss": 0.4925,
      "num_input_tokens_seen": 9638776,
      "step": 14760
    },
    {
      "epoch": 7.738469601677149,
      "grad_norm": 7.53125,
      "learning_rate": 7.396871144360387e-06,
      "loss": 0.5128,
      "num_input_tokens_seen": 9641624,
      "step": 14765
    },
    {
      "epoch": 7.741090146750524,
      "grad_norm": 5.21875,
      "learning_rate": 7.380640091311291e-06,
      "loss": 0.551,
      "num_input_tokens_seen": 9649176,
      "step": 14770
    },
    {
      "epoch": 7.7437106918239,
      "grad_norm": 6.09375,
      "learning_rate": 7.3644237813377535e-06,
      "loss": 0.6899,
      "num_input_tokens_seen": 9651768,
      "step": 14775
    },
    {
      "epoch": 7.746331236897275,
      "grad_norm": 5.28125,
      "learning_rate": 7.348222228008836e-06,
      "loss": 0.3188,
      "num_input_tokens_seen": 9654552,
      "step": 14780
    },
    {
      "epoch": 7.74895178197065,
      "grad_norm": 5.6875,
      "learning_rate": 7.332035444881247e-06,
      "loss": 0.4722,
      "num_input_tokens_seen": 9657112,
      "step": 14785
    },
    {
      "epoch": 7.751572327044025,
      "grad_norm": 3.765625,
      "learning_rate": 7.315863445499366e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 9661176,
      "step": 14790
    },
    {
      "epoch": 7.7541928721174,
      "grad_norm": 5.34375,
      "learning_rate": 7.299706243395177e-06,
      "loss": 0.4578,
      "num_input_tokens_seen": 9663480,
      "step": 14795
    },
    {
      "epoch": 7.756813417190775,
      "grad_norm": 4.15625,
      "learning_rate": 7.283563852088277e-06,
      "loss": 0.4721,
      "num_input_tokens_seen": 9666808,
      "step": 14800
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 8.0625,
      "learning_rate": 7.267436285085905e-06,
      "loss": 0.6247,
      "num_input_tokens_seen": 9669336,
      "step": 14805
    },
    {
      "epoch": 7.762054507337526,
      "grad_norm": 5.40625,
      "learning_rate": 7.251323555882844e-06,
      "loss": 0.5878,
      "num_input_tokens_seen": 9671928,
      "step": 14810
    },
    {
      "epoch": 7.764675052410902,
      "grad_norm": 3.359375,
      "learning_rate": 7.235225677961513e-06,
      "loss": 0.5633,
      "num_input_tokens_seen": 9676120,
      "step": 14815
    },
    {
      "epoch": 7.767295597484277,
      "grad_norm": 7.53125,
      "learning_rate": 7.219142664791872e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 9679032,
      "step": 14820
    },
    {
      "epoch": 7.769916142557652,
      "grad_norm": 7.0625,
      "learning_rate": 7.203074529831444e-06,
      "loss": 0.4471,
      "num_input_tokens_seen": 9682744,
      "step": 14825
    },
    {
      "epoch": 7.772536687631027,
      "grad_norm": 5.59375,
      "learning_rate": 7.187021286525328e-06,
      "loss": 0.6804,
      "num_input_tokens_seen": 9686040,
      "step": 14830
    },
    {
      "epoch": 7.7751572327044025,
      "grad_norm": 8.0625,
      "learning_rate": 7.170982948306135e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 9690168,
      "step": 14835
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 7.53125,
      "learning_rate": 7.154959528594002e-06,
      "loss": 0.5275,
      "num_input_tokens_seen": 9693496,
      "step": 14840
    },
    {
      "epoch": 7.780398322851153,
      "grad_norm": 4.59375,
      "learning_rate": 7.138951040796627e-06,
      "loss": 0.4719,
      "num_input_tokens_seen": 9696760,
      "step": 14845
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 5.125,
      "learning_rate": 7.122957498309143e-06,
      "loss": 0.5508,
      "num_input_tokens_seen": 9699672,
      "step": 14850
    },
    {
      "epoch": 7.785639412997903,
      "grad_norm": 4.5625,
      "learning_rate": 7.1069789145142424e-06,
      "loss": 0.4359,
      "num_input_tokens_seen": 9703384,
      "step": 14855
    },
    {
      "epoch": 7.788259958071279,
      "grad_norm": 3.828125,
      "learning_rate": 7.091015302782064e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 9706328,
      "step": 14860
    },
    {
      "epoch": 7.790880503144654,
      "grad_norm": 5.21875,
      "learning_rate": 7.075066676470224e-06,
      "loss": 0.5556,
      "num_input_tokens_seen": 9709432,
      "step": 14865
    },
    {
      "epoch": 7.79350104821803,
      "grad_norm": 11.375,
      "learning_rate": 7.0591330489238185e-06,
      "loss": 0.592,
      "num_input_tokens_seen": 9712696,
      "step": 14870
    },
    {
      "epoch": 7.796121593291405,
      "grad_norm": 6.3125,
      "learning_rate": 7.0432144334753734e-06,
      "loss": 0.5611,
      "num_input_tokens_seen": 9715704,
      "step": 14875
    },
    {
      "epoch": 7.79874213836478,
      "grad_norm": 5.15625,
      "learning_rate": 7.027310843444846e-06,
      "loss": 0.6971,
      "num_input_tokens_seen": 9720344,
      "step": 14880
    },
    {
      "epoch": 7.801362683438155,
      "grad_norm": 4.59375,
      "learning_rate": 7.011422292139655e-06,
      "loss": 0.4305,
      "num_input_tokens_seen": 9723544,
      "step": 14885
    },
    {
      "epoch": 7.80398322851153,
      "grad_norm": 5.375,
      "learning_rate": 6.995548792854606e-06,
      "loss": 0.7219,
      "num_input_tokens_seen": 9726424,
      "step": 14890
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 8.875,
      "learning_rate": 6.979690358871912e-06,
      "loss": 0.4697,
      "num_input_tokens_seen": 9729144,
      "step": 14895
    },
    {
      "epoch": 7.809224318658281,
      "grad_norm": 4.78125,
      "learning_rate": 6.963847003461188e-06,
      "loss": 0.47,
      "num_input_tokens_seen": 9732504,
      "step": 14900
    },
    {
      "epoch": 7.811844863731656,
      "grad_norm": 5.78125,
      "learning_rate": 6.948018739879439e-06,
      "loss": 0.4629,
      "num_input_tokens_seen": 9737048,
      "step": 14905
    },
    {
      "epoch": 7.814465408805032,
      "grad_norm": 9.625,
      "learning_rate": 6.932205581371026e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 9739960,
      "step": 14910
    },
    {
      "epoch": 7.817085953878407,
      "grad_norm": 4.59375,
      "learning_rate": 6.91640754116768e-06,
      "loss": 0.3899,
      "num_input_tokens_seen": 9743288,
      "step": 14915
    },
    {
      "epoch": 7.819706498951782,
      "grad_norm": 9.5,
      "learning_rate": 6.9006246324884695e-06,
      "loss": 0.6985,
      "num_input_tokens_seen": 9746072,
      "step": 14920
    },
    {
      "epoch": 7.822327044025157,
      "grad_norm": 6.6875,
      "learning_rate": 6.88485686853983e-06,
      "loss": 0.547,
      "num_input_tokens_seen": 9748920,
      "step": 14925
    },
    {
      "epoch": 7.8249475890985325,
      "grad_norm": 6.15625,
      "learning_rate": 6.869104262515494e-06,
      "loss": 0.359,
      "num_input_tokens_seen": 9751864,
      "step": 14930
    },
    {
      "epoch": 7.827568134171908,
      "grad_norm": 6.09375,
      "learning_rate": 6.8533668275965276e-06,
      "loss": 0.4457,
      "num_input_tokens_seen": 9755128,
      "step": 14935
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 12.0,
      "learning_rate": 6.837644576951283e-06,
      "loss": 0.3942,
      "num_input_tokens_seen": 9757976,
      "step": 14940
    },
    {
      "epoch": 7.832809224318658,
      "grad_norm": 4.5,
      "learning_rate": 6.8219375237354445e-06,
      "loss": 0.4213,
      "num_input_tokens_seen": 9761464,
      "step": 14945
    },
    {
      "epoch": 7.835429769392033,
      "grad_norm": 7.625,
      "learning_rate": 6.806245681091944e-06,
      "loss": 0.4939,
      "num_input_tokens_seen": 9768408,
      "step": 14950
    },
    {
      "epoch": 7.838050314465409,
      "grad_norm": 5.25,
      "learning_rate": 6.790569062150992e-06,
      "loss": 0.5363,
      "num_input_tokens_seen": 9771832,
      "step": 14955
    },
    {
      "epoch": 7.840670859538784,
      "grad_norm": 4.4375,
      "learning_rate": 6.774907680030085e-06,
      "loss": 0.5894,
      "num_input_tokens_seen": 9775736,
      "step": 14960
    },
    {
      "epoch": 7.84329140461216,
      "grad_norm": 4.4375,
      "learning_rate": 6.75926154783394e-06,
      "loss": 0.5419,
      "num_input_tokens_seen": 9779096,
      "step": 14965
    },
    {
      "epoch": 7.845911949685535,
      "grad_norm": 9.3125,
      "learning_rate": 6.743630678654519e-06,
      "loss": 0.7524,
      "num_input_tokens_seen": 9782264,
      "step": 14970
    },
    {
      "epoch": 7.84853249475891,
      "grad_norm": 5.15625,
      "learning_rate": 6.728015085571049e-06,
      "loss": 0.4919,
      "num_input_tokens_seen": 9785368,
      "step": 14975
    },
    {
      "epoch": 7.851153039832285,
      "grad_norm": 7.5625,
      "learning_rate": 6.712414781649906e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 9789304,
      "step": 14980
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 7.71875,
      "learning_rate": 6.69682977994473e-06,
      "loss": 0.6101,
      "num_input_tokens_seen": 9792184,
      "step": 14985
    },
    {
      "epoch": 7.856394129979035,
      "grad_norm": 6.40625,
      "learning_rate": 6.681260093496355e-06,
      "loss": 0.5703,
      "num_input_tokens_seen": 9795928,
      "step": 14990
    },
    {
      "epoch": 7.859014675052411,
      "grad_norm": 3.6875,
      "learning_rate": 6.665705735332753e-06,
      "loss": 0.5098,
      "num_input_tokens_seen": 9798776,
      "step": 14995
    },
    {
      "epoch": 7.861635220125786,
      "grad_norm": 6.90625,
      "learning_rate": 6.65016671846912e-06,
      "loss": 0.408,
      "num_input_tokens_seen": 9802008,
      "step": 15000
    },
    {
      "epoch": 7.864255765199162,
      "grad_norm": 15.6875,
      "learning_rate": 6.634643055907791e-06,
      "loss": 0.6231,
      "num_input_tokens_seen": 9805912,
      "step": 15005
    },
    {
      "epoch": 7.866876310272537,
      "grad_norm": 6.4375,
      "learning_rate": 6.619134760638248e-06,
      "loss": 0.4838,
      "num_input_tokens_seen": 9808664,
      "step": 15010
    },
    {
      "epoch": 7.869496855345912,
      "grad_norm": 4.84375,
      "learning_rate": 6.6036418456371516e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 9811256,
      "step": 15015
    },
    {
      "epoch": 7.872117400419287,
      "grad_norm": 4.75,
      "learning_rate": 6.588164323868229e-06,
      "loss": 0.5455,
      "num_input_tokens_seen": 9816440,
      "step": 15020
    },
    {
      "epoch": 7.8747379454926625,
      "grad_norm": 6.59375,
      "learning_rate": 6.572702208282381e-06,
      "loss": 0.4663,
      "num_input_tokens_seen": 9818776,
      "step": 15025
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 4.46875,
      "learning_rate": 6.557255511817617e-06,
      "loss": 0.5156,
      "num_input_tokens_seen": 9822520,
      "step": 15030
    },
    {
      "epoch": 7.879979035639413,
      "grad_norm": 4.6875,
      "learning_rate": 6.5418242473989925e-06,
      "loss": 0.386,
      "num_input_tokens_seen": 9825496,
      "step": 15035
    },
    {
      "epoch": 7.882599580712788,
      "grad_norm": 4.375,
      "learning_rate": 6.52640842793871e-06,
      "loss": 0.6457,
      "num_input_tokens_seen": 9829080,
      "step": 15040
    },
    {
      "epoch": 7.885220125786163,
      "grad_norm": 7.03125,
      "learning_rate": 6.5110080663360165e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 9832056,
      "step": 15045
    },
    {
      "epoch": 7.887840670859539,
      "grad_norm": 8.125,
      "learning_rate": 6.495623175477223e-06,
      "loss": 0.5032,
      "num_input_tokens_seen": 9834680,
      "step": 15050
    },
    {
      "epoch": 7.890461215932914,
      "grad_norm": 3.90625,
      "learning_rate": 6.480253768235714e-06,
      "loss": 0.5683,
      "num_input_tokens_seen": 9838008,
      "step": 15055
    },
    {
      "epoch": 7.8930817610062896,
      "grad_norm": 5.75,
      "learning_rate": 6.464899857471907e-06,
      "loss": 0.588,
      "num_input_tokens_seen": 9841528,
      "step": 15060
    },
    {
      "epoch": 7.895702306079665,
      "grad_norm": 15.75,
      "learning_rate": 6.44956145603324e-06,
      "loss": 0.6266,
      "num_input_tokens_seen": 9844984,
      "step": 15065
    },
    {
      "epoch": 7.89832285115304,
      "grad_norm": 2.5625,
      "learning_rate": 6.4342385767542036e-06,
      "loss": 0.4966,
      "num_input_tokens_seen": 9848568,
      "step": 15070
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 6.3125,
      "learning_rate": 6.418931232456279e-06,
      "loss": 0.3293,
      "num_input_tokens_seen": 9851640,
      "step": 15075
    },
    {
      "epoch": 7.90356394129979,
      "grad_norm": 8.3125,
      "learning_rate": 6.403639435947948e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 9855192,
      "step": 15080
    },
    {
      "epoch": 7.906184486373165,
      "grad_norm": 3.8125,
      "learning_rate": 6.38836320002468e-06,
      "loss": 0.4812,
      "num_input_tokens_seen": 9858072,
      "step": 15085
    },
    {
      "epoch": 7.908805031446541,
      "grad_norm": 5.1875,
      "learning_rate": 6.37310253746895e-06,
      "loss": 0.4535,
      "num_input_tokens_seen": 9861784,
      "step": 15090
    },
    {
      "epoch": 7.911425576519916,
      "grad_norm": 7.53125,
      "learning_rate": 6.357857461050176e-06,
      "loss": 0.5354,
      "num_input_tokens_seen": 9866104,
      "step": 15095
    },
    {
      "epoch": 7.914046121593291,
      "grad_norm": 7.40625,
      "learning_rate": 6.342627983524737e-06,
      "loss": 0.624,
      "num_input_tokens_seen": 9869816,
      "step": 15100
    },
    {
      "epoch": 7.916666666666667,
      "grad_norm": 5.15625,
      "learning_rate": 6.32741411763596e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 9872504,
      "step": 15105
    },
    {
      "epoch": 7.919287211740042,
      "grad_norm": 4.0,
      "learning_rate": 6.312215876114127e-06,
      "loss": 0.4944,
      "num_input_tokens_seen": 9875320,
      "step": 15110
    },
    {
      "epoch": 7.921907756813417,
      "grad_norm": 7.0,
      "learning_rate": 6.297033271676425e-06,
      "loss": 0.4019,
      "num_input_tokens_seen": 9878136,
      "step": 15115
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 7.96875,
      "learning_rate": 6.281866317026966e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 9880792,
      "step": 15120
    },
    {
      "epoch": 7.927148846960168,
      "grad_norm": 5.1875,
      "learning_rate": 6.2667150248567534e-06,
      "loss": 0.4125,
      "num_input_tokens_seen": 9884312,
      "step": 15125
    },
    {
      "epoch": 7.929769392033543,
      "grad_norm": 4.5625,
      "learning_rate": 6.251579407843713e-06,
      "loss": 0.4664,
      "num_input_tokens_seen": 9887416,
      "step": 15130
    },
    {
      "epoch": 7.932389937106918,
      "grad_norm": 6.0,
      "learning_rate": 6.236459478652629e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 9890200,
      "step": 15135
    },
    {
      "epoch": 7.935010482180293,
      "grad_norm": 8.1875,
      "learning_rate": 6.221355249935165e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 9893368,
      "step": 15140
    },
    {
      "epoch": 7.937631027253669,
      "grad_norm": 6.8125,
      "learning_rate": 6.20626673432986e-06,
      "loss": 0.5137,
      "num_input_tokens_seen": 9899224,
      "step": 15145
    },
    {
      "epoch": 7.940251572327044,
      "grad_norm": 6.71875,
      "learning_rate": 6.191193944462087e-06,
      "loss": 0.5196,
      "num_input_tokens_seen": 9901944,
      "step": 15150
    },
    {
      "epoch": 7.9428721174004195,
      "grad_norm": 3.625,
      "learning_rate": 6.176136892944062e-06,
      "loss": 0.5884,
      "num_input_tokens_seen": 9905720,
      "step": 15155
    },
    {
      "epoch": 7.945492662473795,
      "grad_norm": 5.6875,
      "learning_rate": 6.161095592374863e-06,
      "loss": 0.6259,
      "num_input_tokens_seen": 9908440,
      "step": 15160
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 8.6875,
      "learning_rate": 6.1460700553403275e-06,
      "loss": 0.5357,
      "num_input_tokens_seen": 9911256,
      "step": 15165
    },
    {
      "epoch": 7.950733752620545,
      "grad_norm": 7.28125,
      "learning_rate": 6.1310602944131655e-06,
      "loss": 0.5241,
      "num_input_tokens_seen": 9913336,
      "step": 15170
    },
    {
      "epoch": 7.95335429769392,
      "grad_norm": 3.5625,
      "learning_rate": 6.11606632215285e-06,
      "loss": 0.3854,
      "num_input_tokens_seen": 9917048,
      "step": 15175
    },
    {
      "epoch": 7.955974842767295,
      "grad_norm": 8.25,
      "learning_rate": 6.101088151105647e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 9920952,
      "step": 15180
    },
    {
      "epoch": 7.9585953878406706,
      "grad_norm": 7.9375,
      "learning_rate": 6.086125793804618e-06,
      "loss": 0.433,
      "num_input_tokens_seen": 9924376,
      "step": 15185
    },
    {
      "epoch": 7.961215932914046,
      "grad_norm": 4.0625,
      "learning_rate": 6.071179262769572e-06,
      "loss": 0.5971,
      "num_input_tokens_seen": 9927160,
      "step": 15190
    },
    {
      "epoch": 7.963836477987421,
      "grad_norm": 6.125,
      "learning_rate": 6.056248570507078e-06,
      "loss": 0.5541,
      "num_input_tokens_seen": 9930840,
      "step": 15195
    },
    {
      "epoch": 7.966457023060797,
      "grad_norm": 4.96875,
      "learning_rate": 6.041333729510479e-06,
      "loss": 0.6827,
      "num_input_tokens_seen": 9934104,
      "step": 15200
    },
    {
      "epoch": 7.969077568134172,
      "grad_norm": 8.5,
      "learning_rate": 6.026434752259802e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 9937016,
      "step": 15205
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 17.0,
      "learning_rate": 6.011551651221856e-06,
      "loss": 0.5084,
      "num_input_tokens_seen": 9939736,
      "step": 15210
    },
    {
      "epoch": 7.9743186582809225,
      "grad_norm": 6.625,
      "learning_rate": 5.996684438850131e-06,
      "loss": 0.5584,
      "num_input_tokens_seen": 9942744,
      "step": 15215
    },
    {
      "epoch": 7.976939203354298,
      "grad_norm": 11.4375,
      "learning_rate": 5.981833127584824e-06,
      "loss": 0.6307,
      "num_input_tokens_seen": 9946776,
      "step": 15220
    },
    {
      "epoch": 7.979559748427673,
      "grad_norm": 7.28125,
      "learning_rate": 5.966997729852844e-06,
      "loss": 0.5245,
      "num_input_tokens_seen": 9949304,
      "step": 15225
    },
    {
      "epoch": 7.982180293501048,
      "grad_norm": 3.25,
      "learning_rate": 5.952178258067775e-06,
      "loss": 0.4846,
      "num_input_tokens_seen": 9953112,
      "step": 15230
    },
    {
      "epoch": 7.984800838574423,
      "grad_norm": 5.65625,
      "learning_rate": 5.93737472462986e-06,
      "loss": 0.4956,
      "num_input_tokens_seen": 9956056,
      "step": 15235
    },
    {
      "epoch": 7.987421383647799,
      "grad_norm": 8.6875,
      "learning_rate": 5.92258714192604e-06,
      "loss": 0.3806,
      "num_input_tokens_seen": 9959416,
      "step": 15240
    },
    {
      "epoch": 7.990041928721174,
      "grad_norm": 4.625,
      "learning_rate": 5.907815522329877e-06,
      "loss": 0.4192,
      "num_input_tokens_seen": 9962680,
      "step": 15245
    },
    {
      "epoch": 7.9926624737945495,
      "grad_norm": 4.5625,
      "learning_rate": 5.893059878201587e-06,
      "loss": 0.6038,
      "num_input_tokens_seen": 9966616,
      "step": 15250
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 3.984375,
      "learning_rate": 5.878320221888015e-06,
      "loss": 0.465,
      "num_input_tokens_seen": 9971864,
      "step": 15255
    },
    {
      "epoch": 7.9979035639413,
      "grad_norm": 7.03125,
      "learning_rate": 5.8635965657226455e-06,
      "loss": 0.7832,
      "num_input_tokens_seen": 9975864,
      "step": 15260
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.4935668706893921,
      "eval_runtime": 24.3546,
      "eval_samples_per_second": 34.819,
      "eval_steps_per_second": 8.705,
      "num_input_tokens_seen": 9977520,
      "step": 15264
    },
    {
      "epoch": 8.000524109014675,
      "grad_norm": 5.4375,
      "learning_rate": 5.848888922025553e-06,
      "loss": 0.5627,
      "num_input_tokens_seen": 9978032,
      "step": 15265
    },
    {
      "epoch": 8.00314465408805,
      "grad_norm": 7.0625,
      "learning_rate": 5.834197303103414e-06,
      "loss": 0.6575,
      "num_input_tokens_seen": 9981616,
      "step": 15270
    },
    {
      "epoch": 8.005765199161425,
      "grad_norm": 8.0,
      "learning_rate": 5.819521721249524e-06,
      "loss": 0.4174,
      "num_input_tokens_seen": 9984624,
      "step": 15275
    },
    {
      "epoch": 8.0083857442348,
      "grad_norm": 6.03125,
      "learning_rate": 5.804862188743726e-06,
      "loss": 0.4531,
      "num_input_tokens_seen": 9987568,
      "step": 15280
    },
    {
      "epoch": 8.011006289308176,
      "grad_norm": 5.96875,
      "learning_rate": 5.79021871785245e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 9991280,
      "step": 15285
    },
    {
      "epoch": 8.01362683438155,
      "grad_norm": 4.0625,
      "learning_rate": 5.775591320828683e-06,
      "loss": 0.4487,
      "num_input_tokens_seen": 9994288,
      "step": 15290
    },
    {
      "epoch": 8.016247379454926,
      "grad_norm": 9.3125,
      "learning_rate": 5.7609800099119565e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 9997200,
      "step": 15295
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 8.625,
      "learning_rate": 5.746384797328361e-06,
      "loss": 0.6251,
      "num_input_tokens_seen": 10000176,
      "step": 15300
    },
    {
      "epoch": 8.021488469601676,
      "grad_norm": 5.34375,
      "learning_rate": 5.731805695290498e-06,
      "loss": 0.524,
      "num_input_tokens_seen": 10003504,
      "step": 15305
    },
    {
      "epoch": 8.024109014675052,
      "grad_norm": 9.0625,
      "learning_rate": 5.7172427159974865e-06,
      "loss": 0.5057,
      "num_input_tokens_seen": 10008688,
      "step": 15310
    },
    {
      "epoch": 8.026729559748428,
      "grad_norm": 7.0625,
      "learning_rate": 5.702695871634975e-06,
      "loss": 0.6249,
      "num_input_tokens_seen": 10011536,
      "step": 15315
    },
    {
      "epoch": 8.029350104821804,
      "grad_norm": 7.625,
      "learning_rate": 5.688165174375093e-06,
      "loss": 0.4764,
      "num_input_tokens_seen": 10014096,
      "step": 15320
    },
    {
      "epoch": 8.031970649895179,
      "grad_norm": 3.9375,
      "learning_rate": 5.673650636376457e-06,
      "loss": 0.46,
      "num_input_tokens_seen": 10017360,
      "step": 15325
    },
    {
      "epoch": 8.034591194968554,
      "grad_norm": 4.09375,
      "learning_rate": 5.659152269784188e-06,
      "loss": 0.6027,
      "num_input_tokens_seen": 10020560,
      "step": 15330
    },
    {
      "epoch": 8.03721174004193,
      "grad_norm": 10.625,
      "learning_rate": 5.644670086729834e-06,
      "loss": 0.5465,
      "num_input_tokens_seen": 10024176,
      "step": 15335
    },
    {
      "epoch": 8.039832285115304,
      "grad_norm": 6.3125,
      "learning_rate": 5.630204099331432e-06,
      "loss": 0.3692,
      "num_input_tokens_seen": 10028336,
      "step": 15340
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 5.5,
      "learning_rate": 5.615754319693481e-06,
      "loss": 0.4816,
      "num_input_tokens_seen": 10031376,
      "step": 15345
    },
    {
      "epoch": 8.045073375262055,
      "grad_norm": 8.6875,
      "learning_rate": 5.601320759906861e-06,
      "loss": 0.4717,
      "num_input_tokens_seen": 10033904,
      "step": 15350
    },
    {
      "epoch": 8.04769392033543,
      "grad_norm": 5.3125,
      "learning_rate": 5.586903432048943e-06,
      "loss": 0.4878,
      "num_input_tokens_seen": 10036720,
      "step": 15355
    },
    {
      "epoch": 8.050314465408805,
      "grad_norm": 8.125,
      "learning_rate": 5.572502348183475e-06,
      "loss": 0.5515,
      "num_input_tokens_seen": 10040016,
      "step": 15360
    },
    {
      "epoch": 8.05293501048218,
      "grad_norm": 5.25,
      "learning_rate": 5.558117520360623e-06,
      "loss": 0.5406,
      "num_input_tokens_seen": 10043088,
      "step": 15365
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 5.84375,
      "learning_rate": 5.543748960616971e-06,
      "loss": 0.5653,
      "num_input_tokens_seen": 10046160,
      "step": 15370
    },
    {
      "epoch": 8.05817610062893,
      "grad_norm": 5.8125,
      "learning_rate": 5.529396680975457e-06,
      "loss": 0.4971,
      "num_input_tokens_seen": 10048592,
      "step": 15375
    },
    {
      "epoch": 8.060796645702306,
      "grad_norm": 3.0,
      "learning_rate": 5.515060693445418e-06,
      "loss": 0.5073,
      "num_input_tokens_seen": 10051536,
      "step": 15380
    },
    {
      "epoch": 8.06341719077568,
      "grad_norm": 6.1875,
      "learning_rate": 5.500741010022564e-06,
      "loss": 0.5762,
      "num_input_tokens_seen": 10055632,
      "step": 15385
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 4.03125,
      "learning_rate": 5.48643764268893e-06,
      "loss": 0.5447,
      "num_input_tokens_seen": 10060528,
      "step": 15390
    },
    {
      "epoch": 8.068658280922431,
      "grad_norm": 7.75,
      "learning_rate": 5.472150603412937e-06,
      "loss": 0.3722,
      "num_input_tokens_seen": 10063664,
      "step": 15395
    },
    {
      "epoch": 8.071278825995806,
      "grad_norm": 19.0,
      "learning_rate": 5.457879904149327e-06,
      "loss": 0.5126,
      "num_input_tokens_seen": 10066288,
      "step": 15400
    },
    {
      "epoch": 8.073899371069182,
      "grad_norm": 9.9375,
      "learning_rate": 5.4436255568391545e-06,
      "loss": 0.6815,
      "num_input_tokens_seen": 10070064,
      "step": 15405
    },
    {
      "epoch": 8.076519916142558,
      "grad_norm": 8.25,
      "learning_rate": 5.429387573409825e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 10074128,
      "step": 15410
    },
    {
      "epoch": 8.079140461215934,
      "grad_norm": 3.703125,
      "learning_rate": 5.415165965775024e-06,
      "loss": 0.3141,
      "num_input_tokens_seen": 10077936,
      "step": 15415
    },
    {
      "epoch": 8.081761006289309,
      "grad_norm": 6.40625,
      "learning_rate": 5.400960745834735e-06,
      "loss": 0.5336,
      "num_input_tokens_seen": 10080976,
      "step": 15420
    },
    {
      "epoch": 8.084381551362684,
      "grad_norm": 7.0,
      "learning_rate": 5.386771925475256e-06,
      "loss": 0.4878,
      "num_input_tokens_seen": 10083984,
      "step": 15425
    },
    {
      "epoch": 8.08700209643606,
      "grad_norm": 8.25,
      "learning_rate": 5.3725995165691294e-06,
      "loss": 0.636,
      "num_input_tokens_seen": 10086800,
      "step": 15430
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 3.96875,
      "learning_rate": 5.358443530975188e-06,
      "loss": 0.4209,
      "num_input_tokens_seen": 10090544,
      "step": 15435
    },
    {
      "epoch": 8.09224318658281,
      "grad_norm": 16.75,
      "learning_rate": 5.344303980538498e-06,
      "loss": 0.4481,
      "num_input_tokens_seen": 10093360,
      "step": 15440
    },
    {
      "epoch": 8.094863731656185,
      "grad_norm": 6.8125,
      "learning_rate": 5.33018087709041e-06,
      "loss": 0.6286,
      "num_input_tokens_seen": 10097520,
      "step": 15445
    },
    {
      "epoch": 8.09748427672956,
      "grad_norm": 4.65625,
      "learning_rate": 5.316074232448484e-06,
      "loss": 0.4445,
      "num_input_tokens_seen": 10100240,
      "step": 15450
    },
    {
      "epoch": 8.100104821802935,
      "grad_norm": 5.03125,
      "learning_rate": 5.301984058416506e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 10103280,
      "step": 15455
    },
    {
      "epoch": 8.10272536687631,
      "grad_norm": 5.0,
      "learning_rate": 5.2879103667845045e-06,
      "loss": 0.5226,
      "num_input_tokens_seen": 10106288,
      "step": 15460
    },
    {
      "epoch": 8.105345911949685,
      "grad_norm": 5.6875,
      "learning_rate": 5.2738531693286965e-06,
      "loss": 0.4306,
      "num_input_tokens_seen": 10110064,
      "step": 15465
    },
    {
      "epoch": 8.10796645702306,
      "grad_norm": 4.5,
      "learning_rate": 5.2598124778115044e-06,
      "loss": 0.3986,
      "num_input_tokens_seen": 10113360,
      "step": 15470
    },
    {
      "epoch": 8.110587002096436,
      "grad_norm": 2.421875,
      "learning_rate": 5.245788303981533e-06,
      "loss": 0.5399,
      "num_input_tokens_seen": 10119760,
      "step": 15475
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 3.84375,
      "learning_rate": 5.231780659573565e-06,
      "loss": 0.4733,
      "num_input_tokens_seen": 10126864,
      "step": 15480
    },
    {
      "epoch": 8.115828092243186,
      "grad_norm": 6.3125,
      "learning_rate": 5.2177895563085725e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 10129776,
      "step": 15485
    },
    {
      "epoch": 8.118448637316561,
      "grad_norm": 6.0625,
      "learning_rate": 5.203815005893664e-06,
      "loss": 0.5648,
      "num_input_tokens_seen": 10132816,
      "step": 15490
    },
    {
      "epoch": 8.121069182389936,
      "grad_norm": 6.53125,
      "learning_rate": 5.189857020022099e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 10136112,
      "step": 15495
    },
    {
      "epoch": 8.123689727463312,
      "grad_norm": 3.21875,
      "learning_rate": 5.1759156103732946e-06,
      "loss": 0.4806,
      "num_input_tokens_seen": 10138896,
      "step": 15500
    },
    {
      "epoch": 8.126310272536688,
      "grad_norm": 6.09375,
      "learning_rate": 5.161990788612781e-06,
      "loss": 0.4726,
      "num_input_tokens_seen": 10141456,
      "step": 15505
    },
    {
      "epoch": 8.128930817610064,
      "grad_norm": 6.625,
      "learning_rate": 5.148082566392204e-06,
      "loss": 0.5683,
      "num_input_tokens_seen": 10144944,
      "step": 15510
    },
    {
      "epoch": 8.131551362683439,
      "grad_norm": 6.5625,
      "learning_rate": 5.13419095534935e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 10148016,
      "step": 15515
    },
    {
      "epoch": 8.134171907756814,
      "grad_norm": 5.0625,
      "learning_rate": 5.120315967108055e-06,
      "loss": 0.537,
      "num_input_tokens_seen": 10152048,
      "step": 15520
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 4.25,
      "learning_rate": 5.106457613278298e-06,
      "loss": 0.5451,
      "num_input_tokens_seen": 10155600,
      "step": 15525
    },
    {
      "epoch": 8.139412997903564,
      "grad_norm": 7.28125,
      "learning_rate": 5.092615905456111e-06,
      "loss": 0.4977,
      "num_input_tokens_seen": 10158160,
      "step": 15530
    },
    {
      "epoch": 8.14203354297694,
      "grad_norm": 5.0,
      "learning_rate": 5.078790855223595e-06,
      "loss": 0.497,
      "num_input_tokens_seen": 10161296,
      "step": 15535
    },
    {
      "epoch": 8.144654088050315,
      "grad_norm": 4.71875,
      "learning_rate": 5.0649824741489325e-06,
      "loss": 0.5117,
      "num_input_tokens_seen": 10164112,
      "step": 15540
    },
    {
      "epoch": 8.14727463312369,
      "grad_norm": 8.5625,
      "learning_rate": 5.051190773786341e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 10167248,
      "step": 15545
    },
    {
      "epoch": 8.149895178197065,
      "grad_norm": 6.59375,
      "learning_rate": 5.0374157656760786e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 10171088,
      "step": 15550
    },
    {
      "epoch": 8.15251572327044,
      "grad_norm": 3.859375,
      "learning_rate": 5.023657461344456e-06,
      "loss": 0.5578,
      "num_input_tokens_seen": 10174192,
      "step": 15555
    },
    {
      "epoch": 8.155136268343815,
      "grad_norm": 6.28125,
      "learning_rate": 5.009915872303786e-06,
      "loss": 0.5386,
      "num_input_tokens_seen": 10177264,
      "step": 15560
    },
    {
      "epoch": 8.15775681341719,
      "grad_norm": 8.625,
      "learning_rate": 4.996191010052403e-06,
      "loss": 0.6288,
      "num_input_tokens_seen": 10179824,
      "step": 15565
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 4.875,
      "learning_rate": 4.982482886074647e-06,
      "loss": 0.455,
      "num_input_tokens_seen": 10183024,
      "step": 15570
    },
    {
      "epoch": 8.16299790356394,
      "grad_norm": 7.1875,
      "learning_rate": 4.968791511840842e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 10186832,
      "step": 15575
    },
    {
      "epoch": 8.165618448637316,
      "grad_norm": 8.5625,
      "learning_rate": 4.955116898807316e-06,
      "loss": 0.5425,
      "num_input_tokens_seen": 10189360,
      "step": 15580
    },
    {
      "epoch": 8.168238993710691,
      "grad_norm": 3.984375,
      "learning_rate": 4.941459058416356e-06,
      "loss": 0.501,
      "num_input_tokens_seen": 10193360,
      "step": 15585
    },
    {
      "epoch": 8.170859538784066,
      "grad_norm": 7.3125,
      "learning_rate": 4.927818002096213e-06,
      "loss": 0.4861,
      "num_input_tokens_seen": 10197136,
      "step": 15590
    },
    {
      "epoch": 8.173480083857442,
      "grad_norm": 3.671875,
      "learning_rate": 4.9141937412611084e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 10200592,
      "step": 15595
    },
    {
      "epoch": 8.176100628930818,
      "grad_norm": 6.09375,
      "learning_rate": 4.900586287311202e-06,
      "loss": 0.4793,
      "num_input_tokens_seen": 10203088,
      "step": 15600
    },
    {
      "epoch": 8.178721174004194,
      "grad_norm": 5.0625,
      "learning_rate": 4.886995651632584e-06,
      "loss": 0.6239,
      "num_input_tokens_seen": 10205840,
      "step": 15605
    },
    {
      "epoch": 8.181341719077569,
      "grad_norm": 6.71875,
      "learning_rate": 4.873421845597273e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 10209040,
      "step": 15610
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 4.5,
      "learning_rate": 4.859864880563222e-06,
      "loss": 0.4639,
      "num_input_tokens_seen": 10212048,
      "step": 15615
    },
    {
      "epoch": 8.18658280922432,
      "grad_norm": 5.15625,
      "learning_rate": 4.846324767874277e-06,
      "loss": 0.5274,
      "num_input_tokens_seen": 10214896,
      "step": 15620
    },
    {
      "epoch": 8.189203354297694,
      "grad_norm": 9.4375,
      "learning_rate": 4.832801518860175e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 10217872,
      "step": 15625
    },
    {
      "epoch": 8.19182389937107,
      "grad_norm": 5.84375,
      "learning_rate": 4.819295144836566e-06,
      "loss": 0.4832,
      "num_input_tokens_seen": 10221488,
      "step": 15630
    },
    {
      "epoch": 8.194444444444445,
      "grad_norm": 8.6875,
      "learning_rate": 4.805805657104965e-06,
      "loss": 0.3963,
      "num_input_tokens_seen": 10224048,
      "step": 15635
    },
    {
      "epoch": 8.19706498951782,
      "grad_norm": 3.921875,
      "learning_rate": 4.792333066952748e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 10228368,
      "step": 15640
    },
    {
      "epoch": 8.199685534591195,
      "grad_norm": 7.21875,
      "learning_rate": 4.778877385653186e-06,
      "loss": 0.4672,
      "num_input_tokens_seen": 10231728,
      "step": 15645
    },
    {
      "epoch": 8.20230607966457,
      "grad_norm": 6.3125,
      "learning_rate": 4.7654386244653485e-06,
      "loss": 0.4395,
      "num_input_tokens_seen": 10234960,
      "step": 15650
    },
    {
      "epoch": 8.204926624737945,
      "grad_norm": 3.640625,
      "learning_rate": 4.752016794634201e-06,
      "loss": 0.4672,
      "num_input_tokens_seen": 10238448,
      "step": 15655
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 7.8125,
      "learning_rate": 4.738611907390508e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 10240912,
      "step": 15660
    },
    {
      "epoch": 8.210167714884696,
      "grad_norm": 3.96875,
      "learning_rate": 4.725223973950863e-06,
      "loss": 0.4985,
      "num_input_tokens_seen": 10244816,
      "step": 15665
    },
    {
      "epoch": 8.21278825995807,
      "grad_norm": 6.5625,
      "learning_rate": 4.711853005517686e-06,
      "loss": 0.5289,
      "num_input_tokens_seen": 10247728,
      "step": 15670
    },
    {
      "epoch": 8.215408805031446,
      "grad_norm": 4.65625,
      "learning_rate": 4.698499013279189e-06,
      "loss": 0.4314,
      "num_input_tokens_seen": 10251248,
      "step": 15675
    },
    {
      "epoch": 8.218029350104821,
      "grad_norm": 6.5,
      "learning_rate": 4.685162008409374e-06,
      "loss": 0.5002,
      "num_input_tokens_seen": 10254160,
      "step": 15680
    },
    {
      "epoch": 8.220649895178196,
      "grad_norm": 7.8125,
      "learning_rate": 4.671842002068061e-06,
      "loss": 0.4227,
      "num_input_tokens_seen": 10256944,
      "step": 15685
    },
    {
      "epoch": 8.223270440251572,
      "grad_norm": 6.5,
      "learning_rate": 4.658539005400794e-06,
      "loss": 0.535,
      "num_input_tokens_seen": 10259888,
      "step": 15690
    },
    {
      "epoch": 8.225890985324948,
      "grad_norm": 5.46875,
      "learning_rate": 4.645253029538926e-06,
      "loss": 0.613,
      "num_input_tokens_seen": 10262640,
      "step": 15695
    },
    {
      "epoch": 8.228511530398324,
      "grad_norm": 7.78125,
      "learning_rate": 4.631984085599569e-06,
      "loss": 0.5404,
      "num_input_tokens_seen": 10265424,
      "step": 15700
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 4.1875,
      "learning_rate": 4.618732184685542e-06,
      "loss": 0.5171,
      "num_input_tokens_seen": 10267888,
      "step": 15705
    },
    {
      "epoch": 8.233752620545074,
      "grad_norm": 3.515625,
      "learning_rate": 4.60549733788545e-06,
      "loss": 0.3356,
      "num_input_tokens_seen": 10270992,
      "step": 15710
    },
    {
      "epoch": 8.23637316561845,
      "grad_norm": 8.3125,
      "learning_rate": 4.592279556273604e-06,
      "loss": 0.5492,
      "num_input_tokens_seen": 10274128,
      "step": 15715
    },
    {
      "epoch": 8.238993710691824,
      "grad_norm": 7.0,
      "learning_rate": 4.579078850910032e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 10278128,
      "step": 15720
    },
    {
      "epoch": 8.2416142557652,
      "grad_norm": 5.53125,
      "learning_rate": 4.565895232840489e-06,
      "loss": 0.4863,
      "num_input_tokens_seen": 10280720,
      "step": 15725
    },
    {
      "epoch": 8.244234800838575,
      "grad_norm": 5.25,
      "learning_rate": 4.552728713096427e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 10283280,
      "step": 15730
    },
    {
      "epoch": 8.24685534591195,
      "grad_norm": 4.09375,
      "learning_rate": 4.539579302694977e-06,
      "loss": 0.5265,
      "num_input_tokens_seen": 10287600,
      "step": 15735
    },
    {
      "epoch": 8.249475890985325,
      "grad_norm": 8.875,
      "learning_rate": 4.5264470126389765e-06,
      "loss": 0.4439,
      "num_input_tokens_seen": 10289808,
      "step": 15740
    },
    {
      "epoch": 8.2520964360587,
      "grad_norm": 13.9375,
      "learning_rate": 4.5133318539169215e-06,
      "loss": 0.5172,
      "num_input_tokens_seen": 10292656,
      "step": 15745
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 6.53125,
      "learning_rate": 4.500233837502979e-06,
      "loss": 0.4245,
      "num_input_tokens_seen": 10295664,
      "step": 15750
    },
    {
      "epoch": 8.25733752620545,
      "grad_norm": 8.4375,
      "learning_rate": 4.4871529743569675e-06,
      "loss": 0.5278,
      "num_input_tokens_seen": 10299248,
      "step": 15755
    },
    {
      "epoch": 8.259958071278826,
      "grad_norm": 5.5,
      "learning_rate": 4.474089275424351e-06,
      "loss": 0.5295,
      "num_input_tokens_seen": 10302832,
      "step": 15760
    },
    {
      "epoch": 8.2625786163522,
      "grad_norm": 4.5,
      "learning_rate": 4.461042751636252e-06,
      "loss": 0.5008,
      "num_input_tokens_seen": 10307888,
      "step": 15765
    },
    {
      "epoch": 8.265199161425576,
      "grad_norm": 7.1875,
      "learning_rate": 4.448013413909394e-06,
      "loss": 0.4683,
      "num_input_tokens_seen": 10311248,
      "step": 15770
    },
    {
      "epoch": 8.267819706498951,
      "grad_norm": 7.90625,
      "learning_rate": 4.435001273146127e-06,
      "loss": 0.5085,
      "num_input_tokens_seen": 10314320,
      "step": 15775
    },
    {
      "epoch": 8.270440251572326,
      "grad_norm": 6.8125,
      "learning_rate": 4.422006340234433e-06,
      "loss": 0.5408,
      "num_input_tokens_seen": 10317936,
      "step": 15780
    },
    {
      "epoch": 8.273060796645701,
      "grad_norm": 6.65625,
      "learning_rate": 4.4090286260478674e-06,
      "loss": 0.5578,
      "num_input_tokens_seen": 10321136,
      "step": 15785
    },
    {
      "epoch": 8.275681341719078,
      "grad_norm": 8.0,
      "learning_rate": 4.3960681414455864e-06,
      "loss": 0.5824,
      "num_input_tokens_seen": 10323760,
      "step": 15790
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 13.3125,
      "learning_rate": 4.383124897272331e-06,
      "loss": 0.5217,
      "num_input_tokens_seen": 10327088,
      "step": 15795
    },
    {
      "epoch": 8.280922431865829,
      "grad_norm": 6.5625,
      "learning_rate": 4.3701989043584274e-06,
      "loss": 0.6272,
      "num_input_tokens_seen": 10330448,
      "step": 15800
    },
    {
      "epoch": 8.283542976939204,
      "grad_norm": 7.65625,
      "learning_rate": 4.357290173519746e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 10334672,
      "step": 15805
    },
    {
      "epoch": 8.286163522012579,
      "grad_norm": 5.78125,
      "learning_rate": 4.344398715557724e-06,
      "loss": 0.5321,
      "num_input_tokens_seen": 10337904,
      "step": 15810
    },
    {
      "epoch": 8.288784067085954,
      "grad_norm": 7.96875,
      "learning_rate": 4.3315245412593496e-06,
      "loss": 0.44,
      "num_input_tokens_seen": 10341488,
      "step": 15815
    },
    {
      "epoch": 8.29140461215933,
      "grad_norm": 8.3125,
      "learning_rate": 4.318667661397141e-06,
      "loss": 0.4745,
      "num_input_tokens_seen": 10344944,
      "step": 15820
    },
    {
      "epoch": 8.294025157232705,
      "grad_norm": 7.4375,
      "learning_rate": 4.305828086729144e-06,
      "loss": 0.5589,
      "num_input_tokens_seen": 10347248,
      "step": 15825
    },
    {
      "epoch": 8.29664570230608,
      "grad_norm": 7.5625,
      "learning_rate": 4.293005827998942e-06,
      "loss": 0.6014,
      "num_input_tokens_seen": 10349712,
      "step": 15830
    },
    {
      "epoch": 8.299266247379455,
      "grad_norm": 3.734375,
      "learning_rate": 4.280200895935593e-06,
      "loss": 0.4677,
      "num_input_tokens_seen": 10353936,
      "step": 15835
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 18.5,
      "learning_rate": 4.267413301253701e-06,
      "loss": 0.4681,
      "num_input_tokens_seen": 10356432,
      "step": 15840
    },
    {
      "epoch": 8.304507337526205,
      "grad_norm": 7.28125,
      "learning_rate": 4.254643054653329e-06,
      "loss": 0.7386,
      "num_input_tokens_seen": 10360624,
      "step": 15845
    },
    {
      "epoch": 8.30712788259958,
      "grad_norm": 4.46875,
      "learning_rate": 4.241890166820034e-06,
      "loss": 0.5029,
      "num_input_tokens_seen": 10364432,
      "step": 15850
    },
    {
      "epoch": 8.309748427672956,
      "grad_norm": 6.625,
      "learning_rate": 4.22915464842486e-06,
      "loss": 0.4407,
      "num_input_tokens_seen": 10367184,
      "step": 15855
    },
    {
      "epoch": 8.31236897274633,
      "grad_norm": 6.0625,
      "learning_rate": 4.216436510124303e-06,
      "loss": 0.4629,
      "num_input_tokens_seen": 10370192,
      "step": 15860
    },
    {
      "epoch": 8.314989517819706,
      "grad_norm": 4.46875,
      "learning_rate": 4.203735762560312e-06,
      "loss": 0.5637,
      "num_input_tokens_seen": 10373744,
      "step": 15865
    },
    {
      "epoch": 8.317610062893081,
      "grad_norm": 7.375,
      "learning_rate": 4.191052416360314e-06,
      "loss": 0.5223,
      "num_input_tokens_seen": 10376400,
      "step": 15870
    },
    {
      "epoch": 8.320230607966456,
      "grad_norm": 5.625,
      "learning_rate": 4.178386482137126e-06,
      "loss": 0.3916,
      "num_input_tokens_seen": 10379600,
      "step": 15875
    },
    {
      "epoch": 8.322851153039831,
      "grad_norm": 5.09375,
      "learning_rate": 4.165737970489036e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 10383088,
      "step": 15880
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 7.375,
      "learning_rate": 4.153106891999753e-06,
      "loss": 0.4983,
      "num_input_tokens_seen": 10386288,
      "step": 15885
    },
    {
      "epoch": 8.328092243186584,
      "grad_norm": 5.21875,
      "learning_rate": 4.140493257238362e-06,
      "loss": 0.6461,
      "num_input_tokens_seen": 10389744,
      "step": 15890
    },
    {
      "epoch": 8.330712788259959,
      "grad_norm": 7.125,
      "learning_rate": 4.127897076759399e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 10392880,
      "step": 15895
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 6.0,
      "learning_rate": 4.11531836110276e-06,
      "loss": 0.6107,
      "num_input_tokens_seen": 10397168,
      "step": 15900
    },
    {
      "epoch": 8.335953878406709,
      "grad_norm": 6.78125,
      "learning_rate": 4.1027571207937345e-06,
      "loss": 0.5087,
      "num_input_tokens_seen": 10400784,
      "step": 15905
    },
    {
      "epoch": 8.338574423480084,
      "grad_norm": 7.1875,
      "learning_rate": 4.09021336634301e-06,
      "loss": 0.296,
      "num_input_tokens_seen": 10403568,
      "step": 15910
    },
    {
      "epoch": 8.34119496855346,
      "grad_norm": 5.65625,
      "learning_rate": 4.077687108246622e-06,
      "loss": 0.4847,
      "num_input_tokens_seen": 10406800,
      "step": 15915
    },
    {
      "epoch": 8.343815513626835,
      "grad_norm": 4.40625,
      "learning_rate": 4.06517835698596e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 10409744,
      "step": 15920
    },
    {
      "epoch": 8.34643605870021,
      "grad_norm": 9.875,
      "learning_rate": 4.0526871230277905e-06,
      "loss": 0.4967,
      "num_input_tokens_seen": 10412880,
      "step": 15925
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 12.5625,
      "learning_rate": 4.040213416824204e-06,
      "loss": 0.5041,
      "num_input_tokens_seen": 10416432,
      "step": 15930
    },
    {
      "epoch": 8.35167714884696,
      "grad_norm": 9.75,
      "learning_rate": 4.027757248812622e-06,
      "loss": 0.4868,
      "num_input_tokens_seen": 10419824,
      "step": 15935
    },
    {
      "epoch": 8.354297693920335,
      "grad_norm": 4.8125,
      "learning_rate": 4.015318629415804e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 10422800,
      "step": 15940
    },
    {
      "epoch": 8.35691823899371,
      "grad_norm": 6.9375,
      "learning_rate": 4.002897569041808e-06,
      "loss": 0.4141,
      "num_input_tokens_seen": 10425808,
      "step": 15945
    },
    {
      "epoch": 8.359538784067086,
      "grad_norm": 4.125,
      "learning_rate": 3.990494078084022e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 10428944,
      "step": 15950
    },
    {
      "epoch": 8.36215932914046,
      "grad_norm": 4.90625,
      "learning_rate": 3.9781081669211156e-06,
      "loss": 0.524,
      "num_input_tokens_seen": 10433648,
      "step": 15955
    },
    {
      "epoch": 8.364779874213836,
      "grad_norm": 3.84375,
      "learning_rate": 3.965739845917049e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 10436784,
      "step": 15960
    },
    {
      "epoch": 8.367400419287211,
      "grad_norm": 8.4375,
      "learning_rate": 3.953389125421078e-06,
      "loss": 0.4682,
      "num_input_tokens_seen": 10439280,
      "step": 15965
    },
    {
      "epoch": 8.370020964360586,
      "grad_norm": 5.4375,
      "learning_rate": 3.941056015767713e-06,
      "loss": 0.3936,
      "num_input_tokens_seen": 10441744,
      "step": 15970
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 5.875,
      "learning_rate": 3.928740527276745e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 10445424,
      "step": 15975
    },
    {
      "epoch": 8.375262054507338,
      "grad_norm": 8.0,
      "learning_rate": 3.916442670253198e-06,
      "loss": 0.7094,
      "num_input_tokens_seen": 10448688,
      "step": 15980
    },
    {
      "epoch": 8.377882599580714,
      "grad_norm": 4.59375,
      "learning_rate": 3.904162454987373e-06,
      "loss": 0.4957,
      "num_input_tokens_seen": 10453040,
      "step": 15985
    },
    {
      "epoch": 8.380503144654089,
      "grad_norm": 4.5,
      "learning_rate": 3.891899891754788e-06,
      "loss": 0.531,
      "num_input_tokens_seen": 10457072,
      "step": 15990
    },
    {
      "epoch": 8.383123689727464,
      "grad_norm": 22.875,
      "learning_rate": 3.8796549908161864e-06,
      "loss": 0.363,
      "num_input_tokens_seen": 10459280,
      "step": 15995
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 10.5625,
      "learning_rate": 3.867427762417555e-06,
      "loss": 0.4382,
      "num_input_tokens_seen": 10462192,
      "step": 16000
    },
    {
      "epoch": 8.388364779874214,
      "grad_norm": 4.78125,
      "learning_rate": 3.855218216790077e-06,
      "loss": 0.4896,
      "num_input_tokens_seen": 10466672,
      "step": 16005
    },
    {
      "epoch": 8.39098532494759,
      "grad_norm": 4.96875,
      "learning_rate": 3.843026364150132e-06,
      "loss": 0.4651,
      "num_input_tokens_seen": 10469776,
      "step": 16010
    },
    {
      "epoch": 8.393605870020965,
      "grad_norm": 11.125,
      "learning_rate": 3.830852214699326e-06,
      "loss": 0.4079,
      "num_input_tokens_seen": 10472848,
      "step": 16015
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 21.0,
      "learning_rate": 3.818695778624409e-06,
      "loss": 0.5941,
      "num_input_tokens_seen": 10476368,
      "step": 16020
    },
    {
      "epoch": 8.398846960167715,
      "grad_norm": 5.84375,
      "learning_rate": 3.8065570660973436e-06,
      "loss": 0.5153,
      "num_input_tokens_seen": 10479472,
      "step": 16025
    },
    {
      "epoch": 8.40146750524109,
      "grad_norm": 6.8125,
      "learning_rate": 3.7944360872752495e-06,
      "loss": 0.5613,
      "num_input_tokens_seen": 10482000,
      "step": 16030
    },
    {
      "epoch": 8.404088050314465,
      "grad_norm": 9.8125,
      "learning_rate": 3.782332852300402e-06,
      "loss": 0.4882,
      "num_input_tokens_seen": 10484592,
      "step": 16035
    },
    {
      "epoch": 8.40670859538784,
      "grad_norm": 21.125,
      "learning_rate": 3.770247371300242e-06,
      "loss": 0.4421,
      "num_input_tokens_seen": 10487216,
      "step": 16040
    },
    {
      "epoch": 8.409329140461216,
      "grad_norm": 6.53125,
      "learning_rate": 3.7581796543873477e-06,
      "loss": 0.4965,
      "num_input_tokens_seen": 10490512,
      "step": 16045
    },
    {
      "epoch": 8.41194968553459,
      "grad_norm": 7.1875,
      "learning_rate": 3.746129711659424e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 10493552,
      "step": 16050
    },
    {
      "epoch": 8.414570230607966,
      "grad_norm": 7.09375,
      "learning_rate": 3.7340975531993313e-06,
      "loss": 0.5488,
      "num_input_tokens_seen": 10497264,
      "step": 16055
    },
    {
      "epoch": 8.417190775681341,
      "grad_norm": 6.53125,
      "learning_rate": 3.7220831890750067e-06,
      "loss": 0.3767,
      "num_input_tokens_seen": 10499984,
      "step": 16060
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 7.5,
      "learning_rate": 3.7100866293395403e-06,
      "loss": 0.439,
      "num_input_tokens_seen": 10503248,
      "step": 16065
    },
    {
      "epoch": 8.422431865828091,
      "grad_norm": 2.96875,
      "learning_rate": 3.698107884031099e-06,
      "loss": 0.5124,
      "num_input_tokens_seen": 10507088,
      "step": 16070
    },
    {
      "epoch": 8.425052410901468,
      "grad_norm": 5.78125,
      "learning_rate": 3.68614696317294e-06,
      "loss": 0.4437,
      "num_input_tokens_seen": 10510416,
      "step": 16075
    },
    {
      "epoch": 8.427672955974844,
      "grad_norm": 6.59375,
      "learning_rate": 3.6742038767734326e-06,
      "loss": 0.5737,
      "num_input_tokens_seen": 10514384,
      "step": 16080
    },
    {
      "epoch": 8.430293501048219,
      "grad_norm": 4.96875,
      "learning_rate": 3.6622786348259967e-06,
      "loss": 0.5731,
      "num_input_tokens_seen": 10517872,
      "step": 16085
    },
    {
      "epoch": 8.432914046121594,
      "grad_norm": 6.84375,
      "learning_rate": 3.6503712473091257e-06,
      "loss": 0.501,
      "num_input_tokens_seen": 10521200,
      "step": 16090
    },
    {
      "epoch": 8.435534591194969,
      "grad_norm": 5.125,
      "learning_rate": 3.6384817241863877e-06,
      "loss": 0.486,
      "num_input_tokens_seen": 10523856,
      "step": 16095
    },
    {
      "epoch": 8.438155136268344,
      "grad_norm": 3.5625,
      "learning_rate": 3.626610075406389e-06,
      "loss": 0.5342,
      "num_input_tokens_seen": 10527728,
      "step": 16100
    },
    {
      "epoch": 8.44077568134172,
      "grad_norm": 6.0625,
      "learning_rate": 3.614756310902781e-06,
      "loss": 0.4808,
      "num_input_tokens_seen": 10531536,
      "step": 16105
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 6.875,
      "learning_rate": 3.6029204405942485e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 10534352,
      "step": 16110
    },
    {
      "epoch": 8.44601677148847,
      "grad_norm": 6.28125,
      "learning_rate": 3.5911024743845166e-06,
      "loss": 0.4823,
      "num_input_tokens_seen": 10537552,
      "step": 16115
    },
    {
      "epoch": 8.448637316561845,
      "grad_norm": 5.875,
      "learning_rate": 3.5793024221623147e-06,
      "loss": 0.4936,
      "num_input_tokens_seen": 10539952,
      "step": 16120
    },
    {
      "epoch": 8.45125786163522,
      "grad_norm": 6.21875,
      "learning_rate": 3.567520293801388e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 10543408,
      "step": 16125
    },
    {
      "epoch": 8.453878406708595,
      "grad_norm": 2.90625,
      "learning_rate": 3.555756099160476e-06,
      "loss": 0.3905,
      "num_input_tokens_seen": 10546352,
      "step": 16130
    },
    {
      "epoch": 8.45649895178197,
      "grad_norm": 4.75,
      "learning_rate": 3.544009848083335e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 10550352,
      "step": 16135
    },
    {
      "epoch": 8.459119496855346,
      "grad_norm": 5.6875,
      "learning_rate": 3.5322815503986804e-06,
      "loss": 0.6096,
      "num_input_tokens_seen": 10554864,
      "step": 16140
    },
    {
      "epoch": 8.46174004192872,
      "grad_norm": 6.59375,
      "learning_rate": 3.520571215920218e-06,
      "loss": 0.4273,
      "num_input_tokens_seen": 10558960,
      "step": 16145
    },
    {
      "epoch": 8.464360587002096,
      "grad_norm": 5.1875,
      "learning_rate": 3.5088788544466177e-06,
      "loss": 0.4813,
      "num_input_tokens_seen": 10563600,
      "step": 16150
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 8.4375,
      "learning_rate": 3.4972044757615203e-06,
      "loss": 0.4811,
      "num_input_tokens_seen": 10566480,
      "step": 16155
    },
    {
      "epoch": 8.469601677148846,
      "grad_norm": 6.40625,
      "learning_rate": 3.4855480896335084e-06,
      "loss": 0.3822,
      "num_input_tokens_seen": 10569616,
      "step": 16160
    },
    {
      "epoch": 8.472222222222221,
      "grad_norm": 6.25,
      "learning_rate": 3.4739097058161114e-06,
      "loss": 0.5572,
      "num_input_tokens_seen": 10572336,
      "step": 16165
    },
    {
      "epoch": 8.474842767295598,
      "grad_norm": 9.0625,
      "learning_rate": 3.462289334047805e-06,
      "loss": 0.602,
      "num_input_tokens_seen": 10574576,
      "step": 16170
    },
    {
      "epoch": 8.477463312368974,
      "grad_norm": 4.84375,
      "learning_rate": 3.450686984051979e-06,
      "loss": 0.437,
      "num_input_tokens_seen": 10578640,
      "step": 16175
    },
    {
      "epoch": 8.480083857442349,
      "grad_norm": 3.8125,
      "learning_rate": 3.4391026655369474e-06,
      "loss": 0.3964,
      "num_input_tokens_seen": 10581392,
      "step": 16180
    },
    {
      "epoch": 8.482704402515724,
      "grad_norm": 8.0,
      "learning_rate": 3.427536388195954e-06,
      "loss": 0.723,
      "num_input_tokens_seen": 10584144,
      "step": 16185
    },
    {
      "epoch": 8.485324947589099,
      "grad_norm": 18.25,
      "learning_rate": 3.415988161707109e-06,
      "loss": 0.5254,
      "num_input_tokens_seen": 10587600,
      "step": 16190
    },
    {
      "epoch": 8.487945492662474,
      "grad_norm": 11.25,
      "learning_rate": 3.404457995733451e-06,
      "loss": 0.493,
      "num_input_tokens_seen": 10590832,
      "step": 16195
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 6.375,
      "learning_rate": 3.3929458999229113e-06,
      "loss": 0.5571,
      "num_input_tokens_seen": 10593712,
      "step": 16200
    },
    {
      "epoch": 8.493186582809225,
      "grad_norm": 5.9375,
      "learning_rate": 3.381451883908257e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 10596336,
      "step": 16205
    },
    {
      "epoch": 8.4958071278826,
      "grad_norm": 7.25,
      "learning_rate": 3.369975957307178e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 10599536,
      "step": 16210
    },
    {
      "epoch": 8.498427672955975,
      "grad_norm": 6.25,
      "learning_rate": 3.358518129722199e-06,
      "loss": 0.2855,
      "num_input_tokens_seen": 10602576,
      "step": 16215
    },
    {
      "epoch": 8.5,
      "eval_loss": 0.4936758875846863,
      "eval_runtime": 24.2859,
      "eval_samples_per_second": 34.917,
      "eval_steps_per_second": 8.729,
      "num_input_tokens_seen": 10604656,
      "step": 16218
    },
    {
      "epoch": 8.50104821802935,
      "grad_norm": 6.5,
      "learning_rate": 3.3470784107406976e-06,
      "loss": 0.4091,
      "num_input_tokens_seen": 10605616,
      "step": 16220
    },
    {
      "epoch": 8.503668763102725,
      "grad_norm": 6.3125,
      "learning_rate": 3.3356568099349283e-06,
      "loss": 0.6539,
      "num_input_tokens_seen": 10609040,
      "step": 16225
    },
    {
      "epoch": 8.5062893081761,
      "grad_norm": 5.25,
      "learning_rate": 3.3242533368619435e-06,
      "loss": 0.5842,
      "num_input_tokens_seen": 10612304,
      "step": 16230
    },
    {
      "epoch": 8.508909853249476,
      "grad_norm": 6.46875,
      "learning_rate": 3.312868001063654e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 10614928,
      "step": 16235
    },
    {
      "epoch": 8.51153039832285,
      "grad_norm": 5.90625,
      "learning_rate": 3.3015008120668072e-06,
      "loss": 0.6086,
      "num_input_tokens_seen": 10618320,
      "step": 16240
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 11.5625,
      "learning_rate": 3.290151779382922e-06,
      "loss": 0.4912,
      "num_input_tokens_seen": 10621296,
      "step": 16245
    },
    {
      "epoch": 8.516771488469601,
      "grad_norm": 6.21875,
      "learning_rate": 3.2788209125083654e-06,
      "loss": 0.533,
      "num_input_tokens_seen": 10623984,
      "step": 16250
    },
    {
      "epoch": 8.519392033542976,
      "grad_norm": 5.46875,
      "learning_rate": 3.267508220924287e-06,
      "loss": 0.4782,
      "num_input_tokens_seen": 10628560,
      "step": 16255
    },
    {
      "epoch": 8.522012578616351,
      "grad_norm": 11.3125,
      "learning_rate": 3.256213714096623e-06,
      "loss": 0.434,
      "num_input_tokens_seen": 10631856,
      "step": 16260
    },
    {
      "epoch": 8.524633123689728,
      "grad_norm": 4.8125,
      "learning_rate": 3.2449374014761114e-06,
      "loss": 0.4617,
      "num_input_tokens_seen": 10635696,
      "step": 16265
    },
    {
      "epoch": 8.527253668763104,
      "grad_norm": 12.6875,
      "learning_rate": 3.2336792924982514e-06,
      "loss": 0.529,
      "num_input_tokens_seen": 10638384,
      "step": 16270
    },
    {
      "epoch": 8.529874213836479,
      "grad_norm": 6.15625,
      "learning_rate": 3.222439396583307e-06,
      "loss": 0.5033,
      "num_input_tokens_seen": 10641456,
      "step": 16275
    },
    {
      "epoch": 8.532494758909854,
      "grad_norm": 13.1875,
      "learning_rate": 3.2112177231363226e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 10644336,
      "step": 16280
    },
    {
      "epoch": 8.535115303983229,
      "grad_norm": 5.1875,
      "learning_rate": 3.2000142815470756e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 10647440,
      "step": 16285
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 5.75,
      "learning_rate": 3.188829081190095e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 10649872,
      "step": 16290
    },
    {
      "epoch": 8.54035639412998,
      "grad_norm": 3.71875,
      "learning_rate": 3.1776621314246384e-06,
      "loss": 0.5353,
      "num_input_tokens_seen": 10652976,
      "step": 16295
    },
    {
      "epoch": 8.542976939203355,
      "grad_norm": 5.4375,
      "learning_rate": 3.1665134415947125e-06,
      "loss": 0.4509,
      "num_input_tokens_seen": 10656016,
      "step": 16300
    },
    {
      "epoch": 8.54559748427673,
      "grad_norm": 5.75,
      "learning_rate": 3.1553830210290236e-06,
      "loss": 0.5282,
      "num_input_tokens_seen": 10659536,
      "step": 16305
    },
    {
      "epoch": 8.548218029350105,
      "grad_norm": 6.8125,
      "learning_rate": 3.1442708790410002e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 10662416,
      "step": 16310
    },
    {
      "epoch": 8.55083857442348,
      "grad_norm": 3.640625,
      "learning_rate": 3.133177024928771e-06,
      "loss": 0.4355,
      "num_input_tokens_seen": 10666320,
      "step": 16315
    },
    {
      "epoch": 8.553459119496855,
      "grad_norm": 6.09375,
      "learning_rate": 3.1221014679751777e-06,
      "loss": 0.6539,
      "num_input_tokens_seen": 10669136,
      "step": 16320
    },
    {
      "epoch": 8.55607966457023,
      "grad_norm": 5.0,
      "learning_rate": 3.111044217447731e-06,
      "loss": 0.4688,
      "num_input_tokens_seen": 10672240,
      "step": 16325
    },
    {
      "epoch": 8.558700209643606,
      "grad_norm": 8.4375,
      "learning_rate": 3.1000052825986366e-06,
      "loss": 0.4859,
      "num_input_tokens_seen": 10674992,
      "step": 16330
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 7.46875,
      "learning_rate": 3.0889846726647657e-06,
      "loss": 0.5337,
      "num_input_tokens_seen": 10677616,
      "step": 16335
    },
    {
      "epoch": 8.563941299790356,
      "grad_norm": 5.9375,
      "learning_rate": 3.077982396867668e-06,
      "loss": 0.4969,
      "num_input_tokens_seen": 10681136,
      "step": 16340
    },
    {
      "epoch": 8.566561844863731,
      "grad_norm": 9.25,
      "learning_rate": 3.066998464413545e-06,
      "loss": 0.6398,
      "num_input_tokens_seen": 10684368,
      "step": 16345
    },
    {
      "epoch": 8.569182389937106,
      "grad_norm": 4.03125,
      "learning_rate": 3.056032884493243e-06,
      "loss": 0.5396,
      "num_input_tokens_seen": 10687632,
      "step": 16350
    },
    {
      "epoch": 8.571802935010481,
      "grad_norm": 3.75,
      "learning_rate": 3.045085666282266e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 10692336,
      "step": 16355
    },
    {
      "epoch": 8.574423480083858,
      "grad_norm": 3.65625,
      "learning_rate": 3.034156818940745e-06,
      "loss": 0.546,
      "num_input_tokens_seen": 10696048,
      "step": 16360
    },
    {
      "epoch": 8.577044025157234,
      "grad_norm": 20.125,
      "learning_rate": 3.0232463516134317e-06,
      "loss": 0.4529,
      "num_input_tokens_seen": 10700304,
      "step": 16365
    },
    {
      "epoch": 8.579664570230609,
      "grad_norm": 4.59375,
      "learning_rate": 3.0123542734297267e-06,
      "loss": 0.5043,
      "num_input_tokens_seen": 10703120,
      "step": 16370
    },
    {
      "epoch": 8.582285115303984,
      "grad_norm": 2.8125,
      "learning_rate": 3.0014805935035973e-06,
      "loss": 0.4999,
      "num_input_tokens_seen": 10707536,
      "step": 16375
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 3.421875,
      "learning_rate": 2.99062532093366e-06,
      "loss": 0.5657,
      "num_input_tokens_seen": 10710384,
      "step": 16380
    },
    {
      "epoch": 8.587526205450734,
      "grad_norm": 8.125,
      "learning_rate": 2.979788464803107e-06,
      "loss": 0.4808,
      "num_input_tokens_seen": 10713840,
      "step": 16385
    },
    {
      "epoch": 8.59014675052411,
      "grad_norm": 6.5625,
      "learning_rate": 2.968970034179719e-06,
      "loss": 0.4363,
      "num_input_tokens_seen": 10716720,
      "step": 16390
    },
    {
      "epoch": 8.592767295597485,
      "grad_norm": 6.75,
      "learning_rate": 2.9581700381158735e-06,
      "loss": 0.5477,
      "num_input_tokens_seen": 10720720,
      "step": 16395
    },
    {
      "epoch": 8.59538784067086,
      "grad_norm": 7.40625,
      "learning_rate": 2.9473884856485113e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 10723952,
      "step": 16400
    },
    {
      "epoch": 8.598008385744235,
      "grad_norm": 9.1875,
      "learning_rate": 2.936625385799133e-06,
      "loss": 0.5777,
      "num_input_tokens_seen": 10726096,
      "step": 16405
    },
    {
      "epoch": 8.60062893081761,
      "grad_norm": 4.15625,
      "learning_rate": 2.925880747573831e-06,
      "loss": 0.4098,
      "num_input_tokens_seen": 10729456,
      "step": 16410
    },
    {
      "epoch": 8.603249475890985,
      "grad_norm": 6.90625,
      "learning_rate": 2.9151545799632003e-06,
      "loss": 0.3739,
      "num_input_tokens_seen": 10731824,
      "step": 16415
    },
    {
      "epoch": 8.60587002096436,
      "grad_norm": 5.84375,
      "learning_rate": 2.9044468919424305e-06,
      "loss": 0.3632,
      "num_input_tokens_seen": 10734864,
      "step": 16420
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 6.4375,
      "learning_rate": 2.8937576924712133e-06,
      "loss": 0.487,
      "num_input_tokens_seen": 10737744,
      "step": 16425
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 9.375,
      "learning_rate": 2.883086990493783e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 10740656,
      "step": 16430
    },
    {
      "epoch": 8.613731656184486,
      "grad_norm": 10.0,
      "learning_rate": 2.872434794938905e-06,
      "loss": 0.7343,
      "num_input_tokens_seen": 10744688,
      "step": 16435
    },
    {
      "epoch": 8.616352201257861,
      "grad_norm": 6.5,
      "learning_rate": 2.861801114719842e-06,
      "loss": 0.5428,
      "num_input_tokens_seen": 10747184,
      "step": 16440
    },
    {
      "epoch": 8.618972746331236,
      "grad_norm": 5.53125,
      "learning_rate": 2.8511859587343704e-06,
      "loss": 0.6245,
      "num_input_tokens_seen": 10750320,
      "step": 16445
    },
    {
      "epoch": 8.621593291404611,
      "grad_norm": 9.3125,
      "learning_rate": 2.840589335864774e-06,
      "loss": 0.4187,
      "num_input_tokens_seen": 10753328,
      "step": 16450
    },
    {
      "epoch": 8.624213836477988,
      "grad_norm": 7.0625,
      "learning_rate": 2.830011254977821e-06,
      "loss": 0.5124,
      "num_input_tokens_seen": 10756208,
      "step": 16455
    },
    {
      "epoch": 8.626834381551364,
      "grad_norm": 10.8125,
      "learning_rate": 2.819451724924768e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 10759888,
      "step": 16460
    },
    {
      "epoch": 8.629454926624739,
      "grad_norm": 7.625,
      "learning_rate": 2.8089107545413355e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 10763152,
      "step": 16465
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 8.5625,
      "learning_rate": 2.7983883526477433e-06,
      "loss": 0.5109,
      "num_input_tokens_seen": 10767024,
      "step": 16470
    },
    {
      "epoch": 8.634696016771489,
      "grad_norm": 6.40625,
      "learning_rate": 2.7878845280486453e-06,
      "loss": 0.3692,
      "num_input_tokens_seen": 10770096,
      "step": 16475
    },
    {
      "epoch": 8.637316561844864,
      "grad_norm": 8.5625,
      "learning_rate": 2.777399289533164e-06,
      "loss": 0.5706,
      "num_input_tokens_seen": 10774576,
      "step": 16480
    },
    {
      "epoch": 8.63993710691824,
      "grad_norm": 6.65625,
      "learning_rate": 2.766932645874873e-06,
      "loss": 0.3813,
      "num_input_tokens_seen": 10777296,
      "step": 16485
    },
    {
      "epoch": 8.642557651991615,
      "grad_norm": 4.6875,
      "learning_rate": 2.756484605831777e-06,
      "loss": 0.4319,
      "num_input_tokens_seen": 10779696,
      "step": 16490
    },
    {
      "epoch": 8.64517819706499,
      "grad_norm": 9.3125,
      "learning_rate": 2.74605517814632e-06,
      "loss": 0.4127,
      "num_input_tokens_seen": 10783184,
      "step": 16495
    },
    {
      "epoch": 8.647798742138365,
      "grad_norm": 3.34375,
      "learning_rate": 2.7356443715453705e-06,
      "loss": 0.6413,
      "num_input_tokens_seen": 10786512,
      "step": 16500
    },
    {
      "epoch": 8.65041928721174,
      "grad_norm": 6.875,
      "learning_rate": 2.725252194740213e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 10789872,
      "step": 16505
    },
    {
      "epoch": 8.653039832285115,
      "grad_norm": 4.90625,
      "learning_rate": 2.714878656426553e-06,
      "loss": 0.4234,
      "num_input_tokens_seen": 10792944,
      "step": 16510
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 8.75,
      "learning_rate": 2.704523765284489e-06,
      "loss": 0.4988,
      "num_input_tokens_seen": 10796624,
      "step": 16515
    },
    {
      "epoch": 8.658280922431866,
      "grad_norm": 5.8125,
      "learning_rate": 2.6941875299785174e-06,
      "loss": 0.4813,
      "num_input_tokens_seen": 10800144,
      "step": 16520
    },
    {
      "epoch": 8.66090146750524,
      "grad_norm": 9.0625,
      "learning_rate": 2.683869959157534e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 10803152,
      "step": 16525
    },
    {
      "epoch": 8.663522012578616,
      "grad_norm": 11.75,
      "learning_rate": 2.673571061454813e-06,
      "loss": 0.5222,
      "num_input_tokens_seen": 10805552,
      "step": 16530
    },
    {
      "epoch": 8.666142557651991,
      "grad_norm": 8.0,
      "learning_rate": 2.6632908454879898e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 10808880,
      "step": 16535
    },
    {
      "epoch": 8.668763102725366,
      "grad_norm": 8.875,
      "learning_rate": 2.653029319859096e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 10812240,
      "step": 16540
    },
    {
      "epoch": 8.671383647798741,
      "grad_norm": 14.0625,
      "learning_rate": 2.642786493154492e-06,
      "loss": 0.4459,
      "num_input_tokens_seen": 10815600,
      "step": 16545
    },
    {
      "epoch": 8.674004192872118,
      "grad_norm": 4.125,
      "learning_rate": 2.6325623739449108e-06,
      "loss": 0.4729,
      "num_input_tokens_seen": 10819120,
      "step": 16550
    },
    {
      "epoch": 8.676624737945493,
      "grad_norm": 9.0,
      "learning_rate": 2.6223569707854444e-06,
      "loss": 0.4179,
      "num_input_tokens_seen": 10821744,
      "step": 16555
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 12.9375,
      "learning_rate": 2.612170292215482e-06,
      "loss": 0.4619,
      "num_input_tokens_seen": 10824336,
      "step": 16560
    },
    {
      "epoch": 8.681865828092244,
      "grad_norm": 5.28125,
      "learning_rate": 2.6020023467587917e-06,
      "loss": 0.6124,
      "num_input_tokens_seen": 10826672,
      "step": 16565
    },
    {
      "epoch": 8.684486373165619,
      "grad_norm": 6.3125,
      "learning_rate": 2.5918531429234368e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 10830192,
      "step": 16570
    },
    {
      "epoch": 8.687106918238994,
      "grad_norm": 9.0625,
      "learning_rate": 2.5817226892018016e-06,
      "loss": 0.6645,
      "num_input_tokens_seen": 10834096,
      "step": 16575
    },
    {
      "epoch": 8.68972746331237,
      "grad_norm": 8.3125,
      "learning_rate": 2.571610994070603e-06,
      "loss": 0.4774,
      "num_input_tokens_seen": 10837360,
      "step": 16580
    },
    {
      "epoch": 8.692348008385745,
      "grad_norm": 5.8125,
      "learning_rate": 2.561518065990834e-06,
      "loss": 0.5707,
      "num_input_tokens_seen": 10841168,
      "step": 16585
    },
    {
      "epoch": 8.69496855345912,
      "grad_norm": 6.34375,
      "learning_rate": 2.5514439134077945e-06,
      "loss": 0.3655,
      "num_input_tokens_seen": 10844784,
      "step": 16590
    },
    {
      "epoch": 8.697589098532495,
      "grad_norm": 6.8125,
      "learning_rate": 2.541388544751089e-06,
      "loss": 0.4772,
      "num_input_tokens_seen": 10847376,
      "step": 16595
    },
    {
      "epoch": 8.70020964360587,
      "grad_norm": 8.625,
      "learning_rate": 2.53135196843457e-06,
      "loss": 0.4334,
      "num_input_tokens_seen": 10850000,
      "step": 16600
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 6.96875,
      "learning_rate": 2.521334192856403e-06,
      "loss": 0.3957,
      "num_input_tokens_seen": 10852528,
      "step": 16605
    },
    {
      "epoch": 8.70545073375262,
      "grad_norm": 11.125,
      "learning_rate": 2.5113352263990005e-06,
      "loss": 0.4699,
      "num_input_tokens_seen": 10855312,
      "step": 16610
    },
    {
      "epoch": 8.708071278825996,
      "grad_norm": 5.28125,
      "learning_rate": 2.5013550774290322e-06,
      "loss": 0.6211,
      "num_input_tokens_seen": 10859376,
      "step": 16615
    },
    {
      "epoch": 8.71069182389937,
      "grad_norm": 8.3125,
      "learning_rate": 2.491393754297444e-06,
      "loss": 0.5028,
      "num_input_tokens_seen": 10862224,
      "step": 16620
    },
    {
      "epoch": 8.713312368972746,
      "grad_norm": 10.25,
      "learning_rate": 2.48145126533941e-06,
      "loss": 0.6288,
      "num_input_tokens_seen": 10865104,
      "step": 16625
    },
    {
      "epoch": 8.715932914046121,
      "grad_norm": 6.03125,
      "learning_rate": 2.4715276188743476e-06,
      "loss": 0.4191,
      "num_input_tokens_seen": 10867792,
      "step": 16630
    },
    {
      "epoch": 8.718553459119496,
      "grad_norm": 12.625,
      "learning_rate": 2.461622823205917e-06,
      "loss": 0.5128,
      "num_input_tokens_seen": 10870768,
      "step": 16635
    },
    {
      "epoch": 8.721174004192871,
      "grad_norm": 6.625,
      "learning_rate": 2.451736886621997e-06,
      "loss": 0.3478,
      "num_input_tokens_seen": 10875056,
      "step": 16640
    },
    {
      "epoch": 8.723794549266248,
      "grad_norm": 9.4375,
      "learning_rate": 2.4418698173946872e-06,
      "loss": 0.5416,
      "num_input_tokens_seen": 10878288,
      "step": 16645
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 8.6875,
      "learning_rate": 2.432021623780295e-06,
      "loss": 0.5288,
      "num_input_tokens_seen": 10880720,
      "step": 16650
    },
    {
      "epoch": 8.729035639412999,
      "grad_norm": 4.34375,
      "learning_rate": 2.4221923140193477e-06,
      "loss": 0.3064,
      "num_input_tokens_seen": 10884528,
      "step": 16655
    },
    {
      "epoch": 8.731656184486374,
      "grad_norm": 7.21875,
      "learning_rate": 2.41238189633656e-06,
      "loss": 0.5729,
      "num_input_tokens_seen": 10886960,
      "step": 16660
    },
    {
      "epoch": 8.734276729559749,
      "grad_norm": 4.125,
      "learning_rate": 2.402590378940836e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 10889904,
      "step": 16665
    },
    {
      "epoch": 8.736897274633124,
      "grad_norm": 7.71875,
      "learning_rate": 2.3928177700252798e-06,
      "loss": 0.5874,
      "num_input_tokens_seen": 10893072,
      "step": 16670
    },
    {
      "epoch": 8.7395178197065,
      "grad_norm": 5.5,
      "learning_rate": 2.3830640777671583e-06,
      "loss": 0.5167,
      "num_input_tokens_seen": 10897392,
      "step": 16675
    },
    {
      "epoch": 8.742138364779874,
      "grad_norm": 3.453125,
      "learning_rate": 2.3733293103279153e-06,
      "loss": 0.5801,
      "num_input_tokens_seen": 10901008,
      "step": 16680
    },
    {
      "epoch": 8.74475890985325,
      "grad_norm": 4.71875,
      "learning_rate": 2.3636134758531604e-06,
      "loss": 0.5576,
      "num_input_tokens_seen": 10904848,
      "step": 16685
    },
    {
      "epoch": 8.747379454926625,
      "grad_norm": 5.75,
      "learning_rate": 2.3539165824726565e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 10908528,
      "step": 16690
    },
    {
      "epoch": 8.75,
      "grad_norm": 4.625,
      "learning_rate": 2.344238638300328e-06,
      "loss": 0.5918,
      "num_input_tokens_seen": 10912080,
      "step": 16695
    },
    {
      "epoch": 8.752620545073375,
      "grad_norm": 5.875,
      "learning_rate": 2.334579651434235e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 10916528,
      "step": 16700
    },
    {
      "epoch": 8.75524109014675,
      "grad_norm": 7.125,
      "learning_rate": 2.3249396299565683e-06,
      "loss": 0.457,
      "num_input_tokens_seen": 10920688,
      "step": 16705
    },
    {
      "epoch": 8.757861635220126,
      "grad_norm": 11.625,
      "learning_rate": 2.3153185819336705e-06,
      "loss": 0.4674,
      "num_input_tokens_seen": 10923600,
      "step": 16710
    },
    {
      "epoch": 8.7604821802935,
      "grad_norm": 4.5,
      "learning_rate": 2.3057165154159873e-06,
      "loss": 0.408,
      "num_input_tokens_seen": 10926544,
      "step": 16715
    },
    {
      "epoch": 8.763102725366876,
      "grad_norm": 5.28125,
      "learning_rate": 2.296133438438086e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 10929392,
      "step": 16720
    },
    {
      "epoch": 8.765723270440251,
      "grad_norm": 6.40625,
      "learning_rate": 2.2865693590186616e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 10933456,
      "step": 16725
    },
    {
      "epoch": 8.768343815513626,
      "grad_norm": 6.6875,
      "learning_rate": 2.2770242851604813e-06,
      "loss": 0.6182,
      "num_input_tokens_seen": 10936336,
      "step": 16730
    },
    {
      "epoch": 8.770964360587001,
      "grad_norm": 4.75,
      "learning_rate": 2.2674982248504395e-06,
      "loss": 0.5269,
      "num_input_tokens_seen": 10939696,
      "step": 16735
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 7.34375,
      "learning_rate": 2.257991186059502e-06,
      "loss": 0.386,
      "num_input_tokens_seen": 10942192,
      "step": 16740
    },
    {
      "epoch": 8.776205450733752,
      "grad_norm": 9.0,
      "learning_rate": 2.248503176742725e-06,
      "loss": 0.3114,
      "num_input_tokens_seen": 10944624,
      "step": 16745
    },
    {
      "epoch": 8.778825995807129,
      "grad_norm": 5.53125,
      "learning_rate": 2.2390342048392467e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 10947952,
      "step": 16750
    },
    {
      "epoch": 8.781446540880504,
      "grad_norm": 6.34375,
      "learning_rate": 2.229584278272265e-06,
      "loss": 0.4571,
      "num_input_tokens_seen": 10951440,
      "step": 16755
    },
    {
      "epoch": 8.784067085953879,
      "grad_norm": 4.09375,
      "learning_rate": 2.2201534049490436e-06,
      "loss": 0.4293,
      "num_input_tokens_seen": 10954224,
      "step": 16760
    },
    {
      "epoch": 8.786687631027254,
      "grad_norm": 15.6875,
      "learning_rate": 2.2107415927609176e-06,
      "loss": 0.4946,
      "num_input_tokens_seen": 10957392,
      "step": 16765
    },
    {
      "epoch": 8.78930817610063,
      "grad_norm": 4.875,
      "learning_rate": 2.2013488495832542e-06,
      "loss": 0.5263,
      "num_input_tokens_seen": 10961136,
      "step": 16770
    },
    {
      "epoch": 8.791928721174004,
      "grad_norm": 7.125,
      "learning_rate": 2.1919751832754714e-06,
      "loss": 0.5116,
      "num_input_tokens_seen": 10964272,
      "step": 16775
    },
    {
      "epoch": 8.79454926624738,
      "grad_norm": 5.125,
      "learning_rate": 2.182620601681029e-06,
      "loss": 0.4024,
      "num_input_tokens_seen": 10967344,
      "step": 16780
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 10.25,
      "learning_rate": 2.1732851126274047e-06,
      "loss": 0.4305,
      "num_input_tokens_seen": 10970800,
      "step": 16785
    },
    {
      "epoch": 8.79979035639413,
      "grad_norm": 4.25,
      "learning_rate": 2.1639687239261214e-06,
      "loss": 0.5103,
      "num_input_tokens_seen": 10974544,
      "step": 16790
    },
    {
      "epoch": 8.802410901467505,
      "grad_norm": 12.625,
      "learning_rate": 2.1546714433726993e-06,
      "loss": 0.54,
      "num_input_tokens_seen": 10976784,
      "step": 16795
    },
    {
      "epoch": 8.80503144654088,
      "grad_norm": 4.90625,
      "learning_rate": 2.1453932787466767e-06,
      "loss": 0.5443,
      "num_input_tokens_seen": 10980400,
      "step": 16800
    },
    {
      "epoch": 8.807651991614255,
      "grad_norm": 6.78125,
      "learning_rate": 2.1361342378116072e-06,
      "loss": 0.3641,
      "num_input_tokens_seen": 10986160,
      "step": 16805
    },
    {
      "epoch": 8.81027253668763,
      "grad_norm": 7.5625,
      "learning_rate": 2.1268943283150294e-06,
      "loss": 0.5159,
      "num_input_tokens_seen": 10989584,
      "step": 16810
    },
    {
      "epoch": 8.812893081761006,
      "grad_norm": 6.21875,
      "learning_rate": 2.1176735579884753e-06,
      "loss": 0.573,
      "num_input_tokens_seen": 10992464,
      "step": 16815
    },
    {
      "epoch": 8.815513626834381,
      "grad_norm": 11.0,
      "learning_rate": 2.1084719345474597e-06,
      "loss": 0.5649,
      "num_input_tokens_seen": 10995472,
      "step": 16820
    },
    {
      "epoch": 8.818134171907756,
      "grad_norm": 13.1875,
      "learning_rate": 2.0992894656914895e-06,
      "loss": 0.4495,
      "num_input_tokens_seen": 10999088,
      "step": 16825
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 5.4375,
      "learning_rate": 2.0901261591040333e-06,
      "loss": 0.4867,
      "num_input_tokens_seen": 11002000,
      "step": 16830
    },
    {
      "epoch": 8.823375262054507,
      "grad_norm": 9.0,
      "learning_rate": 2.0809820224525213e-06,
      "loss": 0.6488,
      "num_input_tokens_seen": 11005360,
      "step": 16835
    },
    {
      "epoch": 8.825995807127882,
      "grad_norm": 5.71875,
      "learning_rate": 2.0718570633883576e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 11008272,
      "step": 16840
    },
    {
      "epoch": 8.828616352201259,
      "grad_norm": 5.375,
      "learning_rate": 2.0627512895468883e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 11011760,
      "step": 16845
    },
    {
      "epoch": 8.831236897274634,
      "grad_norm": 6.40625,
      "learning_rate": 2.0536647085474037e-06,
      "loss": 0.5298,
      "num_input_tokens_seen": 11014736,
      "step": 16850
    },
    {
      "epoch": 8.833857442348009,
      "grad_norm": 5.875,
      "learning_rate": 2.044597327993153e-06,
      "loss": 0.4029,
      "num_input_tokens_seen": 11018128,
      "step": 16855
    },
    {
      "epoch": 8.836477987421384,
      "grad_norm": 4.53125,
      "learning_rate": 2.035549155471289e-06,
      "loss": 0.4289,
      "num_input_tokens_seen": 11022032,
      "step": 16860
    },
    {
      "epoch": 8.83909853249476,
      "grad_norm": 6.03125,
      "learning_rate": 2.0265201985529226e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 11024944,
      "step": 16865
    },
    {
      "epoch": 8.841719077568134,
      "grad_norm": 5.625,
      "learning_rate": 2.0175104647930655e-06,
      "loss": 0.4477,
      "num_input_tokens_seen": 11028272,
      "step": 16870
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 15.75,
      "learning_rate": 2.008519961730651e-06,
      "loss": 0.5111,
      "num_input_tokens_seen": 11031088,
      "step": 16875
    },
    {
      "epoch": 8.846960167714885,
      "grad_norm": 9.25,
      "learning_rate": 1.9995486968885284e-06,
      "loss": 0.5282,
      "num_input_tokens_seen": 11033456,
      "step": 16880
    },
    {
      "epoch": 8.84958071278826,
      "grad_norm": 5.53125,
      "learning_rate": 1.990596677773435e-06,
      "loss": 0.5179,
      "num_input_tokens_seen": 11036976,
      "step": 16885
    },
    {
      "epoch": 8.852201257861635,
      "grad_norm": 5.4375,
      "learning_rate": 1.981663911876014e-06,
      "loss": 0.4758,
      "num_input_tokens_seen": 11039472,
      "step": 16890
    },
    {
      "epoch": 8.85482180293501,
      "grad_norm": 6.75,
      "learning_rate": 1.972750406670801e-06,
      "loss": 0.4436,
      "num_input_tokens_seen": 11042064,
      "step": 16895
    },
    {
      "epoch": 8.857442348008385,
      "grad_norm": 4.25,
      "learning_rate": 1.9638561696161962e-06,
      "loss": 0.523,
      "num_input_tokens_seen": 11045232,
      "step": 16900
    },
    {
      "epoch": 8.86006289308176,
      "grad_norm": 6.65625,
      "learning_rate": 1.954981208154502e-06,
      "loss": 0.4361,
      "num_input_tokens_seen": 11048784,
      "step": 16905
    },
    {
      "epoch": 8.862683438155136,
      "grad_norm": 5.03125,
      "learning_rate": 1.9461255297118868e-06,
      "loss": 0.5126,
      "num_input_tokens_seen": 11052528,
      "step": 16910
    },
    {
      "epoch": 8.865303983228511,
      "grad_norm": 9.125,
      "learning_rate": 1.937289141698359e-06,
      "loss": 0.7574,
      "num_input_tokens_seen": 11055824,
      "step": 16915
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 7.40625,
      "learning_rate": 1.928472051507821e-06,
      "loss": 0.3368,
      "num_input_tokens_seen": 11059504,
      "step": 16920
    },
    {
      "epoch": 8.870545073375261,
      "grad_norm": 3.1875,
      "learning_rate": 1.919674266518004e-06,
      "loss": 0.4692,
      "num_input_tokens_seen": 11062544,
      "step": 16925
    },
    {
      "epoch": 8.873165618448636,
      "grad_norm": 4.375,
      "learning_rate": 1.910895794090492e-06,
      "loss": 0.4507,
      "num_input_tokens_seen": 11065584,
      "step": 16930
    },
    {
      "epoch": 8.875786163522012,
      "grad_norm": 8.6875,
      "learning_rate": 1.902136641570712e-06,
      "loss": 0.4138,
      "num_input_tokens_seen": 11068336,
      "step": 16935
    },
    {
      "epoch": 8.878406708595389,
      "grad_norm": 6.6875,
      "learning_rate": 1.8933968162879235e-06,
      "loss": 0.5296,
      "num_input_tokens_seen": 11071184,
      "step": 16940
    },
    {
      "epoch": 8.881027253668764,
      "grad_norm": 7.3125,
      "learning_rate": 1.8846763255552097e-06,
      "loss": 0.3148,
      "num_input_tokens_seen": 11074320,
      "step": 16945
    },
    {
      "epoch": 8.883647798742139,
      "grad_norm": 4.15625,
      "learning_rate": 1.8759751766694811e-06,
      "loss": 0.4468,
      "num_input_tokens_seen": 11077168,
      "step": 16950
    },
    {
      "epoch": 8.886268343815514,
      "grad_norm": 9.25,
      "learning_rate": 1.8672933769114636e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 11080080,
      "step": 16955
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 5.46875,
      "learning_rate": 1.8586309335456908e-06,
      "loss": 0.3648,
      "num_input_tokens_seen": 11083408,
      "step": 16960
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 8.8125,
      "learning_rate": 1.8499878538204951e-06,
      "loss": 0.5884,
      "num_input_tokens_seen": 11087184,
      "step": 16965
    },
    {
      "epoch": 8.89412997903564,
      "grad_norm": 4.71875,
      "learning_rate": 1.8413641449680081e-06,
      "loss": 0.4267,
      "num_input_tokens_seen": 11090864,
      "step": 16970
    },
    {
      "epoch": 8.896750524109015,
      "grad_norm": 5.75,
      "learning_rate": 1.8327598142041658e-06,
      "loss": 0.44,
      "num_input_tokens_seen": 11094288,
      "step": 16975
    },
    {
      "epoch": 8.89937106918239,
      "grad_norm": 4.90625,
      "learning_rate": 1.824174868728673e-06,
      "loss": 0.3179,
      "num_input_tokens_seen": 11096656,
      "step": 16980
    },
    {
      "epoch": 8.901991614255765,
      "grad_norm": 4.75,
      "learning_rate": 1.815609315725017e-06,
      "loss": 0.4552,
      "num_input_tokens_seen": 11100144,
      "step": 16985
    },
    {
      "epoch": 8.90461215932914,
      "grad_norm": 4.40625,
      "learning_rate": 1.80706316236047e-06,
      "loss": 0.417,
      "num_input_tokens_seen": 11103440,
      "step": 16990
    },
    {
      "epoch": 8.907232704402515,
      "grad_norm": 6.84375,
      "learning_rate": 1.7985364157860562e-06,
      "loss": 0.4353,
      "num_input_tokens_seen": 11107056,
      "step": 16995
    },
    {
      "epoch": 8.90985324947589,
      "grad_norm": 7.9375,
      "learning_rate": 1.7900290831365713e-06,
      "loss": 0.4605,
      "num_input_tokens_seen": 11110640,
      "step": 17000
    },
    {
      "epoch": 8.912473794549266,
      "grad_norm": 1.0859375,
      "learning_rate": 1.781541171530554e-06,
      "loss": 0.3457,
      "num_input_tokens_seen": 11117008,
      "step": 17005
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 6.78125,
      "learning_rate": 1.7730726880703125e-06,
      "loss": 0.4315,
      "num_input_tokens_seen": 11120176,
      "step": 17010
    },
    {
      "epoch": 8.917714884696016,
      "grad_norm": 10.0,
      "learning_rate": 1.7646236398418835e-06,
      "loss": 0.5152,
      "num_input_tokens_seen": 11123504,
      "step": 17015
    },
    {
      "epoch": 8.920335429769391,
      "grad_norm": 9.875,
      "learning_rate": 1.7561940339150373e-06,
      "loss": 0.5279,
      "num_input_tokens_seen": 11126000,
      "step": 17020
    },
    {
      "epoch": 8.922955974842766,
      "grad_norm": 3.140625,
      "learning_rate": 1.7477838773432926e-06,
      "loss": 0.4424,
      "num_input_tokens_seen": 11129584,
      "step": 17025
    },
    {
      "epoch": 8.925576519916142,
      "grad_norm": 7.28125,
      "learning_rate": 1.7393931771638839e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 11132240,
      "step": 17030
    },
    {
      "epoch": 8.928197064989519,
      "grad_norm": 6.375,
      "learning_rate": 1.7310219403977563e-06,
      "loss": 0.513,
      "num_input_tokens_seen": 11135120,
      "step": 17035
    },
    {
      "epoch": 8.930817610062894,
      "grad_norm": 13.6875,
      "learning_rate": 1.7226701740495926e-06,
      "loss": 0.3609,
      "num_input_tokens_seen": 11137968,
      "step": 17040
    },
    {
      "epoch": 8.933438155136269,
      "grad_norm": 4.125,
      "learning_rate": 1.714337885107753e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 11140368,
      "step": 17045
    },
    {
      "epoch": 8.936058700209644,
      "grad_norm": 7.1875,
      "learning_rate": 1.7060250805443296e-06,
      "loss": 0.5571,
      "num_input_tokens_seen": 11145168,
      "step": 17050
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 4.71875,
      "learning_rate": 1.6977317673150916e-06,
      "loss": 0.508,
      "num_input_tokens_seen": 11149008,
      "step": 17055
    },
    {
      "epoch": 8.941299790356394,
      "grad_norm": 3.3125,
      "learning_rate": 1.6894579523595022e-06,
      "loss": 0.473,
      "num_input_tokens_seen": 11156336,
      "step": 17060
    },
    {
      "epoch": 8.94392033542977,
      "grad_norm": 6.03125,
      "learning_rate": 1.6812036426007176e-06,
      "loss": 0.5782,
      "num_input_tokens_seen": 11159152,
      "step": 17065
    },
    {
      "epoch": 8.946540880503145,
      "grad_norm": 5.75,
      "learning_rate": 1.6729688449455689e-06,
      "loss": 0.416,
      "num_input_tokens_seen": 11162096,
      "step": 17070
    },
    {
      "epoch": 8.94916142557652,
      "grad_norm": 8.125,
      "learning_rate": 1.6647535662845466e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 11165200,
      "step": 17075
    },
    {
      "epoch": 8.951781970649895,
      "grad_norm": 10.5625,
      "learning_rate": 1.656557813491838e-06,
      "loss": 0.6198,
      "num_input_tokens_seen": 11168368,
      "step": 17080
    },
    {
      "epoch": 8.95440251572327,
      "grad_norm": 4.59375,
      "learning_rate": 1.6483815934252578e-06,
      "loss": 0.5486,
      "num_input_tokens_seen": 11172048,
      "step": 17085
    },
    {
      "epoch": 8.957023060796645,
      "grad_norm": 3.171875,
      "learning_rate": 1.6402249129263025e-06,
      "loss": 0.485,
      "num_input_tokens_seen": 11175600,
      "step": 17090
    },
    {
      "epoch": 8.95964360587002,
      "grad_norm": 7.90625,
      "learning_rate": 1.6320877788201127e-06,
      "loss": 0.4543,
      "num_input_tokens_seen": 11178576,
      "step": 17095
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 6.65625,
      "learning_rate": 1.6239701979154614e-06,
      "loss": 0.5365,
      "num_input_tokens_seen": 11181168,
      "step": 17100
    },
    {
      "epoch": 8.964884696016771,
      "grad_norm": 7.65625,
      "learning_rate": 1.6158721770047762e-06,
      "loss": 0.4242,
      "num_input_tokens_seen": 11184400,
      "step": 17105
    },
    {
      "epoch": 8.967505241090146,
      "grad_norm": 7.53125,
      "learning_rate": 1.6077937228641093e-06,
      "loss": 0.5099,
      "num_input_tokens_seen": 11187536,
      "step": 17110
    },
    {
      "epoch": 8.970125786163521,
      "grad_norm": 7.28125,
      "learning_rate": 1.5997348422531395e-06,
      "loss": 0.4542,
      "num_input_tokens_seen": 11191568,
      "step": 17115
    },
    {
      "epoch": 8.972746331236896,
      "grad_norm": 18.125,
      "learning_rate": 1.5916955419151725e-06,
      "loss": 0.4152,
      "num_input_tokens_seen": 11195024,
      "step": 17120
    },
    {
      "epoch": 8.975366876310272,
      "grad_norm": 9.0,
      "learning_rate": 1.5836758285771303e-06,
      "loss": 0.5222,
      "num_input_tokens_seen": 11198480,
      "step": 17125
    },
    {
      "epoch": 8.977987421383649,
      "grad_norm": 3.59375,
      "learning_rate": 1.5756757089495366e-06,
      "loss": 0.5059,
      "num_input_tokens_seen": 11201712,
      "step": 17130
    },
    {
      "epoch": 8.980607966457024,
      "grad_norm": 4.8125,
      "learning_rate": 1.5676951897265313e-06,
      "loss": 0.4204,
      "num_input_tokens_seen": 11204720,
      "step": 17135
    },
    {
      "epoch": 8.983228511530399,
      "grad_norm": 9.625,
      "learning_rate": 1.5597342775858476e-06,
      "loss": 0.4456,
      "num_input_tokens_seen": 11207440,
      "step": 17140
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 4.4375,
      "learning_rate": 1.5517929791888125e-06,
      "loss": 0.4633,
      "num_input_tokens_seen": 11210384,
      "step": 17145
    },
    {
      "epoch": 8.98846960167715,
      "grad_norm": 10.625,
      "learning_rate": 1.5438713011803385e-06,
      "loss": 0.4839,
      "num_input_tokens_seen": 11213392,
      "step": 17150
    },
    {
      "epoch": 8.991090146750524,
      "grad_norm": 7.875,
      "learning_rate": 1.535969250188926e-06,
      "loss": 0.5648,
      "num_input_tokens_seen": 11216016,
      "step": 17155
    },
    {
      "epoch": 8.9937106918239,
      "grad_norm": 5.6875,
      "learning_rate": 1.5280868328266528e-06,
      "loss": 0.5522,
      "num_input_tokens_seen": 11218160,
      "step": 17160
    },
    {
      "epoch": 8.996331236897275,
      "grad_norm": 7.03125,
      "learning_rate": 1.520224055689165e-06,
      "loss": 0.713,
      "num_input_tokens_seen": 11222000,
      "step": 17165
    },
    {
      "epoch": 8.99895178197065,
      "grad_norm": 8.75,
      "learning_rate": 1.5123809253556692e-06,
      "loss": 0.5143,
      "num_input_tokens_seen": 11224816,
      "step": 17170
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.49351298809051514,
      "eval_runtime": 24.2263,
      "eval_samples_per_second": 35.003,
      "eval_steps_per_second": 8.751,
      "num_input_tokens_seen": 11225416,
      "step": 17172
    },
    {
      "epoch": 9.001572327044025,
      "grad_norm": 6.125,
      "learning_rate": 1.5045574483889463e-06,
      "loss": 0.6027,
      "num_input_tokens_seen": 11227112,
      "step": 17175
    },
    {
      "epoch": 9.0041928721174,
      "grad_norm": 4.875,
      "learning_rate": 1.4967536313353237e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 11229544,
      "step": 17180
    },
    {
      "epoch": 9.006813417190775,
      "grad_norm": 4.59375,
      "learning_rate": 1.4889694807246779e-06,
      "loss": 0.5228,
      "num_input_tokens_seen": 11233640,
      "step": 17185
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 7.5,
      "learning_rate": 1.481205003070424e-06,
      "loss": 0.4124,
      "num_input_tokens_seen": 11236552,
      "step": 17190
    },
    {
      "epoch": 9.012054507337526,
      "grad_norm": 9.625,
      "learning_rate": 1.4734602048695312e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 11240008,
      "step": 17195
    },
    {
      "epoch": 9.014675052410901,
      "grad_norm": 7.21875,
      "learning_rate": 1.465735092602491e-06,
      "loss": 0.4866,
      "num_input_tokens_seen": 11243176,
      "step": 17200
    },
    {
      "epoch": 9.017295597484276,
      "grad_norm": 7.0,
      "learning_rate": 1.4580296727333187e-06,
      "loss": 0.4619,
      "num_input_tokens_seen": 11246568,
      "step": 17205
    },
    {
      "epoch": 9.019916142557651,
      "grad_norm": 3.421875,
      "learning_rate": 1.450343951709568e-06,
      "loss": 0.4461,
      "num_input_tokens_seen": 11251080,
      "step": 17210
    },
    {
      "epoch": 9.022536687631026,
      "grad_norm": 9.0,
      "learning_rate": 1.4426779359622916e-06,
      "loss": 0.6844,
      "num_input_tokens_seen": 11253256,
      "step": 17215
    },
    {
      "epoch": 9.025157232704403,
      "grad_norm": 4.34375,
      "learning_rate": 1.4350316319060585e-06,
      "loss": 0.4636,
      "num_input_tokens_seen": 11256936,
      "step": 17220
    },
    {
      "epoch": 9.027777777777779,
      "grad_norm": 5.28125,
      "learning_rate": 1.4274050459389594e-06,
      "loss": 0.5046,
      "num_input_tokens_seen": 11259784,
      "step": 17225
    },
    {
      "epoch": 9.030398322851154,
      "grad_norm": 5.53125,
      "learning_rate": 1.4197981844425583e-06,
      "loss": 0.466,
      "num_input_tokens_seen": 11262728,
      "step": 17230
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 4.96875,
      "learning_rate": 1.4122110537819365e-06,
      "loss": 0.4476,
      "num_input_tokens_seen": 11265640,
      "step": 17235
    },
    {
      "epoch": 9.035639412997904,
      "grad_norm": 3.671875,
      "learning_rate": 1.4046436603056601e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 11270344,
      "step": 17240
    },
    {
      "epoch": 9.03825995807128,
      "grad_norm": 29.375,
      "learning_rate": 1.397096010345772e-06,
      "loss": 0.6366,
      "num_input_tokens_seen": 11272584,
      "step": 17245
    },
    {
      "epoch": 9.040880503144654,
      "grad_norm": 7.65625,
      "learning_rate": 1.3895681102178094e-06,
      "loss": 0.5303,
      "num_input_tokens_seen": 11276872,
      "step": 17250
    },
    {
      "epoch": 9.04350104821803,
      "grad_norm": 5.5625,
      "learning_rate": 1.3820599662207695e-06,
      "loss": 0.4826,
      "num_input_tokens_seen": 11279688,
      "step": 17255
    },
    {
      "epoch": 9.046121593291405,
      "grad_norm": 5.59375,
      "learning_rate": 1.3745715846371244e-06,
      "loss": 0.4404,
      "num_input_tokens_seen": 11282888,
      "step": 17260
    },
    {
      "epoch": 9.04874213836478,
      "grad_norm": 7.25,
      "learning_rate": 1.3671029717328142e-06,
      "loss": 0.5789,
      "num_input_tokens_seen": 11285928,
      "step": 17265
    },
    {
      "epoch": 9.051362683438155,
      "grad_norm": 8.75,
      "learning_rate": 1.3596541337572265e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 11289288,
      "step": 17270
    },
    {
      "epoch": 9.05398322851153,
      "grad_norm": 3.6875,
      "learning_rate": 1.3522250769432115e-06,
      "loss": 0.629,
      "num_input_tokens_seen": 11295976,
      "step": 17275
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 4.78125,
      "learning_rate": 1.3448158075070687e-06,
      "loss": 0.4986,
      "num_input_tokens_seen": 11299816,
      "step": 17280
    },
    {
      "epoch": 9.05922431865828,
      "grad_norm": 4.75,
      "learning_rate": 1.337426331648528e-06,
      "loss": 0.5728,
      "num_input_tokens_seen": 11303176,
      "step": 17285
    },
    {
      "epoch": 9.061844863731656,
      "grad_norm": 5.75,
      "learning_rate": 1.3300566555507709e-06,
      "loss": 0.3682,
      "num_input_tokens_seen": 11306376,
      "step": 17290
    },
    {
      "epoch": 9.064465408805031,
      "grad_norm": 5.09375,
      "learning_rate": 1.3227067853804065e-06,
      "loss": 0.3758,
      "num_input_tokens_seen": 11310696,
      "step": 17295
    },
    {
      "epoch": 9.067085953878406,
      "grad_norm": 12.6875,
      "learning_rate": 1.315376727287465e-06,
      "loss": 0.4249,
      "num_input_tokens_seen": 11314888,
      "step": 17300
    },
    {
      "epoch": 9.069706498951781,
      "grad_norm": 7.9375,
      "learning_rate": 1.3080664874054127e-06,
      "loss": 0.5418,
      "num_input_tokens_seen": 11317576,
      "step": 17305
    },
    {
      "epoch": 9.072327044025156,
      "grad_norm": 5.78125,
      "learning_rate": 1.3007760718511176e-06,
      "loss": 0.5048,
      "num_input_tokens_seen": 11320840,
      "step": 17310
    },
    {
      "epoch": 9.074947589098532,
      "grad_norm": 5.59375,
      "learning_rate": 1.2935054867248692e-06,
      "loss": 0.3231,
      "num_input_tokens_seen": 11324040,
      "step": 17315
    },
    {
      "epoch": 9.077568134171909,
      "grad_norm": 8.8125,
      "learning_rate": 1.2862547381103567e-06,
      "loss": 0.5572,
      "num_input_tokens_seen": 11327784,
      "step": 17320
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 4.46875,
      "learning_rate": 1.2790238320746827e-06,
      "loss": 0.4423,
      "num_input_tokens_seen": 11331304,
      "step": 17325
    },
    {
      "epoch": 9.082809224318659,
      "grad_norm": 6.78125,
      "learning_rate": 1.271812774668335e-06,
      "loss": 0.5896,
      "num_input_tokens_seen": 11334312,
      "step": 17330
    },
    {
      "epoch": 9.085429769392034,
      "grad_norm": 3.53125,
      "learning_rate": 1.2646215719251952e-06,
      "loss": 0.5161,
      "num_input_tokens_seen": 11339144,
      "step": 17335
    },
    {
      "epoch": 9.08805031446541,
      "grad_norm": 5.875,
      "learning_rate": 1.2574502298625334e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 11342664,
      "step": 17340
    },
    {
      "epoch": 9.090670859538784,
      "grad_norm": 6.1875,
      "learning_rate": 1.250298754481008e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 11345768,
      "step": 17345
    },
    {
      "epoch": 9.09329140461216,
      "grad_norm": 4.84375,
      "learning_rate": 1.2431671517646403e-06,
      "loss": 0.6224,
      "num_input_tokens_seen": 11349448,
      "step": 17350
    },
    {
      "epoch": 9.095911949685535,
      "grad_norm": 5.90625,
      "learning_rate": 1.2360554276808295e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 11352744,
      "step": 17355
    },
    {
      "epoch": 9.09853249475891,
      "grad_norm": 22.375,
      "learning_rate": 1.228963588180343e-06,
      "loss": 0.5347,
      "num_input_tokens_seen": 11354920,
      "step": 17360
    },
    {
      "epoch": 9.101153039832285,
      "grad_norm": 7.65625,
      "learning_rate": 1.2218916391973118e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 11359240,
      "step": 17365
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 7.03125,
      "learning_rate": 1.2148395866492135e-06,
      "loss": 0.5367,
      "num_input_tokens_seen": 11362056,
      "step": 17370
    },
    {
      "epoch": 9.106394129979035,
      "grad_norm": 8.3125,
      "learning_rate": 1.2078074364368862e-06,
      "loss": 0.3815,
      "num_input_tokens_seen": 11364648,
      "step": 17375
    },
    {
      "epoch": 9.10901467505241,
      "grad_norm": 5.96875,
      "learning_rate": 1.2007951944445122e-06,
      "loss": 0.4212,
      "num_input_tokens_seen": 11368168,
      "step": 17380
    },
    {
      "epoch": 9.111635220125786,
      "grad_norm": 8.5,
      "learning_rate": 1.1938028665396173e-06,
      "loss": 0.5277,
      "num_input_tokens_seen": 11370856,
      "step": 17385
    },
    {
      "epoch": 9.114255765199161,
      "grad_norm": 6.09375,
      "learning_rate": 1.1868304585730571e-06,
      "loss": 0.473,
      "num_input_tokens_seen": 11373768,
      "step": 17390
    },
    {
      "epoch": 9.116876310272536,
      "grad_norm": 5.78125,
      "learning_rate": 1.1798779763790346e-06,
      "loss": 0.5188,
      "num_input_tokens_seen": 11377192,
      "step": 17395
    },
    {
      "epoch": 9.119496855345911,
      "grad_norm": 6.09375,
      "learning_rate": 1.1729454257750544e-06,
      "loss": 0.4571,
      "num_input_tokens_seen": 11379912,
      "step": 17400
    },
    {
      "epoch": 9.122117400419286,
      "grad_norm": 6.0625,
      "learning_rate": 1.1660328125619652e-06,
      "loss": 0.4889,
      "num_input_tokens_seen": 11383496,
      "step": 17405
    },
    {
      "epoch": 9.124737945492662,
      "grad_norm": 8.875,
      "learning_rate": 1.1591401425239318e-06,
      "loss": 0.6275,
      "num_input_tokens_seen": 11386504,
      "step": 17410
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 25.0,
      "learning_rate": 1.1522674214284158e-06,
      "loss": 0.4105,
      "num_input_tokens_seen": 11388776,
      "step": 17415
    },
    {
      "epoch": 9.129979035639414,
      "grad_norm": 4.6875,
      "learning_rate": 1.145414655026203e-06,
      "loss": 0.6205,
      "num_input_tokens_seen": 11391976,
      "step": 17420
    },
    {
      "epoch": 9.132599580712789,
      "grad_norm": 4.96875,
      "learning_rate": 1.1385818490513733e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 11394856,
      "step": 17425
    },
    {
      "epoch": 9.135220125786164,
      "grad_norm": 9.5625,
      "learning_rate": 1.1317690092213007e-06,
      "loss": 0.5231,
      "num_input_tokens_seen": 11397896,
      "step": 17430
    },
    {
      "epoch": 9.13784067085954,
      "grad_norm": 8.9375,
      "learning_rate": 1.124976141236675e-06,
      "loss": 0.5679,
      "num_input_tokens_seen": 11401096,
      "step": 17435
    },
    {
      "epoch": 9.140461215932914,
      "grad_norm": 10.1875,
      "learning_rate": 1.1182032507814354e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 11404616,
      "step": 17440
    },
    {
      "epoch": 9.14308176100629,
      "grad_norm": 4.34375,
      "learning_rate": 1.1114503435228434e-06,
      "loss": 0.4633,
      "num_input_tokens_seen": 11407848,
      "step": 17445
    },
    {
      "epoch": 9.145702306079665,
      "grad_norm": 9.125,
      "learning_rate": 1.1047174251114234e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 11410760,
      "step": 17450
    },
    {
      "epoch": 9.14832285115304,
      "grad_norm": 7.125,
      "learning_rate": 1.0980045011809604e-06,
      "loss": 0.5424,
      "num_input_tokens_seen": 11413864,
      "step": 17455
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 6.46875,
      "learning_rate": 1.0913115773485388e-06,
      "loss": 0.3975,
      "num_input_tokens_seen": 11416808,
      "step": 17460
    },
    {
      "epoch": 9.15356394129979,
      "grad_norm": 7.53125,
      "learning_rate": 1.084638659214482e-06,
      "loss": 0.5496,
      "num_input_tokens_seen": 11419944,
      "step": 17465
    },
    {
      "epoch": 9.156184486373165,
      "grad_norm": 8.8125,
      "learning_rate": 1.0779857523623815e-06,
      "loss": 0.4363,
      "num_input_tokens_seen": 11422728,
      "step": 17470
    },
    {
      "epoch": 9.15880503144654,
      "grad_norm": 5.15625,
      "learning_rate": 1.071352862359093e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 11430984,
      "step": 17475
    },
    {
      "epoch": 9.161425576519916,
      "grad_norm": 10.5625,
      "learning_rate": 1.0647399947547127e-06,
      "loss": 0.4802,
      "num_input_tokens_seen": 11433672,
      "step": 17480
    },
    {
      "epoch": 9.164046121593291,
      "grad_norm": 9.0,
      "learning_rate": 1.0581471550825812e-06,
      "loss": 0.5024,
      "num_input_tokens_seen": 11436168,
      "step": 17485
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 3.15625,
      "learning_rate": 1.0515743488592939e-06,
      "loss": 0.3572,
      "num_input_tokens_seen": 11439528,
      "step": 17490
    },
    {
      "epoch": 9.169287211740041,
      "grad_norm": 9.25,
      "learning_rate": 1.0450215815846736e-06,
      "loss": 0.523,
      "num_input_tokens_seen": 11442312,
      "step": 17495
    },
    {
      "epoch": 9.171907756813416,
      "grad_norm": 8.25,
      "learning_rate": 1.0384888587417736e-06,
      "loss": 0.4312,
      "num_input_tokens_seen": 11445416,
      "step": 17500
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 9.75,
      "learning_rate": 1.0319761857968735e-06,
      "loss": 0.646,
      "num_input_tokens_seen": 11448712,
      "step": 17505
    },
    {
      "epoch": 9.177148846960169,
      "grad_norm": 6.125,
      "learning_rate": 1.0254835681994895e-06,
      "loss": 0.528,
      "num_input_tokens_seen": 11452104,
      "step": 17510
    },
    {
      "epoch": 9.179769392033544,
      "grad_norm": 5.78125,
      "learning_rate": 1.0190110113823426e-06,
      "loss": 0.3519,
      "num_input_tokens_seen": 11456136,
      "step": 17515
    },
    {
      "epoch": 9.182389937106919,
      "grad_norm": 6.28125,
      "learning_rate": 1.0125585207613752e-06,
      "loss": 0.3958,
      "num_input_tokens_seen": 11459272,
      "step": 17520
    },
    {
      "epoch": 9.185010482180294,
      "grad_norm": 3.703125,
      "learning_rate": 1.0061261017357327e-06,
      "loss": 0.4466,
      "num_input_tokens_seen": 11463208,
      "step": 17525
    },
    {
      "epoch": 9.18763102725367,
      "grad_norm": 8.4375,
      "learning_rate": 9.997137596877732e-07,
      "loss": 0.4597,
      "num_input_tokens_seen": 11466472,
      "step": 17530
    },
    {
      "epoch": 9.190251572327044,
      "grad_norm": 9.875,
      "learning_rate": 9.93321499983052e-07,
      "loss": 0.5655,
      "num_input_tokens_seen": 11469192,
      "step": 17535
    },
    {
      "epoch": 9.19287211740042,
      "grad_norm": 7.5625,
      "learning_rate": 9.869493279703158e-07,
      "loss": 0.4328,
      "num_input_tokens_seen": 11472232,
      "step": 17540
    },
    {
      "epoch": 9.195492662473795,
      "grad_norm": 3.46875,
      "learning_rate": 9.805972489815102e-07,
      "loss": 0.5163,
      "num_input_tokens_seen": 11476040,
      "step": 17545
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 7.28125,
      "learning_rate": 9.742652683317643e-07,
      "loss": 0.4623,
      "num_input_tokens_seen": 11478728,
      "step": 17550
    },
    {
      "epoch": 9.200733752620545,
      "grad_norm": 8.1875,
      "learning_rate": 9.679533913193927e-07,
      "loss": 0.5696,
      "num_input_tokens_seen": 11481448,
      "step": 17555
    },
    {
      "epoch": 9.20335429769392,
      "grad_norm": 4.46875,
      "learning_rate": 9.61661623225879e-07,
      "loss": 0.491,
      "num_input_tokens_seen": 11484424,
      "step": 17560
    },
    {
      "epoch": 9.205974842767295,
      "grad_norm": 4.9375,
      "learning_rate": 9.553899693158951e-07,
      "loss": 0.4625,
      "num_input_tokens_seen": 11487560,
      "step": 17565
    },
    {
      "epoch": 9.20859538784067,
      "grad_norm": 10.9375,
      "learning_rate": 9.491384348372684e-07,
      "loss": 0.4761,
      "num_input_tokens_seen": 11490632,
      "step": 17570
    },
    {
      "epoch": 9.211215932914046,
      "grad_norm": 9.3125,
      "learning_rate": 9.429070250210004e-07,
      "loss": 0.4585,
      "num_input_tokens_seen": 11493032,
      "step": 17575
    },
    {
      "epoch": 9.213836477987421,
      "grad_norm": 5.75,
      "learning_rate": 9.366957450812535e-07,
      "loss": 0.407,
      "num_input_tokens_seen": 11496200,
      "step": 17580
    },
    {
      "epoch": 9.216457023060796,
      "grad_norm": 9.6875,
      "learning_rate": 9.305046002153345e-07,
      "loss": 0.5026,
      "num_input_tokens_seen": 11498696,
      "step": 17585
    },
    {
      "epoch": 9.219077568134171,
      "grad_norm": 5.84375,
      "learning_rate": 9.243335956037186e-07,
      "loss": 0.4041,
      "num_input_tokens_seen": 11502312,
      "step": 17590
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 4.75,
      "learning_rate": 9.181827364100171e-07,
      "loss": 0.4913,
      "num_input_tokens_seen": 11505160,
      "step": 17595
    },
    {
      "epoch": 9.224318658280922,
      "grad_norm": 5.65625,
      "learning_rate": 9.120520277809852e-07,
      "loss": 0.3896,
      "num_input_tokens_seen": 11508456,
      "step": 17600
    },
    {
      "epoch": 9.226939203354299,
      "grad_norm": 6.03125,
      "learning_rate": 9.059414748465278e-07,
      "loss": 0.5624,
      "num_input_tokens_seen": 11511816,
      "step": 17605
    },
    {
      "epoch": 9.229559748427674,
      "grad_norm": 6.5,
      "learning_rate": 8.998510827196715e-07,
      "loss": 0.5264,
      "num_input_tokens_seen": 11514088,
      "step": 17610
    },
    {
      "epoch": 9.232180293501049,
      "grad_norm": 6.21875,
      "learning_rate": 8.937808564965733e-07,
      "loss": 0.4157,
      "num_input_tokens_seen": 11517096,
      "step": 17615
    },
    {
      "epoch": 9.234800838574424,
      "grad_norm": 6.40625,
      "learning_rate": 8.877308012565339e-07,
      "loss": 0.417,
      "num_input_tokens_seen": 11520168,
      "step": 17620
    },
    {
      "epoch": 9.2374213836478,
      "grad_norm": 7.28125,
      "learning_rate": 8.817009220619482e-07,
      "loss": 0.6021,
      "num_input_tokens_seen": 11523944,
      "step": 17625
    },
    {
      "epoch": 9.240041928721174,
      "grad_norm": 3.3125,
      "learning_rate": 8.756912239583554e-07,
      "loss": 0.6396,
      "num_input_tokens_seen": 11527720,
      "step": 17630
    },
    {
      "epoch": 9.24266247379455,
      "grad_norm": 5.09375,
      "learning_rate": 8.697017119743911e-07,
      "loss": 0.5428,
      "num_input_tokens_seen": 11531304,
      "step": 17635
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 5.5625,
      "learning_rate": 8.637323911218048e-07,
      "loss": 0.4414,
      "num_input_tokens_seen": 11535176,
      "step": 17640
    },
    {
      "epoch": 9.2479035639413,
      "grad_norm": 3.703125,
      "learning_rate": 8.577832663954538e-07,
      "loss": 0.5342,
      "num_input_tokens_seen": 11539016,
      "step": 17645
    },
    {
      "epoch": 9.250524109014675,
      "grad_norm": 4.5625,
      "learning_rate": 8.51854342773295e-07,
      "loss": 0.4172,
      "num_input_tokens_seen": 11543752,
      "step": 17650
    },
    {
      "epoch": 9.25314465408805,
      "grad_norm": 8.0,
      "learning_rate": 8.459456252163739e-07,
      "loss": 0.4893,
      "num_input_tokens_seen": 11546664,
      "step": 17655
    },
    {
      "epoch": 9.255765199161425,
      "grad_norm": 13.6875,
      "learning_rate": 8.400571186688466e-07,
      "loss": 0.5014,
      "num_input_tokens_seen": 11549032,
      "step": 17660
    },
    {
      "epoch": 9.2583857442348,
      "grad_norm": 5.59375,
      "learning_rate": 8.341888280579386e-07,
      "loss": 0.5769,
      "num_input_tokens_seen": 11552328,
      "step": 17665
    },
    {
      "epoch": 9.261006289308176,
      "grad_norm": 6.3125,
      "learning_rate": 8.283407582939689e-07,
      "loss": 0.5995,
      "num_input_tokens_seen": 11555464,
      "step": 17670
    },
    {
      "epoch": 9.26362683438155,
      "grad_norm": 7.0625,
      "learning_rate": 8.22512914270332e-07,
      "loss": 0.4845,
      "num_input_tokens_seen": 11558088,
      "step": 17675
    },
    {
      "epoch": 9.266247379454926,
      "grad_norm": 6.53125,
      "learning_rate": 8.167053008635101e-07,
      "loss": 0.5368,
      "num_input_tokens_seen": 11560872,
      "step": 17680
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 6.96875,
      "learning_rate": 8.109179229330438e-07,
      "loss": 0.466,
      "num_input_tokens_seen": 11564264,
      "step": 17685
    },
    {
      "epoch": 9.271488469601676,
      "grad_norm": 4.8125,
      "learning_rate": 8.051507853215401e-07,
      "loss": 0.4621,
      "num_input_tokens_seen": 11567656,
      "step": 17690
    },
    {
      "epoch": 9.274109014675052,
      "grad_norm": 6.625,
      "learning_rate": 7.994038928546887e-07,
      "loss": 0.5394,
      "num_input_tokens_seen": 11571176,
      "step": 17695
    },
    {
      "epoch": 9.276729559748428,
      "grad_norm": 5.9375,
      "learning_rate": 7.93677250341221e-07,
      "loss": 0.5379,
      "num_input_tokens_seen": 11574216,
      "step": 17700
    },
    {
      "epoch": 9.279350104821804,
      "grad_norm": 6.125,
      "learning_rate": 7.879708625729287e-07,
      "loss": 0.3809,
      "num_input_tokens_seen": 11577608,
      "step": 17705
    },
    {
      "epoch": 9.281970649895179,
      "grad_norm": 5.84375,
      "learning_rate": 7.822847343246564e-07,
      "loss": 0.5214,
      "num_input_tokens_seen": 11581000,
      "step": 17710
    },
    {
      "epoch": 9.284591194968554,
      "grad_norm": 7.53125,
      "learning_rate": 7.766188703542954e-07,
      "loss": 0.4988,
      "num_input_tokens_seen": 11584840,
      "step": 17715
    },
    {
      "epoch": 9.28721174004193,
      "grad_norm": 5.5,
      "learning_rate": 7.709732754027866e-07,
      "loss": 0.5136,
      "num_input_tokens_seen": 11587912,
      "step": 17720
    },
    {
      "epoch": 9.289832285115304,
      "grad_norm": 8.3125,
      "learning_rate": 7.653479541941038e-07,
      "loss": 0.4969,
      "num_input_tokens_seen": 11591752,
      "step": 17725
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 9.3125,
      "learning_rate": 7.597429114352572e-07,
      "loss": 0.5691,
      "num_input_tokens_seen": 11594248,
      "step": 17730
    },
    {
      "epoch": 9.295073375262055,
      "grad_norm": 7.9375,
      "learning_rate": 7.541581518162922e-07,
      "loss": 0.4843,
      "num_input_tokens_seen": 11597448,
      "step": 17735
    },
    {
      "epoch": 9.29769392033543,
      "grad_norm": 5.875,
      "learning_rate": 7.485936800102788e-07,
      "loss": 0.7091,
      "num_input_tokens_seen": 11600360,
      "step": 17740
    },
    {
      "epoch": 9.300314465408805,
      "grad_norm": 8.5,
      "learning_rate": 7.430495006733152e-07,
      "loss": 0.5465,
      "num_input_tokens_seen": 11603528,
      "step": 17745
    },
    {
      "epoch": 9.30293501048218,
      "grad_norm": 6.78125,
      "learning_rate": 7.375256184445178e-07,
      "loss": 0.5333,
      "num_input_tokens_seen": 11606600,
      "step": 17750
    },
    {
      "epoch": 9.305555555555555,
      "grad_norm": 4.84375,
      "learning_rate": 7.320220379460146e-07,
      "loss": 0.4328,
      "num_input_tokens_seen": 11609064,
      "step": 17755
    },
    {
      "epoch": 9.30817610062893,
      "grad_norm": 5.53125,
      "learning_rate": 7.265387637829524e-07,
      "loss": 0.4376,
      "num_input_tokens_seen": 11612328,
      "step": 17760
    },
    {
      "epoch": 9.310796645702306,
      "grad_norm": 4.875,
      "learning_rate": 7.210758005434887e-07,
      "loss": 0.5098,
      "num_input_tokens_seen": 11615912,
      "step": 17765
    },
    {
      "epoch": 9.31341719077568,
      "grad_norm": 9.625,
      "learning_rate": 7.156331527987753e-07,
      "loss": 0.542,
      "num_input_tokens_seen": 11618888,
      "step": 17770
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 8.9375,
      "learning_rate": 7.102108251029777e-07,
      "loss": 0.7244,
      "num_input_tokens_seen": 11621544,
      "step": 17775
    },
    {
      "epoch": 9.318658280922431,
      "grad_norm": 7.03125,
      "learning_rate": 7.04808821993247e-07,
      "loss": 0.4227,
      "num_input_tokens_seen": 11624040,
      "step": 17780
    },
    {
      "epoch": 9.321278825995806,
      "grad_norm": 7.0625,
      "learning_rate": 6.994271479897314e-07,
      "loss": 0.5116,
      "num_input_tokens_seen": 11626728,
      "step": 17785
    },
    {
      "epoch": 9.323899371069182,
      "grad_norm": 9.8125,
      "learning_rate": 6.940658075955759e-07,
      "loss": 0.3717,
      "num_input_tokens_seen": 11629832,
      "step": 17790
    },
    {
      "epoch": 9.326519916142558,
      "grad_norm": 5.34375,
      "learning_rate": 6.887248052969003e-07,
      "loss": 0.3974,
      "num_input_tokens_seen": 11633320,
      "step": 17795
    },
    {
      "epoch": 9.329140461215934,
      "grad_norm": 7.53125,
      "learning_rate": 6.834041455628104e-07,
      "loss": 0.4518,
      "num_input_tokens_seen": 11636104,
      "step": 17800
    },
    {
      "epoch": 9.331761006289309,
      "grad_norm": 4.5,
      "learning_rate": 6.781038328454003e-07,
      "loss": 0.4391,
      "num_input_tokens_seen": 11639752,
      "step": 17805
    },
    {
      "epoch": 9.334381551362684,
      "grad_norm": 13.0625,
      "learning_rate": 6.728238715797169e-07,
      "loss": 0.4564,
      "num_input_tokens_seen": 11642664,
      "step": 17810
    },
    {
      "epoch": 9.33700209643606,
      "grad_norm": 6.0,
      "learning_rate": 6.675642661838011e-07,
      "loss": 0.5279,
      "num_input_tokens_seen": 11646024,
      "step": 17815
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 4.625,
      "learning_rate": 6.623250210586463e-07,
      "loss": 0.464,
      "num_input_tokens_seen": 11649192,
      "step": 17820
    },
    {
      "epoch": 9.34224318658281,
      "grad_norm": 4.03125,
      "learning_rate": 6.571061405882095e-07,
      "loss": 0.5668,
      "num_input_tokens_seen": 11653192,
      "step": 17825
    },
    {
      "epoch": 9.344863731656185,
      "grad_norm": 12.5625,
      "learning_rate": 6.519076291394172e-07,
      "loss": 0.4726,
      "num_input_tokens_seen": 11657224,
      "step": 17830
    },
    {
      "epoch": 9.34748427672956,
      "grad_norm": 6.0625,
      "learning_rate": 6.467294910621452e-07,
      "loss": 0.6068,
      "num_input_tokens_seen": 11660328,
      "step": 17835
    },
    {
      "epoch": 9.350104821802935,
      "grad_norm": 4.75,
      "learning_rate": 6.415717306892193e-07,
      "loss": 0.5503,
      "num_input_tokens_seen": 11663400,
      "step": 17840
    },
    {
      "epoch": 9.35272536687631,
      "grad_norm": 8.5625,
      "learning_rate": 6.364343523364263e-07,
      "loss": 0.5181,
      "num_input_tokens_seen": 11665992,
      "step": 17845
    },
    {
      "epoch": 9.355345911949685,
      "grad_norm": 13.0625,
      "learning_rate": 6.313173603024802e-07,
      "loss": 0.4336,
      "num_input_tokens_seen": 11669000,
      "step": 17850
    },
    {
      "epoch": 9.35796645702306,
      "grad_norm": 7.1875,
      "learning_rate": 6.262207588690533e-07,
      "loss": 0.4893,
      "num_input_tokens_seen": 11671976,
      "step": 17855
    },
    {
      "epoch": 9.360587002096436,
      "grad_norm": 4.625,
      "learning_rate": 6.211445523007398e-07,
      "loss": 0.4271,
      "num_input_tokens_seen": 11674984,
      "step": 17860
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 7.65625,
      "learning_rate": 6.160887448450892e-07,
      "loss": 0.4441,
      "num_input_tokens_seen": 11677864,
      "step": 17865
    },
    {
      "epoch": 9.365828092243186,
      "grad_norm": 3.046875,
      "learning_rate": 6.11053340732562e-07,
      "loss": 0.4194,
      "num_input_tokens_seen": 11681128,
      "step": 17870
    },
    {
      "epoch": 9.368448637316561,
      "grad_norm": 10.1875,
      "learning_rate": 6.060383441765544e-07,
      "loss": 0.4865,
      "num_input_tokens_seen": 11683880,
      "step": 17875
    },
    {
      "epoch": 9.371069182389936,
      "grad_norm": 6.84375,
      "learning_rate": 6.01043759373393e-07,
      "loss": 0.4944,
      "num_input_tokens_seen": 11687496,
      "step": 17880
    },
    {
      "epoch": 9.373689727463312,
      "grad_norm": 8.75,
      "learning_rate": 5.960695905023128e-07,
      "loss": 0.4845,
      "num_input_tokens_seen": 11690696,
      "step": 17885
    },
    {
      "epoch": 9.376310272536688,
      "grad_norm": 5.8125,
      "learning_rate": 5.91115841725473e-07,
      "loss": 0.4445,
      "num_input_tokens_seen": 11693544,
      "step": 17890
    },
    {
      "epoch": 9.378930817610064,
      "grad_norm": 6.84375,
      "learning_rate": 5.861825171879415e-07,
      "loss": 0.5512,
      "num_input_tokens_seen": 11696840,
      "step": 17895
    },
    {
      "epoch": 9.381551362683439,
      "grad_norm": 4.9375,
      "learning_rate": 5.812696210177021e-07,
      "loss": 0.503,
      "num_input_tokens_seen": 11700264,
      "step": 17900
    },
    {
      "epoch": 9.384171907756814,
      "grad_norm": 8.5625,
      "learning_rate": 5.763771573256415e-07,
      "loss": 0.3488,
      "num_input_tokens_seen": 11703496,
      "step": 17905
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 8.3125,
      "learning_rate": 5.715051302055491e-07,
      "loss": 0.5982,
      "num_input_tokens_seen": 11706664,
      "step": 17910
    },
    {
      "epoch": 9.389412997903564,
      "grad_norm": 7.125,
      "learning_rate": 5.666535437341108e-07,
      "loss": 0.5115,
      "num_input_tokens_seen": 11709480,
      "step": 17915
    },
    {
      "epoch": 9.39203354297694,
      "grad_norm": 4.90625,
      "learning_rate": 5.618224019709212e-07,
      "loss": 0.3904,
      "num_input_tokens_seen": 11712712,
      "step": 17920
    },
    {
      "epoch": 9.394654088050315,
      "grad_norm": 9.25,
      "learning_rate": 5.570117089584548e-07,
      "loss": 0.5913,
      "num_input_tokens_seen": 11715464,
      "step": 17925
    },
    {
      "epoch": 9.39727463312369,
      "grad_norm": 5.5,
      "learning_rate": 5.522214687220751e-07,
      "loss": 0.563,
      "num_input_tokens_seen": 11718120,
      "step": 17930
    },
    {
      "epoch": 9.399895178197065,
      "grad_norm": 6.625,
      "learning_rate": 5.474516852700451e-07,
      "loss": 0.5187,
      "num_input_tokens_seen": 11721512,
      "step": 17935
    },
    {
      "epoch": 9.40251572327044,
      "grad_norm": 6.875,
      "learning_rate": 5.427023625934946e-07,
      "loss": 0.3463,
      "num_input_tokens_seen": 11724008,
      "step": 17940
    },
    {
      "epoch": 9.405136268343815,
      "grad_norm": 5.46875,
      "learning_rate": 5.379735046664419e-07,
      "loss": 0.4758,
      "num_input_tokens_seen": 11726696,
      "step": 17945
    },
    {
      "epoch": 9.40775681341719,
      "grad_norm": 2.796875,
      "learning_rate": 5.33265115445783e-07,
      "loss": 0.3933,
      "num_input_tokens_seen": 11730408,
      "step": 17950
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 7.75,
      "learning_rate": 5.285771988712746e-07,
      "loss": 0.58,
      "num_input_tokens_seen": 11733320,
      "step": 17955
    },
    {
      "epoch": 9.41299790356394,
      "grad_norm": 7.375,
      "learning_rate": 5.239097588655595e-07,
      "loss": 0.4739,
      "num_input_tokens_seen": 11738472,
      "step": 17960
    },
    {
      "epoch": 9.415618448637316,
      "grad_norm": 6.21875,
      "learning_rate": 5.192627993341359e-07,
      "loss": 0.557,
      "num_input_tokens_seen": 11741032,
      "step": 17965
    },
    {
      "epoch": 9.418238993710691,
      "grad_norm": 10.4375,
      "learning_rate": 5.146363241653657e-07,
      "loss": 0.4831,
      "num_input_tokens_seen": 11744328,
      "step": 17970
    },
    {
      "epoch": 9.420859538784066,
      "grad_norm": 8.5,
      "learning_rate": 5.100303372304716e-07,
      "loss": 0.3752,
      "num_input_tokens_seen": 11747976,
      "step": 17975
    },
    {
      "epoch": 9.423480083857442,
      "grad_norm": 5.875,
      "learning_rate": 5.054448423835373e-07,
      "loss": 0.6,
      "num_input_tokens_seen": 11751144,
      "step": 17980
    },
    {
      "epoch": 9.426100628930818,
      "grad_norm": 6.09375,
      "learning_rate": 5.008798434614908e-07,
      "loss": 0.5124,
      "num_input_tokens_seen": 11754312,
      "step": 17985
    },
    {
      "epoch": 9.428721174004194,
      "grad_norm": 3.890625,
      "learning_rate": 4.963353442841156e-07,
      "loss": 0.3722,
      "num_input_tokens_seen": 11757896,
      "step": 17990
    },
    {
      "epoch": 9.431341719077569,
      "grad_norm": 5.34375,
      "learning_rate": 4.918113486540393e-07,
      "loss": 0.4564,
      "num_input_tokens_seen": 11760840,
      "step": 17995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 8.0,
      "learning_rate": 4.873078603567421e-07,
      "loss": 0.4365,
      "num_input_tokens_seen": 11764136,
      "step": 18000
    },
    {
      "epoch": 9.43658280922432,
      "grad_norm": 23.125,
      "learning_rate": 4.828248831605292e-07,
      "loss": 0.6035,
      "num_input_tokens_seen": 11766696,
      "step": 18005
    },
    {
      "epoch": 9.439203354297694,
      "grad_norm": 15.3125,
      "learning_rate": 4.783624208165554e-07,
      "loss": 0.5351,
      "num_input_tokens_seen": 11769160,
      "step": 18010
    },
    {
      "epoch": 9.44182389937107,
      "grad_norm": 3.375,
      "learning_rate": 4.739204770588035e-07,
      "loss": 0.523,
      "num_input_tokens_seen": 11772360,
      "step": 18015
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 10.125,
      "learning_rate": 4.694990556040918e-07,
      "loss": 0.5305,
      "num_input_tokens_seen": 11774984,
      "step": 18020
    },
    {
      "epoch": 9.44706498951782,
      "grad_norm": 5.6875,
      "learning_rate": 4.65098160152061e-07,
      "loss": 0.468,
      "num_input_tokens_seen": 11777928,
      "step": 18025
    },
    {
      "epoch": 9.449685534591195,
      "grad_norm": 6.8125,
      "learning_rate": 4.6071779438517924e-07,
      "loss": 0.5202,
      "num_input_tokens_seen": 11781096,
      "step": 18030
    },
    {
      "epoch": 9.45230607966457,
      "grad_norm": 3.59375,
      "learning_rate": 4.563579619687369e-07,
      "loss": 0.4382,
      "num_input_tokens_seen": 11784104,
      "step": 18035
    },
    {
      "epoch": 9.454926624737945,
      "grad_norm": 4.3125,
      "learning_rate": 4.5201866655084636e-07,
      "loss": 0.5205,
      "num_input_tokens_seen": 11787304,
      "step": 18040
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 7.96875,
      "learning_rate": 4.4769991176242533e-07,
      "loss": 0.3817,
      "num_input_tokens_seen": 11790024,
      "step": 18045
    },
    {
      "epoch": 9.460167714884696,
      "grad_norm": 6.28125,
      "learning_rate": 4.4340170121721645e-07,
      "loss": 0.4222,
      "num_input_tokens_seen": 11792840,
      "step": 18050
    },
    {
      "epoch": 9.46278825995807,
      "grad_norm": 11.3125,
      "learning_rate": 4.3912403851176234e-07,
      "loss": 0.4449,
      "num_input_tokens_seen": 11795720,
      "step": 18055
    },
    {
      "epoch": 9.465408805031446,
      "grad_norm": 8.25,
      "learning_rate": 4.348669272254163e-07,
      "loss": 0.4584,
      "num_input_tokens_seen": 11798472,
      "step": 18060
    },
    {
      "epoch": 9.468029350104821,
      "grad_norm": 8.75,
      "learning_rate": 4.306303709203374e-07,
      "loss": 0.4792,
      "num_input_tokens_seen": 11801800,
      "step": 18065
    },
    {
      "epoch": 9.470649895178196,
      "grad_norm": 7.09375,
      "learning_rate": 4.264143731414788e-07,
      "loss": 0.5507,
      "num_input_tokens_seen": 11805160,
      "step": 18070
    },
    {
      "epoch": 9.473270440251572,
      "grad_norm": 5.78125,
      "learning_rate": 4.2221893741659636e-07,
      "loss": 0.4477,
      "num_input_tokens_seen": 11811816,
      "step": 18075
    },
    {
      "epoch": 9.475890985324948,
      "grad_norm": 4.71875,
      "learning_rate": 4.180440672562402e-07,
      "loss": 0.5191,
      "num_input_tokens_seen": 11816040,
      "step": 18080
    },
    {
      "epoch": 9.478511530398324,
      "grad_norm": 5.875,
      "learning_rate": 4.1388976615374665e-07,
      "loss": 0.5301,
      "num_input_tokens_seen": 11818920,
      "step": 18085
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 7.65625,
      "learning_rate": 4.097560375852516e-07,
      "loss": 0.5601,
      "num_input_tokens_seen": 11821736,
      "step": 18090
    },
    {
      "epoch": 9.483752620545074,
      "grad_norm": 12.125,
      "learning_rate": 4.056428850096661e-07,
      "loss": 0.5038,
      "num_input_tokens_seen": 11825256,
      "step": 18095
    },
    {
      "epoch": 9.48637316561845,
      "grad_norm": 3.46875,
      "learning_rate": 4.01550311868687e-07,
      "loss": 0.5809,
      "num_input_tokens_seen": 11829416,
      "step": 18100
    },
    {
      "epoch": 9.488993710691824,
      "grad_norm": 5.65625,
      "learning_rate": 3.974783215867972e-07,
      "loss": 0.6124,
      "num_input_tokens_seen": 11832200,
      "step": 18105
    },
    {
      "epoch": 9.4916142557652,
      "grad_norm": 14.8125,
      "learning_rate": 3.9342691757124626e-07,
      "loss": 0.5226,
      "num_input_tokens_seen": 11835592,
      "step": 18110
    },
    {
      "epoch": 9.494234800838575,
      "grad_norm": 7.9375,
      "learning_rate": 3.8939610321206966e-07,
      "loss": 0.4431,
      "num_input_tokens_seen": 11839400,
      "step": 18115
    },
    {
      "epoch": 9.49685534591195,
      "grad_norm": 7.65625,
      "learning_rate": 3.853858818820694e-07,
      "loss": 0.5851,
      "num_input_tokens_seen": 11842216,
      "step": 18120
    },
    {
      "epoch": 9.499475890985325,
      "grad_norm": 7.5,
      "learning_rate": 3.8139625693680847e-07,
      "loss": 0.4405,
      "num_input_tokens_seen": 11844936,
      "step": 18125
    },
    {
      "epoch": 9.5,
      "eval_loss": 0.49346524477005005,
      "eval_runtime": 24.2865,
      "eval_samples_per_second": 34.916,
      "eval_steps_per_second": 8.729,
      "num_input_tokens_seen": 11845704,
      "step": 18126
    },
    {
      "epoch": 9.5020964360587,
      "grad_norm": 9.3125,
      "learning_rate": 3.774272317146277e-07,
      "loss": 0.4939,
      "num_input_tokens_seen": 11847912,
      "step": 18130
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 3.375,
      "learning_rate": 3.7347880953662597e-07,
      "loss": 0.3838,
      "num_input_tokens_seen": 11853192,
      "step": 18135
    },
    {
      "epoch": 9.50733752620545,
      "grad_norm": 7.3125,
      "learning_rate": 3.6955099370666045e-07,
      "loss": 0.5661,
      "num_input_tokens_seen": 11856808,
      "step": 18140
    },
    {
      "epoch": 9.509958071278826,
      "grad_norm": 7.59375,
      "learning_rate": 3.656437875113522e-07,
      "loss": 0.7272,
      "num_input_tokens_seen": 11860360,
      "step": 18145
    },
    {
      "epoch": 9.5125786163522,
      "grad_norm": 5.34375,
      "learning_rate": 3.617571942200693e-07,
      "loss": 0.3784,
      "num_input_tokens_seen": 11863048,
      "step": 18150
    },
    {
      "epoch": 9.515199161425576,
      "grad_norm": 4.875,
      "learning_rate": 3.5789121708493523e-07,
      "loss": 0.4726,
      "num_input_tokens_seen": 11867016,
      "step": 18155
    },
    {
      "epoch": 9.517819706498951,
      "grad_norm": 16.75,
      "learning_rate": 3.5404585934082635e-07,
      "loss": 0.4712,
      "num_input_tokens_seen": 11870728,
      "step": 18160
    },
    {
      "epoch": 9.520440251572326,
      "grad_norm": 5.53125,
      "learning_rate": 3.502211242053577e-07,
      "loss": 0.5026,
      "num_input_tokens_seen": 11873928,
      "step": 18165
    },
    {
      "epoch": 9.523060796645701,
      "grad_norm": 6.875,
      "learning_rate": 3.4641701487889697e-07,
      "loss": 0.4329,
      "num_input_tokens_seen": 11876968,
      "step": 18170
    },
    {
      "epoch": 9.525681341719078,
      "grad_norm": 4.28125,
      "learning_rate": 3.4263353454454806e-07,
      "loss": 0.4659,
      "num_input_tokens_seen": 11880648,
      "step": 18175
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 6.34375,
      "learning_rate": 3.3887068636815346e-07,
      "loss": 0.4903,
      "num_input_tokens_seen": 11883560,
      "step": 18180
    },
    {
      "epoch": 9.530922431865829,
      "grad_norm": 8.0,
      "learning_rate": 3.351284734982918e-07,
      "loss": 0.5316,
      "num_input_tokens_seen": 11888296,
      "step": 18185
    },
    {
      "epoch": 9.533542976939204,
      "grad_norm": 7.5,
      "learning_rate": 3.3140689906628054e-07,
      "loss": 0.4649,
      "num_input_tokens_seen": 11892872,
      "step": 18190
    },
    {
      "epoch": 9.536163522012579,
      "grad_norm": 8.4375,
      "learning_rate": 3.2770596618615645e-07,
      "loss": 0.4385,
      "num_input_tokens_seen": 11895944,
      "step": 18195
    },
    {
      "epoch": 9.538784067085954,
      "grad_norm": 12.625,
      "learning_rate": 3.240256779546952e-07,
      "loss": 0.5187,
      "num_input_tokens_seen": 11898696,
      "step": 18200
    },
    {
      "epoch": 9.54140461215933,
      "grad_norm": 2.46875,
      "learning_rate": 3.2036603745139447e-07,
      "loss": 0.3989,
      "num_input_tokens_seen": 11904744,
      "step": 18205
    },
    {
      "epoch": 9.544025157232705,
      "grad_norm": 5.5625,
      "learning_rate": 3.167270477384743e-07,
      "loss": 0.4408,
      "num_input_tokens_seen": 11907656,
      "step": 18210
    },
    {
      "epoch": 9.54664570230608,
      "grad_norm": 7.5,
      "learning_rate": 3.1310871186086834e-07,
      "loss": 0.5663,
      "num_input_tokens_seen": 11910504,
      "step": 18215
    },
    {
      "epoch": 9.549266247379455,
      "grad_norm": 14.375,
      "learning_rate": 3.095110328462464e-07,
      "loss": 0.5552,
      "num_input_tokens_seen": 11913192,
      "step": 18220
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 6.15625,
      "learning_rate": 3.0593401370497264e-07,
      "loss": 0.6196,
      "num_input_tokens_seen": 11916680,
      "step": 18225
    },
    {
      "epoch": 9.554507337526205,
      "grad_norm": 22.0,
      "learning_rate": 3.0237765743013626e-07,
      "loss": 0.4302,
      "num_input_tokens_seen": 11919656,
      "step": 18230
    },
    {
      "epoch": 9.55712788259958,
      "grad_norm": 6.15625,
      "learning_rate": 2.9884196699753453e-07,
      "loss": 0.4515,
      "num_input_tokens_seen": 11922632,
      "step": 18235
    },
    {
      "epoch": 9.559748427672956,
      "grad_norm": 5.21875,
      "learning_rate": 2.953269453656704e-07,
      "loss": 0.4998,
      "num_input_tokens_seen": 11926280,
      "step": 18240
    },
    {
      "epoch": 9.56236897274633,
      "grad_norm": 2.953125,
      "learning_rate": 2.9183259547575504e-07,
      "loss": 0.3326,
      "num_input_tokens_seen": 11929768,
      "step": 18245
    },
    {
      "epoch": 9.564989517819706,
      "grad_norm": 3.78125,
      "learning_rate": 2.883589202517023e-07,
      "loss": 0.4544,
      "num_input_tokens_seen": 11933480,
      "step": 18250
    },
    {
      "epoch": 9.567610062893081,
      "grad_norm": 6.75,
      "learning_rate": 2.849059226001177e-07,
      "loss": 0.397,
      "num_input_tokens_seen": 11936200,
      "step": 18255
    },
    {
      "epoch": 9.570230607966456,
      "grad_norm": 6.90625,
      "learning_rate": 2.8147360541032065e-07,
      "loss": 0.574,
      "num_input_tokens_seen": 11938472,
      "step": 18260
    },
    {
      "epoch": 9.572851153039831,
      "grad_norm": 4.90625,
      "learning_rate": 2.780619715543109e-07,
      "loss": 0.5555,
      "num_input_tokens_seen": 11942280,
      "step": 18265
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 4.59375,
      "learning_rate": 2.746710238867911e-07,
      "loss": 0.4291,
      "num_input_tokens_seen": 11945800,
      "step": 18270
    },
    {
      "epoch": 9.578092243186584,
      "grad_norm": 4.9375,
      "learning_rate": 2.713007652451499e-07,
      "loss": 0.4556,
      "num_input_tokens_seen": 11948200,
      "step": 18275
    },
    {
      "epoch": 9.580712788259959,
      "grad_norm": 8.5,
      "learning_rate": 2.6795119844946757e-07,
      "loss": 0.5133,
      "num_input_tokens_seen": 11951656,
      "step": 18280
    },
    {
      "epoch": 9.583333333333334,
      "grad_norm": 3.015625,
      "learning_rate": 2.646223263025077e-07,
      "loss": 0.4866,
      "num_input_tokens_seen": 11955208,
      "step": 18285
    },
    {
      "epoch": 9.585953878406709,
      "grad_norm": 6.25,
      "learning_rate": 2.6131415158971993e-07,
      "loss": 0.3932,
      "num_input_tokens_seen": 11957768,
      "step": 18290
    },
    {
      "epoch": 9.588574423480084,
      "grad_norm": 6.875,
      "learning_rate": 2.5802667707922887e-07,
      "loss": 0.457,
      "num_input_tokens_seen": 11960552,
      "step": 18295
    },
    {
      "epoch": 9.59119496855346,
      "grad_norm": 5.0625,
      "learning_rate": 2.54759905521848e-07,
      "loss": 0.5494,
      "num_input_tokens_seen": 11964552,
      "step": 18300
    },
    {
      "epoch": 9.593815513626835,
      "grad_norm": 15.6875,
      "learning_rate": 2.51513839651063e-07,
      "loss": 0.4582,
      "num_input_tokens_seen": 11966824,
      "step": 18305
    },
    {
      "epoch": 9.59643605870021,
      "grad_norm": 9.375,
      "learning_rate": 2.4828848218302615e-07,
      "loss": 0.4692,
      "num_input_tokens_seen": 11969832,
      "step": 18310
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 6.25,
      "learning_rate": 2.450838358165786e-07,
      "loss": 0.4238,
      "num_input_tokens_seen": 11972424,
      "step": 18315
    },
    {
      "epoch": 9.60167714884696,
      "grad_norm": 10.4375,
      "learning_rate": 2.41899903233217e-07,
      "loss": 0.4993,
      "num_input_tokens_seen": 11976552,
      "step": 18320
    },
    {
      "epoch": 9.604297693920335,
      "grad_norm": 7.25,
      "learning_rate": 2.387366870971103e-07,
      "loss": 0.4614,
      "num_input_tokens_seen": 11979720,
      "step": 18325
    },
    {
      "epoch": 9.60691823899371,
      "grad_norm": 6.15625,
      "learning_rate": 2.3559419005509675e-07,
      "loss": 0.5552,
      "num_input_tokens_seen": 11982536,
      "step": 18330
    },
    {
      "epoch": 9.609538784067086,
      "grad_norm": 3.921875,
      "learning_rate": 2.3247241473667026e-07,
      "loss": 0.3758,
      "num_input_tokens_seen": 11985384,
      "step": 18335
    },
    {
      "epoch": 9.61215932914046,
      "grad_norm": 6.40625,
      "learning_rate": 2.2937136375399126e-07,
      "loss": 0.4599,
      "num_input_tokens_seen": 11988712,
      "step": 18340
    },
    {
      "epoch": 9.614779874213836,
      "grad_norm": 3.53125,
      "learning_rate": 2.2629103970188137e-07,
      "loss": 0.4535,
      "num_input_tokens_seen": 11993352,
      "step": 18345
    },
    {
      "epoch": 9.617400419287211,
      "grad_norm": 6.71875,
      "learning_rate": 2.2323144515780935e-07,
      "loss": 0.4869,
      "num_input_tokens_seen": 11995848,
      "step": 18350
    },
    {
      "epoch": 9.620020964360586,
      "grad_norm": 7.34375,
      "learning_rate": 2.201925826819079e-07,
      "loss": 0.5491,
      "num_input_tokens_seen": 11999336,
      "step": 18355
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 6.9375,
      "learning_rate": 2.1717445481695408e-07,
      "loss": 0.499,
      "num_input_tokens_seen": 12003080,
      "step": 18360
    },
    {
      "epoch": 9.625262054507338,
      "grad_norm": 3.21875,
      "learning_rate": 2.1417706408838333e-07,
      "loss": 0.3016,
      "num_input_tokens_seen": 12007240,
      "step": 18365
    },
    {
      "epoch": 9.627882599580714,
      "grad_norm": 7.0,
      "learning_rate": 2.112004130042755e-07,
      "loss": 0.5111,
      "num_input_tokens_seen": 12009928,
      "step": 18370
    },
    {
      "epoch": 9.630503144654089,
      "grad_norm": 6.4375,
      "learning_rate": 2.082445040553549e-07,
      "loss": 0.4439,
      "num_input_tokens_seen": 12013384,
      "step": 18375
    },
    {
      "epoch": 9.633123689727464,
      "grad_norm": 6.9375,
      "learning_rate": 2.053093397149902e-07,
      "loss": 0.5225,
      "num_input_tokens_seen": 12016520,
      "step": 18380
    },
    {
      "epoch": 9.635744234800839,
      "grad_norm": 9.5,
      "learning_rate": 2.0239492243919467e-07,
      "loss": 0.4522,
      "num_input_tokens_seen": 12020040,
      "step": 18385
    },
    {
      "epoch": 9.638364779874214,
      "grad_norm": 4.625,
      "learning_rate": 1.9950125466662028e-07,
      "loss": 0.4685,
      "num_input_tokens_seen": 12023464,
      "step": 18390
    },
    {
      "epoch": 9.64098532494759,
      "grad_norm": 8.6875,
      "learning_rate": 1.9662833881855248e-07,
      "loss": 0.3587,
      "num_input_tokens_seen": 12026664,
      "step": 18395
    },
    {
      "epoch": 9.643605870020965,
      "grad_norm": 5.3125,
      "learning_rate": 1.9377617729891828e-07,
      "loss": 0.4043,
      "num_input_tokens_seen": 12030440,
      "step": 18400
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 7.0625,
      "learning_rate": 1.9094477249427534e-07,
      "loss": 0.514,
      "num_input_tokens_seen": 12032872,
      "step": 18405
    },
    {
      "epoch": 9.648846960167715,
      "grad_norm": 5.0625,
      "learning_rate": 1.8813412677381737e-07,
      "loss": 0.41,
      "num_input_tokens_seen": 12037864,
      "step": 18410
    },
    {
      "epoch": 9.65146750524109,
      "grad_norm": 7.40625,
      "learning_rate": 1.8534424248935756e-07,
      "loss": 0.3506,
      "num_input_tokens_seen": 12040904,
      "step": 18415
    },
    {
      "epoch": 9.654088050314465,
      "grad_norm": 3.96875,
      "learning_rate": 1.8257512197535076e-07,
      "loss": 0.4699,
      "num_input_tokens_seen": 12044296,
      "step": 18420
    },
    {
      "epoch": 9.65670859538784,
      "grad_norm": 6.09375,
      "learning_rate": 1.7982676754886574e-07,
      "loss": 0.3848,
      "num_input_tokens_seen": 12047208,
      "step": 18425
    },
    {
      "epoch": 9.659329140461216,
      "grad_norm": 5.25,
      "learning_rate": 1.7709918150959904e-07,
      "loss": 0.5437,
      "num_input_tokens_seen": 12050696,
      "step": 18430
    },
    {
      "epoch": 9.66194968553459,
      "grad_norm": 6.65625,
      "learning_rate": 1.7439236613987775e-07,
      "loss": 0.5252,
      "num_input_tokens_seen": 12054536,
      "step": 18435
    },
    {
      "epoch": 9.664570230607966,
      "grad_norm": 11.1875,
      "learning_rate": 1.717063237046318e-07,
      "loss": 0.5061,
      "num_input_tokens_seen": 12056776,
      "step": 18440
    },
    {
      "epoch": 9.667190775681341,
      "grad_norm": 5.5,
      "learning_rate": 1.6904105645142444e-07,
      "loss": 0.5389,
      "num_input_tokens_seen": 12059720,
      "step": 18445
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 6.9375,
      "learning_rate": 1.6639656661043e-07,
      "loss": 0.4766,
      "num_input_tokens_seen": 12063304,
      "step": 18450
    },
    {
      "epoch": 9.672431865828091,
      "grad_norm": 3.796875,
      "learning_rate": 1.6377285639443407e-07,
      "loss": 0.5299,
      "num_input_tokens_seen": 12067592,
      "step": 18455
    },
    {
      "epoch": 9.675052410901468,
      "grad_norm": 5.0,
      "learning_rate": 1.61169927998836e-07,
      "loss": 0.4822,
      "num_input_tokens_seen": 12070856,
      "step": 18460
    },
    {
      "epoch": 9.677672955974844,
      "grad_norm": 5.8125,
      "learning_rate": 1.5858778360165195e-07,
      "loss": 0.5095,
      "num_input_tokens_seen": 12074280,
      "step": 18465
    },
    {
      "epoch": 9.680293501048219,
      "grad_norm": 7.78125,
      "learning_rate": 1.5602642536350075e-07,
      "loss": 0.4208,
      "num_input_tokens_seen": 12077288,
      "step": 18470
    },
    {
      "epoch": 9.682914046121594,
      "grad_norm": 6.5,
      "learning_rate": 1.5348585542760974e-07,
      "loss": 0.404,
      "num_input_tokens_seen": 12082056,
      "step": 18475
    },
    {
      "epoch": 9.685534591194969,
      "grad_norm": 5.53125,
      "learning_rate": 1.5096607591980894e-07,
      "loss": 0.5608,
      "num_input_tokens_seen": 12085128,
      "step": 18480
    },
    {
      "epoch": 9.688155136268344,
      "grad_norm": 3.90625,
      "learning_rate": 1.4846708894853955e-07,
      "loss": 0.6203,
      "num_input_tokens_seen": 12089032,
      "step": 18485
    },
    {
      "epoch": 9.69077568134172,
      "grad_norm": 4.5625,
      "learning_rate": 1.459888966048373e-07,
      "loss": 0.5655,
      "num_input_tokens_seen": 12091496,
      "step": 18490
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 5.75,
      "learning_rate": 1.4353150096234058e-07,
      "loss": 0.587,
      "num_input_tokens_seen": 12095208,
      "step": 18495
    },
    {
      "epoch": 9.69601677148847,
      "grad_norm": 5.21875,
      "learning_rate": 1.410949040772852e-07,
      "loss": 0.5365,
      "num_input_tokens_seen": 12099176,
      "step": 18500
    },
    {
      "epoch": 9.698637316561845,
      "grad_norm": 6.5625,
      "learning_rate": 1.3867910798850692e-07,
      "loss": 0.5739,
      "num_input_tokens_seen": 12102408,
      "step": 18505
    },
    {
      "epoch": 9.70125786163522,
      "grad_norm": 6.03125,
      "learning_rate": 1.3628411471742764e-07,
      "loss": 0.5749,
      "num_input_tokens_seen": 12105704,
      "step": 18510
    },
    {
      "epoch": 9.703878406708595,
      "grad_norm": 3.984375,
      "learning_rate": 1.3390992626807485e-07,
      "loss": 0.5614,
      "num_input_tokens_seen": 12108520,
      "step": 18515
    },
    {
      "epoch": 9.70649895178197,
      "grad_norm": 5.625,
      "learning_rate": 1.315565446270567e-07,
      "loss": 0.4575,
      "num_input_tokens_seen": 12111912,
      "step": 18520
    },
    {
      "epoch": 9.709119496855346,
      "grad_norm": 7.15625,
      "learning_rate": 1.292239717635785e-07,
      "loss": 0.4493,
      "num_input_tokens_seen": 12116040,
      "step": 18525
    },
    {
      "epoch": 9.71174004192872,
      "grad_norm": 4.59375,
      "learning_rate": 1.269122096294262e-07,
      "loss": 0.4364,
      "num_input_tokens_seen": 12118632,
      "step": 18530
    },
    {
      "epoch": 9.714360587002096,
      "grad_norm": 6.84375,
      "learning_rate": 1.24621260158983e-07,
      "loss": 0.3493,
      "num_input_tokens_seen": 12121192,
      "step": 18535
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 6.75,
      "learning_rate": 1.2235112526920723e-07,
      "loss": 0.4641,
      "num_input_tokens_seen": 12123976,
      "step": 18540
    },
    {
      "epoch": 9.719601677148846,
      "grad_norm": 6.5,
      "learning_rate": 1.2010180685964324e-07,
      "loss": 0.6101,
      "num_input_tokens_seen": 12127816,
      "step": 18545
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 3.453125,
      "learning_rate": 1.1787330681241881e-07,
      "loss": 0.5649,
      "num_input_tokens_seen": 12131848,
      "step": 18550
    },
    {
      "epoch": 9.724842767295598,
      "grad_norm": 7.75,
      "learning_rate": 1.156656269922396e-07,
      "loss": 0.4704,
      "num_input_tokens_seen": 12135432,
      "step": 18555
    },
    {
      "epoch": 9.727463312368974,
      "grad_norm": 6.34375,
      "learning_rate": 1.1347876924639455e-07,
      "loss": 0.5224,
      "num_input_tokens_seen": 12138376,
      "step": 18560
    },
    {
      "epoch": 9.730083857442349,
      "grad_norm": 6.21875,
      "learning_rate": 1.1131273540474496e-07,
      "loss": 0.5892,
      "num_input_tokens_seen": 12141480,
      "step": 18565
    },
    {
      "epoch": 9.732704402515724,
      "grad_norm": 5.625,
      "learning_rate": 1.091675272797299e-07,
      "loss": 0.4339,
      "num_input_tokens_seen": 12144168,
      "step": 18570
    },
    {
      "epoch": 9.735324947589099,
      "grad_norm": 3.328125,
      "learning_rate": 1.0704314666635795e-07,
      "loss": 0.5897,
      "num_input_tokens_seen": 12148168,
      "step": 18575
    },
    {
      "epoch": 9.737945492662474,
      "grad_norm": 4.90625,
      "learning_rate": 1.0493959534221832e-07,
      "loss": 0.3555,
      "num_input_tokens_seen": 12150696,
      "step": 18580
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 5.59375,
      "learning_rate": 1.0285687506746133e-07,
      "loss": 0.3857,
      "num_input_tokens_seen": 12153928,
      "step": 18585
    },
    {
      "epoch": 9.743186582809225,
      "grad_norm": 5.96875,
      "learning_rate": 1.0079498758481798e-07,
      "loss": 0.4415,
      "num_input_tokens_seen": 12157384,
      "step": 18590
    },
    {
      "epoch": 9.7458071278826,
      "grad_norm": 5.375,
      "learning_rate": 9.87539346195776e-08,
      "loss": 0.4572,
      "num_input_tokens_seen": 12160200,
      "step": 18595
    },
    {
      "epoch": 9.748427672955975,
      "grad_norm": 7.65625,
      "learning_rate": 9.673371787960183e-08,
      "loss": 0.6747,
      "num_input_tokens_seen": 12162504,
      "step": 18600
    },
    {
      "epoch": 9.75104821802935,
      "grad_norm": 6.5,
      "learning_rate": 9.473433905531626e-08,
      "loss": 0.4962,
      "num_input_tokens_seen": 12165288,
      "step": 18605
    },
    {
      "epoch": 9.753668763102725,
      "grad_norm": 14.4375,
      "learning_rate": 9.275579981970483e-08,
      "loss": 0.4674,
      "num_input_tokens_seen": 12167912,
      "step": 18610
    },
    {
      "epoch": 9.7562893081761,
      "grad_norm": 33.0,
      "learning_rate": 9.07981018283266e-08,
      "loss": 0.5648,
      "num_input_tokens_seen": 12171624,
      "step": 18615
    },
    {
      "epoch": 9.758909853249476,
      "grad_norm": 7.65625,
      "learning_rate": 8.886124671928786e-08,
      "loss": 0.5095,
      "num_input_tokens_seen": 12174632,
      "step": 18620
    },
    {
      "epoch": 9.76153039832285,
      "grad_norm": 14.375,
      "learning_rate": 8.694523611326444e-08,
      "loss": 0.6197,
      "num_input_tokens_seen": 12177896,
      "step": 18625
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 7.125,
      "learning_rate": 8.505007161348222e-08,
      "loss": 0.5257,
      "num_input_tokens_seen": 12181256,
      "step": 18630
    },
    {
      "epoch": 9.766771488469601,
      "grad_norm": 6.65625,
      "learning_rate": 8.31757548057338e-08,
      "loss": 0.5319,
      "num_input_tokens_seen": 12183848,
      "step": 18635
    },
    {
      "epoch": 9.769392033542976,
      "grad_norm": 4.8125,
      "learning_rate": 8.132228725835634e-08,
      "loss": 0.4589,
      "num_input_tokens_seen": 12187208,
      "step": 18640
    },
    {
      "epoch": 9.772012578616351,
      "grad_norm": 4.75,
      "learning_rate": 7.948967052225087e-08,
      "loss": 0.4996,
      "num_input_tokens_seen": 12190568,
      "step": 18645
    },
    {
      "epoch": 9.774633123689728,
      "grad_norm": 7.40625,
      "learning_rate": 7.767790613086301e-08,
      "loss": 0.3631,
      "num_input_tokens_seen": 12193224,
      "step": 18650
    },
    {
      "epoch": 9.777253668763104,
      "grad_norm": 6.125,
      "learning_rate": 7.588699560019952e-08,
      "loss": 0.4754,
      "num_input_tokens_seen": 12195816,
      "step": 18655
    },
    {
      "epoch": 9.779874213836479,
      "grad_norm": 8.1875,
      "learning_rate": 7.411694042881168e-08,
      "loss": 0.4728,
      "num_input_tokens_seen": 12198856,
      "step": 18660
    },
    {
      "epoch": 9.782494758909854,
      "grad_norm": 9.5625,
      "learning_rate": 7.23677420977953e-08,
      "loss": 0.5516,
      "num_input_tokens_seen": 12201992,
      "step": 18665
    },
    {
      "epoch": 9.785115303983229,
      "grad_norm": 6.53125,
      "learning_rate": 7.063940207080733e-08,
      "loss": 0.4992,
      "num_input_tokens_seen": 12205608,
      "step": 18670
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 5.09375,
      "learning_rate": 6.893192179403817e-08,
      "loss": 0.5809,
      "num_input_tokens_seen": 12209352,
      "step": 18675
    },
    {
      "epoch": 9.79035639412998,
      "grad_norm": 9.3125,
      "learning_rate": 6.724530269623108e-08,
      "loss": 0.4863,
      "num_input_tokens_seen": 12213768,
      "step": 18680
    },
    {
      "epoch": 9.792976939203355,
      "grad_norm": 6.90625,
      "learning_rate": 6.557954618867102e-08,
      "loss": 0.4849,
      "num_input_tokens_seen": 12216776,
      "step": 18685
    },
    {
      "epoch": 9.79559748427673,
      "grad_norm": 6.09375,
      "learning_rate": 6.393465366519024e-08,
      "loss": 0.4061,
      "num_input_tokens_seen": 12221000,
      "step": 18690
    },
    {
      "epoch": 9.798218029350105,
      "grad_norm": 14.9375,
      "learning_rate": 6.231062650215724e-08,
      "loss": 0.4713,
      "num_input_tokens_seen": 12224200,
      "step": 18695
    },
    {
      "epoch": 9.80083857442348,
      "grad_norm": 4.84375,
      "learning_rate": 6.070746605848221e-08,
      "loss": 0.4868,
      "num_input_tokens_seen": 12226920,
      "step": 18700
    },
    {
      "epoch": 9.803459119496855,
      "grad_norm": 5.875,
      "learning_rate": 5.912517367561987e-08,
      "loss": 0.5285,
      "num_input_tokens_seen": 12229960,
      "step": 18705
    },
    {
      "epoch": 9.80607966457023,
      "grad_norm": 8.875,
      "learning_rate": 5.756375067755837e-08,
      "loss": 0.4468,
      "num_input_tokens_seen": 12232040,
      "step": 18710
    },
    {
      "epoch": 9.808700209643606,
      "grad_norm": 5.625,
      "learning_rate": 5.602319837082481e-08,
      "loss": 0.4096,
      "num_input_tokens_seen": 12234472,
      "step": 18715
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 9.4375,
      "learning_rate": 5.450351804448528e-08,
      "loss": 0.3662,
      "num_input_tokens_seen": 12237448,
      "step": 18720
    },
    {
      "epoch": 9.813941299790356,
      "grad_norm": 4.5625,
      "learning_rate": 5.3004710970133705e-08,
      "loss": 0.5035,
      "num_input_tokens_seen": 12240264,
      "step": 18725
    },
    {
      "epoch": 9.816561844863731,
      "grad_norm": 4.46875,
      "learning_rate": 5.1526778401911334e-08,
      "loss": 0.4455,
      "num_input_tokens_seen": 12243176,
      "step": 18730
    },
    {
      "epoch": 9.819182389937106,
      "grad_norm": 11.375,
      "learning_rate": 5.0069721576476156e-08,
      "loss": 0.5173,
      "num_input_tokens_seen": 12245864,
      "step": 18735
    },
    {
      "epoch": 9.821802935010481,
      "grad_norm": 5.03125,
      "learning_rate": 4.863354171303347e-08,
      "loss": 0.4505,
      "num_input_tokens_seen": 12248712,
      "step": 18740
    },
    {
      "epoch": 9.824423480083858,
      "grad_norm": 13.75,
      "learning_rate": 4.72182400133081e-08,
      "loss": 0.5472,
      "num_input_tokens_seen": 12251144,
      "step": 18745
    },
    {
      "epoch": 9.827044025157234,
      "grad_norm": 6.5,
      "learning_rate": 4.582381766156385e-08,
      "loss": 0.4718,
      "num_input_tokens_seen": 12255336,
      "step": 18750
    },
    {
      "epoch": 9.829664570230609,
      "grad_norm": 10.5625,
      "learning_rate": 4.445027582458683e-08,
      "loss": 0.5339,
      "num_input_tokens_seen": 12257672,
      "step": 18755
    },
    {
      "epoch": 9.832285115303984,
      "grad_norm": 4.65625,
      "learning_rate": 4.309761565169379e-08,
      "loss": 0.4722,
      "num_input_tokens_seen": 12261032,
      "step": 18760
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 6.0625,
      "learning_rate": 4.1765838274732125e-08,
      "loss": 0.4682,
      "num_input_tokens_seen": 12264488,
      "step": 18765
    },
    {
      "epoch": 9.837526205450734,
      "grad_norm": 9.5625,
      "learning_rate": 4.045494480807155e-08,
      "loss": 0.493,
      "num_input_tokens_seen": 12267432,
      "step": 18770
    },
    {
      "epoch": 9.84014675052411,
      "grad_norm": 10.5625,
      "learning_rate": 3.916493634860407e-08,
      "loss": 0.3779,
      "num_input_tokens_seen": 12270888,
      "step": 18775
    },
    {
      "epoch": 9.842767295597485,
      "grad_norm": 3.96875,
      "learning_rate": 3.789581397575515e-08,
      "loss": 0.4699,
      "num_input_tokens_seen": 12273896,
      "step": 18780
    },
    {
      "epoch": 9.84538784067086,
      "grad_norm": 5.78125,
      "learning_rate": 3.664757875146418e-08,
      "loss": 0.5473,
      "num_input_tokens_seen": 12276328,
      "step": 18785
    },
    {
      "epoch": 9.848008385744235,
      "grad_norm": 7.53125,
      "learning_rate": 3.5420231720198485e-08,
      "loss": 0.4228,
      "num_input_tokens_seen": 12279016,
      "step": 18790
    },
    {
      "epoch": 9.85062893081761,
      "grad_norm": 10.1875,
      "learning_rate": 3.421377390894764e-08,
      "loss": 0.4009,
      "num_input_tokens_seen": 12281512,
      "step": 18795
    },
    {
      "epoch": 9.853249475890985,
      "grad_norm": 8.125,
      "learning_rate": 3.3028206327218035e-08,
      "loss": 0.5596,
      "num_input_tokens_seen": 12285160,
      "step": 18800
    },
    {
      "epoch": 9.85587002096436,
      "grad_norm": 4.75,
      "learning_rate": 3.1863529967041117e-08,
      "loss": 0.4887,
      "num_input_tokens_seen": 12288616,
      "step": 18805
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 4.9375,
      "learning_rate": 3.071974580296233e-08,
      "loss": 0.57,
      "num_input_tokens_seen": 12292680,
      "step": 18810
    },
    {
      "epoch": 9.86111111111111,
      "grad_norm": 4.6875,
      "learning_rate": 2.9596854792052207e-08,
      "loss": 0.424,
      "num_input_tokens_seen": 12297160,
      "step": 18815
    },
    {
      "epoch": 9.863731656184486,
      "grad_norm": 4.4375,
      "learning_rate": 2.8494857873889724e-08,
      "loss": 0.648,
      "num_input_tokens_seen": 12300520,
      "step": 18820
    },
    {
      "epoch": 9.866352201257861,
      "grad_norm": 7.34375,
      "learning_rate": 2.741375597057616e-08,
      "loss": 0.5104,
      "num_input_tokens_seen": 12303496,
      "step": 18825
    },
    {
      "epoch": 9.868972746331236,
      "grad_norm": 5.25,
      "learning_rate": 2.6353549986729566e-08,
      "loss": 0.558,
      "num_input_tokens_seen": 12306568,
      "step": 18830
    },
    {
      "epoch": 9.871593291404611,
      "grad_norm": 6.09375,
      "learning_rate": 2.531424080948197e-08,
      "loss": 0.5724,
      "num_input_tokens_seen": 12309160,
      "step": 18835
    },
    {
      "epoch": 9.874213836477988,
      "grad_norm": 4.5625,
      "learning_rate": 2.4295829308482176e-08,
      "loss": 0.4531,
      "num_input_tokens_seen": 12312776,
      "step": 18840
    },
    {
      "epoch": 9.876834381551364,
      "grad_norm": 7.1875,
      "learning_rate": 2.329831633588464e-08,
      "loss": 0.4212,
      "num_input_tokens_seen": 12316104,
      "step": 18845
    },
    {
      "epoch": 9.879454926624739,
      "grad_norm": 10.0625,
      "learning_rate": 2.232170272636891e-08,
      "loss": 0.5838,
      "num_input_tokens_seen": 12318568,
      "step": 18850
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 10.125,
      "learning_rate": 2.136598929711464e-08,
      "loss": 0.538,
      "num_input_tokens_seen": 12321736,
      "step": 18855
    },
    {
      "epoch": 9.884696016771489,
      "grad_norm": 5.8125,
      "learning_rate": 2.0431176847823807e-08,
      "loss": 0.51,
      "num_input_tokens_seen": 12325032,
      "step": 18860
    },
    {
      "epoch": 9.887316561844864,
      "grad_norm": 3.3125,
      "learning_rate": 1.9517266160704038e-08,
      "loss": 0.5172,
      "num_input_tokens_seen": 12329160,
      "step": 18865
    },
    {
      "epoch": 9.88993710691824,
      "grad_norm": 5.4375,
      "learning_rate": 1.8624258000471405e-08,
      "loss": 0.7113,
      "num_input_tokens_seen": 12332392,
      "step": 18870
    },
    {
      "epoch": 9.892557651991615,
      "grad_norm": 4.53125,
      "learning_rate": 1.7752153114358737e-08,
      "loss": 0.4527,
      "num_input_tokens_seen": 12336456,
      "step": 18875
    },
    {
      "epoch": 9.89517819706499,
      "grad_norm": 6.0,
      "learning_rate": 1.6900952232098977e-08,
      "loss": 0.4309,
      "num_input_tokens_seen": 12339080,
      "step": 18880
    },
    {
      "epoch": 9.897798742138365,
      "grad_norm": 8.875,
      "learning_rate": 1.6070656065939048e-08,
      "loss": 0.4131,
      "num_input_tokens_seen": 12341768,
      "step": 18885
    },
    {
      "epoch": 9.90041928721174,
      "grad_norm": 8.0625,
      "learning_rate": 1.526126531063432e-08,
      "loss": 0.414,
      "num_input_tokens_seen": 12344936,
      "step": 18890
    },
    {
      "epoch": 9.903039832285115,
      "grad_norm": 6.46875,
      "learning_rate": 1.4472780643445817e-08,
      "loss": 0.4877,
      "num_input_tokens_seen": 12349416,
      "step": 18895
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 9.0,
      "learning_rate": 1.3705202724142996e-08,
      "loss": 0.5503,
      "num_input_tokens_seen": 12352360,
      "step": 18900
    },
    {
      "epoch": 9.908280922431866,
      "grad_norm": 6.1875,
      "learning_rate": 1.2958532194995432e-08,
      "loss": 0.5392,
      "num_input_tokens_seen": 12355688,
      "step": 18905
    },
    {
      "epoch": 9.91090146750524,
      "grad_norm": 2.8125,
      "learning_rate": 1.2232769680789457e-08,
      "loss": 0.4929,
      "num_input_tokens_seen": 12359560,
      "step": 18910
    },
    {
      "epoch": 9.913522012578616,
      "grad_norm": 4.96875,
      "learning_rate": 1.152791578880319e-08,
      "loss": 0.5336,
      "num_input_tokens_seen": 12363656,
      "step": 18915
    },
    {
      "epoch": 9.916142557651991,
      "grad_norm": 4.25,
      "learning_rate": 1.0843971108828732e-08,
      "loss": 0.4124,
      "num_input_tokens_seen": 12367688,
      "step": 18920
    },
    {
      "epoch": 9.918763102725366,
      "grad_norm": 7.5625,
      "learning_rate": 1.018093621316385e-08,
      "loss": 0.4846,
      "num_input_tokens_seen": 12371400,
      "step": 18925
    },
    {
      "epoch": 9.921383647798741,
      "grad_norm": 5.125,
      "learning_rate": 9.53881165659809e-09,
      "loss": 0.4637,
      "num_input_tokens_seen": 12374152,
      "step": 18930
    },
    {
      "epoch": 9.924004192872118,
      "grad_norm": 8.0625,
      "learning_rate": 8.91759797644054e-09,
      "loss": 0.3924,
      "num_input_tokens_seen": 12378472,
      "step": 18935
    },
    {
      "epoch": 9.926624737945493,
      "grad_norm": 5.25,
      "learning_rate": 8.317295692486516e-09,
      "loss": 0.5325,
      "num_input_tokens_seen": 12381480,
      "step": 18940
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 3.640625,
      "learning_rate": 7.737905307045323e-09,
      "loss": 0.4508,
      "num_input_tokens_seen": 12384648,
      "step": 18945
    },
    {
      "epoch": 9.931865828092244,
      "grad_norm": 3.453125,
      "learning_rate": 7.179427304926378e-09,
      "loss": 0.391,
      "num_input_tokens_seen": 12387432,
      "step": 18950
    },
    {
      "epoch": 9.934486373165619,
      "grad_norm": 5.5625,
      "learning_rate": 6.641862153433653e-09,
      "loss": 0.4593,
      "num_input_tokens_seen": 12390984,
      "step": 18955
    },
    {
      "epoch": 9.937106918238994,
      "grad_norm": 4.8125,
      "learning_rate": 6.125210302382333e-09,
      "loss": 0.4211,
      "num_input_tokens_seen": 12394760,
      "step": 18960
    },
    {
      "epoch": 9.93972746331237,
      "grad_norm": 7.59375,
      "learning_rate": 5.629472184079387e-09,
      "loss": 0.5481,
      "num_input_tokens_seen": 12397768,
      "step": 18965
    },
    {
      "epoch": 9.942348008385745,
      "grad_norm": 7.5625,
      "learning_rate": 5.154648213334668e-09,
      "loss": 0.4215,
      "num_input_tokens_seen": 12400968,
      "step": 18970
    },
    {
      "epoch": 9.94496855345912,
      "grad_norm": 5.3125,
      "learning_rate": 4.700738787466463e-09,
      "loss": 0.5535,
      "num_input_tokens_seen": 12406664,
      "step": 18975
    },
    {
      "epoch": 9.947589098532495,
      "grad_norm": 8.5,
      "learning_rate": 4.26774428627652e-09,
      "loss": 0.4668,
      "num_input_tokens_seen": 12409448,
      "step": 18980
    },
    {
      "epoch": 9.95020964360587,
      "grad_norm": 5.4375,
      "learning_rate": 3.855665072080572e-09,
      "loss": 0.3727,
      "num_input_tokens_seen": 12412744,
      "step": 18985
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 5.125,
      "learning_rate": 3.464501489683358e-09,
      "loss": 0.656,
      "num_input_tokens_seen": 12415656,
      "step": 18990
    },
    {
      "epoch": 9.95545073375262,
      "grad_norm": 6.375,
      "learning_rate": 3.094253866398056e-09,
      "loss": 0.5815,
      "num_input_tokens_seen": 12418472,
      "step": 18995
    },
    {
      "epoch": 9.958071278825996,
      "grad_norm": 7.0,
      "learning_rate": 2.7449225120268484e-09,
      "loss": 0.4373,
      "num_input_tokens_seen": 12421768,
      "step": 19000
    },
    {
      "epoch": 9.96069182389937,
      "grad_norm": 4.59375,
      "learning_rate": 2.416507718877581e-09,
      "loss": 0.4565,
      "num_input_tokens_seen": 12425800,
      "step": 19005
    },
    {
      "epoch": 9.963312368972746,
      "grad_norm": 6.65625,
      "learning_rate": 2.109009761747105e-09,
      "loss": 0.5965,
      "num_input_tokens_seen": 12429288,
      "step": 19010
    },
    {
      "epoch": 9.965932914046121,
      "grad_norm": 4.6875,
      "learning_rate": 1.8224288979434844e-09,
      "loss": 0.4221,
      "num_input_tokens_seen": 12433160,
      "step": 19015
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 9.4375,
      "learning_rate": 1.5567653672554638e-09,
      "loss": 0.4764,
      "num_input_tokens_seen": 12435944,
      "step": 19020
    },
    {
      "epoch": 9.971174004192871,
      "grad_norm": 5.625,
      "learning_rate": 1.3120193919857748e-09,
      "loss": 0.4212,
      "num_input_tokens_seen": 12438216,
      "step": 19025
    },
    {
      "epoch": 9.973794549266248,
      "grad_norm": 9.3125,
      "learning_rate": 1.0881911769261565e-09,
      "loss": 0.4396,
      "num_input_tokens_seen": 12440904,
      "step": 19030
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 9.25,
      "learning_rate": 8.852809093601311e-10,
      "loss": 0.5479,
      "num_input_tokens_seen": 12443752,
      "step": 19035
    },
    {
      "epoch": 9.979035639412999,
      "grad_norm": 12.0,
      "learning_rate": 7.03288759076881e-10,
      "loss": 0.708,
      "num_input_tokens_seen": 12446152,
      "step": 19040
    },
    {
      "epoch": 9.981656184486374,
      "grad_norm": 4.6875,
      "learning_rate": 5.422148783629233e-10,
      "loss": 0.496,
      "num_input_tokens_seen": 12449160,
      "step": 19045
    },
    {
      "epoch": 9.984276729559749,
      "grad_norm": 4.625,
      "learning_rate": 4.0205940199100623e-10,
      "loss": 0.5216,
      "num_input_tokens_seen": 12453064,
      "step": 19050
    },
    {
      "epoch": 9.986897274633124,
      "grad_norm": 10.0,
      "learning_rate": 2.828224472395391e-10,
      "loss": 0.5004,
      "num_input_tokens_seen": 12455944,
      "step": 19055
    },
    {
      "epoch": 9.9895178197065,
      "grad_norm": 6.46875,
      "learning_rate": 1.8450411388426515e-10,
      "loss": 0.5452,
      "num_input_tokens_seen": 12459528,
      "step": 19060
    },
    {
      "epoch": 9.992138364779874,
      "grad_norm": 4.375,
      "learning_rate": 1.0710448418715935e-10,
      "loss": 0.5028,
      "num_input_tokens_seen": 12463048,
      "step": 19065
    },
    {
      "epoch": 9.99475890985325,
      "grad_norm": 5.65625,
      "learning_rate": 5.062362291585743e-11,
      "loss": 0.4727,
      "num_input_tokens_seen": 12467240,
      "step": 19070
    },
    {
      "epoch": 9.997379454926625,
      "grad_norm": 6.375,
      "learning_rate": 1.5061577329777976e-11,
      "loss": 0.683,
      "num_input_tokens_seen": 12470216,
      "step": 19075
    },
    {
      "epoch": 10.0,
      "grad_norm": 13.0,
      "learning_rate": 4.183771884491705e-13,
      "loss": 0.672,
      "num_input_tokens_seen": 12472912,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.49317190051078796,
      "eval_runtime": 24.2883,
      "eval_samples_per_second": 34.914,
      "eval_steps_per_second": 8.728,
      "num_input_tokens_seen": 12472912,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "num_input_tokens_seen": 12472912,
      "step": 19080,
      "total_flos": 5.616495900029092e+17,
      "train_loss": 0.5190783202710642,
      "train_runtime": 5198.2272,
      "train_samples_per_second": 14.676,
      "train_steps_per_second": 3.67
    }
  ],
  "logging_steps": 5,
  "max_steps": 19080,
  "num_input_tokens_seen": 12472912,
  "num_train_epochs": 10,
  "save_steps": 954,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.616495900029092e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}