{
  "best_global_step": 13356,
  "best_metric": 0.45409154891967773,
  "best_model_checkpoint": "saves_multiple/prompt-tuning/llama-3-8b-instruct/train_codealpacapy_42_1760638811/checkpoint-13356",
  "epoch": 20.0,
  "eval_steps": 1908,
  "global_step": 38160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002620545073375262,
      "grad_norm": 0.80078125,
      "learning_rate": 3.144654088050314e-05,
      "loss": 1.1844,
      "num_input_tokens_seen": 3264,
      "step": 5
    },
    {
      "epoch": 0.005241090146750524,
      "grad_norm": 1.65625,
      "learning_rate": 7.075471698113207e-05,
      "loss": 1.2352,
      "num_input_tokens_seen": 5888,
      "step": 10
    },
    {
      "epoch": 0.007861635220125786,
      "grad_norm": 1.640625,
      "learning_rate": 0.00011006289308176101,
      "loss": 1.3124,
      "num_input_tokens_seen": 8480,
      "step": 15
    },
    {
      "epoch": 0.010482180293501049,
      "grad_norm": 9.5,
      "learning_rate": 0.00014937106918238994,
      "loss": 1.0964,
      "num_input_tokens_seen": 11488,
      "step": 20
    },
    {
      "epoch": 0.01310272536687631,
      "grad_norm": 2.875,
      "learning_rate": 0.00018867924528301886,
      "loss": 0.7146,
      "num_input_tokens_seen": 14880,
      "step": 25
    },
    {
      "epoch": 0.015723270440251572,
      "grad_norm": 3.328125,
      "learning_rate": 0.00022798742138364778,
      "loss": 0.5595,
      "num_input_tokens_seen": 18592,
      "step": 30
    },
    {
      "epoch": 0.018343815513626835,
      "grad_norm": 7.34375,
      "learning_rate": 0.0002672955974842767,
      "loss": 0.6301,
      "num_input_tokens_seen": 21568,
      "step": 35
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 5.59375,
      "learning_rate": 0.0003066037735849056,
      "loss": 0.6314,
      "num_input_tokens_seen": 24160,
      "step": 40
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 8.0,
      "learning_rate": 0.00034591194968553455,
      "loss": 0.6348,
      "num_input_tokens_seen": 27168,
      "step": 45
    },
    {
      "epoch": 0.02620545073375262,
      "grad_norm": 3.6875,
      "learning_rate": 0.0003852201257861635,
      "loss": 0.5616,
      "num_input_tokens_seen": 30080,
      "step": 50
    },
    {
      "epoch": 0.028825995807127882,
      "grad_norm": 8.875,
      "learning_rate": 0.00042452830188679245,
      "loss": 0.5855,
      "num_input_tokens_seen": 32608,
      "step": 55
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 1.734375,
      "learning_rate": 0.00046383647798742137,
      "loss": 0.6173,
      "num_input_tokens_seen": 36448,
      "step": 60
    },
    {
      "epoch": 0.034067085953878404,
      "grad_norm": 3.609375,
      "learning_rate": 0.0005031446540880503,
      "loss": 0.4004,
      "num_input_tokens_seen": 38944,
      "step": 65
    },
    {
      "epoch": 0.03668763102725367,
      "grad_norm": 10.0,
      "learning_rate": 0.0005424528301886792,
      "loss": 0.5282,
      "num_input_tokens_seen": 41824,
      "step": 70
    },
    {
      "epoch": 0.03930817610062893,
      "grad_norm": 7.8125,
      "learning_rate": 0.0005817610062893082,
      "loss": 0.5162,
      "num_input_tokens_seen": 44896,
      "step": 75
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 5.0,
      "learning_rate": 0.0006210691823899371,
      "loss": 0.5314,
      "num_input_tokens_seen": 47616,
      "step": 80
    },
    {
      "epoch": 0.044549266247379454,
      "grad_norm": 2.421875,
      "learning_rate": 0.000660377358490566,
      "loss": 0.6236,
      "num_input_tokens_seen": 50368,
      "step": 85
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 2.296875,
      "learning_rate": 0.000699685534591195,
      "loss": 0.664,
      "num_input_tokens_seen": 54016,
      "step": 90
    },
    {
      "epoch": 0.04979035639412998,
      "grad_norm": 10.8125,
      "learning_rate": 0.0007389937106918238,
      "loss": 0.5132,
      "num_input_tokens_seen": 57888,
      "step": 95
    },
    {
      "epoch": 0.05241090146750524,
      "grad_norm": 3.609375,
      "learning_rate": 0.0007783018867924529,
      "loss": 0.5627,
      "num_input_tokens_seen": 61024,
      "step": 100
    },
    {
      "epoch": 0.055031446540880505,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0008176100628930818,
      "loss": 0.6611,
      "num_input_tokens_seen": 64672,
      "step": 105
    },
    {
      "epoch": 0.057651991614255764,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0008569182389937106,
      "loss": 0.5128,
      "num_input_tokens_seen": 67616,
      "step": 110
    },
    {
      "epoch": 0.06027253668763103,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008962264150943396,
      "loss": 0.5436,
      "num_input_tokens_seen": 71232,
      "step": 115
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 19.25,
      "learning_rate": 0.0009355345911949684,
      "loss": 0.4524,
      "num_input_tokens_seen": 74880,
      "step": 120
    },
    {
      "epoch": 0.06551362683438156,
      "grad_norm": 10.3125,
      "learning_rate": 0.0009748427672955974,
      "loss": 0.555,
      "num_input_tokens_seen": 78432,
      "step": 125
    },
    {
      "epoch": 0.06813417190775681,
      "grad_norm": 1.46875,
      "learning_rate": 0.0010141509433962263,
      "loss": 0.4996,
      "num_input_tokens_seen": 81088,
      "step": 130
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0010534591194968554,
      "loss": 0.5435,
      "num_input_tokens_seen": 84896,
      "step": 135
    },
    {
      "epoch": 0.07337526205450734,
      "grad_norm": 1.0625,
      "learning_rate": 0.0010927672955974841,
      "loss": 0.6155,
      "num_input_tokens_seen": 88736,
      "step": 140
    },
    {
      "epoch": 0.0759958071278826,
      "grad_norm": 1.8046875,
      "learning_rate": 0.001132075471698113,
      "loss": 0.844,
      "num_input_tokens_seen": 92224,
      "step": 145
    },
    {
      "epoch": 0.07861635220125786,
      "grad_norm": 3.765625,
      "learning_rate": 0.0011713836477987422,
      "loss": 0.5394,
      "num_input_tokens_seen": 95168,
      "step": 150
    },
    {
      "epoch": 0.08123689727463312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.001210691823899371,
      "loss": 0.5812,
      "num_input_tokens_seen": 98976,
      "step": 155
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 338.0,
      "learning_rate": 0.0012499999999999998,
      "loss": 1.0217,
      "num_input_tokens_seen": 102176,
      "step": 160
    },
    {
      "epoch": 0.08647798742138364,
      "grad_norm": 0.84765625,
      "learning_rate": 0.001289308176100629,
      "loss": 0.5575,
      "num_input_tokens_seen": 105824,
      "step": 165
    },
    {
      "epoch": 0.08909853249475891,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0013286163522012579,
      "loss": 0.5285,
      "num_input_tokens_seen": 108512,
      "step": 170
    },
    {
      "epoch": 0.09171907756813417,
      "grad_norm": 0.921875,
      "learning_rate": 0.0013679245283018866,
      "loss": 1.1701,
      "num_input_tokens_seen": 111872,
      "step": 175
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0014072327044025157,
      "loss": 0.4581,
      "num_input_tokens_seen": 115552,
      "step": 180
    },
    {
      "epoch": 0.09696016771488469,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0014465408805031446,
      "loss": 0.6527,
      "num_input_tokens_seen": 118784,
      "step": 185
    },
    {
      "epoch": 0.09958071278825996,
      "grad_norm": 0.84375,
      "learning_rate": 0.0014858490566037736,
      "loss": 0.4345,
      "num_input_tokens_seen": 122528,
      "step": 190
    },
    {
      "epoch": 0.10220125786163523,
      "grad_norm": 0.578125,
      "learning_rate": 0.0015251572327044025,
      "loss": 0.5036,
      "num_input_tokens_seen": 125024,
      "step": 195
    },
    {
      "epoch": 0.10482180293501048,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0015644654088050314,
      "loss": 0.5946,
      "num_input_tokens_seen": 128096,
      "step": 200
    },
    {
      "epoch": 0.10744234800838574,
      "grad_norm": 1.359375,
      "learning_rate": 0.0016037735849056603,
      "loss": 0.6472,
      "num_input_tokens_seen": 130496,
      "step": 205
    },
    {
      "epoch": 0.11006289308176101,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0016430817610062892,
      "loss": 0.5741,
      "num_input_tokens_seen": 133152,
      "step": 210
    },
    {
      "epoch": 0.11268343815513626,
      "grad_norm": 2.625,
      "learning_rate": 0.0016823899371069182,
      "loss": 0.7318,
      "num_input_tokens_seen": 136384,
      "step": 215
    },
    {
      "epoch": 0.11530398322851153,
      "grad_norm": 0.6796875,
      "learning_rate": 0.001721698113207547,
      "loss": 0.6247,
      "num_input_tokens_seen": 139456,
      "step": 220
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 2.0625,
      "learning_rate": 0.0017610062893081762,
      "loss": 0.6002,
      "num_input_tokens_seen": 142816,
      "step": 225
    },
    {
      "epoch": 0.12054507337526206,
      "grad_norm": 0.431640625,
      "learning_rate": 0.001800314465408805,
      "loss": 0.5176,
      "num_input_tokens_seen": 145696,
      "step": 230
    },
    {
      "epoch": 0.12316561844863731,
      "grad_norm": 1.703125,
      "learning_rate": 0.0018396226415094339,
      "loss": 0.645,
      "num_input_tokens_seen": 149024,
      "step": 235
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0018789308176100628,
      "loss": 0.5637,
      "num_input_tokens_seen": 151840,
      "step": 240
    },
    {
      "epoch": 0.12840670859538783,
      "grad_norm": 22.375,
      "learning_rate": 0.001918238993710692,
      "loss": 0.7413,
      "num_input_tokens_seen": 155456,
      "step": 245
    },
    {
      "epoch": 0.1310272536687631,
      "grad_norm": 2.734375,
      "learning_rate": 0.001957547169811321,
      "loss": 0.6253,
      "num_input_tokens_seen": 158528,
      "step": 250
    },
    {
      "epoch": 0.13364779874213836,
      "grad_norm": 0.443359375,
      "learning_rate": 0.0019968553459119493,
      "loss": 0.511,
      "num_input_tokens_seen": 166496,
      "step": 255
    },
    {
      "epoch": 0.13626834381551362,
      "grad_norm": 0.390625,
      "learning_rate": 0.0020361635220125787,
      "loss": 0.4821,
      "num_input_tokens_seen": 169568,
      "step": 260
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.55078125,
      "learning_rate": 0.0020754716981132076,
      "loss": 0.5612,
      "num_input_tokens_seen": 172704,
      "step": 265
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 0.9296875,
      "learning_rate": 0.002114779874213836,
      "loss": 0.722,
      "num_input_tokens_seen": 176480,
      "step": 270
    },
    {
      "epoch": 0.1441299790356394,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0021540880503144655,
      "loss": 0.5341,
      "num_input_tokens_seen": 181408,
      "step": 275
    },
    {
      "epoch": 0.14675052410901468,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.0021933962264150944,
      "loss": 0.5993,
      "num_input_tokens_seen": 184736,
      "step": 280
    },
    {
      "epoch": 0.14937106918238993,
      "grad_norm": 0.30078125,
      "learning_rate": 0.0022327044025157233,
      "loss": 0.5156,
      "num_input_tokens_seen": 188352,
      "step": 285
    },
    {
      "epoch": 0.1519916142557652,
      "grad_norm": 0.546875,
      "learning_rate": 0.0022720125786163522,
      "loss": 0.4667,
      "num_input_tokens_seen": 193408,
      "step": 290
    },
    {
      "epoch": 0.15461215932914046,
      "grad_norm": 0.412109375,
      "learning_rate": 0.002311320754716981,
      "loss": 0.417,
      "num_input_tokens_seen": 196384,
      "step": 295
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 0.62109375,
      "learning_rate": 0.00235062893081761,
      "loss": 0.6198,
      "num_input_tokens_seen": 199360,
      "step": 300
    },
    {
      "epoch": 0.159853249475891,
      "grad_norm": 0.2158203125,
      "learning_rate": 0.002389937106918239,
      "loss": 0.5278,
      "num_input_tokens_seen": 202848,
      "step": 305
    },
    {
      "epoch": 0.16247379454926625,
      "grad_norm": 1.96875,
      "learning_rate": 0.002429245283018868,
      "loss": 0.4694,
      "num_input_tokens_seen": 207104,
      "step": 310
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 2.390625,
      "learning_rate": 0.002468553459119497,
      "loss": 0.6627,
      "num_input_tokens_seen": 209600,
      "step": 315
    },
    {
      "epoch": 0.16771488469601678,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0025078616352201257,
      "loss": 0.6067,
      "num_input_tokens_seen": 213152,
      "step": 320
    },
    {
      "epoch": 0.17033542976939203,
      "grad_norm": 0.375,
      "learning_rate": 0.0025471698113207547,
      "loss": 0.6694,
      "num_input_tokens_seen": 215616,
      "step": 325
    },
    {
      "epoch": 0.17295597484276728,
      "grad_norm": 0.19921875,
      "learning_rate": 0.0025864779874213836,
      "loss": 0.6111,
      "num_input_tokens_seen": 218528,
      "step": 330
    },
    {
      "epoch": 0.17557651991614256,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0026257861635220125,
      "loss": 0.422,
      "num_input_tokens_seen": 221824,
      "step": 335
    },
    {
      "epoch": 0.17819706498951782,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0026650943396226414,
      "loss": 0.4998,
      "num_input_tokens_seen": 225280,
      "step": 340
    },
    {
      "epoch": 0.18081761006289307,
      "grad_norm": 0.283203125,
      "learning_rate": 0.0027044025157232704,
      "loss": 0.4446,
      "num_input_tokens_seen": 230912,
      "step": 345
    },
    {
      "epoch": 0.18343815513626835,
      "grad_norm": 0.20703125,
      "learning_rate": 0.0027437106918238993,
      "loss": 0.5223,
      "num_input_tokens_seen": 234592,
      "step": 350
    },
    {
      "epoch": 0.1860587002096436,
      "grad_norm": 0.177734375,
      "learning_rate": 0.002783018867924528,
      "loss": 0.4735,
      "num_input_tokens_seen": 237600,
      "step": 355
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 0.224609375,
      "learning_rate": 0.002822327044025157,
      "loss": 0.6054,
      "num_input_tokens_seen": 241088,
      "step": 360
    },
    {
      "epoch": 0.19129979035639413,
      "grad_norm": 0.369140625,
      "learning_rate": 0.002861635220125786,
      "loss": 0.5513,
      "num_input_tokens_seen": 243520,
      "step": 365
    },
    {
      "epoch": 0.19392033542976939,
      "grad_norm": 1.8125,
      "learning_rate": 0.002900943396226415,
      "loss": 0.6999,
      "num_input_tokens_seen": 247456,
      "step": 370
    },
    {
      "epoch": 0.19654088050314467,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.002940251572327044,
      "loss": 0.4726,
      "num_input_tokens_seen": 250784,
      "step": 375
    },
    {
      "epoch": 0.19916142557651992,
      "grad_norm": 0.39453125,
      "learning_rate": 0.002979559748427673,
      "loss": 0.5151,
      "num_input_tokens_seen": 255040,
      "step": 380
    },
    {
      "epoch": 0.20178197064989517,
      "grad_norm": 0.236328125,
      "learning_rate": 0.0030188679245283017,
      "loss": 0.5961,
      "num_input_tokens_seen": 258240,
      "step": 385
    },
    {
      "epoch": 0.20440251572327045,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0030581761006289307,
      "loss": 0.4508,
      "num_input_tokens_seen": 261152,
      "step": 390
    },
    {
      "epoch": 0.2070230607966457,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0030974842767295596,
      "loss": 0.5334,
      "num_input_tokens_seen": 264544,
      "step": 395
    },
    {
      "epoch": 0.20964360587002095,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0031367924528301885,
      "loss": 0.5972,
      "num_input_tokens_seen": 267136,
      "step": 400
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 0.220703125,
      "learning_rate": 0.0031761006289308174,
      "loss": 0.4627,
      "num_input_tokens_seen": 270464,
      "step": 405
    },
    {
      "epoch": 0.2148846960167715,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0032154088050314463,
      "loss": 0.4958,
      "num_input_tokens_seen": 274528,
      "step": 410
    },
    {
      "epoch": 0.21750524109014674,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0032547169811320753,
      "loss": 0.4783,
      "num_input_tokens_seen": 277376,
      "step": 415
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 0.376953125,
      "learning_rate": 0.003294025157232704,
      "loss": 0.3616,
      "num_input_tokens_seen": 279616,
      "step": 420
    },
    {
      "epoch": 0.22274633123689727,
      "grad_norm": 0.65625,
      "learning_rate": 0.003333333333333333,
      "loss": 0.4586,
      "num_input_tokens_seen": 282656,
      "step": 425
    },
    {
      "epoch": 0.22536687631027252,
      "grad_norm": 0.275390625,
      "learning_rate": 0.003372641509433962,
      "loss": 0.5508,
      "num_input_tokens_seen": 285600,
      "step": 430
    },
    {
      "epoch": 0.2279874213836478,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.003411949685534591,
      "loss": 0.3919,
      "num_input_tokens_seen": 289568,
      "step": 435
    },
    {
      "epoch": 0.23060796645702306,
      "grad_norm": 54.25,
      "learning_rate": 0.0034512578616352203,
      "loss": 4.6963,
      "num_input_tokens_seen": 292768,
      "step": 440
    },
    {
      "epoch": 0.23322851153039834,
      "grad_norm": 0.234375,
      "learning_rate": 0.003490566037735849,
      "loss": 1.3091,
      "num_input_tokens_seen": 295456,
      "step": 445
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0035298742138364777,
      "loss": 0.5661,
      "num_input_tokens_seen": 299488,
      "step": 450
    },
    {
      "epoch": 0.23846960167714884,
      "grad_norm": 0.33984375,
      "learning_rate": 0.003569182389937107,
      "loss": 0.455,
      "num_input_tokens_seen": 302976,
      "step": 455
    },
    {
      "epoch": 0.24109014675052412,
      "grad_norm": 109.5,
      "learning_rate": 0.0036084905660377356,
      "loss": 0.5935,
      "num_input_tokens_seen": 306560,
      "step": 460
    },
    {
      "epoch": 0.24371069182389937,
      "grad_norm": 0.216796875,
      "learning_rate": 0.0036477987421383645,
      "loss": 0.439,
      "num_input_tokens_seen": 309536,
      "step": 465
    },
    {
      "epoch": 0.24633123689727462,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.003687106918238994,
      "loss": 0.4878,
      "num_input_tokens_seen": 312576,
      "step": 470
    },
    {
      "epoch": 0.2489517819706499,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0037264150943396223,
      "loss": 0.4764,
      "num_input_tokens_seen": 315616,
      "step": 475
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0037657232704402517,
      "loss": 0.5019,
      "num_input_tokens_seen": 319136,
      "step": 480
    },
    {
      "epoch": 0.25419287211740044,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.00380503144654088,
      "loss": 0.6124,
      "num_input_tokens_seen": 321376,
      "step": 485
    },
    {
      "epoch": 0.25681341719077566,
      "grad_norm": 0.35546875,
      "learning_rate": 0.003844339622641509,
      "loss": 0.5429,
      "num_input_tokens_seen": 324928,
      "step": 490
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0038836477987421385,
      "loss": 0.6511,
      "num_input_tokens_seen": 329248,
      "step": 495
    },
    {
      "epoch": 0.2620545073375262,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.003922955974842767,
      "loss": 0.625,
      "num_input_tokens_seen": 332288,
      "step": 500
    },
    {
      "epoch": 0.26467505241090145,
      "grad_norm": 0.197265625,
      "learning_rate": 0.003962264150943396,
      "loss": 0.5014,
      "num_input_tokens_seen": 334752,
      "step": 505
    },
    {
      "epoch": 0.2672955974842767,
      "grad_norm": 0.150390625,
      "learning_rate": 0.004001572327044026,
      "loss": 0.543,
      "num_input_tokens_seen": 337856,
      "step": 510
    },
    {
      "epoch": 0.269916142557652,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.004040880503144654,
      "loss": 0.6324,
      "num_input_tokens_seen": 341792,
      "step": 515
    },
    {
      "epoch": 0.27253668763102723,
      "grad_norm": 0.291015625,
      "learning_rate": 0.004080188679245283,
      "loss": 1.4137,
      "num_input_tokens_seen": 344768,
      "step": 520
    },
    {
      "epoch": 0.2751572327044025,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.004119496855345912,
      "loss": 0.4416,
      "num_input_tokens_seen": 347360,
      "step": 525
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0041588050314465405,
      "loss": 0.4899,
      "num_input_tokens_seen": 350592,
      "step": 530
    },
    {
      "epoch": 0.280398322851153,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.004198113207547169,
      "loss": 0.4846,
      "num_input_tokens_seen": 353568,
      "step": 535
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.004237421383647798,
      "loss": 0.6225,
      "num_input_tokens_seen": 356320,
      "step": 540
    },
    {
      "epoch": 0.2856394129979036,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.004276729559748427,
      "loss": 0.5091,
      "num_input_tokens_seen": 359968,
      "step": 545
    },
    {
      "epoch": 0.2882599580712788,
      "grad_norm": 0.2578125,
      "learning_rate": 0.004316037735849056,
      "loss": 0.428,
      "num_input_tokens_seen": 362752,
      "step": 550
    },
    {
      "epoch": 0.2908805031446541,
      "grad_norm": 0.22265625,
      "learning_rate": 0.004355345911949685,
      "loss": 0.5728,
      "num_input_tokens_seen": 365312,
      "step": 555
    },
    {
      "epoch": 0.29350104821802936,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.004394654088050315,
      "loss": 0.4618,
      "num_input_tokens_seen": 368896,
      "step": 560
    },
    {
      "epoch": 0.29612159329140464,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.004433962264150943,
      "loss": 0.5231,
      "num_input_tokens_seen": 371648,
      "step": 565
    },
    {
      "epoch": 0.29874213836477986,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.004473270440251572,
      "loss": 0.5661,
      "num_input_tokens_seen": 374848,
      "step": 570
    },
    {
      "epoch": 0.30136268343815514,
      "grad_norm": 0.162109375,
      "learning_rate": 0.004512578616352202,
      "loss": 0.5291,
      "num_input_tokens_seen": 377184,
      "step": 575
    },
    {
      "epoch": 0.3039832285115304,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.00455188679245283,
      "loss": 0.6348,
      "num_input_tokens_seen": 381504,
      "step": 580
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 0.171875,
      "learning_rate": 0.004591194968553459,
      "loss": 0.5365,
      "num_input_tokens_seen": 384672,
      "step": 585
    },
    {
      "epoch": 0.30922431865828093,
      "grad_norm": 0.14453125,
      "learning_rate": 0.004630503144654088,
      "loss": 0.515,
      "num_input_tokens_seen": 389376,
      "step": 590
    },
    {
      "epoch": 0.3118448637316562,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0046698113207547165,
      "loss": 0.5701,
      "num_input_tokens_seen": 392512,
      "step": 595
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 0.1328125,
      "learning_rate": 0.004709119496855345,
      "loss": 0.5009,
      "num_input_tokens_seen": 396864,
      "step": 600
    },
    {
      "epoch": 0.3170859538784067,
      "grad_norm": 0.25390625,
      "learning_rate": 0.004748427672955975,
      "loss": 0.4783,
      "num_input_tokens_seen": 399520,
      "step": 605
    },
    {
      "epoch": 0.319706498951782,
      "grad_norm": 1.7109375,
      "learning_rate": 0.004787735849056603,
      "loss": 0.5085,
      "num_input_tokens_seen": 402560,
      "step": 610
    },
    {
      "epoch": 0.3223270440251572,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.004827044025157232,
      "loss": 0.6263,
      "num_input_tokens_seen": 406240,
      "step": 615
    },
    {
      "epoch": 0.3249475890985325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.004866352201257862,
      "loss": 2.4162,
      "num_input_tokens_seen": 408960,
      "step": 620
    },
    {
      "epoch": 0.3275681341719078,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.00490566037735849,
      "loss": 0.658,
      "num_input_tokens_seen": 412224,
      "step": 625
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 0.119140625,
      "learning_rate": 0.004944968553459119,
      "loss": 0.4752,
      "num_input_tokens_seen": 415296,
      "step": 630
    },
    {
      "epoch": 0.3328092243186583,
      "grad_norm": 0.22265625,
      "learning_rate": 0.004984276729559749,
      "loss": 0.4522,
      "num_input_tokens_seen": 417504,
      "step": 635
    },
    {
      "epoch": 0.33542976939203356,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.005023584905660377,
      "loss": 0.5249,
      "num_input_tokens_seen": 420288,
      "step": 640
    },
    {
      "epoch": 0.3380503144654088,
      "grad_norm": 0.328125,
      "learning_rate": 0.005062893081761006,
      "loss": 0.6741,
      "num_input_tokens_seen": 422560,
      "step": 645
    },
    {
      "epoch": 0.34067085953878407,
      "grad_norm": 0.314453125,
      "learning_rate": 0.0051022012578616355,
      "loss": 0.6016,
      "num_input_tokens_seen": 425472,
      "step": 650
    },
    {
      "epoch": 0.34329140461215935,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0051415094339622635,
      "loss": 0.515,
      "num_input_tokens_seen": 429056,
      "step": 655
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0051808176100628925,
      "loss": 0.4824,
      "num_input_tokens_seen": 433024,
      "step": 660
    },
    {
      "epoch": 0.34853249475890985,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.005220125786163522,
      "loss": 0.5512,
      "num_input_tokens_seen": 435776,
      "step": 665
    },
    {
      "epoch": 0.35115303983228513,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.00525943396226415,
      "loss": 0.5034,
      "num_input_tokens_seen": 438432,
      "step": 670
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 0.12890625,
      "learning_rate": 0.005298742138364779,
      "loss": 0.5035,
      "num_input_tokens_seen": 442400,
      "step": 675
    },
    {
      "epoch": 0.35639412997903563,
      "grad_norm": 0.25,
      "learning_rate": 0.005338050314465409,
      "loss": 0.5874,
      "num_input_tokens_seen": 445888,
      "step": 680
    },
    {
      "epoch": 0.3590146750524109,
      "grad_norm": 0.07421875,
      "learning_rate": 0.005377358490566037,
      "loss": 0.4941,
      "num_input_tokens_seen": 448864,
      "step": 685
    },
    {
      "epoch": 0.36163522012578614,
      "grad_norm": 0.109375,
      "learning_rate": 0.005416666666666666,
      "loss": 0.6583,
      "num_input_tokens_seen": 452640,
      "step": 690
    },
    {
      "epoch": 0.3642557651991614,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.005455974842767296,
      "loss": 0.6004,
      "num_input_tokens_seen": 455296,
      "step": 695
    },
    {
      "epoch": 0.3668763102725367,
      "grad_norm": 0.271484375,
      "learning_rate": 0.005495283018867924,
      "loss": 0.5327,
      "num_input_tokens_seen": 457888,
      "step": 700
    },
    {
      "epoch": 0.3694968553459119,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.005534591194968553,
      "loss": 0.5981,
      "num_input_tokens_seen": 461664,
      "step": 705
    },
    {
      "epoch": 0.3721174004192872,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0055738993710691825,
      "loss": 0.5958,
      "num_input_tokens_seen": 464352,
      "step": 710
    },
    {
      "epoch": 0.3747379454926625,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.005613207547169811,
      "loss": 0.6053,
      "num_input_tokens_seen": 468256,
      "step": 715
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 0.30859375,
      "learning_rate": 0.0056525157232704395,
      "loss": 0.4738,
      "num_input_tokens_seen": 471104,
      "step": 720
    },
    {
      "epoch": 0.379979035639413,
      "grad_norm": 3.328125,
      "learning_rate": 0.005691823899371069,
      "loss": 0.6215,
      "num_input_tokens_seen": 473696,
      "step": 725
    },
    {
      "epoch": 0.38259958071278827,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.005731132075471698,
      "loss": 0.5059,
      "num_input_tokens_seen": 476512,
      "step": 730
    },
    {
      "epoch": 0.38522012578616355,
      "grad_norm": 0.126953125,
      "learning_rate": 0.005770440251572326,
      "loss": 0.5783,
      "num_input_tokens_seen": 480128,
      "step": 735
    },
    {
      "epoch": 0.38784067085953877,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.005809748427672956,
      "loss": 0.6165,
      "num_input_tokens_seen": 483808,
      "step": 740
    },
    {
      "epoch": 0.39046121593291405,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.005849056603773585,
      "loss": 0.5926,
      "num_input_tokens_seen": 486240,
      "step": 745
    },
    {
      "epoch": 0.39308176100628933,
      "grad_norm": 0.14453125,
      "learning_rate": 0.005888364779874213,
      "loss": 0.4745,
      "num_input_tokens_seen": 488448,
      "step": 750
    },
    {
      "epoch": 0.39570230607966456,
      "grad_norm": 0.23046875,
      "learning_rate": 0.005927672955974843,
      "loss": 0.5552,
      "num_input_tokens_seen": 491392,
      "step": 755
    },
    {
      "epoch": 0.39832285115303984,
      "grad_norm": 0.162109375,
      "learning_rate": 0.005966981132075472,
      "loss": 0.5121,
      "num_input_tokens_seen": 495232,
      "step": 760
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0060062893081761,
      "loss": 0.4764,
      "num_input_tokens_seen": 498016,
      "step": 765
    },
    {
      "epoch": 0.40356394129979034,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.00604559748427673,
      "loss": 0.4979,
      "num_input_tokens_seen": 501152,
      "step": 770
    },
    {
      "epoch": 0.4061844863731656,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0060849056603773585,
      "loss": 0.4842,
      "num_input_tokens_seen": 503872,
      "step": 775
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.006124213836477987,
      "loss": 0.5081,
      "num_input_tokens_seen": 507328,
      "step": 780
    },
    {
      "epoch": 0.4114255765199161,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.006163522012578616,
      "loss": 0.4033,
      "num_input_tokens_seen": 510912,
      "step": 785
    },
    {
      "epoch": 0.4140461215932914,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.006202830188679245,
      "loss": 0.5732,
      "num_input_tokens_seen": 514304,
      "step": 790
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.140625,
      "learning_rate": 0.006242138364779873,
      "loss": 0.4255,
      "num_input_tokens_seen": 517440,
      "step": 795
    },
    {
      "epoch": 0.4192872117400419,
      "grad_norm": 0.14453125,
      "learning_rate": 0.006281446540880503,
      "loss": 0.645,
      "num_input_tokens_seen": 520416,
      "step": 800
    },
    {
      "epoch": 0.4219077568134172,
      "grad_norm": 0.12109375,
      "learning_rate": 0.006320754716981132,
      "loss": 0.6101,
      "num_input_tokens_seen": 523456,
      "step": 805
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.00636006289308176,
      "loss": 0.4931,
      "num_input_tokens_seen": 526720,
      "step": 810
    },
    {
      "epoch": 0.4271488469601677,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.00639937106918239,
      "loss": 0.4125,
      "num_input_tokens_seen": 530272,
      "step": 815
    },
    {
      "epoch": 0.429769392033543,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.006438679245283019,
      "loss": 0.6165,
      "num_input_tokens_seen": 533408,
      "step": 820
    },
    {
      "epoch": 0.43238993710691825,
      "grad_norm": 0.34765625,
      "learning_rate": 0.006477987421383647,
      "loss": 0.6289,
      "num_input_tokens_seen": 536768,
      "step": 825
    },
    {
      "epoch": 0.4350104821802935,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.006517295597484277,
      "loss": 0.5924,
      "num_input_tokens_seen": 541184,
      "step": 830
    },
    {
      "epoch": 0.43763102725366876,
      "grad_norm": 0.1640625,
      "learning_rate": 0.006556603773584906,
      "loss": 0.4428,
      "num_input_tokens_seen": 544576,
      "step": 835
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.006595911949685534,
      "loss": 0.5669,
      "num_input_tokens_seen": 549408,
      "step": 840
    },
    {
      "epoch": 0.44287211740041926,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0066352201257861634,
      "loss": 0.4527,
      "num_input_tokens_seen": 552032,
      "step": 845
    },
    {
      "epoch": 0.44549266247379454,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.006674528301886792,
      "loss": 0.4904,
      "num_input_tokens_seen": 555872,
      "step": 850
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.00671383647798742,
      "loss": 0.6025,
      "num_input_tokens_seen": 559136,
      "step": 855
    },
    {
      "epoch": 0.45073375262054505,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00675314465408805,
      "loss": 0.5232,
      "num_input_tokens_seen": 561984,
      "step": 860
    },
    {
      "epoch": 0.4533542976939203,
      "grad_norm": 12.0625,
      "learning_rate": 0.006792452830188679,
      "loss": 0.651,
      "num_input_tokens_seen": 564448,
      "step": 865
    },
    {
      "epoch": 0.4559748427672956,
      "grad_norm": 0.15625,
      "learning_rate": 0.006831761006289307,
      "loss": 0.6018,
      "num_input_tokens_seen": 566752,
      "step": 870
    },
    {
      "epoch": 0.4585953878406709,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.006871069182389937,
      "loss": 0.5102,
      "num_input_tokens_seen": 569856,
      "step": 875
    },
    {
      "epoch": 0.4612159329140461,
      "grad_norm": 0.11328125,
      "learning_rate": 0.006910377358490566,
      "loss": 0.4938,
      "num_input_tokens_seen": 573056,
      "step": 880
    },
    {
      "epoch": 0.4638364779874214,
      "grad_norm": 0.1953125,
      "learning_rate": 0.006949685534591195,
      "loss": 0.5621,
      "num_input_tokens_seen": 575712,
      "step": 885
    },
    {
      "epoch": 0.46645702306079667,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.006988993710691824,
      "loss": 0.439,
      "num_input_tokens_seen": 578880,
      "step": 890
    },
    {
      "epoch": 0.4690775681341719,
      "grad_norm": 1.265625,
      "learning_rate": 0.007028301886792453,
      "loss": 0.4717,
      "num_input_tokens_seen": 581920,
      "step": 895
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.007067610062893082,
      "loss": 0.5117,
      "num_input_tokens_seen": 584672,
      "step": 900
    },
    {
      "epoch": 0.47431865828092246,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0071069182389937105,
      "loss": 0.4639,
      "num_input_tokens_seen": 588704,
      "step": 905
    },
    {
      "epoch": 0.4769392033542977,
      "grad_norm": 0.17578125,
      "learning_rate": 0.007146226415094339,
      "loss": 0.4913,
      "num_input_tokens_seen": 591712,
      "step": 910
    },
    {
      "epoch": 0.47955974842767296,
      "grad_norm": 0.435546875,
      "learning_rate": 0.007185534591194968,
      "loss": 0.5585,
      "num_input_tokens_seen": 595104,
      "step": 915
    },
    {
      "epoch": 0.48218029350104824,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.007224842767295597,
      "loss": 0.4923,
      "num_input_tokens_seen": 599808,
      "step": 920
    },
    {
      "epoch": 0.48480083857442346,
      "grad_norm": 0.1875,
      "learning_rate": 0.007264150943396226,
      "loss": 0.4673,
      "num_input_tokens_seen": 602720,
      "step": 925
    },
    {
      "epoch": 0.48742138364779874,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.007303459119496855,
      "loss": 0.4877,
      "num_input_tokens_seen": 606176,
      "step": 930
    },
    {
      "epoch": 0.490041928721174,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.007342767295597484,
      "loss": 0.4617,
      "num_input_tokens_seen": 608864,
      "step": 935
    },
    {
      "epoch": 0.49266247379454925,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.007382075471698113,
      "loss": 0.4893,
      "num_input_tokens_seen": 611968,
      "step": 940
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 0.130859375,
      "learning_rate": 0.007421383647798742,
      "loss": 0.6029,
      "num_input_tokens_seen": 614848,
      "step": 945
    },
    {
      "epoch": 0.4979035639412998,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.007460691823899371,
      "loss": 0.6554,
      "num_input_tokens_seen": 618080,
      "step": 950
    },
    {
      "epoch": 0.500524109014675,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0075,
      "loss": 0.3577,
      "num_input_tokens_seen": 620736,
      "step": 955
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 0.119140625,
      "learning_rate": 0.007539308176100629,
      "loss": 0.4388,
      "num_input_tokens_seen": 624480,
      "step": 960
    },
    {
      "epoch": 0.5057651991614256,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0075786163522012584,
      "loss": 0.4379,
      "num_input_tokens_seen": 628448,
      "step": 965
    },
    {
      "epoch": 0.5083857442348009,
      "grad_norm": 0.216796875,
      "learning_rate": 0.007617924528301887,
      "loss": 0.5789,
      "num_input_tokens_seen": 632000,
      "step": 970
    },
    {
      "epoch": 0.5110062893081762,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0076572327044025145,
      "loss": 0.6722,
      "num_input_tokens_seen": 635680,
      "step": 975
    },
    {
      "epoch": 0.5136268343815513,
      "grad_norm": 0.134765625,
      "learning_rate": 0.007696540880503144,
      "loss": 0.7385,
      "num_input_tokens_seen": 638496,
      "step": 980
    },
    {
      "epoch": 0.5162473794549266,
      "grad_norm": 0.107421875,
      "learning_rate": 0.007735849056603773,
      "loss": 0.4547,
      "num_input_tokens_seen": 642272,
      "step": 985
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.007775157232704402,
      "loss": 0.5163,
      "num_input_tokens_seen": 645888,
      "step": 990
    },
    {
      "epoch": 0.5214884696016772,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.007814465408805031,
      "loss": 0.3653,
      "num_input_tokens_seen": 651808,
      "step": 995
    },
    {
      "epoch": 0.5241090146750524,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.00785377358490566,
      "loss": 0.5253,
      "num_input_tokens_seen": 655488,
      "step": 1000
    },
    {
      "epoch": 0.5267295597484277,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.007893081761006289,
      "loss": 0.4522,
      "num_input_tokens_seen": 658976,
      "step": 1005
    },
    {
      "epoch": 0.5293501048218029,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.007932389937106918,
      "loss": 0.4503,
      "num_input_tokens_seen": 662048,
      "step": 1010
    },
    {
      "epoch": 0.5319706498951782,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.007971698113207547,
      "loss": 0.5725,
      "num_input_tokens_seen": 664640,
      "step": 1015
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.008011006289308176,
      "loss": 0.4823,
      "num_input_tokens_seen": 669056,
      "step": 1020
    },
    {
      "epoch": 0.5372117400419287,
      "grad_norm": 0.126953125,
      "learning_rate": 0.008050314465408805,
      "loss": 0.5155,
      "num_input_tokens_seen": 672128,
      "step": 1025
    },
    {
      "epoch": 0.539832285115304,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.008089622641509434,
      "loss": 0.4882,
      "num_input_tokens_seen": 675136,
      "step": 1030
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.008128930817610062,
      "loss": 0.5913,
      "num_input_tokens_seen": 678976,
      "step": 1035
    },
    {
      "epoch": 0.5450733752620545,
      "grad_norm": 0.1640625,
      "learning_rate": 0.008168238993710691,
      "loss": 0.6392,
      "num_input_tokens_seen": 682176,
      "step": 1040
    },
    {
      "epoch": 0.5476939203354297,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00820754716981132,
      "loss": 0.5489,
      "num_input_tokens_seen": 684768,
      "step": 1045
    },
    {
      "epoch": 0.550314465408805,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00824685534591195,
      "loss": 0.5386,
      "num_input_tokens_seen": 687392,
      "step": 1050
    },
    {
      "epoch": 0.5529350104821803,
      "grad_norm": 0.302734375,
      "learning_rate": 0.008286163522012578,
      "loss": 0.4921,
      "num_input_tokens_seen": 690496,
      "step": 1055
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.008325471698113207,
      "loss": 0.6507,
      "num_input_tokens_seen": 693376,
      "step": 1060
    },
    {
      "epoch": 0.5581761006289309,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.008364779874213836,
      "loss": 0.4586,
      "num_input_tokens_seen": 695808,
      "step": 1065
    },
    {
      "epoch": 0.560796645702306,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.008404088050314465,
      "loss": 0.482,
      "num_input_tokens_seen": 699424,
      "step": 1070
    },
    {
      "epoch": 0.5634171907756813,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.008443396226415094,
      "loss": 0.5269,
      "num_input_tokens_seen": 703584,
      "step": 1075
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.008482704402515723,
      "loss": 0.5047,
      "num_input_tokens_seen": 706144,
      "step": 1080
    },
    {
      "epoch": 0.5686582809224319,
      "grad_norm": 0.087890625,
      "learning_rate": 0.008522012578616352,
      "loss": 0.5123,
      "num_input_tokens_seen": 708864,
      "step": 1085
    },
    {
      "epoch": 0.5712788259958071,
      "grad_norm": 0.125,
      "learning_rate": 0.00856132075471698,
      "loss": 0.546,
      "num_input_tokens_seen": 712352,
      "step": 1090
    },
    {
      "epoch": 0.5738993710691824,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00860062893081761,
      "loss": 0.7096,
      "num_input_tokens_seen": 715296,
      "step": 1095
    },
    {
      "epoch": 0.5765199161425576,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.008639937106918238,
      "loss": 0.4296,
      "num_input_tokens_seen": 718528,
      "step": 1100
    },
    {
      "epoch": 0.5791404612159329,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.008679245283018867,
      "loss": 0.4475,
      "num_input_tokens_seen": 722752,
      "step": 1105
    },
    {
      "epoch": 0.5817610062893082,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.008718553459119496,
      "loss": 0.456,
      "num_input_tokens_seen": 725824,
      "step": 1110
    },
    {
      "epoch": 0.5843815513626834,
      "grad_norm": 0.064453125,
      "learning_rate": 0.008757861635220125,
      "loss": 0.5754,
      "num_input_tokens_seen": 730080,
      "step": 1115
    },
    {
      "epoch": 0.5870020964360587,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.008797169811320754,
      "loss": 0.4993,
      "num_input_tokens_seen": 733216,
      "step": 1120
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.008836477987421385,
      "loss": 0.6518,
      "num_input_tokens_seen": 736416,
      "step": 1125
    },
    {
      "epoch": 0.5922431865828093,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.008875786163522012,
      "loss": 0.4539,
      "num_input_tokens_seen": 739808,
      "step": 1130
    },
    {
      "epoch": 0.5948637316561844,
      "grad_norm": 0.1328125,
      "learning_rate": 0.008915094339622641,
      "loss": 0.5387,
      "num_input_tokens_seen": 742400,
      "step": 1135
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00895440251572327,
      "loss": 0.3931,
      "num_input_tokens_seen": 745760,
      "step": 1140
    },
    {
      "epoch": 0.600104821802935,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.008993710691823899,
      "loss": 0.5699,
      "num_input_tokens_seen": 748736,
      "step": 1145
    },
    {
      "epoch": 0.6027253668763103,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.009033018867924528,
      "loss": 0.5237,
      "num_input_tokens_seen": 752032,
      "step": 1150
    },
    {
      "epoch": 0.6053459119496856,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.009072327044025158,
      "loss": 0.485,
      "num_input_tokens_seen": 756736,
      "step": 1155
    },
    {
      "epoch": 0.6079664570230608,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.009111635220125786,
      "loss": 0.4642,
      "num_input_tokens_seen": 759680,
      "step": 1160
    },
    {
      "epoch": 0.610587002096436,
      "grad_norm": 0.162109375,
      "learning_rate": 0.009150943396226414,
      "loss": 0.5145,
      "num_input_tokens_seen": 762560,
      "step": 1165
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.009190251572327043,
      "loss": 0.4225,
      "num_input_tokens_seen": 765600,
      "step": 1170
    },
    {
      "epoch": 0.6158280922431866,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.009229559748427672,
      "loss": 0.3942,
      "num_input_tokens_seen": 768736,
      "step": 1175
    },
    {
      "epoch": 0.6184486373165619,
      "grad_norm": 0.1015625,
      "learning_rate": 0.009268867924528301,
      "loss": 0.6584,
      "num_input_tokens_seen": 771872,
      "step": 1180
    },
    {
      "epoch": 0.6210691823899371,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.009308176100628932,
      "loss": 0.602,
      "num_input_tokens_seen": 775008,
      "step": 1185
    },
    {
      "epoch": 0.6236897274633124,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.009347484276729559,
      "loss": 0.4026,
      "num_input_tokens_seen": 777920,
      "step": 1190
    },
    {
      "epoch": 0.6263102725366876,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.009386792452830188,
      "loss": 0.5107,
      "num_input_tokens_seen": 781056,
      "step": 1195
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.009426100628930817,
      "loss": 0.5497,
      "num_input_tokens_seen": 784128,
      "step": 1200
    },
    {
      "epoch": 0.6315513626834381,
      "grad_norm": 0.078125,
      "learning_rate": 0.009465408805031446,
      "loss": 0.5161,
      "num_input_tokens_seen": 786624,
      "step": 1205
    },
    {
      "epoch": 0.6341719077568134,
      "grad_norm": 0.099609375,
      "learning_rate": 0.009504716981132075,
      "loss": 0.5164,
      "num_input_tokens_seen": 789664,
      "step": 1210
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.009544025157232705,
      "loss": 0.4304,
      "num_input_tokens_seen": 793824,
      "step": 1215
    },
    {
      "epoch": 0.639412997903564,
      "grad_norm": 0.1484375,
      "learning_rate": 0.009583333333333333,
      "loss": 0.6198,
      "num_input_tokens_seen": 796544,
      "step": 1220
    },
    {
      "epoch": 0.6420335429769392,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.009622641509433962,
      "loss": 0.3987,
      "num_input_tokens_seen": 799520,
      "step": 1225
    },
    {
      "epoch": 0.6446540880503144,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.00966194968553459,
      "loss": 0.3319,
      "num_input_tokens_seen": 802464,
      "step": 1230
    },
    {
      "epoch": 0.6472746331236897,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.00970125786163522,
      "loss": 0.5831,
      "num_input_tokens_seen": 805536,
      "step": 1235
    },
    {
      "epoch": 0.649895178197065,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.009740566037735848,
      "loss": 0.358,
      "num_input_tokens_seen": 808448,
      "step": 1240
    },
    {
      "epoch": 0.6525157232704403,
      "grad_norm": 0.08203125,
      "learning_rate": 0.009779874213836479,
      "loss": 0.4355,
      "num_input_tokens_seen": 812160,
      "step": 1245
    },
    {
      "epoch": 0.6551362683438156,
      "grad_norm": 0.173828125,
      "learning_rate": 0.009819182389937106,
      "loss": 0.6224,
      "num_input_tokens_seen": 814752,
      "step": 1250
    },
    {
      "epoch": 0.6577568134171907,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.009858490566037735,
      "loss": 0.4116,
      "num_input_tokens_seen": 817568,
      "step": 1255
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.009897798742138364,
      "loss": 0.5022,
      "num_input_tokens_seen": 820384,
      "step": 1260
    },
    {
      "epoch": 0.6629979035639413,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.009937106918238993,
      "loss": 0.3212,
      "num_input_tokens_seen": 823040,
      "step": 1265
    },
    {
      "epoch": 0.6656184486373166,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.009976415094339622,
      "loss": 0.5961,
      "num_input_tokens_seen": 825600,
      "step": 1270
    },
    {
      "epoch": 0.6682389937106918,
      "grad_norm": 0.09765625,
      "learning_rate": 0.010015723270440252,
      "loss": 0.5848,
      "num_input_tokens_seen": 828032,
      "step": 1275
    },
    {
      "epoch": 0.6708595387840671,
      "grad_norm": 0.11328125,
      "learning_rate": 0.010055031446540881,
      "loss": 0.4395,
      "num_input_tokens_seen": 831296,
      "step": 1280
    },
    {
      "epoch": 0.6734800838574424,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010094339622641509,
      "loss": 0.6058,
      "num_input_tokens_seen": 834208,
      "step": 1285
    },
    {
      "epoch": 0.6761006289308176,
      "grad_norm": 0.09765625,
      "learning_rate": 0.010133647798742138,
      "loss": 0.4562,
      "num_input_tokens_seen": 838048,
      "step": 1290
    },
    {
      "epoch": 0.6787211740041929,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.010172955974842766,
      "loss": 0.5725,
      "num_input_tokens_seen": 841536,
      "step": 1295
    },
    {
      "epoch": 0.6813417190775681,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.010212264150943395,
      "loss": 0.4319,
      "num_input_tokens_seen": 845184,
      "step": 1300
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 0.052734375,
      "learning_rate": 0.010251572327044026,
      "loss": 0.4297,
      "num_input_tokens_seen": 848000,
      "step": 1305
    },
    {
      "epoch": 0.6865828092243187,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.010290880503144655,
      "loss": 0.5753,
      "num_input_tokens_seen": 851040,
      "step": 1310
    },
    {
      "epoch": 0.689203354297694,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.010330188679245282,
      "loss": 0.5846,
      "num_input_tokens_seen": 853920,
      "step": 1315
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.010369496855345911,
      "loss": 0.5106,
      "num_input_tokens_seen": 857312,
      "step": 1320
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01040880503144654,
      "loss": 0.4088,
      "num_input_tokens_seen": 859904,
      "step": 1325
    },
    {
      "epoch": 0.6970649895178197,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.010448113207547169,
      "loss": 0.5478,
      "num_input_tokens_seen": 863680,
      "step": 1330
    },
    {
      "epoch": 0.699685534591195,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0104874213836478,
      "loss": 0.5661,
      "num_input_tokens_seen": 866880,
      "step": 1335
    },
    {
      "epoch": 0.7023060796645703,
      "grad_norm": 0.130859375,
      "learning_rate": 0.010526729559748428,
      "loss": 0.4115,
      "num_input_tokens_seen": 869856,
      "step": 1340
    },
    {
      "epoch": 0.7049266247379455,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.010566037735849056,
      "loss": 0.5106,
      "num_input_tokens_seen": 873472,
      "step": 1345
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.010605345911949685,
      "loss": 0.4421,
      "num_input_tokens_seen": 876704,
      "step": 1350
    },
    {
      "epoch": 0.710167714884696,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.010644654088050313,
      "loss": 0.4662,
      "num_input_tokens_seen": 880576,
      "step": 1355
    },
    {
      "epoch": 0.7127882599580713,
      "grad_norm": 0.177734375,
      "learning_rate": 0.010683962264150942,
      "loss": 0.4679,
      "num_input_tokens_seen": 883680,
      "step": 1360
    },
    {
      "epoch": 0.7154088050314465,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.010723270440251573,
      "loss": 0.5471,
      "num_input_tokens_seen": 886464,
      "step": 1365
    },
    {
      "epoch": 0.7180293501048218,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.010762578616352202,
      "loss": 0.52,
      "num_input_tokens_seen": 889536,
      "step": 1370
    },
    {
      "epoch": 0.7206498951781971,
      "grad_norm": 0.078125,
      "learning_rate": 0.01080188679245283,
      "loss": 0.496,
      "num_input_tokens_seen": 894272,
      "step": 1375
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.010841194968553458,
      "loss": 0.4534,
      "num_input_tokens_seen": 897120,
      "step": 1380
    },
    {
      "epoch": 0.7258909853249476,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.010880503144654087,
      "loss": 0.5005,
      "num_input_tokens_seen": 899808,
      "step": 1385
    },
    {
      "epoch": 0.7285115303983228,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.010919811320754716,
      "loss": 0.4824,
      "num_input_tokens_seen": 903520,
      "step": 1390
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.010959119496855347,
      "loss": 0.5387,
      "num_input_tokens_seen": 907200,
      "step": 1395
    },
    {
      "epoch": 0.7337526205450734,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010998427672955976,
      "loss": 0.4733,
      "num_input_tokens_seen": 911392,
      "step": 1400
    },
    {
      "epoch": 0.7363731656184487,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.011037735849056603,
      "loss": 0.4995,
      "num_input_tokens_seen": 913440,
      "step": 1405
    },
    {
      "epoch": 0.7389937106918238,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.011077044025157232,
      "loss": 0.4774,
      "num_input_tokens_seen": 917888,
      "step": 1410
    },
    {
      "epoch": 0.7416142557651991,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.01111635220125786,
      "loss": 0.5317,
      "num_input_tokens_seen": 921376,
      "step": 1415
    },
    {
      "epoch": 0.7442348008385744,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.01115566037735849,
      "loss": 0.5585,
      "num_input_tokens_seen": 924800,
      "step": 1420
    },
    {
      "epoch": 0.7468553459119497,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.01119496855345912,
      "loss": 0.6391,
      "num_input_tokens_seen": 927648,
      "step": 1425
    },
    {
      "epoch": 0.749475890985325,
      "grad_norm": 0.08203125,
      "learning_rate": 0.011234276729559749,
      "loss": 0.3898,
      "num_input_tokens_seen": 930144,
      "step": 1430
    },
    {
      "epoch": 0.7520964360587002,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.011273584905660378,
      "loss": 0.4167,
      "num_input_tokens_seen": 933184,
      "step": 1435
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.011312893081761005,
      "loss": 0.5347,
      "num_input_tokens_seen": 937376,
      "step": 1440
    },
    {
      "epoch": 0.7573375262054507,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.011352201257861634,
      "loss": 0.455,
      "num_input_tokens_seen": 940160,
      "step": 1445
    },
    {
      "epoch": 0.759958071278826,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.011391509433962263,
      "loss": 0.5272,
      "num_input_tokens_seen": 942816,
      "step": 1450
    },
    {
      "epoch": 0.7625786163522013,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.011430817610062894,
      "loss": 0.5643,
      "num_input_tokens_seen": 945952,
      "step": 1455
    },
    {
      "epoch": 0.7651991614255765,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.011470125786163523,
      "loss": 0.3971,
      "num_input_tokens_seen": 949472,
      "step": 1460
    },
    {
      "epoch": 0.7678197064989518,
      "grad_norm": 0.138671875,
      "learning_rate": 0.011509433962264152,
      "loss": 0.5313,
      "num_input_tokens_seen": 952672,
      "step": 1465
    },
    {
      "epoch": 0.7704402515723271,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.011548742138364779,
      "loss": 0.5401,
      "num_input_tokens_seen": 955520,
      "step": 1470
    },
    {
      "epoch": 0.7730607966457023,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.011588050314465408,
      "loss": 0.558,
      "num_input_tokens_seen": 958464,
      "step": 1475
    },
    {
      "epoch": 0.7756813417190775,
      "grad_norm": 0.13671875,
      "learning_rate": 0.011627358490566037,
      "loss": 0.4907,
      "num_input_tokens_seen": 960768,
      "step": 1480
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.011666666666666667,
      "loss": 0.6935,
      "num_input_tokens_seen": 963616,
      "step": 1485
    },
    {
      "epoch": 0.7809224318658281,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.011705974842767296,
      "loss": 0.6883,
      "num_input_tokens_seen": 967936,
      "step": 1490
    },
    {
      "epoch": 0.7835429769392034,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.011745283018867925,
      "loss": 0.4696,
      "num_input_tokens_seen": 971744,
      "step": 1495
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.011784591194968552,
      "loss": 0.5127,
      "num_input_tokens_seen": 974080,
      "step": 1500
    },
    {
      "epoch": 0.7887840670859538,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.011823899371069181,
      "loss": 0.4763,
      "num_input_tokens_seen": 976992,
      "step": 1505
    },
    {
      "epoch": 0.7914046121593291,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.01186320754716981,
      "loss": 0.5018,
      "num_input_tokens_seen": 979648,
      "step": 1510
    },
    {
      "epoch": 0.7940251572327044,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.01190251572327044,
      "loss": 0.5514,
      "num_input_tokens_seen": 983136,
      "step": 1515
    },
    {
      "epoch": 0.7966457023060797,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01194182389937107,
      "loss": 0.4851,
      "num_input_tokens_seen": 986976,
      "step": 1520
    },
    {
      "epoch": 0.799266247379455,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.011981132075471699,
      "loss": 0.4619,
      "num_input_tokens_seen": 990272,
      "step": 1525
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.012020440251572326,
      "loss": 0.4672,
      "num_input_tokens_seen": 994368,
      "step": 1530
    },
    {
      "epoch": 0.8045073375262054,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.012059748427672955,
      "loss": 0.435,
      "num_input_tokens_seen": 997504,
      "step": 1535
    },
    {
      "epoch": 0.8071278825995807,
      "grad_norm": 0.091796875,
      "learning_rate": 0.012099056603773584,
      "loss": 0.5701,
      "num_input_tokens_seen": 1000352,
      "step": 1540
    },
    {
      "epoch": 0.809748427672956,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.012138364779874214,
      "loss": 0.5745,
      "num_input_tokens_seen": 1003616,
      "step": 1545
    },
    {
      "epoch": 0.8123689727463312,
      "grad_norm": 0.20703125,
      "learning_rate": 0.012177672955974843,
      "loss": 0.5539,
      "num_input_tokens_seen": 1006240,
      "step": 1550
    },
    {
      "epoch": 0.8149895178197065,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.012216981132075472,
      "loss": 0.5437,
      "num_input_tokens_seen": 1009440,
      "step": 1555
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0122562893081761,
      "loss": 0.5416,
      "num_input_tokens_seen": 1011936,
      "step": 1560
    },
    {
      "epoch": 0.820230607966457,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.012295597484276728,
      "loss": 0.4451,
      "num_input_tokens_seen": 1015040,
      "step": 1565
    },
    {
      "epoch": 0.8228511530398323,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.012334905660377357,
      "loss": 0.6014,
      "num_input_tokens_seen": 1018240,
      "step": 1570
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.012374213836477988,
      "loss": 0.4991,
      "num_input_tokens_seen": 1022336,
      "step": 1575
    },
    {
      "epoch": 0.8280922431865828,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.012413522012578617,
      "loss": 0.509,
      "num_input_tokens_seen": 1025184,
      "step": 1580
    },
    {
      "epoch": 0.8307127882599581,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.012452830188679246,
      "loss": 0.5656,
      "num_input_tokens_seen": 1028448,
      "step": 1585
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.068359375,
      "learning_rate": 0.012492138364779873,
      "loss": 0.5379,
      "num_input_tokens_seen": 1031072,
      "step": 1590
    },
    {
      "epoch": 0.8359538784067087,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.012531446540880502,
      "loss": 0.4713,
      "num_input_tokens_seen": 1034208,
      "step": 1595
    },
    {
      "epoch": 0.8385744234800838,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.01257075471698113,
      "loss": 0.4328,
      "num_input_tokens_seen": 1038624,
      "step": 1600
    },
    {
      "epoch": 0.8411949685534591,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.012610062893081761,
      "loss": 0.5458,
      "num_input_tokens_seen": 1041152,
      "step": 1605
    },
    {
      "epoch": 0.8438155136268344,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.01264937106918239,
      "loss": 0.3988,
      "num_input_tokens_seen": 1043904,
      "step": 1610
    },
    {
      "epoch": 0.8464360587002097,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01268867924528302,
      "loss": 0.5117,
      "num_input_tokens_seen": 1047456,
      "step": 1615
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.012727987421383648,
      "loss": 0.5619,
      "num_input_tokens_seen": 1050304,
      "step": 1620
    },
    {
      "epoch": 0.8516771488469602,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.012767295597484275,
      "loss": 0.523,
      "num_input_tokens_seen": 1053600,
      "step": 1625
    },
    {
      "epoch": 0.8542976939203354,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.012806603773584904,
      "loss": 0.3434,
      "num_input_tokens_seen": 1056672,
      "step": 1630
    },
    {
      "epoch": 0.8569182389937107,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.012845911949685535,
      "loss": 0.3873,
      "num_input_tokens_seen": 1059904,
      "step": 1635
    },
    {
      "epoch": 0.859538784067086,
      "grad_norm": 0.05078125,
      "learning_rate": 0.012885220125786164,
      "loss": 0.4933,
      "num_input_tokens_seen": 1063360,
      "step": 1640
    },
    {
      "epoch": 0.8621593291404612,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.012924528301886793,
      "loss": 0.4898,
      "num_input_tokens_seen": 1065728,
      "step": 1645
    },
    {
      "epoch": 0.8647798742138365,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.012963836477987422,
      "loss": 0.4752,
      "num_input_tokens_seen": 1068448,
      "step": 1650
    },
    {
      "epoch": 0.8674004192872118,
      "grad_norm": 0.0625,
      "learning_rate": 0.013003144654088049,
      "loss": 0.506,
      "num_input_tokens_seen": 1071104,
      "step": 1655
    },
    {
      "epoch": 0.870020964360587,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.013042452830188678,
      "loss": 0.4131,
      "num_input_tokens_seen": 1075072,
      "step": 1660
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.013081761006289308,
      "loss": 0.5212,
      "num_input_tokens_seen": 1077824,
      "step": 1665
    },
    {
      "epoch": 0.8752620545073375,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.013121069182389937,
      "loss": 0.4705,
      "num_input_tokens_seen": 1080832,
      "step": 1670
    },
    {
      "epoch": 0.8778825995807128,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.013160377358490566,
      "loss": 0.5259,
      "num_input_tokens_seen": 1084096,
      "step": 1675
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.013199685534591195,
      "loss": 0.5005,
      "num_input_tokens_seen": 1087104,
      "step": 1680
    },
    {
      "epoch": 0.8831236897274634,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.013238993710691822,
      "loss": 0.5732,
      "num_input_tokens_seen": 1090304,
      "step": 1685
    },
    {
      "epoch": 0.8857442348008385,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.013278301886792451,
      "loss": 0.7743,
      "num_input_tokens_seen": 1094112,
      "step": 1690
    },
    {
      "epoch": 0.8883647798742138,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.013317610062893082,
      "loss": 0.5296,
      "num_input_tokens_seen": 1098496,
      "step": 1695
    },
    {
      "epoch": 0.8909853249475891,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.01335691823899371,
      "loss": 0.434,
      "num_input_tokens_seen": 1100992,
      "step": 1700
    },
    {
      "epoch": 0.8936058700209644,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.01339622641509434,
      "loss": 0.5345,
      "num_input_tokens_seen": 1103776,
      "step": 1705
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.013435534591194969,
      "loss": 0.4531,
      "num_input_tokens_seen": 1106368,
      "step": 1710
    },
    {
      "epoch": 0.8988469601677149,
      "grad_norm": 0.076171875,
      "learning_rate": 0.013474842767295596,
      "loss": 0.4342,
      "num_input_tokens_seen": 1109376,
      "step": 1715
    },
    {
      "epoch": 0.9014675052410901,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.013514150943396225,
      "loss": 0.4681,
      "num_input_tokens_seen": 1112448,
      "step": 1720
    },
    {
      "epoch": 0.9040880503144654,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.013553459119496855,
      "loss": 0.4732,
      "num_input_tokens_seen": 1115520,
      "step": 1725
    },
    {
      "epoch": 0.9067085953878407,
      "grad_norm": 0.087890625,
      "learning_rate": 0.013592767295597484,
      "loss": 0.6396,
      "num_input_tokens_seen": 1118784,
      "step": 1730
    },
    {
      "epoch": 0.9093291404612159,
      "grad_norm": 0.166015625,
      "learning_rate": 0.013632075471698113,
      "loss": 0.7068,
      "num_input_tokens_seen": 1122112,
      "step": 1735
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.013671383647798742,
      "loss": 0.4633,
      "num_input_tokens_seen": 1125952,
      "step": 1740
    },
    {
      "epoch": 0.9145702306079665,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.01371069182389937,
      "loss": 0.6304,
      "num_input_tokens_seen": 1129024,
      "step": 1745
    },
    {
      "epoch": 0.9171907756813418,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.013749999999999998,
      "loss": 0.4501,
      "num_input_tokens_seen": 1131488,
      "step": 1750
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 0.0703125,
      "learning_rate": 0.013789308176100629,
      "loss": 0.4538,
      "num_input_tokens_seen": 1134176,
      "step": 1755
    },
    {
      "epoch": 0.9224318658280922,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.013828616352201258,
      "loss": 0.5062,
      "num_input_tokens_seen": 1137888,
      "step": 1760
    },
    {
      "epoch": 0.9250524109014675,
      "grad_norm": 0.16796875,
      "learning_rate": 0.013867924528301887,
      "loss": 0.5929,
      "num_input_tokens_seen": 1141184,
      "step": 1765
    },
    {
      "epoch": 0.9276729559748428,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.013907232704402516,
      "loss": 0.5522,
      "num_input_tokens_seen": 1144064,
      "step": 1770
    },
    {
      "epoch": 0.9302935010482181,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.013946540880503145,
      "loss": 0.4806,
      "num_input_tokens_seen": 1147584,
      "step": 1775
    },
    {
      "epoch": 0.9329140461215933,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.013985849056603772,
      "loss": 0.4733,
      "num_input_tokens_seen": 1154784,
      "step": 1780
    },
    {
      "epoch": 0.9355345911949685,
      "grad_norm": 0.126953125,
      "learning_rate": 0.014025157232704403,
      "loss": 0.4707,
      "num_input_tokens_seen": 1157440,
      "step": 1785
    },
    {
      "epoch": 0.9381551362683438,
      "grad_norm": 0.04296875,
      "learning_rate": 0.014064465408805031,
      "loss": 0.6205,
      "num_input_tokens_seen": 1160224,
      "step": 1790
    },
    {
      "epoch": 0.9407756813417191,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.01410377358490566,
      "loss": 0.5156,
      "num_input_tokens_seen": 1164480,
      "step": 1795
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.01414308176100629,
      "loss": 0.4002,
      "num_input_tokens_seen": 1168704,
      "step": 1800
    },
    {
      "epoch": 0.9460167714884696,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.014182389937106918,
      "loss": 0.4235,
      "num_input_tokens_seen": 1172384,
      "step": 1805
    },
    {
      "epoch": 0.9486373165618449,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.014221698113207545,
      "loss": 0.4774,
      "num_input_tokens_seen": 1175392,
      "step": 1810
    },
    {
      "epoch": 0.9512578616352201,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.014261006289308174,
      "loss": 0.4598,
      "num_input_tokens_seen": 1178368,
      "step": 1815
    },
    {
      "epoch": 0.9538784067085954,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.014300314465408805,
      "loss": 0.3852,
      "num_input_tokens_seen": 1182240,
      "step": 1820
    },
    {
      "epoch": 0.9564989517819706,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014339622641509434,
      "loss": 0.4595,
      "num_input_tokens_seen": 1185504,
      "step": 1825
    },
    {
      "epoch": 0.9591194968553459,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.014378930817610063,
      "loss": 0.4594,
      "num_input_tokens_seen": 1188288,
      "step": 1830
    },
    {
      "epoch": 0.9617400419287212,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.014418238993710692,
      "loss": 0.4507,
      "num_input_tokens_seen": 1192064,
      "step": 1835
    },
    {
      "epoch": 0.9643605870020965,
      "grad_norm": 0.044921875,
      "learning_rate": 0.014457547169811319,
      "loss": 0.4579,
      "num_input_tokens_seen": 1195104,
      "step": 1840
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.014496855345911948,
      "loss": 0.5971,
      "num_input_tokens_seen": 1197920,
      "step": 1845
    },
    {
      "epoch": 0.9696016771488469,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.014536163522012579,
      "loss": 0.4384,
      "num_input_tokens_seen": 1200896,
      "step": 1850
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.014575471698113207,
      "loss": 0.4753,
      "num_input_tokens_seen": 1204032,
      "step": 1855
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.014614779874213836,
      "loss": 0.4062,
      "num_input_tokens_seen": 1207264,
      "step": 1860
    },
    {
      "epoch": 0.9774633123689728,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.014654088050314465,
      "loss": 0.5518,
      "num_input_tokens_seen": 1209984,
      "step": 1865
    },
    {
      "epoch": 0.980083857442348,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.014693396226415092,
      "loss": 0.421,
      "num_input_tokens_seen": 1213856,
      "step": 1870
    },
    {
      "epoch": 0.9827044025157232,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.014732704402515721,
      "loss": 0.4342,
      "num_input_tokens_seen": 1217408,
      "step": 1875
    },
    {
      "epoch": 0.9853249475890985,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.014772012578616352,
      "loss": 0.5046,
      "num_input_tokens_seen": 1221024,
      "step": 1880
    },
    {
      "epoch": 0.9879454926624738,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.014811320754716981,
      "loss": 0.4007,
      "num_input_tokens_seen": 1224320,
      "step": 1885
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.01485062893081761,
      "loss": 0.3838,
      "num_input_tokens_seen": 1228000,
      "step": 1890
    },
    {
      "epoch": 0.9931865828092243,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.014889937106918239,
      "loss": 0.4887,
      "num_input_tokens_seen": 1232320,
      "step": 1895
    },
    {
      "epoch": 0.9958071278825996,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.014929245283018866,
      "loss": 0.593,
      "num_input_tokens_seen": 1238272,
      "step": 1900
    },
    {
      "epoch": 0.9984276729559748,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.014968553459119495,
      "loss": 0.4965,
      "num_input_tokens_seen": 1241344,
      "step": 1905
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.46922871470451355,
      "eval_runtime": 13.6028,
      "eval_samples_per_second": 62.34,
      "eval_steps_per_second": 15.585,
      "num_input_tokens_seen": 1243048,
      "step": 1908
    },
    {
      "epoch": 1.00104821802935,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.015007861635220126,
      "loss": 0.4196,
      "num_input_tokens_seen": 1244520,
      "step": 1910
    },
    {
      "epoch": 1.0036687631027255,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.015047169811320756,
      "loss": 0.4914,
      "num_input_tokens_seen": 1247528,
      "step": 1915
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.015086477987421383,
      "loss": 0.5832,
      "num_input_tokens_seen": 1250824,
      "step": 1920
    },
    {
      "epoch": 1.0089098532494758,
      "grad_norm": 0.052734375,
      "learning_rate": 0.01512578616352201,
      "loss": 0.5006,
      "num_input_tokens_seen": 1254024,
      "step": 1925
    },
    {
      "epoch": 1.0115303983228512,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.015165094339622641,
      "loss": 0.446,
      "num_input_tokens_seen": 1260328,
      "step": 1930
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.015204402515723268,
      "loss": 0.679,
      "num_input_tokens_seen": 1263656,
      "step": 1935
    },
    {
      "epoch": 1.0167714884696017,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0152437106918239,
      "loss": 0.5201,
      "num_input_tokens_seen": 1267528,
      "step": 1940
    },
    {
      "epoch": 1.019392033542977,
      "grad_norm": 0.099609375,
      "learning_rate": 0.015283018867924528,
      "loss": 0.5707,
      "num_input_tokens_seen": 1270536,
      "step": 1945
    },
    {
      "epoch": 1.0220125786163523,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.015322327044025155,
      "loss": 0.3842,
      "num_input_tokens_seen": 1273416,
      "step": 1950
    },
    {
      "epoch": 1.0246331236897275,
      "grad_norm": 0.11328125,
      "learning_rate": 0.015361635220125786,
      "loss": 0.3703,
      "num_input_tokens_seen": 1276232,
      "step": 1955
    },
    {
      "epoch": 1.0272536687631026,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.015400943396226413,
      "loss": 0.3792,
      "num_input_tokens_seen": 1279400,
      "step": 1960
    },
    {
      "epoch": 1.029874213836478,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.015440251572327044,
      "loss": 0.4411,
      "num_input_tokens_seen": 1283752,
      "step": 1965
    },
    {
      "epoch": 1.0324947589098532,
      "grad_norm": 0.06640625,
      "learning_rate": 0.015479559748427673,
      "loss": 0.4256,
      "num_input_tokens_seen": 1287048,
      "step": 1970
    },
    {
      "epoch": 1.0351153039832286,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.015518867924528303,
      "loss": 0.4991,
      "num_input_tokens_seen": 1289544,
      "step": 1975
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.01555817610062893,
      "loss": 0.4775,
      "num_input_tokens_seen": 1292392,
      "step": 1980
    },
    {
      "epoch": 1.040356394129979,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.015597484276729558,
      "loss": 0.5677,
      "num_input_tokens_seen": 1295400,
      "step": 1985
    },
    {
      "epoch": 1.0429769392033543,
      "grad_norm": 0.048828125,
      "learning_rate": 0.01563679245283019,
      "loss": 0.491,
      "num_input_tokens_seen": 1298152,
      "step": 1990
    },
    {
      "epoch": 1.0455974842767295,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.015676100628930816,
      "loss": 0.5795,
      "num_input_tokens_seen": 1301384,
      "step": 1995
    },
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.015715408805031448,
      "loss": 0.5282,
      "num_input_tokens_seen": 1304296,
      "step": 2000
    },
    {
      "epoch": 1.05083857442348,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.015754716981132073,
      "loss": 0.5834,
      "num_input_tokens_seen": 1307048,
      "step": 2005
    },
    {
      "epoch": 1.0534591194968554,
      "grad_norm": 0.044921875,
      "learning_rate": 0.015794025157232702,
      "loss": 0.4502,
      "num_input_tokens_seen": 1310792,
      "step": 2010
    },
    {
      "epoch": 1.0560796645702306,
      "grad_norm": 0.06640625,
      "learning_rate": 0.015833333333333335,
      "loss": 0.6049,
      "num_input_tokens_seen": 1313448,
      "step": 2015
    },
    {
      "epoch": 1.0587002096436058,
      "grad_norm": 0.087890625,
      "learning_rate": 0.01587264150943396,
      "loss": 0.3941,
      "num_input_tokens_seen": 1316200,
      "step": 2020
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.015911949685534593,
      "loss": 0.7858,
      "num_input_tokens_seen": 1319720,
      "step": 2025
    },
    {
      "epoch": 1.0639412997903563,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.015951257861635218,
      "loss": 0.5189,
      "num_input_tokens_seen": 1323432,
      "step": 2030
    },
    {
      "epoch": 1.0665618448637317,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01599056603773585,
      "loss": 0.4759,
      "num_input_tokens_seen": 1326760,
      "step": 2035
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.016029874213836476,
      "loss": 0.4531,
      "num_input_tokens_seen": 1330120,
      "step": 2040
    },
    {
      "epoch": 1.0718029350104823,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.016069182389937105,
      "loss": 0.5232,
      "num_input_tokens_seen": 1333032,
      "step": 2045
    },
    {
      "epoch": 1.0744234800838575,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.016108490566037737,
      "loss": 0.5808,
      "num_input_tokens_seen": 1336296,
      "step": 2050
    },
    {
      "epoch": 1.0770440251572326,
      "grad_norm": 0.03515625,
      "learning_rate": 0.016147798742138363,
      "loss": 0.365,
      "num_input_tokens_seen": 1339592,
      "step": 2055
    },
    {
      "epoch": 1.079664570230608,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.016187106918238995,
      "loss": 0.4898,
      "num_input_tokens_seen": 1342664,
      "step": 2060
    },
    {
      "epoch": 1.0822851153039832,
      "grad_norm": 0.046875,
      "learning_rate": 0.01622641509433962,
      "loss": 0.5935,
      "num_input_tokens_seen": 1345288,
      "step": 2065
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.016265723270440253,
      "loss": 0.4585,
      "num_input_tokens_seen": 1348488,
      "step": 2070
    },
    {
      "epoch": 1.0875262054507338,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.016305031446540882,
      "loss": 0.5971,
      "num_input_tokens_seen": 1352712,
      "step": 2075
    },
    {
      "epoch": 1.090146750524109,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.016344339622641507,
      "loss": 0.4416,
      "num_input_tokens_seen": 1355496,
      "step": 2080
    },
    {
      "epoch": 1.0927672955974843,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01638364779874214,
      "loss": 0.5155,
      "num_input_tokens_seen": 1360552,
      "step": 2085
    },
    {
      "epoch": 1.0953878406708595,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.016422955974842765,
      "loss": 0.4171,
      "num_input_tokens_seen": 1363432,
      "step": 2090
    },
    {
      "epoch": 1.0980083857442349,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.016462264150943397,
      "loss": 0.3745,
      "num_input_tokens_seen": 1366824,
      "step": 2095
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 0.08984375,
      "learning_rate": 0.016501572327044023,
      "loss": 0.5526,
      "num_input_tokens_seen": 1370312,
      "step": 2100
    },
    {
      "epoch": 1.1032494758909852,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.016540880503144652,
      "loss": 0.704,
      "num_input_tokens_seen": 1373160,
      "step": 2105
    },
    {
      "epoch": 1.1058700209643606,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.016580188679245284,
      "loss": 0.5425,
      "num_input_tokens_seen": 1376008,
      "step": 2110
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.01661949685534591,
      "loss": 0.5397,
      "num_input_tokens_seen": 1378440,
      "step": 2115
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.016658805031446542,
      "loss": 0.527,
      "num_input_tokens_seen": 1382152,
      "step": 2120
    },
    {
      "epoch": 1.1137316561844863,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.016698113207547167,
      "loss": 0.6447,
      "num_input_tokens_seen": 1385416,
      "step": 2125
    },
    {
      "epoch": 1.1163522012578617,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0167374213836478,
      "loss": 0.4456,
      "num_input_tokens_seen": 1389288,
      "step": 2130
    },
    {
      "epoch": 1.118972746331237,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.01677672955974843,
      "loss": 0.5077,
      "num_input_tokens_seen": 1392136,
      "step": 2135
    },
    {
      "epoch": 1.121593291404612,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.016816037735849054,
      "loss": 0.5042,
      "num_input_tokens_seen": 1395208,
      "step": 2140
    },
    {
      "epoch": 1.1242138364779874,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.016855345911949687,
      "loss": 0.3947,
      "num_input_tokens_seen": 1398024,
      "step": 2145
    },
    {
      "epoch": 1.1268343815513626,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.016894654088050312,
      "loss": 0.4354,
      "num_input_tokens_seen": 1401256,
      "step": 2150
    },
    {
      "epoch": 1.129454926624738,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.016933962264150944,
      "loss": 0.3545,
      "num_input_tokens_seen": 1403848,
      "step": 2155
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.01697327044025157,
      "loss": 0.5786,
      "num_input_tokens_seen": 1408040,
      "step": 2160
    },
    {
      "epoch": 1.1346960167714886,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0170125786163522,
      "loss": 0.5891,
      "num_input_tokens_seen": 1411688,
      "step": 2165
    },
    {
      "epoch": 1.1373165618448637,
      "grad_norm": 0.0703125,
      "learning_rate": 0.01705188679245283,
      "loss": 0.581,
      "num_input_tokens_seen": 1415464,
      "step": 2170
    },
    {
      "epoch": 1.139937106918239,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.017091194968553457,
      "loss": 0.3927,
      "num_input_tokens_seen": 1418664,
      "step": 2175
    },
    {
      "epoch": 1.1425576519916143,
      "grad_norm": 0.0703125,
      "learning_rate": 0.01713050314465409,
      "loss": 0.4216,
      "num_input_tokens_seen": 1422088,
      "step": 2180
    },
    {
      "epoch": 1.1451781970649895,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.017169811320754715,
      "loss": 0.4233,
      "num_input_tokens_seen": 1424872,
      "step": 2185
    },
    {
      "epoch": 1.1477987421383649,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.017209119496855347,
      "loss": 0.6044,
      "num_input_tokens_seen": 1429320,
      "step": 2190
    },
    {
      "epoch": 1.15041928721174,
      "grad_norm": 0.046875,
      "learning_rate": 0.017248427672955976,
      "loss": 0.4644,
      "num_input_tokens_seen": 1433704,
      "step": 2195
    },
    {
      "epoch": 1.1530398322851152,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0172877358490566,
      "loss": 0.5018,
      "num_input_tokens_seen": 1436872,
      "step": 2200
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 0.080078125,
      "learning_rate": 0.017327044025157234,
      "loss": 0.488,
      "num_input_tokens_seen": 1440104,
      "step": 2205
    },
    {
      "epoch": 1.1582809224318658,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.01736635220125786,
      "loss": 0.6385,
      "num_input_tokens_seen": 1443176,
      "step": 2210
    },
    {
      "epoch": 1.1609014675052411,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01740566037735849,
      "loss": 0.4521,
      "num_input_tokens_seen": 1446440,
      "step": 2215
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.017444968553459117,
      "loss": 0.4628,
      "num_input_tokens_seen": 1450120,
      "step": 2220
    },
    {
      "epoch": 1.1661425576519917,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.017484276729559746,
      "loss": 0.4539,
      "num_input_tokens_seen": 1453608,
      "step": 2225
    },
    {
      "epoch": 1.1687631027253669,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.01752358490566038,
      "loss": 0.6107,
      "num_input_tokens_seen": 1457832,
      "step": 2230
    },
    {
      "epoch": 1.171383647798742,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.017562893081761004,
      "loss": 0.5327,
      "num_input_tokens_seen": 1462280,
      "step": 2235
    },
    {
      "epoch": 1.1740041928721174,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.017602201257861636,
      "loss": 0.4768,
      "num_input_tokens_seen": 1466024,
      "step": 2240
    },
    {
      "epoch": 1.1766247379454926,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01764150943396226,
      "loss": 0.4564,
      "num_input_tokens_seen": 1468872,
      "step": 2245
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.017680817610062894,
      "loss": 0.48,
      "num_input_tokens_seen": 1471336,
      "step": 2250
    },
    {
      "epoch": 1.1818658280922432,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.017720125786163523,
      "loss": 0.6082,
      "num_input_tokens_seen": 1473928,
      "step": 2255
    },
    {
      "epoch": 1.1844863731656186,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.01775943396226415,
      "loss": 0.4225,
      "num_input_tokens_seen": 1478728,
      "step": 2260
    },
    {
      "epoch": 1.1871069182389937,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.01779874213836478,
      "loss": 0.5064,
      "num_input_tokens_seen": 1481576,
      "step": 2265
    },
    {
      "epoch": 1.189727463312369,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.017838050314465406,
      "loss": 0.5285,
      "num_input_tokens_seen": 1484104,
      "step": 2270
    },
    {
      "epoch": 1.1923480083857443,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.01787735849056604,
      "loss": 0.5633,
      "num_input_tokens_seen": 1487048,
      "step": 2275
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.017916666666666664,
      "loss": 0.4564,
      "num_input_tokens_seen": 1490152,
      "step": 2280
    },
    {
      "epoch": 1.1975890985324948,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.017955974842767296,
      "loss": 0.4105,
      "num_input_tokens_seen": 1492648,
      "step": 2285
    },
    {
      "epoch": 1.20020964360587,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.017995283018867925,
      "loss": 0.5767,
      "num_input_tokens_seen": 1495272,
      "step": 2290
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 0.05859375,
      "learning_rate": 0.01803459119496855,
      "loss": 0.4979,
      "num_input_tokens_seen": 1497832,
      "step": 2295
    },
    {
      "epoch": 1.2054507337526206,
      "grad_norm": 0.064453125,
      "learning_rate": 0.018073899371069183,
      "loss": 0.3826,
      "num_input_tokens_seen": 1501896,
      "step": 2300
    },
    {
      "epoch": 1.2080712788259957,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.01811320754716981,
      "loss": 0.466,
      "num_input_tokens_seen": 1504328,
      "step": 2305
    },
    {
      "epoch": 1.2106918238993711,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.01815251572327044,
      "loss": 0.5271,
      "num_input_tokens_seen": 1508264,
      "step": 2310
    },
    {
      "epoch": 1.2133123689727463,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.01819182389937107,
      "loss": 0.5916,
      "num_input_tokens_seen": 1511048,
      "step": 2315
    },
    {
      "epoch": 1.2159329140461215,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.018231132075471695,
      "loss": 0.6652,
      "num_input_tokens_seen": 1513928,
      "step": 2320
    },
    {
      "epoch": 1.2185534591194969,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.018270440251572328,
      "loss": 0.431,
      "num_input_tokens_seen": 1517512,
      "step": 2325
    },
    {
      "epoch": 1.221174004192872,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.018309748427672953,
      "loss": 0.3834,
      "num_input_tokens_seen": 1521256,
      "step": 2330
    },
    {
      "epoch": 1.2237945492662474,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.018349056603773586,
      "loss": 0.6301,
      "num_input_tokens_seen": 1523144,
      "step": 2335
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01838836477987421,
      "loss": 0.4735,
      "num_input_tokens_seen": 1527624,
      "step": 2340
    },
    {
      "epoch": 1.229035639412998,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.018427672955974844,
      "loss": 0.4714,
      "num_input_tokens_seen": 1531560,
      "step": 2345
    },
    {
      "epoch": 1.2316561844863732,
      "grad_norm": 0.07421875,
      "learning_rate": 0.018466981132075472,
      "loss": 0.4789,
      "num_input_tokens_seen": 1533832,
      "step": 2350
    },
    {
      "epoch": 1.2342767295597485,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.018506289308176098,
      "loss": 0.5036,
      "num_input_tokens_seen": 1536584,
      "step": 2355
    },
    {
      "epoch": 1.2368972746331237,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01854559748427673,
      "loss": 0.471,
      "num_input_tokens_seen": 1539720,
      "step": 2360
    },
    {
      "epoch": 1.2395178197064989,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.018584905660377356,
      "loss": 0.368,
      "num_input_tokens_seen": 1543112,
      "step": 2365
    },
    {
      "epoch": 1.2421383647798743,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.018624213836477988,
      "loss": 0.5017,
      "num_input_tokens_seen": 1546568,
      "step": 2370
    },
    {
      "epoch": 1.2447589098532494,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.018663522012578617,
      "loss": 0.5051,
      "num_input_tokens_seen": 1549864,
      "step": 2375
    },
    {
      "epoch": 1.2473794549266248,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.018702830188679243,
      "loss": 0.4318,
      "num_input_tokens_seen": 1552456,
      "step": 2380
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.018742138364779875,
      "loss": 0.566,
      "num_input_tokens_seen": 1555240,
      "step": 2385
    },
    {
      "epoch": 1.2526205450733752,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0187814465408805,
      "loss": 0.4103,
      "num_input_tokens_seen": 1558568,
      "step": 2390
    },
    {
      "epoch": 1.2552410901467506,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.018820754716981133,
      "loss": 0.5267,
      "num_input_tokens_seen": 1561288,
      "step": 2395
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.018860062893081758,
      "loss": 0.5612,
      "num_input_tokens_seen": 1564200,
      "step": 2400
    },
    {
      "epoch": 1.2604821802935011,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01889937106918239,
      "loss": 0.5089,
      "num_input_tokens_seen": 1567208,
      "step": 2405
    },
    {
      "epoch": 1.2631027253668763,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.01893867924528302,
      "loss": 0.4349,
      "num_input_tokens_seen": 1571240,
      "step": 2410
    },
    {
      "epoch": 1.2657232704402515,
      "grad_norm": 0.12890625,
      "learning_rate": 0.018977987421383645,
      "loss": 0.4972,
      "num_input_tokens_seen": 1574376,
      "step": 2415
    },
    {
      "epoch": 1.2683438155136268,
      "grad_norm": 0.056640625,
      "learning_rate": 0.019017295597484277,
      "loss": 0.5574,
      "num_input_tokens_seen": 1577672,
      "step": 2420
    },
    {
      "epoch": 1.270964360587002,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.019056603773584903,
      "loss": 0.4395,
      "num_input_tokens_seen": 1581192,
      "step": 2425
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.019095911949685535,
      "loss": 0.4197,
      "num_input_tokens_seen": 1584424,
      "step": 2430
    },
    {
      "epoch": 1.2762054507337526,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.019135220125786164,
      "loss": 0.4841,
      "num_input_tokens_seen": 1587912,
      "step": 2435
    },
    {
      "epoch": 1.2788259958071277,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.019174528301886793,
      "loss": 0.3602,
      "num_input_tokens_seen": 1591624,
      "step": 2440
    },
    {
      "epoch": 1.2814465408805031,
      "grad_norm": 0.041015625,
      "learning_rate": 0.019213836477987422,
      "loss": 0.5116,
      "num_input_tokens_seen": 1594376,
      "step": 2445
    },
    {
      "epoch": 1.2840670859538785,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.019253144654088047,
      "loss": 0.4637,
      "num_input_tokens_seen": 1597608,
      "step": 2450
    },
    {
      "epoch": 1.2866876310272537,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.01929245283018868,
      "loss": 0.4538,
      "num_input_tokens_seen": 1600136,
      "step": 2455
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.019331761006289305,
      "loss": 0.4897,
      "num_input_tokens_seen": 1603592,
      "step": 2460
    },
    {
      "epoch": 1.2919287211740043,
      "grad_norm": 0.056640625,
      "learning_rate": 0.019371069182389938,
      "loss": 0.6012,
      "num_input_tokens_seen": 1606504,
      "step": 2465
    },
    {
      "epoch": 1.2945492662473794,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.019410377358490567,
      "loss": 0.4312,
      "num_input_tokens_seen": 1613064,
      "step": 2470
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019449685534591192,
      "loss": 0.4672,
      "num_input_tokens_seen": 1616136,
      "step": 2475
    },
    {
      "epoch": 1.29979035639413,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.019488993710691824,
      "loss": 0.4516,
      "num_input_tokens_seen": 1618792,
      "step": 2480
    },
    {
      "epoch": 1.3024109014675052,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.01952830188679245,
      "loss": 0.3875,
      "num_input_tokens_seen": 1621096,
      "step": 2485
    },
    {
      "epoch": 1.3050314465408805,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.019567610062893082,
      "loss": 0.4077,
      "num_input_tokens_seen": 1623720,
      "step": 2490
    },
    {
      "epoch": 1.3076519916142557,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.01960691823899371,
      "loss": 0.5137,
      "num_input_tokens_seen": 1626888,
      "step": 2495
    },
    {
      "epoch": 1.310272536687631,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.01964622641509434,
      "loss": 0.6176,
      "num_input_tokens_seen": 1630472,
      "step": 2500
    },
    {
      "epoch": 1.3128930817610063,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01968553459119497,
      "loss": 0.445,
      "num_input_tokens_seen": 1633192,
      "step": 2505
    },
    {
      "epoch": 1.3155136268343814,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.019724842767295594,
      "loss": 0.539,
      "num_input_tokens_seen": 1637128,
      "step": 2510
    },
    {
      "epoch": 1.3181341719077568,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.019764150943396227,
      "loss": 0.5083,
      "num_input_tokens_seen": 1640456,
      "step": 2515
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.019803459119496852,
      "loss": 0.6007,
      "num_input_tokens_seen": 1644008,
      "step": 2520
    },
    {
      "epoch": 1.3233752620545074,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.019842767295597485,
      "loss": 0.4623,
      "num_input_tokens_seen": 1647720,
      "step": 2525
    },
    {
      "epoch": 1.3259958071278826,
      "grad_norm": 0.10546875,
      "learning_rate": 0.019882075471698114,
      "loss": 0.6123,
      "num_input_tokens_seen": 1650440,
      "step": 2530
    },
    {
      "epoch": 1.3286163522012577,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.01992138364779874,
      "loss": 0.5926,
      "num_input_tokens_seen": 1653352,
      "step": 2535
    },
    {
      "epoch": 1.3312368972746331,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01996069182389937,
      "loss": 0.4683,
      "num_input_tokens_seen": 1656360,
      "step": 2540
    },
    {
      "epoch": 1.3338574423480085,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.019999999999999997,
      "loss": 0.6073,
      "num_input_tokens_seen": 1659336,
      "step": 2545
    },
    {
      "epoch": 1.3364779874213837,
      "grad_norm": 0.150390625,
      "learning_rate": 0.02003930817610063,
      "loss": 0.6892,
      "num_input_tokens_seen": 1661896,
      "step": 2550
    },
    {
      "epoch": 1.3390985324947589,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.020078616352201258,
      "loss": 0.3779,
      "num_input_tokens_seen": 1665832,
      "step": 2555
    },
    {
      "epoch": 1.3417190775681342,
      "grad_norm": 0.08203125,
      "learning_rate": 0.020117924528301887,
      "loss": 0.5413,
      "num_input_tokens_seen": 1668520,
      "step": 2560
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.020157232704402516,
      "loss": 0.4809,
      "num_input_tokens_seen": 1671912,
      "step": 2565
    },
    {
      "epoch": 1.3469601677148848,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02019654088050314,
      "loss": 0.4796,
      "num_input_tokens_seen": 1675080,
      "step": 2570
    },
    {
      "epoch": 1.34958071278826,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.020235849056603774,
      "loss": 0.3478,
      "num_input_tokens_seen": 1681512,
      "step": 2575
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0202751572327044,
      "loss": 0.482,
      "num_input_tokens_seen": 1685000,
      "step": 2580
    },
    {
      "epoch": 1.3548218029350105,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.020314465408805032,
      "loss": 0.4375,
      "num_input_tokens_seen": 1687528,
      "step": 2585
    },
    {
      "epoch": 1.3574423480083857,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.02035377358490566,
      "loss": 0.6043,
      "num_input_tokens_seen": 1690280,
      "step": 2590
    },
    {
      "epoch": 1.360062893081761,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.02039308176100629,
      "loss": 0.4059,
      "num_input_tokens_seen": 1693352,
      "step": 2595
    },
    {
      "epoch": 1.3626834381551363,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.02043238993710692,
      "loss": 0.4735,
      "num_input_tokens_seen": 1696456,
      "step": 2600
    },
    {
      "epoch": 1.3653039832285114,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020471698113207544,
      "loss": 0.4296,
      "num_input_tokens_seen": 1700488,
      "step": 2605
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 0.078125,
      "learning_rate": 0.020511006289308176,
      "loss": 0.4457,
      "num_input_tokens_seen": 1703304,
      "step": 2610
    },
    {
      "epoch": 1.370545073375262,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.020550314465408805,
      "loss": 0.5289,
      "num_input_tokens_seen": 1705864,
      "step": 2615
    },
    {
      "epoch": 1.3731656184486374,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.020589622641509434,
      "loss": 0.5959,
      "num_input_tokens_seen": 1708616,
      "step": 2620
    },
    {
      "epoch": 1.3757861635220126,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.020628930817610063,
      "loss": 0.4738,
      "num_input_tokens_seen": 1712232,
      "step": 2625
    },
    {
      "epoch": 1.3784067085953877,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.02066823899371069,
      "loss": 0.3287,
      "num_input_tokens_seen": 1715144,
      "step": 2630
    },
    {
      "epoch": 1.381027253668763,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.02070754716981132,
      "loss": 0.4492,
      "num_input_tokens_seen": 1717800,
      "step": 2635
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.020746855345911946,
      "loss": 0.4798,
      "num_input_tokens_seen": 1721128,
      "step": 2640
    },
    {
      "epoch": 1.3862683438155137,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02078616352201258,
      "loss": 0.4454,
      "num_input_tokens_seen": 1724328,
      "step": 2645
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.020825471698113208,
      "loss": 0.5416,
      "num_input_tokens_seen": 1728328,
      "step": 2650
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.020864779874213837,
      "loss": 0.4588,
      "num_input_tokens_seen": 1731080,
      "step": 2655
    },
    {
      "epoch": 1.3941299790356394,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.020904088050314466,
      "loss": 0.5333,
      "num_input_tokens_seen": 1736008,
      "step": 2660
    },
    {
      "epoch": 1.3967505241090148,
      "grad_norm": 0.064453125,
      "learning_rate": 0.02094339622641509,
      "loss": 0.3978,
      "num_input_tokens_seen": 1739368,
      "step": 2665
    },
    {
      "epoch": 1.39937106918239,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.020982704402515723,
      "loss": 0.6928,
      "num_input_tokens_seen": 1742408,
      "step": 2670
    },
    {
      "epoch": 1.4019916142557651,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.02102201257861635,
      "loss": 0.5211,
      "num_input_tokens_seen": 1746120,
      "step": 2675
    },
    {
      "epoch": 1.4046121593291405,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.02106132075471698,
      "loss": 0.5158,
      "num_input_tokens_seen": 1749096,
      "step": 2680
    },
    {
      "epoch": 1.4072327044025157,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.02110062893081761,
      "loss": 0.5483,
      "num_input_tokens_seen": 1751624,
      "step": 2685
    },
    {
      "epoch": 1.409853249475891,
      "grad_norm": 0.06640625,
      "learning_rate": 0.021139937106918236,
      "loss": 0.4328,
      "num_input_tokens_seen": 1754216,
      "step": 2690
    },
    {
      "epoch": 1.4124737945492662,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.021179245283018868,
      "loss": 0.5662,
      "num_input_tokens_seen": 1758280,
      "step": 2695
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.021218553459119494,
      "loss": 0.4805,
      "num_input_tokens_seen": 1761352,
      "step": 2700
    },
    {
      "epoch": 1.4177148846960168,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.021257861635220126,
      "loss": 0.4354,
      "num_input_tokens_seen": 1764648,
      "step": 2705
    },
    {
      "epoch": 1.420335429769392,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.021297169811320755,
      "loss": 0.6126,
      "num_input_tokens_seen": 1767592,
      "step": 2710
    },
    {
      "epoch": 1.4229559748427674,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.021336477987421384,
      "loss": 0.4533,
      "num_input_tokens_seen": 1774600,
      "step": 2715
    },
    {
      "epoch": 1.4255765199161425,
      "grad_norm": 0.0703125,
      "learning_rate": 0.021375786163522013,
      "loss": 0.5642,
      "num_input_tokens_seen": 1778248,
      "step": 2720
    },
    {
      "epoch": 1.4281970649895177,
      "grad_norm": 0.037109375,
      "learning_rate": 0.021415094339622638,
      "loss": 0.4634,
      "num_input_tokens_seen": 1781608,
      "step": 2725
    },
    {
      "epoch": 1.430817610062893,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02145440251572327,
      "loss": 0.6245,
      "num_input_tokens_seen": 1784776,
      "step": 2730
    },
    {
      "epoch": 1.4334381551362683,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.021493710691823896,
      "loss": 0.5584,
      "num_input_tokens_seen": 1787688,
      "step": 2735
    },
    {
      "epoch": 1.4360587002096437,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02153301886792453,
      "loss": 0.5751,
      "num_input_tokens_seen": 1791016,
      "step": 2740
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.021572327044025157,
      "loss": 0.5095,
      "num_input_tokens_seen": 1794280,
      "step": 2745
    },
    {
      "epoch": 1.441299790356394,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.021611635220125786,
      "loss": 0.4013,
      "num_input_tokens_seen": 1797480,
      "step": 2750
    },
    {
      "epoch": 1.4439203354297694,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.021650943396226415,
      "loss": 0.6094,
      "num_input_tokens_seen": 1801160,
      "step": 2755
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.02169025157232704,
      "loss": 0.5878,
      "num_input_tokens_seen": 1803464,
      "step": 2760
    },
    {
      "epoch": 1.44916142557652,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.021729559748427673,
      "loss": 0.4513,
      "num_input_tokens_seen": 1806728,
      "step": 2765
    },
    {
      "epoch": 1.4517819706498951,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.021768867924528302,
      "loss": 0.4277,
      "num_input_tokens_seen": 1810088,
      "step": 2770
    },
    {
      "epoch": 1.4544025157232705,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02180817610062893,
      "loss": 0.4232,
      "num_input_tokens_seen": 1815144,
      "step": 2775
    },
    {
      "epoch": 1.4570230607966457,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02184748427672956,
      "loss": 0.4091,
      "num_input_tokens_seen": 1818792,
      "step": 2780
    },
    {
      "epoch": 1.459643605870021,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.021886792452830185,
      "loss": 0.5931,
      "num_input_tokens_seen": 1821768,
      "step": 2785
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.021926100628930818,
      "loss": 0.6062,
      "num_input_tokens_seen": 1825512,
      "step": 2790
    },
    {
      "epoch": 1.4648846960167714,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.021965408805031443,
      "loss": 0.3677,
      "num_input_tokens_seen": 1828968,
      "step": 2795
    },
    {
      "epoch": 1.4675052410901468,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.022004716981132075,
      "loss": 0.534,
      "num_input_tokens_seen": 1831336,
      "step": 2800
    },
    {
      "epoch": 1.470125786163522,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.022044025157232704,
      "loss": 0.5222,
      "num_input_tokens_seen": 1834536,
      "step": 2805
    },
    {
      "epoch": 1.4727463312368974,
      "grad_norm": 0.033203125,
      "learning_rate": 0.022083333333333333,
      "loss": 0.4641,
      "num_input_tokens_seen": 1837480,
      "step": 2810
    },
    {
      "epoch": 1.4753668763102725,
      "grad_norm": 0.09375,
      "learning_rate": 0.022122641509433962,
      "loss": 0.5596,
      "num_input_tokens_seen": 1840808,
      "step": 2815
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 0.0390625,
      "learning_rate": 0.022161949685534588,
      "loss": 0.604,
      "num_input_tokens_seen": 1844392,
      "step": 2820
    },
    {
      "epoch": 1.480607966457023,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02220125786163522,
      "loss": 0.3949,
      "num_input_tokens_seen": 1848008,
      "step": 2825
    },
    {
      "epoch": 1.4832285115303983,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02224056603773585,
      "loss": 0.416,
      "num_input_tokens_seen": 1850888,
      "step": 2830
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.022279874213836478,
      "loss": 0.6783,
      "num_input_tokens_seen": 1854312,
      "step": 2835
    },
    {
      "epoch": 1.4884696016771488,
      "grad_norm": 0.04296875,
      "learning_rate": 0.022319182389937107,
      "loss": 0.4931,
      "num_input_tokens_seen": 1858792,
      "step": 2840
    },
    {
      "epoch": 1.491090146750524,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.022358490566037732,
      "loss": 0.4695,
      "num_input_tokens_seen": 1861672,
      "step": 2845
    },
    {
      "epoch": 1.4937106918238994,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.022397798742138365,
      "loss": 0.4306,
      "num_input_tokens_seen": 1864360,
      "step": 2850
    },
    {
      "epoch": 1.4963312368972748,
      "grad_norm": 0.083984375,
      "learning_rate": 0.02243710691823899,
      "loss": 0.4724,
      "num_input_tokens_seen": 1867048,
      "step": 2855
    },
    {
      "epoch": 1.49895178197065,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.022476415094339623,
      "loss": 0.4558,
      "num_input_tokens_seen": 1870280,
      "step": 2860
    },
    {
      "epoch": 1.501572327044025,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.02251572327044025,
      "loss": 0.6523,
      "num_input_tokens_seen": 1873896,
      "step": 2865
    },
    {
      "epoch": 1.5041928721174003,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02255503144654088,
      "loss": 0.5002,
      "num_input_tokens_seen": 1876808,
      "step": 2870
    },
    {
      "epoch": 1.5068134171907757,
      "grad_norm": 0.04296875,
      "learning_rate": 0.02259433962264151,
      "loss": 0.5206,
      "num_input_tokens_seen": 1880168,
      "step": 2875
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.022633647798742135,
      "loss": 0.4057,
      "num_input_tokens_seen": 1882824,
      "step": 2880
    },
    {
      "epoch": 1.5120545073375262,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.022672955974842767,
      "loss": 0.4509,
      "num_input_tokens_seen": 1886184,
      "step": 2885
    },
    {
      "epoch": 1.5146750524109014,
      "grad_norm": 0.0390625,
      "learning_rate": 0.022712264150943396,
      "loss": 0.535,
      "num_input_tokens_seen": 1888968,
      "step": 2890
    },
    {
      "epoch": 1.5172955974842768,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.022751572327044025,
      "loss": 0.5797,
      "num_input_tokens_seen": 1892232,
      "step": 2895
    },
    {
      "epoch": 1.519916142557652,
      "grad_norm": 0.048828125,
      "learning_rate": 0.022790880503144654,
      "loss": 0.4603,
      "num_input_tokens_seen": 1895240,
      "step": 2900
    },
    {
      "epoch": 1.5225366876310273,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.022830188679245283,
      "loss": 0.3465,
      "num_input_tokens_seen": 1899144,
      "step": 2905
    },
    {
      "epoch": 1.5251572327044025,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02286949685534591,
      "loss": 0.4519,
      "num_input_tokens_seen": 1901896,
      "step": 2910
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 0.0390625,
      "learning_rate": 0.022908805031446537,
      "loss": 0.4626,
      "num_input_tokens_seen": 1904936,
      "step": 2915
    },
    {
      "epoch": 1.530398322851153,
      "grad_norm": 0.046875,
      "learning_rate": 0.02294811320754717,
      "loss": 0.5387,
      "num_input_tokens_seen": 1908328,
      "step": 2920
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0229874213836478,
      "loss": 0.5642,
      "num_input_tokens_seen": 1911496,
      "step": 2925
    },
    {
      "epoch": 1.5356394129979036,
      "grad_norm": 0.064453125,
      "learning_rate": 0.023026729559748427,
      "loss": 0.415,
      "num_input_tokens_seen": 1913992,
      "step": 2930
    },
    {
      "epoch": 1.5382599580712788,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.023066037735849056,
      "loss": 0.5507,
      "num_input_tokens_seen": 1917640,
      "step": 2935
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.023105345911949682,
      "loss": 0.3968,
      "num_input_tokens_seen": 1920616,
      "step": 2940
    },
    {
      "epoch": 1.5435010482180294,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.023144654088050314,
      "loss": 0.4436,
      "num_input_tokens_seen": 1923208,
      "step": 2945
    },
    {
      "epoch": 1.5461215932914047,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.023183962264150943,
      "loss": 0.4023,
      "num_input_tokens_seen": 1926952,
      "step": 2950
    },
    {
      "epoch": 1.54874213836478,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.023223270440251572,
      "loss": 0.4594,
      "num_input_tokens_seen": 1930696,
      "step": 2955
    },
    {
      "epoch": 1.551362683438155,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0232625786163522,
      "loss": 0.4736,
      "num_input_tokens_seen": 1933864,
      "step": 2960
    },
    {
      "epoch": 1.5539832285115303,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.02330188679245283,
      "loss": 0.4923,
      "num_input_tokens_seen": 1938440,
      "step": 2965
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02334119496855346,
      "loss": 0.5657,
      "num_input_tokens_seen": 1942184,
      "step": 2970
    },
    {
      "epoch": 1.559224318658281,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.023380503144654084,
      "loss": 0.4257,
      "num_input_tokens_seen": 1945928,
      "step": 2975
    },
    {
      "epoch": 1.5618448637316562,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.023419811320754717,
      "loss": 0.5502,
      "num_input_tokens_seen": 1949416,
      "step": 2980
    },
    {
      "epoch": 1.5644654088050314,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.023459119496855346,
      "loss": 0.4443,
      "num_input_tokens_seen": 1952008,
      "step": 2985
    },
    {
      "epoch": 1.5670859538784065,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.023498427672955974,
      "loss": 0.4395,
      "num_input_tokens_seen": 1955656,
      "step": 2990
    },
    {
      "epoch": 1.569706498951782,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.023537735849056603,
      "loss": 0.6587,
      "num_input_tokens_seen": 1958632,
      "step": 2995
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02357704402515723,
      "loss": 0.4768,
      "num_input_tokens_seen": 1961256,
      "step": 3000
    },
    {
      "epoch": 1.5749475890985325,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02361635220125786,
      "loss": 0.4164,
      "num_input_tokens_seen": 1965480,
      "step": 3005
    },
    {
      "epoch": 1.5775681341719077,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.02365566037735849,
      "loss": 0.4746,
      "num_input_tokens_seen": 1968552,
      "step": 3010
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02369496855345912,
      "loss": 0.4572,
      "num_input_tokens_seen": 1971400,
      "step": 3015
    },
    {
      "epoch": 1.5828092243186582,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.023734276729559748,
      "loss": 0.3991,
      "num_input_tokens_seen": 1974312,
      "step": 3020
    },
    {
      "epoch": 1.5854297693920336,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.023773584905660377,
      "loss": 0.3689,
      "num_input_tokens_seen": 1976904,
      "step": 3025
    },
    {
      "epoch": 1.5880503144654088,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.023812893081761006,
      "loss": 0.5497,
      "num_input_tokens_seen": 1979432,
      "step": 3030
    },
    {
      "epoch": 1.590670859538784,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02385220125786163,
      "loss": 0.4969,
      "num_input_tokens_seen": 1982952,
      "step": 3035
    },
    {
      "epoch": 1.5932914046121593,
      "grad_norm": 0.095703125,
      "learning_rate": 0.023891509433962264,
      "loss": 0.4722,
      "num_input_tokens_seen": 1986408,
      "step": 3040
    },
    {
      "epoch": 1.5959119496855347,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.023930817610062893,
      "loss": 0.4466,
      "num_input_tokens_seen": 1988936,
      "step": 3045
    },
    {
      "epoch": 1.59853249475891,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.02397012578616352,
      "loss": 0.6176,
      "num_input_tokens_seen": 1992264,
      "step": 3050
    },
    {
      "epoch": 1.601153039832285,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.02400943396226415,
      "loss": 0.5888,
      "num_input_tokens_seen": 1995240,
      "step": 3055
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02404874213836478,
      "loss": 0.5304,
      "num_input_tokens_seen": 1998408,
      "step": 3060
    },
    {
      "epoch": 1.6063941299790356,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02408805031446541,
      "loss": 0.4596,
      "num_input_tokens_seen": 2001416,
      "step": 3065
    },
    {
      "epoch": 1.609014675052411,
      "grad_norm": 0.033203125,
      "learning_rate": 0.024127358490566037,
      "loss": 0.4593,
      "num_input_tokens_seen": 2003752,
      "step": 3070
    },
    {
      "epoch": 1.6116352201257862,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.024166666666666666,
      "loss": 0.4586,
      "num_input_tokens_seen": 2006728,
      "step": 3075
    },
    {
      "epoch": 1.6142557651991614,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.024205974842767295,
      "loss": 0.456,
      "num_input_tokens_seen": 2009928,
      "step": 3080
    },
    {
      "epoch": 1.6168763102725365,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.024245283018867924,
      "loss": 0.3392,
      "num_input_tokens_seen": 2012648,
      "step": 3085
    },
    {
      "epoch": 1.619496855345912,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.024284591194968553,
      "loss": 0.3951,
      "num_input_tokens_seen": 2016936,
      "step": 3090
    },
    {
      "epoch": 1.6221174004192873,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.02432389937106918,
      "loss": 0.3308,
      "num_input_tokens_seen": 2019720,
      "step": 3095
    },
    {
      "epoch": 1.6247379454926625,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.02436320754716981,
      "loss": 0.4376,
      "num_input_tokens_seen": 2022760,
      "step": 3100
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02440251572327044,
      "loss": 0.4252,
      "num_input_tokens_seen": 2025544,
      "step": 3105
    },
    {
      "epoch": 1.629979035639413,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.02444182389937107,
      "loss": 0.4254,
      "num_input_tokens_seen": 2028872,
      "step": 3110
    },
    {
      "epoch": 1.6325995807127882,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.024481132075471698,
      "loss": 0.4143,
      "num_input_tokens_seen": 2031816,
      "step": 3115
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.024520440251572326,
      "loss": 0.5799,
      "num_input_tokens_seen": 2034280,
      "step": 3120
    },
    {
      "epoch": 1.6378406708595388,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.024559748427672955,
      "loss": 0.4868,
      "num_input_tokens_seen": 2038056,
      "step": 3125
    },
    {
      "epoch": 1.640461215932914,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.024599056603773584,
      "loss": 0.5244,
      "num_input_tokens_seen": 2041192,
      "step": 3130
    },
    {
      "epoch": 1.6430817610062893,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.024638364779874213,
      "loss": 0.4607,
      "num_input_tokens_seen": 2043720,
      "step": 3135
    },
    {
      "epoch": 1.6457023060796647,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.024677672955974842,
      "loss": 0.5499,
      "num_input_tokens_seen": 2046792,
      "step": 3140
    },
    {
      "epoch": 1.64832285115304,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02471698113207547,
      "loss": 0.6285,
      "num_input_tokens_seen": 2049768,
      "step": 3145
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0247562893081761,
      "loss": 0.5107,
      "num_input_tokens_seen": 2052840,
      "step": 3150
    },
    {
      "epoch": 1.6535639412997902,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.024795597484276725,
      "loss": 0.5216,
      "num_input_tokens_seen": 2055432,
      "step": 3155
    },
    {
      "epoch": 1.6561844863731656,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.024834905660377358,
      "loss": 0.6007,
      "num_input_tokens_seen": 2058792,
      "step": 3160
    },
    {
      "epoch": 1.658805031446541,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.024874213836477987,
      "loss": 0.4629,
      "num_input_tokens_seen": 2061960,
      "step": 3165
    },
    {
      "epoch": 1.6614255765199162,
      "grad_norm": 0.041015625,
      "learning_rate": 0.024913522012578616,
      "loss": 0.6032,
      "num_input_tokens_seen": 2064712,
      "step": 3170
    },
    {
      "epoch": 1.6640461215932913,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.024952830188679245,
      "loss": 0.4311,
      "num_input_tokens_seen": 2068360,
      "step": 3175
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.024992138364779874,
      "loss": 0.4676,
      "num_input_tokens_seen": 2071304,
      "step": 3180
    },
    {
      "epoch": 1.669287211740042,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.025031446540880502,
      "loss": 0.6951,
      "num_input_tokens_seen": 2075016,
      "step": 3185
    },
    {
      "epoch": 1.6719077568134173,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02507075471698113,
      "loss": 0.4479,
      "num_input_tokens_seen": 2079016,
      "step": 3190
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02511006289308176,
      "loss": 0.372,
      "num_input_tokens_seen": 2081960,
      "step": 3195
    },
    {
      "epoch": 1.6771488469601676,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02514937106918239,
      "loss": 0.4108,
      "num_input_tokens_seen": 2084648,
      "step": 3200
    },
    {
      "epoch": 1.679769392033543,
      "grad_norm": 0.0390625,
      "learning_rate": 0.025188679245283018,
      "loss": 0.4379,
      "num_input_tokens_seen": 2088840,
      "step": 3205
    },
    {
      "epoch": 1.6823899371069182,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.025227987421383647,
      "loss": 0.4945,
      "num_input_tokens_seen": 2091944,
      "step": 3210
    },
    {
      "epoch": 1.6850104821802936,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.025267295597484276,
      "loss": 0.5266,
      "num_input_tokens_seen": 2095144,
      "step": 3215
    },
    {
      "epoch": 1.6876310272536688,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.025306603773584905,
      "loss": 0.5453,
      "num_input_tokens_seen": 2098984,
      "step": 3220
    },
    {
      "epoch": 1.690251572327044,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025345911949685534,
      "loss": 0.3579,
      "num_input_tokens_seen": 2101736,
      "step": 3225
    },
    {
      "epoch": 1.6928721174004193,
      "grad_norm": 0.03125,
      "learning_rate": 0.025385220125786163,
      "loss": 0.5292,
      "num_input_tokens_seen": 2104904,
      "step": 3230
    },
    {
      "epoch": 1.6954926624737947,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.02542452830188679,
      "loss": 0.5222,
      "num_input_tokens_seen": 2108392,
      "step": 3235
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02546383647798742,
      "loss": 0.5996,
      "num_input_tokens_seen": 2112360,
      "step": 3240
    },
    {
      "epoch": 1.700733752620545,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.02550314465408805,
      "loss": 0.5147,
      "num_input_tokens_seen": 2117992,
      "step": 3245
    },
    {
      "epoch": 1.7033542976939202,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02554245283018868,
      "loss": 0.4206,
      "num_input_tokens_seen": 2121096,
      "step": 3250
    },
    {
      "epoch": 1.7059748427672956,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.025581761006289307,
      "loss": 0.48,
      "num_input_tokens_seen": 2125064,
      "step": 3255
    },
    {
      "epoch": 1.708595387840671,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.025621069182389936,
      "loss": 0.3687,
      "num_input_tokens_seen": 2129032,
      "step": 3260
    },
    {
      "epoch": 1.7112159329140462,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.025660377358490565,
      "loss": 0.571,
      "num_input_tokens_seen": 2131656,
      "step": 3265
    },
    {
      "epoch": 1.7138364779874213,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.025699685534591194,
      "loss": 0.5832,
      "num_input_tokens_seen": 2134760,
      "step": 3270
    },
    {
      "epoch": 1.7164570230607965,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.025738993710691823,
      "loss": 0.5801,
      "num_input_tokens_seen": 2137960,
      "step": 3275
    },
    {
      "epoch": 1.719077568134172,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.025778301886792452,
      "loss": 0.4604,
      "num_input_tokens_seen": 2140552,
      "step": 3280
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.02581761006289308,
      "loss": 0.3667,
      "num_input_tokens_seen": 2143272,
      "step": 3285
    },
    {
      "epoch": 1.7243186582809225,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.02585691823899371,
      "loss": 0.5207,
      "num_input_tokens_seen": 2147304,
      "step": 3290
    },
    {
      "epoch": 1.7269392033542976,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02589622641509434,
      "loss": 0.6547,
      "num_input_tokens_seen": 2151624,
      "step": 3295
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.025935534591194968,
      "loss": 0.5422,
      "num_input_tokens_seen": 2154248,
      "step": 3300
    },
    {
      "epoch": 1.7321802935010482,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.025974842767295597,
      "loss": 0.546,
      "num_input_tokens_seen": 2157000,
      "step": 3305
    },
    {
      "epoch": 1.7348008385744236,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.026014150943396225,
      "loss": 0.4257,
      "num_input_tokens_seen": 2160904,
      "step": 3310
    },
    {
      "epoch": 1.7374213836477987,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.026053459119496854,
      "loss": 0.5387,
      "num_input_tokens_seen": 2163720,
      "step": 3315
    },
    {
      "epoch": 1.740041928721174,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.026092767295597483,
      "loss": 0.4179,
      "num_input_tokens_seen": 2167912,
      "step": 3320
    },
    {
      "epoch": 1.7426624737945493,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.026132075471698112,
      "loss": 0.4589,
      "num_input_tokens_seen": 2171592,
      "step": 3325
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02617138364779874,
      "loss": 0.5383,
      "num_input_tokens_seen": 2173864,
      "step": 3330
    },
    {
      "epoch": 1.7479035639412999,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02621069182389937,
      "loss": 0.4397,
      "num_input_tokens_seen": 2176744,
      "step": 3335
    },
    {
      "epoch": 1.750524109014675,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02625,
      "loss": 0.6475,
      "num_input_tokens_seen": 2179528,
      "step": 3340
    },
    {
      "epoch": 1.7531446540880502,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.026289308176100628,
      "loss": 0.4176,
      "num_input_tokens_seen": 2182856,
      "step": 3345
    },
    {
      "epoch": 1.7557651991614256,
      "grad_norm": 0.03515625,
      "learning_rate": 0.026328616352201257,
      "loss": 0.434,
      "num_input_tokens_seen": 2185832,
      "step": 3350
    },
    {
      "epoch": 1.758385744234801,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.026367924528301886,
      "loss": 0.502,
      "num_input_tokens_seen": 2189352,
      "step": 3355
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 0.072265625,
      "learning_rate": 0.026407232704402515,
      "loss": 0.4611,
      "num_input_tokens_seen": 2192264,
      "step": 3360
    },
    {
      "epoch": 1.7636268343815513,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.026446540880503144,
      "loss": 0.4674,
      "num_input_tokens_seen": 2195560,
      "step": 3365
    },
    {
      "epoch": 1.7662473794549265,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.026485849056603773,
      "loss": 0.407,
      "num_input_tokens_seen": 2199080,
      "step": 3370
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.0265251572327044,
      "loss": 0.3877,
      "num_input_tokens_seen": 2202408,
      "step": 3375
    },
    {
      "epoch": 1.7714884696016773,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02656446540880503,
      "loss": 0.5156,
      "num_input_tokens_seen": 2206056,
      "step": 3380
    },
    {
      "epoch": 1.7741090146750524,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.02660377358490566,
      "loss": 0.4183,
      "num_input_tokens_seen": 2208456,
      "step": 3385
    },
    {
      "epoch": 1.7767295597484276,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.026643081761006288,
      "loss": 0.3657,
      "num_input_tokens_seen": 2211848,
      "step": 3390
    },
    {
      "epoch": 1.7793501048218028,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.026682389937106917,
      "loss": 0.3912,
      "num_input_tokens_seen": 2215112,
      "step": 3395
    },
    {
      "epoch": 1.7819706498951782,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.026721698113207546,
      "loss": 0.5298,
      "num_input_tokens_seen": 2219528,
      "step": 3400
    },
    {
      "epoch": 1.7845911949685536,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.026761006289308175,
      "loss": 0.5331,
      "num_input_tokens_seen": 2223624,
      "step": 3405
    },
    {
      "epoch": 1.7872117400419287,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.026800314465408804,
      "loss": 0.4873,
      "num_input_tokens_seen": 2227464,
      "step": 3410
    },
    {
      "epoch": 1.789832285115304,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.026839622641509433,
      "loss": 0.4241,
      "num_input_tokens_seen": 2231240,
      "step": 3415
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.026878930817610062,
      "loss": 0.5762,
      "num_input_tokens_seen": 2234600,
      "step": 3420
    },
    {
      "epoch": 1.7950733752620545,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02691823899371069,
      "loss": 0.5021,
      "num_input_tokens_seen": 2238056,
      "step": 3425
    },
    {
      "epoch": 1.7976939203354299,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02695754716981132,
      "loss": 0.4147,
      "num_input_tokens_seen": 2241640,
      "step": 3430
    },
    {
      "epoch": 1.800314465408805,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02699685534591195,
      "loss": 0.5187,
      "num_input_tokens_seen": 2245832,
      "step": 3435
    },
    {
      "epoch": 1.8029350104821802,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.027036163522012577,
      "loss": 0.7107,
      "num_input_tokens_seen": 2248552,
      "step": 3440
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.027075471698113206,
      "loss": 0.552,
      "num_input_tokens_seen": 2251432,
      "step": 3445
    },
    {
      "epoch": 1.808176100628931,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.027114779874213835,
      "loss": 0.5567,
      "num_input_tokens_seen": 2254408,
      "step": 3450
    },
    {
      "epoch": 1.8107966457023061,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.027154088050314464,
      "loss": 0.6669,
      "num_input_tokens_seen": 2257032,
      "step": 3455
    },
    {
      "epoch": 1.8134171907756813,
      "grad_norm": 0.068359375,
      "learning_rate": 0.027193396226415093,
      "loss": 0.4143,
      "num_input_tokens_seen": 2259560,
      "step": 3460
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.027232704402515722,
      "loss": 0.5158,
      "num_input_tokens_seen": 2263240,
      "step": 3465
    },
    {
      "epoch": 1.8186582809224319,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02727201257861635,
      "loss": 0.4518,
      "num_input_tokens_seen": 2267144,
      "step": 3470
    },
    {
      "epoch": 1.8212788259958073,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.02731132075471698,
      "loss": 0.5498,
      "num_input_tokens_seen": 2270856,
      "step": 3475
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.02735062893081761,
      "loss": 0.5716,
      "num_input_tokens_seen": 2273672,
      "step": 3480
    },
    {
      "epoch": 1.8265199161425576,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.027389937106918238,
      "loss": 0.4675,
      "num_input_tokens_seen": 2276520,
      "step": 3485
    },
    {
      "epoch": 1.8291404612159328,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.027429245283018867,
      "loss": 0.4117,
      "num_input_tokens_seen": 2279368,
      "step": 3490
    },
    {
      "epoch": 1.8317610062893082,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.027468553459119496,
      "loss": 0.7183,
      "num_input_tokens_seen": 2281992,
      "step": 3495
    },
    {
      "epoch": 1.8343815513626835,
      "grad_norm": 0.029296875,
      "learning_rate": 0.027507861635220125,
      "loss": 0.439,
      "num_input_tokens_seen": 2285576,
      "step": 3500
    },
    {
      "epoch": 1.8370020964360587,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.027547169811320753,
      "loss": 0.4198,
      "num_input_tokens_seen": 2288072,
      "step": 3505
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027586477987421382,
      "loss": 0.3883,
      "num_input_tokens_seen": 2291336,
      "step": 3510
    },
    {
      "epoch": 1.8422431865828093,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02762578616352201,
      "loss": 0.533,
      "num_input_tokens_seen": 2294152,
      "step": 3515
    },
    {
      "epoch": 1.8448637316561844,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02766509433962264,
      "loss": 0.3847,
      "num_input_tokens_seen": 2296872,
      "step": 3520
    },
    {
      "epoch": 1.8474842767295598,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.027704402515723273,
      "loss": 0.4793,
      "num_input_tokens_seen": 2299528,
      "step": 3525
    },
    {
      "epoch": 1.850104821802935,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.027743710691823898,
      "loss": 0.6397,
      "num_input_tokens_seen": 2301864,
      "step": 3530
    },
    {
      "epoch": 1.8527253668763102,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.027783018867924527,
      "loss": 0.4364,
      "num_input_tokens_seen": 2306088,
      "step": 3535
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.027822327044025156,
      "loss": 0.5367,
      "num_input_tokens_seen": 2309416,
      "step": 3540
    },
    {
      "epoch": 1.857966457023061,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.027861635220125785,
      "loss": 0.5237,
      "num_input_tokens_seen": 2312328,
      "step": 3545
    },
    {
      "epoch": 1.8605870020964361,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.027900943396226414,
      "loss": 0.538,
      "num_input_tokens_seen": 2314472,
      "step": 3550
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.027940251572327043,
      "loss": 0.386,
      "num_input_tokens_seen": 2318664,
      "step": 3555
    },
    {
      "epoch": 1.8658280922431865,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02797955974842767,
      "loss": 0.5493,
      "num_input_tokens_seen": 2323784,
      "step": 3560
    },
    {
      "epoch": 1.8684486373165619,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0280188679245283,
      "loss": 0.5283,
      "num_input_tokens_seen": 2326536,
      "step": 3565
    },
    {
      "epoch": 1.8710691823899372,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02805817610062893,
      "loss": 0.4685,
      "num_input_tokens_seen": 2330440,
      "step": 3570
    },
    {
      "epoch": 1.8736897274633124,
      "grad_norm": 0.00933837890625,
      "learning_rate": 0.02809748427672956,
      "loss": 0.5615,
      "num_input_tokens_seen": 2335688,
      "step": 3575
    },
    {
      "epoch": 1.8763102725366876,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.028136792452830187,
      "loss": 0.4362,
      "num_input_tokens_seen": 2338568,
      "step": 3580
    },
    {
      "epoch": 1.8789308176100628,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02817610062893082,
      "loss": 0.5005,
      "num_input_tokens_seen": 2342120,
      "step": 3585
    },
    {
      "epoch": 1.8815513626834381,
      "grad_norm": 0.029296875,
      "learning_rate": 0.028215408805031445,
      "loss": 0.4138,
      "num_input_tokens_seen": 2346184,
      "step": 3590
    },
    {
      "epoch": 1.8841719077568135,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.028254716981132074,
      "loss": 0.6041,
      "num_input_tokens_seen": 2348936,
      "step": 3595
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.028294025157232703,
      "loss": 0.4944,
      "num_input_tokens_seen": 2351976,
      "step": 3600
    },
    {
      "epoch": 1.8894129979035639,
      "grad_norm": 0.025390625,
      "learning_rate": 0.028333333333333332,
      "loss": 0.4862,
      "num_input_tokens_seen": 2357192,
      "step": 3605
    },
    {
      "epoch": 1.892033542976939,
      "grad_norm": 0.046875,
      "learning_rate": 0.02837264150943396,
      "loss": 0.6277,
      "num_input_tokens_seen": 2360072,
      "step": 3610
    },
    {
      "epoch": 1.8946540880503144,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02841194968553459,
      "loss": 0.4751,
      "num_input_tokens_seen": 2363656,
      "step": 3615
    },
    {
      "epoch": 1.8972746331236898,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.02845125786163522,
      "loss": 0.5326,
      "num_input_tokens_seen": 2366216,
      "step": 3620
    },
    {
      "epoch": 1.899895178197065,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.028490566037735848,
      "loss": 0.5135,
      "num_input_tokens_seen": 2369448,
      "step": 3625
    },
    {
      "epoch": 1.9025157232704402,
      "grad_norm": 0.037109375,
      "learning_rate": 0.028529874213836477,
      "loss": 0.6448,
      "num_input_tokens_seen": 2372744,
      "step": 3630
    },
    {
      "epoch": 1.9051362683438156,
      "grad_norm": 0.048828125,
      "learning_rate": 0.028569182389937105,
      "loss": 0.4176,
      "num_input_tokens_seen": 2375080,
      "step": 3635
    },
    {
      "epoch": 1.9077568134171907,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.028608490566037734,
      "loss": 0.5113,
      "num_input_tokens_seen": 2377864,
      "step": 3640
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.028647798742138367,
      "loss": 0.4849,
      "num_input_tokens_seen": 2381224,
      "step": 3645
    },
    {
      "epoch": 1.9129979035639413,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.028687106918238992,
      "loss": 0.5262,
      "num_input_tokens_seen": 2384424,
      "step": 3650
    },
    {
      "epoch": 1.9156184486373165,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02872641509433962,
      "loss": 0.4495,
      "num_input_tokens_seen": 2388936,
      "step": 3655
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02876572327044025,
      "loss": 0.52,
      "num_input_tokens_seen": 2392744,
      "step": 3660
    },
    {
      "epoch": 1.9208595387840672,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02880503144654088,
      "loss": 0.5866,
      "num_input_tokens_seen": 2395112,
      "step": 3665
    },
    {
      "epoch": 1.9234800838574424,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.028844339622641508,
      "loss": 0.4118,
      "num_input_tokens_seen": 2398344,
      "step": 3670
    },
    {
      "epoch": 1.9261006289308176,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.028883647798742137,
      "loss": 0.4979,
      "num_input_tokens_seen": 2401704,
      "step": 3675
    },
    {
      "epoch": 1.9287211740041927,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.02892295597484277,
      "loss": 0.3759,
      "num_input_tokens_seen": 2404168,
      "step": 3680
    },
    {
      "epoch": 1.9313417190775681,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028962264150943395,
      "loss": 0.3838,
      "num_input_tokens_seen": 2407688,
      "step": 3685
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029001572327044024,
      "loss": 0.404,
      "num_input_tokens_seen": 2411528,
      "step": 3690
    },
    {
      "epoch": 1.9365828092243187,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.029040880503144652,
      "loss": 0.53,
      "num_input_tokens_seen": 2414440,
      "step": 3695
    },
    {
      "epoch": 1.9392033542976939,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02908018867924528,
      "loss": 0.4281,
      "num_input_tokens_seen": 2417448,
      "step": 3700
    },
    {
      "epoch": 1.941823899371069,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.029119496855345914,
      "loss": 0.466,
      "num_input_tokens_seen": 2420680,
      "step": 3705
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02915880503144654,
      "loss": 0.4693,
      "num_input_tokens_seen": 2423752,
      "step": 3710
    },
    {
      "epoch": 1.9470649895178198,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029198113207547168,
      "loss": 0.493,
      "num_input_tokens_seen": 2426216,
      "step": 3715
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029237421383647797,
      "loss": 0.5414,
      "num_input_tokens_seen": 2429640,
      "step": 3720
    },
    {
      "epoch": 1.9523060796645701,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.029276729559748426,
      "loss": 0.4289,
      "num_input_tokens_seen": 2433224,
      "step": 3725
    },
    {
      "epoch": 1.9549266247379455,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.029316037735849055,
      "loss": 0.5463,
      "num_input_tokens_seen": 2437576,
      "step": 3730
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.029355345911949684,
      "loss": 0.4656,
      "num_input_tokens_seen": 2439976,
      "step": 3735
    },
    {
      "epoch": 1.960167714884696,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029394654088050316,
      "loss": 0.4571,
      "num_input_tokens_seen": 2443272,
      "step": 3740
    },
    {
      "epoch": 1.9627882599580713,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02943396226415094,
      "loss": 0.5562,
      "num_input_tokens_seen": 2445832,
      "step": 3745
    },
    {
      "epoch": 1.9654088050314464,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02947327044025157,
      "loss": 0.3767,
      "num_input_tokens_seen": 2448456,
      "step": 3750
    },
    {
      "epoch": 1.9680293501048218,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0295125786163522,
      "loss": 0.5551,
      "num_input_tokens_seen": 2451208,
      "step": 3755
    },
    {
      "epoch": 1.9706498951781972,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.02955188679245283,
      "loss": 0.4464,
      "num_input_tokens_seen": 2454152,
      "step": 3760
    },
    {
      "epoch": 1.9732704402515724,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.02959119496855346,
      "loss": 0.3867,
      "num_input_tokens_seen": 2457736,
      "step": 3765
    },
    {
      "epoch": 1.9758909853249476,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029630503144654086,
      "loss": 0.5914,
      "num_input_tokens_seen": 2461384,
      "step": 3770
    },
    {
      "epoch": 1.9785115303983227,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.029669811320754715,
      "loss": 0.3773,
      "num_input_tokens_seen": 2464776,
      "step": 3775
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.029709119496855344,
      "loss": 0.4475,
      "num_input_tokens_seen": 2467144,
      "step": 3780
    },
    {
      "epoch": 1.9837526205450735,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029748427672955973,
      "loss": 0.5695,
      "num_input_tokens_seen": 2469736,
      "step": 3785
    },
    {
      "epoch": 1.9863731656184487,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029787735849056602,
      "loss": 0.3758,
      "num_input_tokens_seen": 2474536,
      "step": 3790
    },
    {
      "epoch": 1.9889937106918238,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02982704402515723,
      "loss": 0.504,
      "num_input_tokens_seen": 2477704,
      "step": 3795
    },
    {
      "epoch": 1.991614255765199,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029866352201257863,
      "loss": 0.4619,
      "num_input_tokens_seen": 2480296,
      "step": 3800
    },
    {
      "epoch": 1.9942348008385744,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.02990566037735849,
      "loss": 0.5497,
      "num_input_tokens_seen": 2483368,
      "step": 3805
    },
    {
      "epoch": 1.9968553459119498,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.029944968553459118,
      "loss": 0.3818,
      "num_input_tokens_seen": 2486440,
      "step": 3810
    },
    {
      "epoch": 1.999475890985325,
      "grad_norm": 0.0390625,
      "learning_rate": 0.029984276729559747,
      "loss": 0.3516,
      "num_input_tokens_seen": 2489352,
      "step": 3815
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4636456370353699,
      "eval_runtime": 13.604,
      "eval_samples_per_second": 62.334,
      "eval_steps_per_second": 15.584,
      "num_input_tokens_seen": 2489456,
      "step": 3816
    },
    {
      "epoch": 2.0020964360587,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.029999999435190796,
      "loss": 0.5472,
      "num_input_tokens_seen": 2492112,
      "step": 3820
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.029999995983579154,
      "loss": 0.3871,
      "num_input_tokens_seen": 2494800,
      "step": 3825
    },
    {
      "epoch": 2.007337526205451,
      "grad_norm": 0.048828125,
      "learning_rate": 0.029999989394139483,
      "loss": 0.4762,
      "num_input_tokens_seen": 2498896,
      "step": 3830
    },
    {
      "epoch": 2.009958071278826,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029999979666873157,
      "loss": 0.5307,
      "num_input_tokens_seen": 2502160,
      "step": 3835
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02999996680178221,
      "loss": 0.4732,
      "num_input_tokens_seen": 2504752,
      "step": 3840
    },
    {
      "epoch": 2.0151991614255764,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029999950798869343,
      "loss": 0.4699,
      "num_input_tokens_seen": 2507536,
      "step": 3845
    },
    {
      "epoch": 2.0178197064989516,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.029999931658137895,
      "loss": 0.4921,
      "num_input_tokens_seen": 2510352,
      "step": 3850
    },
    {
      "epoch": 2.020440251572327,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02999990937959187,
      "loss": 0.4296,
      "num_input_tokens_seen": 2512784,
      "step": 3855
    },
    {
      "epoch": 2.0230607966457024,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029999883963235933,
      "loss": 0.4175,
      "num_input_tokens_seen": 2516304,
      "step": 3860
    },
    {
      "epoch": 2.0256813417190775,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.029999855409075397,
      "loss": 0.5152,
      "num_input_tokens_seen": 2519184,
      "step": 3865
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029999823717116243,
      "loss": 0.5501,
      "num_input_tokens_seen": 2522448,
      "step": 3870
    },
    {
      "epoch": 2.030922431865828,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.029999788887365087,
      "loss": 0.4378,
      "num_input_tokens_seen": 2525744,
      "step": 3875
    },
    {
      "epoch": 2.0335429769392035,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.029999750919829225,
      "loss": 0.4259,
      "num_input_tokens_seen": 2528240,
      "step": 3880
    },
    {
      "epoch": 2.0361635220125787,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0299997098145166,
      "loss": 0.4463,
      "num_input_tokens_seen": 2530928,
      "step": 3885
    },
    {
      "epoch": 2.038784067085954,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029999665571435803,
      "loss": 0.4613,
      "num_input_tokens_seen": 2533936,
      "step": 3890
    },
    {
      "epoch": 2.041404612159329,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0299996181905961,
      "loss": 0.5761,
      "num_input_tokens_seen": 2536720,
      "step": 3895
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029999567672007385,
      "loss": 0.4541,
      "num_input_tokens_seen": 2540656,
      "step": 3900
    },
    {
      "epoch": 2.04664570230608,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029999514015680247,
      "loss": 0.435,
      "num_input_tokens_seen": 2545040,
      "step": 3905
    },
    {
      "epoch": 2.049266247379455,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0299994572216259,
      "loss": 0.7105,
      "num_input_tokens_seen": 2547888,
      "step": 3910
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02999939728985622,
      "loss": 0.3574,
      "num_input_tokens_seen": 2550256,
      "step": 3915
    },
    {
      "epoch": 2.0545073375262053,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029999334220383752,
      "loss": 0.3658,
      "num_input_tokens_seen": 2553712,
      "step": 3920
    },
    {
      "epoch": 2.057127882599581,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029999268013221683,
      "loss": 0.5116,
      "num_input_tokens_seen": 2557392,
      "step": 3925
    },
    {
      "epoch": 2.059748427672956,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02999919866838387,
      "loss": 0.6396,
      "num_input_tokens_seen": 2560656,
      "step": 3930
    },
    {
      "epoch": 2.0623689727463312,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029999126185884812,
      "loss": 0.4048,
      "num_input_tokens_seen": 2563984,
      "step": 3935
    },
    {
      "epoch": 2.0649895178197064,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02999905056573968,
      "loss": 0.5729,
      "num_input_tokens_seen": 2567088,
      "step": 3940
    },
    {
      "epoch": 2.0676100628930816,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029998971807964285,
      "loss": 0.4179,
      "num_input_tokens_seen": 2570288,
      "step": 3945
    },
    {
      "epoch": 2.070230607966457,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.02999888991257511,
      "loss": 0.4958,
      "num_input_tokens_seen": 2573136,
      "step": 3950
    },
    {
      "epoch": 2.0728511530398324,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029998804879589277,
      "loss": 0.4853,
      "num_input_tokens_seen": 2576912,
      "step": 3955
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.08984375,
      "learning_rate": 0.029998716709024577,
      "loss": 0.4848,
      "num_input_tokens_seen": 2579824,
      "step": 3960
    },
    {
      "epoch": 2.0780922431865827,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02999862540089946,
      "loss": 0.476,
      "num_input_tokens_seen": 2583664,
      "step": 3965
    },
    {
      "epoch": 2.080712788259958,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02999853095523302,
      "loss": 0.4707,
      "num_input_tokens_seen": 2587344,
      "step": 3970
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.029998433372045014,
      "loss": 0.4647,
      "num_input_tokens_seen": 2590800,
      "step": 3975
    },
    {
      "epoch": 2.0859538784067087,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.02999833265135586,
      "loss": 0.6003,
      "num_input_tokens_seen": 2593872,
      "step": 3980
    },
    {
      "epoch": 2.088574423480084,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02999822879318663,
      "loss": 0.5551,
      "num_input_tokens_seen": 2596976,
      "step": 3985
    },
    {
      "epoch": 2.091194968553459,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.02999812179755904,
      "loss": 0.477,
      "num_input_tokens_seen": 2599824,
      "step": 3990
    },
    {
      "epoch": 2.0938155136268346,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029998011664495474,
      "loss": 0.5234,
      "num_input_tokens_seen": 2602992,
      "step": 3995
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.029997898394018983,
      "loss": 0.3539,
      "num_input_tokens_seen": 2606448,
      "step": 4000
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02999778198615325,
      "loss": 0.3453,
      "num_input_tokens_seen": 2610864,
      "step": 4005
    },
    {
      "epoch": 2.10167714884696,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02999766244092263,
      "loss": 0.414,
      "num_input_tokens_seen": 2614256,
      "step": 4010
    },
    {
      "epoch": 2.1042976939203353,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029997539758352128,
      "loss": 0.6266,
      "num_input_tokens_seen": 2617072,
      "step": 4015
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.029997413938467405,
      "loss": 0.457,
      "num_input_tokens_seen": 2619792,
      "step": 4020
    },
    {
      "epoch": 2.109538784067086,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02999728498129479,
      "loss": 0.3709,
      "num_input_tokens_seen": 2622448,
      "step": 4025
    },
    {
      "epoch": 2.1121593291404612,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.029997152886861257,
      "loss": 0.5139,
      "num_input_tokens_seen": 2625328,
      "step": 4030
    },
    {
      "epoch": 2.1147798742138364,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02999701765519443,
      "loss": 0.5053,
      "num_input_tokens_seen": 2628336,
      "step": 4035
    },
    {
      "epoch": 2.1174004192872116,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.029996879286322612,
      "loss": 0.4798,
      "num_input_tokens_seen": 2634128,
      "step": 4040
    },
    {
      "epoch": 2.120020964360587,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02999673778027474,
      "loss": 0.407,
      "num_input_tokens_seen": 2638672,
      "step": 4045
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029996593137080415,
      "loss": 0.5125,
      "num_input_tokens_seen": 2642768,
      "step": 4050
    },
    {
      "epoch": 2.1252620545073375,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.029996445356769895,
      "loss": 0.4559,
      "num_input_tokens_seen": 2645552,
      "step": 4055
    },
    {
      "epoch": 2.1278825995807127,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.029996294439374096,
      "loss": 0.4259,
      "num_input_tokens_seen": 2647728,
      "step": 4060
    },
    {
      "epoch": 2.130503144654088,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.029996140384924583,
      "loss": 0.557,
      "num_input_tokens_seen": 2650992,
      "step": 4065
    },
    {
      "epoch": 2.1331236897274635,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02999598319345359,
      "loss": 0.486,
      "num_input_tokens_seen": 2653744,
      "step": 4070
    },
    {
      "epoch": 2.1357442348008386,
      "grad_norm": 0.06640625,
      "learning_rate": 0.029995822864993995,
      "loss": 0.303,
      "num_input_tokens_seen": 2656336,
      "step": 4075
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02999565939957934,
      "loss": 0.7278,
      "num_input_tokens_seen": 2659216,
      "step": 4080
    },
    {
      "epoch": 2.140985324947589,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029995492797243818,
      "loss": 0.5341,
      "num_input_tokens_seen": 2662224,
      "step": 4085
    },
    {
      "epoch": 2.1436058700209646,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02999532305802228,
      "loss": 0.3648,
      "num_input_tokens_seen": 2665936,
      "step": 4090
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029995150181950227,
      "loss": 0.3555,
      "num_input_tokens_seen": 2669648,
      "step": 4095
    },
    {
      "epoch": 2.148846960167715,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.02999497416906384,
      "loss": 0.707,
      "num_input_tokens_seen": 2673552,
      "step": 4100
    },
    {
      "epoch": 2.15146750524109,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.029994795019399927,
      "loss": 0.6107,
      "num_input_tokens_seen": 2675824,
      "step": 4105
    },
    {
      "epoch": 2.1540880503144653,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.02999461273299596,
      "loss": 0.5344,
      "num_input_tokens_seen": 2678256,
      "step": 4110
    },
    {
      "epoch": 2.156708595387841,
      "grad_norm": 0.03125,
      "learning_rate": 0.02999442730989008,
      "loss": 0.4057,
      "num_input_tokens_seen": 2682224,
      "step": 4115
    },
    {
      "epoch": 2.159329140461216,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029994238750121072,
      "loss": 0.3387,
      "num_input_tokens_seen": 2684848,
      "step": 4120
    },
    {
      "epoch": 2.161949685534591,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02999404705372838,
      "loss": 0.5051,
      "num_input_tokens_seen": 2688816,
      "step": 4125
    },
    {
      "epoch": 2.1645702306079664,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02999385222075211,
      "loss": 0.8886,
      "num_input_tokens_seen": 2691728,
      "step": 4130
    },
    {
      "epoch": 2.1671907756813416,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.029993654251233008,
      "loss": 0.5833,
      "num_input_tokens_seen": 2694608,
      "step": 4135
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029993453145212495,
      "loss": 0.5375,
      "num_input_tokens_seen": 2697872,
      "step": 4140
    },
    {
      "epoch": 2.1724318658280923,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02999324890273264,
      "loss": 0.4248,
      "num_input_tokens_seen": 2701040,
      "step": 4145
    },
    {
      "epoch": 2.1750524109014675,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.029993041523836167,
      "loss": 0.6242,
      "num_input_tokens_seen": 2703760,
      "step": 4150
    },
    {
      "epoch": 2.1776729559748427,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.029992831008566454,
      "loss": 0.4661,
      "num_input_tokens_seen": 2706480,
      "step": 4155
    },
    {
      "epoch": 2.180293501048218,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029992617356967545,
      "loss": 0.3182,
      "num_input_tokens_seen": 2709488,
      "step": 4160
    },
    {
      "epoch": 2.1829140461215935,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02999240056908413,
      "loss": 0.4127,
      "num_input_tokens_seen": 2712880,
      "step": 4165
    },
    {
      "epoch": 2.1855345911949686,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029992180644961553,
      "loss": 0.4769,
      "num_input_tokens_seen": 2716112,
      "step": 4170
    },
    {
      "epoch": 2.188155136268344,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.029991957584645828,
      "loss": 0.506,
      "num_input_tokens_seen": 2719088,
      "step": 4175
    },
    {
      "epoch": 2.190775681341719,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.029991731388183614,
      "loss": 0.4602,
      "num_input_tokens_seen": 2722064,
      "step": 4180
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.02999150205562223,
      "loss": 0.3348,
      "num_input_tokens_seen": 2725072,
      "step": 4185
    },
    {
      "epoch": 2.1960167714884697,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029991269587009643,
      "loss": 0.425,
      "num_input_tokens_seen": 2728368,
      "step": 4190
    },
    {
      "epoch": 2.198637316561845,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029991033982394492,
      "loss": 0.5223,
      "num_input_tokens_seen": 2731984,
      "step": 4195
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.029990795241826056,
      "loss": 0.4526,
      "num_input_tokens_seen": 2735632,
      "step": 4200
    },
    {
      "epoch": 2.2038784067085953,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.029990553365354278,
      "loss": 0.5164,
      "num_input_tokens_seen": 2738928,
      "step": 4205
    },
    {
      "epoch": 2.2064989517819704,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.029990308353029762,
      "loss": 0.4733,
      "num_input_tokens_seen": 2743248,
      "step": 4210
    },
    {
      "epoch": 2.209119496855346,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.029990060204903756,
      "loss": 0.5574,
      "num_input_tokens_seen": 2745744,
      "step": 4215
    },
    {
      "epoch": 2.211740041928721,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.029989808921028167,
      "loss": 0.4743,
      "num_input_tokens_seen": 2748432,
      "step": 4220
    },
    {
      "epoch": 2.2143605870020964,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.029989554501455573,
      "loss": 0.5743,
      "num_input_tokens_seen": 2751280,
      "step": 4225
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02998929694623918,
      "loss": 0.5425,
      "num_input_tokens_seen": 2754096,
      "step": 4230
    },
    {
      "epoch": 2.219601677148847,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029989036255432876,
      "loss": 0.5983,
      "num_input_tokens_seen": 2757200,
      "step": 4235
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.02998877242909119,
      "loss": 0.5106,
      "num_input_tokens_seen": 2760240,
      "step": 4240
    },
    {
      "epoch": 2.2248427672955975,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.029988505467269312,
      "loss": 0.4599,
      "num_input_tokens_seen": 2763248,
      "step": 4245
    },
    {
      "epoch": 2.2274633123689727,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029988235370023084,
      "loss": 0.4614,
      "num_input_tokens_seen": 2766192,
      "step": 4250
    },
    {
      "epoch": 2.230083857442348,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029987962137409018,
      "loss": 0.4837,
      "num_input_tokens_seen": 2770160,
      "step": 4255
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029987685769484265,
      "loss": 0.402,
      "num_input_tokens_seen": 2773424,
      "step": 4260
    },
    {
      "epoch": 2.2353249475890986,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.029987406266306634,
      "loss": 0.4656,
      "num_input_tokens_seen": 2776240,
      "step": 4265
    },
    {
      "epoch": 2.237945492662474,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.029987123627934598,
      "loss": 0.4869,
      "num_input_tokens_seen": 2779088,
      "step": 4270
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02998683785442728,
      "loss": 0.4111,
      "num_input_tokens_seen": 2782032,
      "step": 4275
    },
    {
      "epoch": 2.243186582809224,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029986548945844467,
      "loss": 0.6289,
      "num_input_tokens_seen": 2786000,
      "step": 4280
    },
    {
      "epoch": 2.2458071278825997,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029986256902246586,
      "loss": 0.3771,
      "num_input_tokens_seen": 2789584,
      "step": 4285
    },
    {
      "epoch": 2.248427672955975,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.029985961723694735,
      "loss": 0.5276,
      "num_input_tokens_seen": 2792976,
      "step": 4290
    },
    {
      "epoch": 2.25104821802935,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.02998566341025066,
      "loss": 0.5043,
      "num_input_tokens_seen": 2795856,
      "step": 4295
    },
    {
      "epoch": 2.2536687631027252,
      "grad_norm": 0.060546875,
      "learning_rate": 0.02998536196197676,
      "loss": 0.4782,
      "num_input_tokens_seen": 2798384,
      "step": 4300
    },
    {
      "epoch": 2.2562893081761004,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029985057378936107,
      "loss": 0.3614,
      "num_input_tokens_seen": 2801552,
      "step": 4305
    },
    {
      "epoch": 2.258909853249476,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.029984749661192404,
      "loss": 0.5482,
      "num_input_tokens_seen": 2804816,
      "step": 4310
    },
    {
      "epoch": 2.261530398322851,
      "grad_norm": 0.083984375,
      "learning_rate": 0.029984438808810028,
      "loss": 0.7371,
      "num_input_tokens_seen": 2807536,
      "step": 4315
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029984124821854006,
      "loss": 0.4624,
      "num_input_tokens_seen": 2812080,
      "step": 4320
    },
    {
      "epoch": 2.2667714884696015,
      "grad_norm": 0.087890625,
      "learning_rate": 0.029983807700390018,
      "loss": 0.5051,
      "num_input_tokens_seen": 2814768,
      "step": 4325
    },
    {
      "epoch": 2.269392033542977,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.029983487444484407,
      "loss": 0.5346,
      "num_input_tokens_seen": 2817328,
      "step": 4330
    },
    {
      "epoch": 2.2720125786163523,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02998316405420416,
      "loss": 0.5952,
      "num_input_tokens_seen": 2820624,
      "step": 4335
    },
    {
      "epoch": 2.2746331236897275,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.029982837529616932,
      "loss": 0.3826,
      "num_input_tokens_seen": 2824720,
      "step": 4340
    },
    {
      "epoch": 2.2772536687631026,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029982507870791025,
      "loss": 0.5743,
      "num_input_tokens_seen": 2828560,
      "step": 4345
    },
    {
      "epoch": 2.279874213836478,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029982175077795395,
      "loss": 0.5434,
      "num_input_tokens_seen": 2832080,
      "step": 4350
    },
    {
      "epoch": 2.2824947589098534,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.02998183915069967,
      "loss": 0.5687,
      "num_input_tokens_seen": 2834640,
      "step": 4355
    },
    {
      "epoch": 2.2851153039832286,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02998150008957412,
      "loss": 0.4937,
      "num_input_tokens_seen": 2837136,
      "step": 4360
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.029981157894489664,
      "loss": 0.4765,
      "num_input_tokens_seen": 2841040,
      "step": 4365
    },
    {
      "epoch": 2.290356394129979,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.029980812565517892,
      "loss": 0.5238,
      "num_input_tokens_seen": 2843984,
      "step": 4370
    },
    {
      "epoch": 2.2929769392033545,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02998046410273104,
      "loss": 0.6429,
      "num_input_tokens_seen": 2846832,
      "step": 4375
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02998011250620201,
      "loss": 0.3944,
      "num_input_tokens_seen": 2850512,
      "step": 4380
    },
    {
      "epoch": 2.298218029350105,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02997975777600434,
      "loss": 0.3617,
      "num_input_tokens_seen": 2854800,
      "step": 4385
    },
    {
      "epoch": 2.30083857442348,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02997939991221224,
      "loss": 0.3972,
      "num_input_tokens_seen": 2857552,
      "step": 4390
    },
    {
      "epoch": 2.3034591194968552,
      "grad_norm": 0.0546875,
      "learning_rate": 0.029979038914900575,
      "loss": 0.4889,
      "num_input_tokens_seen": 2860432,
      "step": 4395
    },
    {
      "epoch": 2.3060796645702304,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029978674784144857,
      "loss": 0.5561,
      "num_input_tokens_seen": 2863568,
      "step": 4400
    },
    {
      "epoch": 2.308700209643606,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02997830752002126,
      "loss": 0.3991,
      "num_input_tokens_seen": 2865968,
      "step": 4405
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.029977937122606615,
      "loss": 0.6362,
      "num_input_tokens_seen": 2868848,
      "step": 4410
    },
    {
      "epoch": 2.3139412997903563,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.029977563591978403,
      "loss": 0.4233,
      "num_input_tokens_seen": 2871792,
      "step": 4415
    },
    {
      "epoch": 2.3165618448637315,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029977186928214754,
      "loss": 0.4769,
      "num_input_tokens_seen": 2874832,
      "step": 4420
    },
    {
      "epoch": 2.319182389937107,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.029976807131394476,
      "loss": 0.4925,
      "num_input_tokens_seen": 2877648,
      "step": 4425
    },
    {
      "epoch": 2.3218029350104823,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.029976424201597004,
      "loss": 0.5112,
      "num_input_tokens_seen": 2880016,
      "step": 4430
    },
    {
      "epoch": 2.3244234800838575,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.029976038138902452,
      "loss": 0.5999,
      "num_input_tokens_seen": 2883184,
      "step": 4435
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02997564894339157,
      "loss": 0.3938,
      "num_input_tokens_seen": 2887408,
      "step": 4440
    },
    {
      "epoch": 2.329664570230608,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02997525661514579,
      "loss": 0.2845,
      "num_input_tokens_seen": 2892912,
      "step": 4445
    },
    {
      "epoch": 2.3322851153039834,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02997486115424717,
      "loss": 0.5118,
      "num_input_tokens_seen": 2895504,
      "step": 4450
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.029974462560778434,
      "loss": 0.5792,
      "num_input_tokens_seen": 2900208,
      "step": 4455
    },
    {
      "epoch": 2.3375262054507338,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.029974060834822975,
      "loss": 0.3833,
      "num_input_tokens_seen": 2903568,
      "step": 4460
    },
    {
      "epoch": 2.340146750524109,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.029973655976464824,
      "loss": 0.5759,
      "num_input_tokens_seen": 2906864,
      "step": 4465
    },
    {
      "epoch": 2.342767295597484,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.029973247985788664,
      "loss": 0.4274,
      "num_input_tokens_seen": 2910480,
      "step": 4470
    },
    {
      "epoch": 2.3453878406708597,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02997283686287985,
      "loss": 0.4237,
      "num_input_tokens_seen": 2913552,
      "step": 4475
    },
    {
      "epoch": 2.348008385744235,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029972422607824383,
      "loss": 0.5446,
      "num_input_tokens_seen": 2916368,
      "step": 4480
    },
    {
      "epoch": 2.35062893081761,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029972005220708926,
      "loss": 0.497,
      "num_input_tokens_seen": 2918960,
      "step": 4485
    },
    {
      "epoch": 2.353249475890985,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02997158470162078,
      "loss": 0.4806,
      "num_input_tokens_seen": 2922768,
      "step": 4490
    },
    {
      "epoch": 2.3558700209643604,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029971161050647926,
      "loss": 0.4907,
      "num_input_tokens_seen": 2925840,
      "step": 4495
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02997073426787898,
      "loss": 0.4018,
      "num_input_tokens_seen": 2929008,
      "step": 4500
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.029970304353403216,
      "loss": 0.4895,
      "num_input_tokens_seen": 2931632,
      "step": 4505
    },
    {
      "epoch": 2.3637316561844863,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02996987130731057,
      "loss": 0.5524,
      "num_input_tokens_seen": 2935824,
      "step": 4510
    },
    {
      "epoch": 2.3663522012578615,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.02996943512969164,
      "loss": 0.3676,
      "num_input_tokens_seen": 2938832,
      "step": 4515
    },
    {
      "epoch": 2.368972746331237,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029968995820637653,
      "loss": 0.4352,
      "num_input_tokens_seen": 2942096,
      "step": 4520
    },
    {
      "epoch": 2.3715932914046123,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02996855338024052,
      "loss": 0.4644,
      "num_input_tokens_seen": 2945968,
      "step": 4525
    },
    {
      "epoch": 2.3742138364779874,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.029968107808592787,
      "loss": 0.539,
      "num_input_tokens_seen": 2949456,
      "step": 4530
    },
    {
      "epoch": 2.3768343815513626,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02996765910578767,
      "loss": 0.4954,
      "num_input_tokens_seen": 2952592,
      "step": 4535
    },
    {
      "epoch": 2.379454926624738,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029967207271919025,
      "loss": 0.349,
      "num_input_tokens_seen": 2955760,
      "step": 4540
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.029966752307081373,
      "loss": 0.2953,
      "num_input_tokens_seen": 2958768,
      "step": 4545
    },
    {
      "epoch": 2.3846960167714886,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.029966294211369888,
      "loss": 0.3892,
      "num_input_tokens_seen": 2961552,
      "step": 4550
    },
    {
      "epoch": 2.3873165618448637,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0299658329848804,
      "loss": 0.4279,
      "num_input_tokens_seen": 2964464,
      "step": 4555
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 0.04296875,
      "learning_rate": 0.029965368627709394,
      "loss": 0.5385,
      "num_input_tokens_seen": 2966832,
      "step": 4560
    },
    {
      "epoch": 2.392557651991614,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029964901139954,
      "loss": 0.4626,
      "num_input_tokens_seen": 2969968,
      "step": 4565
    },
    {
      "epoch": 2.3951781970649897,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02996443052171202,
      "loss": 0.424,
      "num_input_tokens_seen": 2973040,
      "step": 4570
    },
    {
      "epoch": 2.397798742138365,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029963956773081897,
      "loss": 0.5263,
      "num_input_tokens_seen": 2977520,
      "step": 4575
    },
    {
      "epoch": 2.40041928721174,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.02996347989416273,
      "loss": 0.3948,
      "num_input_tokens_seen": 2980880,
      "step": 4580
    },
    {
      "epoch": 2.403039832285115,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02996299988505429,
      "loss": 0.5989,
      "num_input_tokens_seen": 2983632,
      "step": 4585
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.029962516745856975,
      "loss": 0.4583,
      "num_input_tokens_seen": 2986256,
      "step": 4590
    },
    {
      "epoch": 2.408280922431866,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02996203047667186,
      "loss": 0.4384,
      "num_input_tokens_seen": 2991056,
      "step": 4595
    },
    {
      "epoch": 2.410901467505241,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029961541077600663,
      "loss": 0.366,
      "num_input_tokens_seen": 2994320,
      "step": 4600
    },
    {
      "epoch": 2.4135220125786163,
      "grad_norm": 0.0625,
      "learning_rate": 0.029961048548745768,
      "loss": 0.3979,
      "num_input_tokens_seen": 2997104,
      "step": 4605
    },
    {
      "epoch": 2.4161425576519915,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029960552890210195,
      "loss": 0.4912,
      "num_input_tokens_seen": 3000112,
      "step": 4610
    },
    {
      "epoch": 2.418763102725367,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02996005410209764,
      "loss": 0.4338,
      "num_input_tokens_seen": 3003568,
      "step": 4615
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029959552184512437,
      "loss": 0.4138,
      "num_input_tokens_seen": 3007696,
      "step": 4620
    },
    {
      "epoch": 2.4240041928721174,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029959047137559588,
      "loss": 0.5331,
      "num_input_tokens_seen": 3011472,
      "step": 4625
    },
    {
      "epoch": 2.4266247379454926,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02995853896134474,
      "loss": 0.4367,
      "num_input_tokens_seen": 3014320,
      "step": 4630
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.029958027655974193,
      "loss": 0.597,
      "num_input_tokens_seen": 3018224,
      "step": 4635
    },
    {
      "epoch": 2.431865828092243,
      "grad_norm": 0.044921875,
      "learning_rate": 0.02995751322155492,
      "loss": 0.5204,
      "num_input_tokens_seen": 3021072,
      "step": 4640
    },
    {
      "epoch": 2.4344863731656186,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029956995658194516,
      "loss": 0.3565,
      "num_input_tokens_seen": 3023600,
      "step": 4645
    },
    {
      "epoch": 2.4371069182389937,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02995647496600126,
      "loss": 0.4756,
      "num_input_tokens_seen": 3026768,
      "step": 4650
    },
    {
      "epoch": 2.439727463312369,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029955951145084078,
      "loss": 0.5042,
      "num_input_tokens_seen": 3033424,
      "step": 4655
    },
    {
      "epoch": 2.442348008385744,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029955424195552537,
      "loss": 0.4781,
      "num_input_tokens_seen": 3035952,
      "step": 4660
    },
    {
      "epoch": 2.4449685534591197,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.02995489411751688,
      "loss": 0.4717,
      "num_input_tokens_seen": 3039216,
      "step": 4665
    },
    {
      "epoch": 2.447589098532495,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.029954360911087986,
      "loss": 0.6898,
      "num_input_tokens_seen": 3042224,
      "step": 4670
    },
    {
      "epoch": 2.45020964360587,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029953824576377394,
      "loss": 0.4116,
      "num_input_tokens_seen": 3044752,
      "step": 4675
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02995328511349731,
      "loss": 0.5622,
      "num_input_tokens_seen": 3048336,
      "step": 4680
    },
    {
      "epoch": 2.4554507337526204,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.029952742522560576,
      "loss": 0.4357,
      "num_input_tokens_seen": 3050864,
      "step": 4685
    },
    {
      "epoch": 2.458071278825996,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.02995219680368069,
      "loss": 0.465,
      "num_input_tokens_seen": 3053712,
      "step": 4690
    },
    {
      "epoch": 2.460691823899371,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029951647956971823,
      "loss": 0.6642,
      "num_input_tokens_seen": 3056944,
      "step": 4695
    },
    {
      "epoch": 2.4633123689727463,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.029951095982548772,
      "loss": 0.5046,
      "num_input_tokens_seen": 3060176,
      "step": 4700
    },
    {
      "epoch": 2.4659329140461215,
      "grad_norm": 0.056640625,
      "learning_rate": 0.02995054088052702,
      "loss": 0.6933,
      "num_input_tokens_seen": 3062352,
      "step": 4705
    },
    {
      "epoch": 2.468553459119497,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.02994998265102268,
      "loss": 0.4724,
      "num_input_tokens_seen": 3064848,
      "step": 4710
    },
    {
      "epoch": 2.4711740041928723,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02994942129415253,
      "loss": 0.5329,
      "num_input_tokens_seen": 3068368,
      "step": 4715
    },
    {
      "epoch": 2.4737945492662474,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.02994885681003399,
      "loss": 0.5756,
      "num_input_tokens_seen": 3071536,
      "step": 4720
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.029948289198785154,
      "loss": 0.3691,
      "num_input_tokens_seen": 3074224,
      "step": 4725
    },
    {
      "epoch": 2.4790356394129978,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02994771846052476,
      "loss": 0.5271,
      "num_input_tokens_seen": 3077744,
      "step": 4730
    },
    {
      "epoch": 2.481656184486373,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.029947144595372194,
      "loss": 0.477,
      "num_input_tokens_seen": 3080656,
      "step": 4735
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.0299465676034475,
      "loss": 0.4918,
      "num_input_tokens_seen": 3084624,
      "step": 4740
    },
    {
      "epoch": 2.4868972746331237,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.02994598748487139,
      "loss": 0.4435,
      "num_input_tokens_seen": 3087760,
      "step": 4745
    },
    {
      "epoch": 2.489517819706499,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029945404239765206,
      "loss": 0.5189,
      "num_input_tokens_seen": 3091440,
      "step": 4750
    },
    {
      "epoch": 2.492138364779874,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.029944817868250966,
      "loss": 0.4671,
      "num_input_tokens_seen": 3094192,
      "step": 4755
    },
    {
      "epoch": 2.4947589098532497,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02994422837045133,
      "loss": 0.4062,
      "num_input_tokens_seen": 3097584,
      "step": 4760
    },
    {
      "epoch": 2.497379454926625,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.029943635746489602,
      "loss": 0.4883,
      "num_input_tokens_seen": 3101232,
      "step": 4765
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.029943039996489767,
      "loss": 0.3233,
      "num_input_tokens_seen": 3103888,
      "step": 4770
    },
    {
      "epoch": 2.502620545073375,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.029942441120576445,
      "loss": 0.3744,
      "num_input_tokens_seen": 3111408,
      "step": 4775
    },
    {
      "epoch": 2.5052410901467503,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02994183911887491,
      "loss": 0.3954,
      "num_input_tokens_seen": 3114288,
      "step": 4780
    },
    {
      "epoch": 2.507861635220126,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0299412339915111,
      "loss": 0.5478,
      "num_input_tokens_seen": 3117296,
      "step": 4785
    },
    {
      "epoch": 2.510482180293501,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.029940625738611594,
      "loss": 0.3563,
      "num_input_tokens_seen": 3119792,
      "step": 4790
    },
    {
      "epoch": 2.5131027253668763,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02994001436030364,
      "loss": 0.5217,
      "num_input_tokens_seen": 3122544,
      "step": 4795
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02993939985671512,
      "loss": 0.4204,
      "num_input_tokens_seen": 3125648,
      "step": 4800
    },
    {
      "epoch": 2.518343815513627,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.029938782227974595,
      "loss": 0.4266,
      "num_input_tokens_seen": 3128400,
      "step": 4805
    },
    {
      "epoch": 2.5209643605870022,
      "grad_norm": 0.064453125,
      "learning_rate": 0.029938161474211255,
      "loss": 0.4271,
      "num_input_tokens_seen": 3131280,
      "step": 4810
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029937537595554956,
      "loss": 0.4984,
      "num_input_tokens_seen": 3134288,
      "step": 4815
    },
    {
      "epoch": 2.5262054507337526,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.02993691059213621,
      "loss": 0.4947,
      "num_input_tokens_seen": 3136784,
      "step": 4820
    },
    {
      "epoch": 2.5288259958071277,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02993628046408618,
      "loss": 0.4949,
      "num_input_tokens_seen": 3139664,
      "step": 4825
    },
    {
      "epoch": 2.531446540880503,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.029935647211536676,
      "loss": 0.708,
      "num_input_tokens_seen": 3142896,
      "step": 4830
    },
    {
      "epoch": 2.5340670859538785,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02993501083462017,
      "loss": 0.524,
      "num_input_tokens_seen": 3146224,
      "step": 4835
    },
    {
      "epoch": 2.5366876310272537,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029934371333469784,
      "loss": 0.5242,
      "num_input_tokens_seen": 3149136,
      "step": 4840
    },
    {
      "epoch": 2.539308176100629,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0299337287082193,
      "loss": 0.5164,
      "num_input_tokens_seen": 3151984,
      "step": 4845
    },
    {
      "epoch": 2.541928721174004,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02993308295900314,
      "loss": 0.5267,
      "num_input_tokens_seen": 3155440,
      "step": 4850
    },
    {
      "epoch": 2.5445492662473796,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.029932434085956392,
      "loss": 0.367,
      "num_input_tokens_seen": 3159216,
      "step": 4855
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.02993178208921479,
      "loss": 0.5807,
      "num_input_tokens_seen": 3162640,
      "step": 4860
    },
    {
      "epoch": 2.54979035639413,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02993112696891473,
      "loss": 0.4312,
      "num_input_tokens_seen": 3165968,
      "step": 4865
    },
    {
      "epoch": 2.552410901467505,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029930468725193244,
      "loss": 0.4218,
      "num_input_tokens_seen": 3168688,
      "step": 4870
    },
    {
      "epoch": 2.5550314465408803,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02992980735818804,
      "loss": 0.5589,
      "num_input_tokens_seen": 3172400,
      "step": 4875
    },
    {
      "epoch": 2.5576519916142555,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.029929142868037467,
      "loss": 0.5846,
      "num_input_tokens_seen": 3175472,
      "step": 4880
    },
    {
      "epoch": 2.560272536687631,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.029928475254880522,
      "loss": 0.715,
      "num_input_tokens_seen": 3178032,
      "step": 4885
    },
    {
      "epoch": 2.5628930817610063,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029927804518856873,
      "loss": 0.3621,
      "num_input_tokens_seen": 3181744,
      "step": 4890
    },
    {
      "epoch": 2.5655136268343814,
      "grad_norm": 0.04296875,
      "learning_rate": 0.02992713066010682,
      "loss": 0.4486,
      "num_input_tokens_seen": 3184656,
      "step": 4895
    },
    {
      "epoch": 2.568134171907757,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02992645367877133,
      "loss": 0.4591,
      "num_input_tokens_seen": 3187504,
      "step": 4900
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.029925773574992025,
      "loss": 0.5069,
      "num_input_tokens_seen": 3190608,
      "step": 4905
    },
    {
      "epoch": 2.5733752620545074,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029925090348911165,
      "loss": 0.4923,
      "num_input_tokens_seen": 3193584,
      "step": 4910
    },
    {
      "epoch": 2.5759958071278826,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02992440400067168,
      "loss": 0.6106,
      "num_input_tokens_seen": 3197488,
      "step": 4915
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.029923714530417145,
      "loss": 0.5733,
      "num_input_tokens_seen": 3200592,
      "step": 4920
    },
    {
      "epoch": 2.581236897274633,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.02992302193829179,
      "loss": 0.4519,
      "num_input_tokens_seen": 3204208,
      "step": 4925
    },
    {
      "epoch": 2.5838574423480085,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.029922326224440497,
      "loss": 0.5686,
      "num_input_tokens_seen": 3208048,
      "step": 4930
    },
    {
      "epoch": 2.5864779874213837,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0299216273890088,
      "loss": 0.4355,
      "num_input_tokens_seen": 3210576,
      "step": 4935
    },
    {
      "epoch": 2.589098532494759,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029920925432142885,
      "loss": 0.3789,
      "num_input_tokens_seen": 3213648,
      "step": 4940
    },
    {
      "epoch": 2.591719077568134,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.029920220353989595,
      "loss": 0.6727,
      "num_input_tokens_seen": 3216592,
      "step": 4945
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.029919512154696435,
      "loss": 0.4487,
      "num_input_tokens_seen": 3219760,
      "step": 4950
    },
    {
      "epoch": 2.596960167714885,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.029918800834411532,
      "loss": 0.6883,
      "num_input_tokens_seen": 3223600,
      "step": 4955
    },
    {
      "epoch": 2.59958071278826,
      "grad_norm": 0.052734375,
      "learning_rate": 0.029918086393283708,
      "loss": 0.588,
      "num_input_tokens_seen": 3226416,
      "step": 4960
    },
    {
      "epoch": 2.602201257861635,
      "grad_norm": 0.02734375,
      "learning_rate": 0.0299173688314624,
      "loss": 0.3942,
      "num_input_tokens_seen": 3230288,
      "step": 4965
    },
    {
      "epoch": 2.6048218029350103,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.029916648149097714,
      "loss": 0.5145,
      "num_input_tokens_seen": 3233104,
      "step": 4970
    },
    {
      "epoch": 2.6074423480083855,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02991592434634042,
      "loss": 0.5239,
      "num_input_tokens_seen": 3237872,
      "step": 4975
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02991519742334192,
      "loss": 0.4564,
      "num_input_tokens_seen": 3242064,
      "step": 4980
    },
    {
      "epoch": 2.6126834381551363,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02991446738025428,
      "loss": 0.4618,
      "num_input_tokens_seen": 3245712,
      "step": 4985
    },
    {
      "epoch": 2.6153039832285114,
      "grad_norm": 0.04296875,
      "learning_rate": 0.02991373421723022,
      "loss": 0.4821,
      "num_input_tokens_seen": 3248560,
      "step": 4990
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.029912997934423105,
      "loss": 0.4944,
      "num_input_tokens_seen": 3251440,
      "step": 4995
    },
    {
      "epoch": 2.620545073375262,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029912258531986962,
      "loss": 0.4663,
      "num_input_tokens_seen": 3254640,
      "step": 5000
    },
    {
      "epoch": 2.6231656184486374,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029911516010076456,
      "loss": 0.4026,
      "num_input_tokens_seen": 3258320,
      "step": 5005
    },
    {
      "epoch": 2.6257861635220126,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02991077036884692,
      "loss": 0.434,
      "num_input_tokens_seen": 3261168,
      "step": 5010
    },
    {
      "epoch": 2.6284067085953877,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02991002160845434,
      "loss": 0.5593,
      "num_input_tokens_seen": 3264816,
      "step": 5015
    },
    {
      "epoch": 2.631027253668763,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02990926972905534,
      "loss": 0.3663,
      "num_input_tokens_seen": 3268080,
      "step": 5020
    },
    {
      "epoch": 2.6336477987421385,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029908514730807207,
      "loss": 0.515,
      "num_input_tokens_seen": 3272528,
      "step": 5025
    },
    {
      "epoch": 2.6362683438155137,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.029907756613867874,
      "loss": 0.5018,
      "num_input_tokens_seen": 3275888,
      "step": 5030
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.029906995378395935,
      "loss": 0.4129,
      "num_input_tokens_seen": 3281072,
      "step": 5035
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02990623102455063,
      "loss": 0.4867,
      "num_input_tokens_seen": 3285808,
      "step": 5040
    },
    {
      "epoch": 2.6441299790356396,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02990546355249186,
      "loss": 0.4116,
      "num_input_tokens_seen": 3289168,
      "step": 5045
    },
    {
      "epoch": 2.646750524109015,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.029904692962380162,
      "loss": 0.6347,
      "num_input_tokens_seen": 3292336,
      "step": 5050
    },
    {
      "epoch": 2.64937106918239,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029903919254376736,
      "loss": 0.4329,
      "num_input_tokens_seen": 3295024,
      "step": 5055
    },
    {
      "epoch": 2.651991614255765,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.029903142428643434,
      "loss": 0.6418,
      "num_input_tokens_seen": 3297776,
      "step": 5060
    },
    {
      "epoch": 2.6546121593291403,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029902362485342764,
      "loss": 0.3843,
      "num_input_tokens_seen": 3300560,
      "step": 5065
    },
    {
      "epoch": 2.6572327044025155,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029901579424637878,
      "loss": 0.4849,
      "num_input_tokens_seen": 3304432,
      "step": 5070
    },
    {
      "epoch": 2.659853249475891,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02990079324669258,
      "loss": 0.4609,
      "num_input_tokens_seen": 3307344,
      "step": 5075
    },
    {
      "epoch": 2.6624737945492662,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029900003951671333,
      "loss": 0.4902,
      "num_input_tokens_seen": 3311056,
      "step": 5080
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029899211539739245,
      "loss": 0.4987,
      "num_input_tokens_seen": 3313712,
      "step": 5085
    },
    {
      "epoch": 2.667714884696017,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029898416011062082,
      "loss": 0.4508,
      "num_input_tokens_seen": 3317232,
      "step": 5090
    },
    {
      "epoch": 2.670335429769392,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029897617365806265,
      "loss": 0.6138,
      "num_input_tokens_seen": 3320048,
      "step": 5095
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02989681560413885,
      "loss": 0.4501,
      "num_input_tokens_seen": 3324720,
      "step": 5100
    },
    {
      "epoch": 2.6755765199161425,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.029896010726227565,
      "loss": 0.4662,
      "num_input_tokens_seen": 3327696,
      "step": 5105
    },
    {
      "epoch": 2.6781970649895177,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029895202732240778,
      "loss": 0.363,
      "num_input_tokens_seen": 3330608,
      "step": 5110
    },
    {
      "epoch": 2.680817610062893,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.029894391622347513,
      "loss": 0.4771,
      "num_input_tokens_seen": 3333296,
      "step": 5115
    },
    {
      "epoch": 2.6834381551362685,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.029893577396717448,
      "loss": 0.4415,
      "num_input_tokens_seen": 3337552,
      "step": 5120
    },
    {
      "epoch": 2.6860587002096437,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029892760055520905,
      "loss": 0.4681,
      "num_input_tokens_seen": 3341296,
      "step": 5125
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02989193959892886,
      "loss": 0.5093,
      "num_input_tokens_seen": 3343920,
      "step": 5130
    },
    {
      "epoch": 2.691299790356394,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.02989111602711295,
      "loss": 0.4427,
      "num_input_tokens_seen": 3346736,
      "step": 5135
    },
    {
      "epoch": 2.6939203354297696,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029890289340245454,
      "loss": 0.4841,
      "num_input_tokens_seen": 3350320,
      "step": 5140
    },
    {
      "epoch": 2.6965408805031448,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.029889459538499304,
      "loss": 0.4179,
      "num_input_tokens_seen": 3353552,
      "step": 5145
    },
    {
      "epoch": 2.69916142557652,
      "grad_norm": 0.056640625,
      "learning_rate": 0.029888626622048088,
      "loss": 0.5132,
      "num_input_tokens_seen": 3357136,
      "step": 5150
    },
    {
      "epoch": 2.701781970649895,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.029887790591066044,
      "loss": 0.4378,
      "num_input_tokens_seen": 3360048,
      "step": 5155
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.029886951445728054,
      "loss": 0.4034,
      "num_input_tokens_seen": 3363280,
      "step": 5160
    },
    {
      "epoch": 2.7070230607966455,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.02988610918620966,
      "loss": 0.6168,
      "num_input_tokens_seen": 3366000,
      "step": 5165
    },
    {
      "epoch": 2.709643605870021,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029885263812687055,
      "loss": 0.5106,
      "num_input_tokens_seen": 3369648,
      "step": 5170
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.02988441532533708,
      "loss": 0.4467,
      "num_input_tokens_seen": 3372528,
      "step": 5175
    },
    {
      "epoch": 2.7148846960167714,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.029883563724337227,
      "loss": 0.5084,
      "num_input_tokens_seen": 3375728,
      "step": 5180
    },
    {
      "epoch": 2.717505241090147,
      "grad_norm": 0.064453125,
      "learning_rate": 0.02988270900986565,
      "loss": 0.4275,
      "num_input_tokens_seen": 3378864,
      "step": 5185
    },
    {
      "epoch": 2.720125786163522,
      "grad_norm": 0.044921875,
      "learning_rate": 0.029881851182101134,
      "loss": 0.4144,
      "num_input_tokens_seen": 3381936,
      "step": 5190
    },
    {
      "epoch": 2.7227463312368974,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02988099024122313,
      "loss": 0.5174,
      "num_input_tokens_seen": 3384464,
      "step": 5195
    },
    {
      "epoch": 2.7253668763102725,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029880126187411746,
      "loss": 0.5606,
      "num_input_tokens_seen": 3387312,
      "step": 5200
    },
    {
      "epoch": 2.7279874213836477,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.02987925902084772,
      "loss": 0.4397,
      "num_input_tokens_seen": 3391376,
      "step": 5205
    },
    {
      "epoch": 2.730607966457023,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02987838874171246,
      "loss": 0.4945,
      "num_input_tokens_seen": 3394352,
      "step": 5210
    },
    {
      "epoch": 2.7332285115303985,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.029877515350188018,
      "loss": 0.5368,
      "num_input_tokens_seen": 3397456,
      "step": 5215
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.029876638846457096,
      "loss": 0.5436,
      "num_input_tokens_seen": 3403792,
      "step": 5220
    },
    {
      "epoch": 2.738469601677149,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.029875759230703045,
      "loss": 0.4954,
      "num_input_tokens_seen": 3407184,
      "step": 5225
    },
    {
      "epoch": 2.741090146750524,
      "grad_norm": 0.03125,
      "learning_rate": 0.02987487650310988,
      "loss": 0.5204,
      "num_input_tokens_seen": 3410256,
      "step": 5230
    },
    {
      "epoch": 2.7437106918238996,
      "grad_norm": 0.044921875,
      "learning_rate": 0.029873990663862246,
      "loss": 0.5519,
      "num_input_tokens_seen": 3413168,
      "step": 5235
    },
    {
      "epoch": 2.7463312368972748,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.02987310171314546,
      "loss": 0.6418,
      "num_input_tokens_seen": 3415728,
      "step": 5240
    },
    {
      "epoch": 2.74895178197065,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.029872209651145478,
      "loss": 0.4071,
      "num_input_tokens_seen": 3419568,
      "step": 5245
    },
    {
      "epoch": 2.751572327044025,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02987131447804891,
      "loss": 0.3641,
      "num_input_tokens_seen": 3422736,
      "step": 5250
    },
    {
      "epoch": 2.7541928721174003,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029870416194043008,
      "loss": 0.4326,
      "num_input_tokens_seen": 3427472,
      "step": 5255
    },
    {
      "epoch": 2.7568134171907754,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02986951479931569,
      "loss": 0.3951,
      "num_input_tokens_seen": 3430256,
      "step": 5260
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02986861029405552,
      "loss": 0.58,
      "num_input_tokens_seen": 3433328,
      "step": 5265
    },
    {
      "epoch": 2.762054507337526,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.029867702678451704,
      "loss": 0.4312,
      "num_input_tokens_seen": 3437456,
      "step": 5270
    },
    {
      "epoch": 2.7646750524109014,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.02986679195269411,
      "loss": 0.3778,
      "num_input_tokens_seen": 3440240,
      "step": 5275
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.029865878116973246,
      "loss": 0.4677,
      "num_input_tokens_seen": 3443824,
      "step": 5280
    },
    {
      "epoch": 2.769916142557652,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02986496117148028,
      "loss": 0.5402,
      "num_input_tokens_seen": 3447184,
      "step": 5285
    },
    {
      "epoch": 2.7725366876310273,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.029864041116407027,
      "loss": 0.4211,
      "num_input_tokens_seen": 3450320,
      "step": 5290
    },
    {
      "epoch": 2.7751572327044025,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029863117951945947,
      "loss": 0.5906,
      "num_input_tokens_seen": 3454640,
      "step": 5295
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02986219167829016,
      "loss": 0.4572,
      "num_input_tokens_seen": 3458000,
      "step": 5300
    },
    {
      "epoch": 2.780398322851153,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.029861262295633434,
      "loss": 0.5872,
      "num_input_tokens_seen": 3461296,
      "step": 5305
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02986032980417018,
      "loss": 0.4065,
      "num_input_tokens_seen": 3464336,
      "step": 5310
    },
    {
      "epoch": 2.7856394129979036,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.029859394204095464,
      "loss": 0.572,
      "num_input_tokens_seen": 3467184,
      "step": 5315
    },
    {
      "epoch": 2.788259958071279,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.029858455495605007,
      "loss": 0.5497,
      "num_input_tokens_seen": 3470320,
      "step": 5320
    },
    {
      "epoch": 2.790880503144654,
      "grad_norm": 0.0546875,
      "learning_rate": 0.029857513678895175,
      "loss": 0.399,
      "num_input_tokens_seen": 3474128,
      "step": 5325
    },
    {
      "epoch": 2.7935010482180296,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.029856568754162986,
      "loss": 0.6136,
      "num_input_tokens_seen": 3476336,
      "step": 5330
    },
    {
      "epoch": 2.7961215932914047,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.029855620721606104,
      "loss": 0.5256,
      "num_input_tokens_seen": 3480016,
      "step": 5335
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.029854669581422848,
      "loss": 0.4232,
      "num_input_tokens_seen": 3482608,
      "step": 5340
    },
    {
      "epoch": 2.801362683438155,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02985371533381219,
      "loss": 0.7033,
      "num_input_tokens_seen": 3485200,
      "step": 5345
    },
    {
      "epoch": 2.8039832285115303,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029852757978973743,
      "loss": 0.3626,
      "num_input_tokens_seen": 3488336,
      "step": 5350
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029851797517107776,
      "loss": 0.5462,
      "num_input_tokens_seen": 3492080,
      "step": 5355
    },
    {
      "epoch": 2.809224318658281,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029850833948415208,
      "loss": 0.4292,
      "num_input_tokens_seen": 3495952,
      "step": 5360
    },
    {
      "epoch": 2.811844863731656,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0298498672730976,
      "loss": 0.4308,
      "num_input_tokens_seen": 3498224,
      "step": 5365
    },
    {
      "epoch": 2.8144654088050314,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.02984889749135718,
      "loss": 0.5542,
      "num_input_tokens_seen": 3501872,
      "step": 5370
    },
    {
      "epoch": 2.8170859538784065,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02984792460339681,
      "loss": 0.4018,
      "num_input_tokens_seen": 3505008,
      "step": 5375
    },
    {
      "epoch": 2.819706498951782,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029846948609420006,
      "loss": 0.3729,
      "num_input_tokens_seen": 3508016,
      "step": 5380
    },
    {
      "epoch": 2.8223270440251573,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.029845969509630935,
      "loss": 0.5165,
      "num_input_tokens_seen": 3512912,
      "step": 5385
    },
    {
      "epoch": 2.8249475890985325,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029844987304234415,
      "loss": 0.4689,
      "num_input_tokens_seen": 3516112,
      "step": 5390
    },
    {
      "epoch": 2.8275681341719077,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029844001993435912,
      "loss": 0.5824,
      "num_input_tokens_seen": 3518800,
      "step": 5395
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02984301357744154,
      "loss": 0.3976,
      "num_input_tokens_seen": 3522608,
      "step": 5400
    },
    {
      "epoch": 2.832809224318658,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.029842022056458067,
      "loss": 0.5026,
      "num_input_tokens_seen": 3526512,
      "step": 5405
    },
    {
      "epoch": 2.8354297693920336,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.029841027430692908,
      "loss": 0.4231,
      "num_input_tokens_seen": 3529456,
      "step": 5410
    },
    {
      "epoch": 2.838050314465409,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02984002970035412,
      "loss": 0.4337,
      "num_input_tokens_seen": 3532464,
      "step": 5415
    },
    {
      "epoch": 2.840670859538784,
      "grad_norm": 0.056640625,
      "learning_rate": 0.029839028865650426,
      "loss": 0.4662,
      "num_input_tokens_seen": 3535824,
      "step": 5420
    },
    {
      "epoch": 2.8432914046121596,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029838024926791187,
      "loss": 0.4882,
      "num_input_tokens_seen": 3539696,
      "step": 5425
    },
    {
      "epoch": 2.8459119496855347,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029837017883986416,
      "loss": 0.4928,
      "num_input_tokens_seen": 3543408,
      "step": 5430
    },
    {
      "epoch": 2.84853249475891,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.029836007737446768,
      "loss": 0.3987,
      "num_input_tokens_seen": 3546864,
      "step": 5435
    },
    {
      "epoch": 2.851153039832285,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.029834994487383564,
      "loss": 0.4554,
      "num_input_tokens_seen": 3549904,
      "step": 5440
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02983397813400876,
      "loss": 0.4428,
      "num_input_tokens_seen": 3553072,
      "step": 5445
    },
    {
      "epoch": 2.8563941299790354,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029832958677534962,
      "loss": 0.3354,
      "num_input_tokens_seen": 3556784,
      "step": 5450
    },
    {
      "epoch": 2.859014675052411,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029831936118175435,
      "loss": 0.5025,
      "num_input_tokens_seen": 3561232,
      "step": 5455
    },
    {
      "epoch": 2.861635220125786,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02983091045614408,
      "loss": 0.5025,
      "num_input_tokens_seen": 3564048,
      "step": 5460
    },
    {
      "epoch": 2.8642557651991614,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02982988169165546,
      "loss": 0.3866,
      "num_input_tokens_seen": 3567664,
      "step": 5465
    },
    {
      "epoch": 2.8668763102725365,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02982884982492478,
      "loss": 0.4685,
      "num_input_tokens_seen": 3571344,
      "step": 5470
    },
    {
      "epoch": 2.869496855345912,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029827814856167888,
      "loss": 0.4363,
      "num_input_tokens_seen": 3574128,
      "step": 5475
    },
    {
      "epoch": 2.8721174004192873,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.0298267767856013,
      "loss": 0.4435,
      "num_input_tokens_seen": 3577552,
      "step": 5480
    },
    {
      "epoch": 2.8747379454926625,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029825735613442154,
      "loss": 0.6547,
      "num_input_tokens_seen": 3581040,
      "step": 5485
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02982469133990826,
      "loss": 0.4362,
      "num_input_tokens_seen": 3583760,
      "step": 5490
    },
    {
      "epoch": 2.879979035639413,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.029823643965218073,
      "loss": 0.4954,
      "num_input_tokens_seen": 3587632,
      "step": 5495
    },
    {
      "epoch": 2.882599580712788,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.029822593489590682,
      "loss": 0.4955,
      "num_input_tokens_seen": 3590960,
      "step": 5500
    },
    {
      "epoch": 2.8852201257861636,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029821539913245838,
      "loss": 0.3855,
      "num_input_tokens_seen": 3594672,
      "step": 5505
    },
    {
      "epoch": 2.8878406708595388,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029820483236403933,
      "loss": 0.4264,
      "num_input_tokens_seen": 3597872,
      "step": 5510
    },
    {
      "epoch": 2.890461215932914,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029819423459286024,
      "loss": 0.3604,
      "num_input_tokens_seen": 3600528,
      "step": 5515
    },
    {
      "epoch": 2.8930817610062896,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029818360582113786,
      "loss": 0.5146,
      "num_input_tokens_seen": 3603344,
      "step": 5520
    },
    {
      "epoch": 2.8957023060796647,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02981729460510958,
      "loss": 0.4524,
      "num_input_tokens_seen": 3606448,
      "step": 5525
    },
    {
      "epoch": 2.89832285115304,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029816225528496384,
      "loss": 0.6585,
      "num_input_tokens_seen": 3609776,
      "step": 5530
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.02981515335249784,
      "loss": 0.4881,
      "num_input_tokens_seen": 3612080,
      "step": 5535
    },
    {
      "epoch": 2.9035639412997902,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.029814078077338234,
      "loss": 0.5327,
      "num_input_tokens_seen": 3615504,
      "step": 5540
    },
    {
      "epoch": 2.9061844863731654,
      "grad_norm": 0.072265625,
      "learning_rate": 0.029812999703242502,
      "loss": 0.6683,
      "num_input_tokens_seen": 3618096,
      "step": 5545
    },
    {
      "epoch": 2.908805031446541,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02981191823043623,
      "loss": 0.4242,
      "num_input_tokens_seen": 3621296,
      "step": 5550
    },
    {
      "epoch": 2.911425576519916,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.029810833659145646,
      "loss": 0.4679,
      "num_input_tokens_seen": 3624784,
      "step": 5555
    },
    {
      "epoch": 2.9140461215932913,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02980974598959763,
      "loss": 0.5537,
      "num_input_tokens_seen": 3628336,
      "step": 5560
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.029808655222019715,
      "loss": 0.5799,
      "num_input_tokens_seen": 3631280,
      "step": 5565
    },
    {
      "epoch": 2.919287211740042,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02980756135664007,
      "loss": 0.6284,
      "num_input_tokens_seen": 3634448,
      "step": 5570
    },
    {
      "epoch": 2.9219077568134173,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029806464393687525,
      "loss": 0.4633,
      "num_input_tokens_seen": 3637072,
      "step": 5575
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02980536433339155,
      "loss": 0.6316,
      "num_input_tokens_seen": 3640880,
      "step": 5580
    },
    {
      "epoch": 2.9271488469601676,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029804261175982262,
      "loss": 0.497,
      "num_input_tokens_seen": 3643312,
      "step": 5585
    },
    {
      "epoch": 2.929769392033543,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029803154921690438,
      "loss": 0.4928,
      "num_input_tokens_seen": 3647152,
      "step": 5590
    },
    {
      "epoch": 2.932389937106918,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.029802045570747483,
      "loss": 0.4836,
      "num_input_tokens_seen": 3649904,
      "step": 5595
    },
    {
      "epoch": 2.9350104821802936,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029800933123385463,
      "loss": 0.3899,
      "num_input_tokens_seen": 3653680,
      "step": 5600
    },
    {
      "epoch": 2.9376310272536688,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029799817579837094,
      "loss": 0.5662,
      "num_input_tokens_seen": 3657328,
      "step": 5605
    },
    {
      "epoch": 2.940251572327044,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029798698940335735,
      "loss": 0.4828,
      "num_input_tokens_seen": 3661008,
      "step": 5610
    },
    {
      "epoch": 2.9428721174004195,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02979757720511539,
      "loss": 0.4746,
      "num_input_tokens_seen": 3664144,
      "step": 5615
    },
    {
      "epoch": 2.9454926624737947,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.029796452374410712,
      "loss": 0.3994,
      "num_input_tokens_seen": 3666800,
      "step": 5620
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029795324448457002,
      "loss": 0.4508,
      "num_input_tokens_seen": 3669936,
      "step": 5625
    },
    {
      "epoch": 2.950733752620545,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029794193427490213,
      "loss": 0.4052,
      "num_input_tokens_seen": 3673584,
      "step": 5630
    },
    {
      "epoch": 2.95335429769392,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02979305931174694,
      "loss": 0.4537,
      "num_input_tokens_seen": 3676816,
      "step": 5635
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.029791922101464426,
      "loss": 0.5643,
      "num_input_tokens_seen": 3679792,
      "step": 5640
    },
    {
      "epoch": 2.958595387840671,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029790781796880567,
      "loss": 0.4764,
      "num_input_tokens_seen": 3682704,
      "step": 5645
    },
    {
      "epoch": 2.961215932914046,
      "grad_norm": 0.0625,
      "learning_rate": 0.029789638398233897,
      "loss": 0.4578,
      "num_input_tokens_seen": 3685968,
      "step": 5650
    },
    {
      "epoch": 2.9638364779874213,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.029788491905763603,
      "loss": 0.4967,
      "num_input_tokens_seen": 3688528,
      "step": 5655
    },
    {
      "epoch": 2.9664570230607965,
      "grad_norm": 0.05078125,
      "learning_rate": 0.029787342319709517,
      "loss": 0.4894,
      "num_input_tokens_seen": 3692464,
      "step": 5660
    },
    {
      "epoch": 2.969077568134172,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029786189640312124,
      "loss": 0.3908,
      "num_input_tokens_seen": 3695920,
      "step": 5665
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029785033867812543,
      "loss": 0.6254,
      "num_input_tokens_seen": 3699600,
      "step": 5670
    },
    {
      "epoch": 2.9743186582809225,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029783875002452557,
      "loss": 0.512,
      "num_input_tokens_seen": 3703024,
      "step": 5675
    },
    {
      "epoch": 2.9769392033542976,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.029782713044474583,
      "loss": 0.3729,
      "num_input_tokens_seen": 3706544,
      "step": 5680
    },
    {
      "epoch": 2.979559748427673,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02978154799412169,
      "loss": 0.5955,
      "num_input_tokens_seen": 3709584,
      "step": 5685
    },
    {
      "epoch": 2.982180293501048,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029780379851637593,
      "loss": 0.5127,
      "num_input_tokens_seen": 3713616,
      "step": 5690
    },
    {
      "epoch": 2.9848008385744236,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029779208617266656,
      "loss": 0.5045,
      "num_input_tokens_seen": 3716656,
      "step": 5695
    },
    {
      "epoch": 2.9874213836477987,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029778034291253884,
      "loss": 0.3629,
      "num_input_tokens_seen": 3719504,
      "step": 5700
    },
    {
      "epoch": 2.990041928721174,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.029776856873844935,
      "loss": 0.4532,
      "num_input_tokens_seen": 3722672,
      "step": 5705
    },
    {
      "epoch": 2.9926624737945495,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029775676365286115,
      "loss": 0.4705,
      "num_input_tokens_seen": 3725104,
      "step": 5710
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029774492765824365,
      "loss": 0.6153,
      "num_input_tokens_seen": 3728272,
      "step": 5715
    },
    {
      "epoch": 2.9979035639413,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029773306075707287,
      "loss": 0.4189,
      "num_input_tokens_seen": 3731344,
      "step": 5720
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.4576379656791687,
      "eval_runtime": 13.6505,
      "eval_samples_per_second": 62.122,
      "eval_steps_per_second": 15.531,
      "num_input_tokens_seen": 3733736,
      "step": 5724
    },
    {
      "epoch": 3.000524109014675,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.029772116295183122,
      "loss": 0.4397,
      "num_input_tokens_seen": 3734280,
      "step": 5725
    },
    {
      "epoch": 3.00314465408805,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029770923424500752,
      "loss": 0.3955,
      "num_input_tokens_seen": 3736776,
      "step": 5730
    },
    {
      "epoch": 3.0057651991614254,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029769727463909722,
      "loss": 0.4551,
      "num_input_tokens_seen": 3739496,
      "step": 5735
    },
    {
      "epoch": 3.008385744234801,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.029768528413660206,
      "loss": 0.3876,
      "num_input_tokens_seen": 3741992,
      "step": 5740
    },
    {
      "epoch": 3.011006289308176,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.029767326274003036,
      "loss": 0.5296,
      "num_input_tokens_seen": 3744488,
      "step": 5745
    },
    {
      "epoch": 3.0136268343815513,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02976612104518968,
      "loss": 0.5327,
      "num_input_tokens_seen": 3747688,
      "step": 5750
    },
    {
      "epoch": 3.0162473794549265,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02976491272747227,
      "loss": 0.4643,
      "num_input_tokens_seen": 3753064,
      "step": 5755
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02976370132110356,
      "loss": 0.529,
      "num_input_tokens_seen": 3756136,
      "step": 5760
    },
    {
      "epoch": 3.0214884696016773,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02976248682633697,
      "loss": 0.5269,
      "num_input_tokens_seen": 3758888,
      "step": 5765
    },
    {
      "epoch": 3.0241090146750524,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02976126924342655,
      "loss": 0.3632,
      "num_input_tokens_seen": 3761832,
      "step": 5770
    },
    {
      "epoch": 3.0267295597484276,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029760048572627013,
      "loss": 0.4608,
      "num_input_tokens_seen": 3765032,
      "step": 5775
    },
    {
      "epoch": 3.029350104821803,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.029758824814193705,
      "loss": 0.5529,
      "num_input_tokens_seen": 3767944,
      "step": 5780
    },
    {
      "epoch": 3.0319706498951784,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.02975759796838262,
      "loss": 0.4461,
      "num_input_tokens_seen": 3771112,
      "step": 5785
    },
    {
      "epoch": 3.0345911949685536,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02975636803545041,
      "loss": 0.5739,
      "num_input_tokens_seen": 3773960,
      "step": 5790
    },
    {
      "epoch": 3.0372117400419287,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029755135015654348,
      "loss": 0.5551,
      "num_input_tokens_seen": 3777192,
      "step": 5795
    },
    {
      "epoch": 3.039832285115304,
      "grad_norm": 0.052734375,
      "learning_rate": 0.029753898909252386,
      "loss": 0.5129,
      "num_input_tokens_seen": 3780488,
      "step": 5800
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02975265971650309,
      "loss": 0.4646,
      "num_input_tokens_seen": 3784328,
      "step": 5805
    },
    {
      "epoch": 3.0450733752620547,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.029751417437665687,
      "loss": 0.478,
      "num_input_tokens_seen": 3786696,
      "step": 5810
    },
    {
      "epoch": 3.04769392033543,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02975017207300005,
      "loss": 0.5414,
      "num_input_tokens_seen": 3789768,
      "step": 5815
    },
    {
      "epoch": 3.050314465408805,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029748923622766696,
      "loss": 0.4923,
      "num_input_tokens_seen": 3792936,
      "step": 5820
    },
    {
      "epoch": 3.05293501048218,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029747672087226777,
      "loss": 0.5234,
      "num_input_tokens_seen": 3796168,
      "step": 5825
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.056640625,
      "learning_rate": 0.029746417466642115,
      "loss": 0.4837,
      "num_input_tokens_seen": 3798696,
      "step": 5830
    },
    {
      "epoch": 3.058176100628931,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.029745159761275158,
      "loss": 0.4163,
      "num_input_tokens_seen": 3801992,
      "step": 5835
    },
    {
      "epoch": 3.060796645702306,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029743898971388995,
      "loss": 0.3773,
      "num_input_tokens_seen": 3804648,
      "step": 5840
    },
    {
      "epoch": 3.0634171907756813,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029742635097247373,
      "loss": 0.6427,
      "num_input_tokens_seen": 3808296,
      "step": 5845
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 0.029296875,
      "learning_rate": 0.029741368139114686,
      "loss": 0.4999,
      "num_input_tokens_seen": 3811688,
      "step": 5850
    },
    {
      "epoch": 3.068658280922432,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02974009809725596,
      "loss": 0.4414,
      "num_input_tokens_seen": 3817416,
      "step": 5855
    },
    {
      "epoch": 3.0712788259958073,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02973882497193688,
      "loss": 0.4684,
      "num_input_tokens_seen": 3820168,
      "step": 5860
    },
    {
      "epoch": 3.0738993710691824,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02973754876342376,
      "loss": 0.4221,
      "num_input_tokens_seen": 3823720,
      "step": 5865
    },
    {
      "epoch": 3.0765199161425576,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.029736269471983576,
      "loss": 0.4857,
      "num_input_tokens_seen": 3827432,
      "step": 5870
    },
    {
      "epoch": 3.0791404612159328,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029734987097883938,
      "loss": 0.507,
      "num_input_tokens_seen": 3829832,
      "step": 5875
    },
    {
      "epoch": 3.0817610062893084,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.029733701641393107,
      "loss": 0.4459,
      "num_input_tokens_seen": 3832040,
      "step": 5880
    },
    {
      "epoch": 3.0843815513626835,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029732413102779978,
      "loss": 0.4699,
      "num_input_tokens_seen": 3834984,
      "step": 5885
    },
    {
      "epoch": 3.0870020964360587,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02973112148231411,
      "loss": 0.4919,
      "num_input_tokens_seen": 3841800,
      "step": 5890
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.029729826780265686,
      "loss": 0.4554,
      "num_input_tokens_seen": 3844456,
      "step": 5895
    },
    {
      "epoch": 3.092243186582809,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02972852899690555,
      "loss": 0.3547,
      "num_input_tokens_seen": 3847912,
      "step": 5900
    },
    {
      "epoch": 3.0948637316561847,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029727228132505175,
      "loss": 0.3392,
      "num_input_tokens_seen": 3851208,
      "step": 5905
    },
    {
      "epoch": 3.09748427672956,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029725924187336696,
      "loss": 0.3562,
      "num_input_tokens_seen": 3854920,
      "step": 5910
    },
    {
      "epoch": 3.100104821802935,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029724617161672875,
      "loss": 0.4602,
      "num_input_tokens_seen": 3858408,
      "step": 5915
    },
    {
      "epoch": 3.10272536687631,
      "grad_norm": 0.0859375,
      "learning_rate": 0.029723307055787136,
      "loss": 0.5061,
      "num_input_tokens_seen": 3861352,
      "step": 5920
    },
    {
      "epoch": 3.1053459119496853,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.02972199386995353,
      "loss": 0.3849,
      "num_input_tokens_seen": 3864680,
      "step": 5925
    },
    {
      "epoch": 3.107966457023061,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029720677604446764,
      "loss": 0.4532,
      "num_input_tokens_seen": 3867176,
      "step": 5930
    },
    {
      "epoch": 3.110587002096436,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.029719358259542188,
      "loss": 0.463,
      "num_input_tokens_seen": 3869864,
      "step": 5935
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02971803583551579,
      "loss": 0.4491,
      "num_input_tokens_seen": 3872744,
      "step": 5940
    },
    {
      "epoch": 3.1158280922431865,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.029716710332644204,
      "loss": 0.4644,
      "num_input_tokens_seen": 3875880,
      "step": 5945
    },
    {
      "epoch": 3.1184486373165616,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02971538175120472,
      "loss": 0.3823,
      "num_input_tokens_seen": 3879048,
      "step": 5950
    },
    {
      "epoch": 3.1210691823899372,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02971405009147525,
      "loss": 0.3183,
      "num_input_tokens_seen": 3882376,
      "step": 5955
    },
    {
      "epoch": 3.1236897274633124,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029712715353734372,
      "loss": 0.3891,
      "num_input_tokens_seen": 3885576,
      "step": 5960
    },
    {
      "epoch": 3.1263102725366876,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.02971137753826129,
      "loss": 0.4473,
      "num_input_tokens_seen": 3888808,
      "step": 5965
    },
    {
      "epoch": 3.1289308176100628,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.029710036645335863,
      "loss": 0.4473,
      "num_input_tokens_seen": 3891976,
      "step": 5970
    },
    {
      "epoch": 3.131551362683438,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029708692675238593,
      "loss": 0.4623,
      "num_input_tokens_seen": 3895592,
      "step": 5975
    },
    {
      "epoch": 3.1341719077568135,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02970734562825062,
      "loss": 0.6225,
      "num_input_tokens_seen": 3899336,
      "step": 5980
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.029705995504653734,
      "loss": 0.4176,
      "num_input_tokens_seen": 3902280,
      "step": 5985
    },
    {
      "epoch": 3.139412997903564,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029704642304730362,
      "loss": 0.4515,
      "num_input_tokens_seen": 3905672,
      "step": 5990
    },
    {
      "epoch": 3.142033542976939,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029703286028763577,
      "loss": 0.4147,
      "num_input_tokens_seen": 3908392,
      "step": 5995
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0297019266770371,
      "loss": 0.5341,
      "num_input_tokens_seen": 3912200,
      "step": 6000
    },
    {
      "epoch": 3.14727463312369,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02970056424983529,
      "loss": 0.5432,
      "num_input_tokens_seen": 3915304,
      "step": 6005
    },
    {
      "epoch": 3.149895178197065,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.029699198747443156,
      "loss": 0.5186,
      "num_input_tokens_seen": 3918952,
      "step": 6010
    },
    {
      "epoch": 3.15251572327044,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029697830170146336,
      "loss": 0.5993,
      "num_input_tokens_seen": 3921864,
      "step": 6015
    },
    {
      "epoch": 3.1551362683438153,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02969645851823113,
      "loss": 0.405,
      "num_input_tokens_seen": 3925256,
      "step": 6020
    },
    {
      "epoch": 3.157756813417191,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029695083791984465,
      "loss": 0.4099,
      "num_input_tokens_seen": 3927432,
      "step": 6025
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029693705991693922,
      "loss": 0.4345,
      "num_input_tokens_seen": 3931368,
      "step": 6030
    },
    {
      "epoch": 3.1629979035639413,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.029692325117647723,
      "loss": 0.4172,
      "num_input_tokens_seen": 3934504,
      "step": 6035
    },
    {
      "epoch": 3.1656184486373165,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029690941170134724,
      "loss": 0.4298,
      "num_input_tokens_seen": 3937384,
      "step": 6040
    },
    {
      "epoch": 3.1682389937106916,
      "grad_norm": 0.0546875,
      "learning_rate": 0.02968955414944444,
      "loss": 0.584,
      "num_input_tokens_seen": 3940168,
      "step": 6045
    },
    {
      "epoch": 3.1708595387840672,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029688164055867012,
      "loss": 0.4088,
      "num_input_tokens_seen": 3942504,
      "step": 6050
    },
    {
      "epoch": 3.1734800838574424,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029686770889693233,
      "loss": 0.4857,
      "num_input_tokens_seen": 3945480,
      "step": 6055
    },
    {
      "epoch": 3.1761006289308176,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02968537465121454,
      "loss": 0.4576,
      "num_input_tokens_seen": 3948520,
      "step": 6060
    },
    {
      "epoch": 3.1787211740041927,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029683975340723014,
      "loss": 0.5516,
      "num_input_tokens_seen": 3952904,
      "step": 6065
    },
    {
      "epoch": 3.181341719077568,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02968257295851137,
      "loss": 0.468,
      "num_input_tokens_seen": 3955816,
      "step": 6070
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 0.15625,
      "learning_rate": 0.029681167504872968,
      "loss": 0.4804,
      "num_input_tokens_seen": 3958120,
      "step": 6075
    },
    {
      "epoch": 3.1865828092243187,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.029679758980101816,
      "loss": 0.4949,
      "num_input_tokens_seen": 3960520,
      "step": 6080
    },
    {
      "epoch": 3.189203354297694,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.02967834738449256,
      "loss": 0.4795,
      "num_input_tokens_seen": 3963208,
      "step": 6085
    },
    {
      "epoch": 3.191823899371069,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.02967693271834049,
      "loss": 0.5295,
      "num_input_tokens_seen": 3966568,
      "step": 6090
    },
    {
      "epoch": 3.1944444444444446,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02967551498194154,
      "loss": 0.5336,
      "num_input_tokens_seen": 3970184,
      "step": 6095
    },
    {
      "epoch": 3.19706498951782,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.029674094175592285,
      "loss": 0.5097,
      "num_input_tokens_seen": 3973192,
      "step": 6100
    },
    {
      "epoch": 3.199685534591195,
      "grad_norm": 0.048828125,
      "learning_rate": 0.029672670299589934,
      "loss": 0.4211,
      "num_input_tokens_seen": 3975656,
      "step": 6105
    },
    {
      "epoch": 3.20230607966457,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02967124335423236,
      "loss": 0.4189,
      "num_input_tokens_seen": 3979240,
      "step": 6110
    },
    {
      "epoch": 3.2049266247379453,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029669813339818045,
      "loss": 0.4449,
      "num_input_tokens_seen": 3983464,
      "step": 6115
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.029668380256646146,
      "loss": 0.5122,
      "num_input_tokens_seen": 3986568,
      "step": 6120
    },
    {
      "epoch": 3.210167714884696,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029666944105016444,
      "loss": 0.4492,
      "num_input_tokens_seen": 3989224,
      "step": 6125
    },
    {
      "epoch": 3.2127882599580713,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029665504885229364,
      "loss": 0.6062,
      "num_input_tokens_seen": 3992648,
      "step": 6130
    },
    {
      "epoch": 3.2154088050314464,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.029664062597585972,
      "loss": 0.4138,
      "num_input_tokens_seen": 3995528,
      "step": 6135
    },
    {
      "epoch": 3.2180293501048216,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029662617242387983,
      "loss": 0.452,
      "num_input_tokens_seen": 3998536,
      "step": 6140
    },
    {
      "epoch": 3.220649895178197,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.029661168819937748,
      "loss": 0.4861,
      "num_input_tokens_seen": 4001224,
      "step": 6145
    },
    {
      "epoch": 3.2232704402515724,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02965971733053826,
      "loss": 0.4214,
      "num_input_tokens_seen": 4004328,
      "step": 6150
    },
    {
      "epoch": 3.2258909853249476,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029658262774493148,
      "loss": 0.4527,
      "num_input_tokens_seen": 4007944,
      "step": 6155
    },
    {
      "epoch": 3.2285115303983227,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0296568051521067,
      "loss": 0.4671,
      "num_input_tokens_seen": 4011144,
      "step": 6160
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029655344463683828,
      "loss": 0.4137,
      "num_input_tokens_seen": 4014376,
      "step": 6165
    },
    {
      "epoch": 3.2337526205450735,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029653880709530088,
      "loss": 0.469,
      "num_input_tokens_seen": 4018216,
      "step": 6170
    },
    {
      "epoch": 3.2363731656184487,
      "grad_norm": 0.052734375,
      "learning_rate": 0.02965241388995168,
      "loss": 0.498,
      "num_input_tokens_seen": 4021032,
      "step": 6175
    },
    {
      "epoch": 3.238993710691824,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.029650944005255458,
      "loss": 0.5773,
      "num_input_tokens_seen": 4024040,
      "step": 6180
    },
    {
      "epoch": 3.241614255765199,
      "grad_norm": 0.046875,
      "learning_rate": 0.029649471055748892,
      "loss": 0.3273,
      "num_input_tokens_seen": 4027336,
      "step": 6185
    },
    {
      "epoch": 3.2442348008385746,
      "grad_norm": 0.044921875,
      "learning_rate": 0.029647995041740115,
      "loss": 0.5372,
      "num_input_tokens_seen": 4030472,
      "step": 6190
    },
    {
      "epoch": 3.24685534591195,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029646515963537887,
      "loss": 0.4573,
      "num_input_tokens_seen": 4034312,
      "step": 6195
    },
    {
      "epoch": 3.249475890985325,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02964503382145161,
      "loss": 0.4542,
      "num_input_tokens_seen": 4038632,
      "step": 6200
    },
    {
      "epoch": 3.2520964360587,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02964354861579134,
      "loss": 0.5335,
      "num_input_tokens_seen": 4042824,
      "step": 6205
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 0.07421875,
      "learning_rate": 0.02964206034686777,
      "loss": 0.4707,
      "num_input_tokens_seen": 4045896,
      "step": 6210
    },
    {
      "epoch": 3.257337526205451,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02964056901499221,
      "loss": 0.517,
      "num_input_tokens_seen": 4048712,
      "step": 6215
    },
    {
      "epoch": 3.259958071278826,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029639074620476647,
      "loss": 0.5094,
      "num_input_tokens_seen": 4051176,
      "step": 6220
    },
    {
      "epoch": 3.2625786163522013,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02963757716363368,
      "loss": 0.3513,
      "num_input_tokens_seen": 4054056,
      "step": 6225
    },
    {
      "epoch": 3.2651991614255764,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02963607664477657,
      "loss": 0.4219,
      "num_input_tokens_seen": 4058120,
      "step": 6230
    },
    {
      "epoch": 3.2678197064989516,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.029634573064219197,
      "loss": 0.5121,
      "num_input_tokens_seen": 4060840,
      "step": 6235
    },
    {
      "epoch": 3.270440251572327,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029633066422276106,
      "loss": 0.532,
      "num_input_tokens_seen": 4064136,
      "step": 6240
    },
    {
      "epoch": 3.2730607966457024,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029631556719262455,
      "loss": 0.4675,
      "num_input_tokens_seen": 4066824,
      "step": 6245
    },
    {
      "epoch": 3.2756813417190775,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.02963004395549407,
      "loss": 0.6084,
      "num_input_tokens_seen": 4069864,
      "step": 6250
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029628528131287395,
      "loss": 0.3644,
      "num_input_tokens_seen": 4072936,
      "step": 6255
    },
    {
      "epoch": 3.280922431865828,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02962700924695953,
      "loss": 0.6302,
      "num_input_tokens_seen": 4075400,
      "step": 6260
    },
    {
      "epoch": 3.2835429769392035,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.029625487302828196,
      "loss": 0.4956,
      "num_input_tokens_seen": 4078696,
      "step": 6265
    },
    {
      "epoch": 3.2861635220125787,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.029623962299211784,
      "loss": 0.486,
      "num_input_tokens_seen": 4082824,
      "step": 6270
    },
    {
      "epoch": 3.288784067085954,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.029622434236429292,
      "loss": 0.5961,
      "num_input_tokens_seen": 4085448,
      "step": 6275
    },
    {
      "epoch": 3.291404612159329,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.029620903114800378,
      "loss": 0.5204,
      "num_input_tokens_seen": 4088424,
      "step": 6280
    },
    {
      "epoch": 3.2940251572327046,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02961936893464534,
      "loss": 0.5605,
      "num_input_tokens_seen": 4091752,
      "step": 6285
    },
    {
      "epoch": 3.29664570230608,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.02961783169628511,
      "loss": 0.4722,
      "num_input_tokens_seen": 4094568,
      "step": 6290
    },
    {
      "epoch": 3.299266247379455,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029616291400041253,
      "loss": 0.4873,
      "num_input_tokens_seen": 4097352,
      "step": 6295
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02961474804623599,
      "loss": 0.3933,
      "num_input_tokens_seen": 4100328,
      "step": 6300
    },
    {
      "epoch": 3.3045073375262053,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.029613201635192165,
      "loss": 0.4216,
      "num_input_tokens_seen": 4102728,
      "step": 6305
    },
    {
      "epoch": 3.307127882599581,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.029611652167233278,
      "loss": 0.3725,
      "num_input_tokens_seen": 4105352,
      "step": 6310
    },
    {
      "epoch": 3.309748427672956,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029610099642683456,
      "loss": 0.4069,
      "num_input_tokens_seen": 4108744,
      "step": 6315
    },
    {
      "epoch": 3.3123689727463312,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029608544061867467,
      "loss": 0.3599,
      "num_input_tokens_seen": 4112168,
      "step": 6320
    },
    {
      "epoch": 3.3149895178197064,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029606985425110727,
      "loss": 0.634,
      "num_input_tokens_seen": 4115240,
      "step": 6325
    },
    {
      "epoch": 3.3176100628930816,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029605423732739278,
      "loss": 0.4714,
      "num_input_tokens_seen": 4117864,
      "step": 6330
    },
    {
      "epoch": 3.320230607966457,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.029603858985079817,
      "loss": 0.4714,
      "num_input_tokens_seen": 4121960,
      "step": 6335
    },
    {
      "epoch": 3.3228511530398324,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02960229118245966,
      "loss": 0.5396,
      "num_input_tokens_seen": 4125160,
      "step": 6340
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.029600720325206782,
      "loss": 0.5338,
      "num_input_tokens_seen": 4128168,
      "step": 6345
    },
    {
      "epoch": 3.3280922431865827,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.029599146413649784,
      "loss": 0.5623,
      "num_input_tokens_seen": 4134632,
      "step": 6350
    },
    {
      "epoch": 3.330712788259958,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029597569448117915,
      "loss": 0.5175,
      "num_input_tokens_seen": 4138088,
      "step": 6355
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.052734375,
      "learning_rate": 0.029595989428941056,
      "loss": 0.4234,
      "num_input_tokens_seen": 4141608,
      "step": 6360
    },
    {
      "epoch": 3.3359538784067087,
      "grad_norm": 0.03125,
      "learning_rate": 0.029594406356449724,
      "loss": 0.4084,
      "num_input_tokens_seen": 4144712,
      "step": 6365
    },
    {
      "epoch": 3.338574423480084,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.029592820230975085,
      "loss": 0.3968,
      "num_input_tokens_seen": 4147624,
      "step": 6370
    },
    {
      "epoch": 3.341194968553459,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029591231052848936,
      "loss": 0.3731,
      "num_input_tokens_seen": 4150728,
      "step": 6375
    },
    {
      "epoch": 3.3438155136268346,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02958963882240372,
      "loss": 0.4415,
      "num_input_tokens_seen": 4153928,
      "step": 6380
    },
    {
      "epoch": 3.3464360587002098,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.029588043539972513,
      "loss": 0.4015,
      "num_input_tokens_seen": 4157608,
      "step": 6385
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02958644520588902,
      "loss": 0.4346,
      "num_input_tokens_seen": 4160648,
      "step": 6390
    },
    {
      "epoch": 3.35167714884696,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0295848438204876,
      "loss": 0.5015,
      "num_input_tokens_seen": 4163720,
      "step": 6395
    },
    {
      "epoch": 3.3542976939203353,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02958323938410325,
      "loss": 0.4512,
      "num_input_tokens_seen": 4166920,
      "step": 6400
    },
    {
      "epoch": 3.3569182389937104,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029581631897071595,
      "loss": 0.3406,
      "num_input_tokens_seen": 4170024,
      "step": 6405
    },
    {
      "epoch": 3.359538784067086,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0295800213597289,
      "loss": 0.3909,
      "num_input_tokens_seen": 4173288,
      "step": 6410
    },
    {
      "epoch": 3.3621593291404612,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.029578407772412076,
      "loss": 0.3652,
      "num_input_tokens_seen": 4175688,
      "step": 6415
    },
    {
      "epoch": 3.3647798742138364,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.029576791135458667,
      "loss": 0.416,
      "num_input_tokens_seen": 4178504,
      "step": 6420
    },
    {
      "epoch": 3.3674004192872116,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.02957517144920685,
      "loss": 0.3962,
      "num_input_tokens_seen": 4180840,
      "step": 6425
    },
    {
      "epoch": 3.370020964360587,
      "grad_norm": 0.029296875,
      "learning_rate": 0.029573548713995455,
      "loss": 0.4452,
      "num_input_tokens_seen": 4184328,
      "step": 6430
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029571922930163926,
      "loss": 0.5871,
      "num_input_tokens_seen": 4188072,
      "step": 6435
    },
    {
      "epoch": 3.3752620545073375,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029570294098052365,
      "loss": 0.4413,
      "num_input_tokens_seen": 4191848,
      "step": 6440
    },
    {
      "epoch": 3.3778825995807127,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029568662218001508,
      "loss": 0.4215,
      "num_input_tokens_seen": 4194440,
      "step": 6445
    },
    {
      "epoch": 3.380503144654088,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029567027290352724,
      "loss": 0.5271,
      "num_input_tokens_seen": 4197256,
      "step": 6450
    },
    {
      "epoch": 3.3831236897274635,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.029565389315448016,
      "loss": 0.4638,
      "num_input_tokens_seen": 4200872,
      "step": 6455
    },
    {
      "epoch": 3.3857442348008386,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02956374829363004,
      "loss": 0.4277,
      "num_input_tokens_seen": 4204232,
      "step": 6460
    },
    {
      "epoch": 3.388364779874214,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.029562104225242067,
      "loss": 0.4407,
      "num_input_tokens_seen": 4207304,
      "step": 6465
    },
    {
      "epoch": 3.390985324947589,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.029560457110628026,
      "loss": 0.4182,
      "num_input_tokens_seen": 4209832,
      "step": 6470
    },
    {
      "epoch": 3.3936058700209646,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029558806950132473,
      "loss": 0.4363,
      "num_input_tokens_seen": 4213128,
      "step": 6475
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0295571537441006,
      "loss": 0.3698,
      "num_input_tokens_seen": 4215976,
      "step": 6480
    },
    {
      "epoch": 3.398846960167715,
      "grad_norm": 0.0546875,
      "learning_rate": 0.02955549749287824,
      "loss": 0.6401,
      "num_input_tokens_seen": 4218632,
      "step": 6485
    },
    {
      "epoch": 3.40146750524109,
      "grad_norm": 0.0703125,
      "learning_rate": 0.029553838196811868,
      "loss": 0.498,
      "num_input_tokens_seen": 4221416,
      "step": 6490
    },
    {
      "epoch": 3.4040880503144653,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02955217585624858,
      "loss": 0.3636,
      "num_input_tokens_seen": 4224392,
      "step": 6495
    },
    {
      "epoch": 3.4067085953878404,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.02955051047153612,
      "loss": 0.4361,
      "num_input_tokens_seen": 4227176,
      "step": 6500
    },
    {
      "epoch": 3.409329140461216,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029548842043022872,
      "loss": 0.514,
      "num_input_tokens_seen": 4229960,
      "step": 6505
    },
    {
      "epoch": 3.411949685534591,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.029547170571057855,
      "loss": 0.4132,
      "num_input_tokens_seen": 4232136,
      "step": 6510
    },
    {
      "epoch": 3.4145702306079664,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029545496055990713,
      "loss": 0.5321,
      "num_input_tokens_seen": 4235624,
      "step": 6515
    },
    {
      "epoch": 3.4171907756813416,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02954381849817174,
      "loss": 0.5313,
      "num_input_tokens_seen": 4240616,
      "step": 6520
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029542137897951862,
      "loss": 0.4475,
      "num_input_tokens_seen": 4243528,
      "step": 6525
    },
    {
      "epoch": 3.4224318658280923,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029540454255682642,
      "loss": 0.5,
      "num_input_tokens_seen": 4247208,
      "step": 6530
    },
    {
      "epoch": 3.4250524109014675,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.029538767571716276,
      "loss": 0.3623,
      "num_input_tokens_seen": 4252424,
      "step": 6535
    },
    {
      "epoch": 3.4276729559748427,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.029537077846405604,
      "loss": 0.504,
      "num_input_tokens_seen": 4255752,
      "step": 6540
    },
    {
      "epoch": 3.430293501048218,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029535385080104094,
      "loss": 0.5099,
      "num_input_tokens_seen": 4258472,
      "step": 6545
    },
    {
      "epoch": 3.4329140461215935,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.029533689273165853,
      "loss": 0.3812,
      "num_input_tokens_seen": 4260904,
      "step": 6550
    },
    {
      "epoch": 3.4355345911949686,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029531990425945627,
      "loss": 0.5651,
      "num_input_tokens_seen": 4264104,
      "step": 6555
    },
    {
      "epoch": 3.438155136268344,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029530288538798792,
      "loss": 0.5494,
      "num_input_tokens_seen": 4268648,
      "step": 6560
    },
    {
      "epoch": 3.440775681341719,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.029528583612081362,
      "loss": 0.5224,
      "num_input_tokens_seen": 4272776,
      "step": 6565
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029526875646149992,
      "loss": 0.4786,
      "num_input_tokens_seen": 4275784,
      "step": 6570
    },
    {
      "epoch": 3.4460167714884697,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.029525164641361974,
      "loss": 0.5905,
      "num_input_tokens_seen": 4279400,
      "step": 6575
    },
    {
      "epoch": 3.448637316561845,
      "grad_norm": 0.03125,
      "learning_rate": 0.02952345059807522,
      "loss": 0.4038,
      "num_input_tokens_seen": 4282920,
      "step": 6580
    },
    {
      "epoch": 3.45125786163522,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02952173351664829,
      "loss": 0.5032,
      "num_input_tokens_seen": 4286856,
      "step": 6585
    },
    {
      "epoch": 3.4538784067085953,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029520013397440387,
      "loss": 0.5181,
      "num_input_tokens_seen": 4290216,
      "step": 6590
    },
    {
      "epoch": 3.4564989517819704,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.029518290240811332,
      "loss": 0.6348,
      "num_input_tokens_seen": 4293768,
      "step": 6595
    },
    {
      "epoch": 3.459119496855346,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.029516564047121593,
      "loss": 0.4509,
      "num_input_tokens_seen": 4297288,
      "step": 6600
    },
    {
      "epoch": 3.461740041928721,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.02951483481673227,
      "loss": 0.448,
      "num_input_tokens_seen": 4299592,
      "step": 6605
    },
    {
      "epoch": 3.4643605870020964,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.029513102550005093,
      "loss": 0.463,
      "num_input_tokens_seen": 4302600,
      "step": 6610
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.029511367247302443,
      "loss": 0.5902,
      "num_input_tokens_seen": 4305640,
      "step": 6615
    },
    {
      "epoch": 3.469601677148847,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.029509628908987316,
      "loss": 0.4424,
      "num_input_tokens_seen": 4308168,
      "step": 6620
    },
    {
      "epoch": 3.4722222222222223,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.029507887535423355,
      "loss": 0.4315,
      "num_input_tokens_seen": 4311208,
      "step": 6625
    },
    {
      "epoch": 3.4748427672955975,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02950614312697484,
      "loss": 0.5607,
      "num_input_tokens_seen": 4314312,
      "step": 6630
    },
    {
      "epoch": 3.4774633123689727,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.029504395684006673,
      "loss": 0.521,
      "num_input_tokens_seen": 4316936,
      "step": 6635
    },
    {
      "epoch": 3.480083857442348,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.029502645206884406,
      "loss": 0.425,
      "num_input_tokens_seen": 4320232,
      "step": 6640
    },
    {
      "epoch": 3.4827044025157234,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029500891695974214,
      "loss": 0.5696,
      "num_input_tokens_seen": 4324104,
      "step": 6645
    },
    {
      "epoch": 3.4853249475890986,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029499135151642916,
      "loss": 0.6551,
      "num_input_tokens_seen": 4327304,
      "step": 6650
    },
    {
      "epoch": 3.487945492662474,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02949737557425796,
      "loss": 0.5137,
      "num_input_tokens_seen": 4332008,
      "step": 6655
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02949561296418743,
      "loss": 0.4803,
      "num_input_tokens_seen": 4334728,
      "step": 6660
    },
    {
      "epoch": 3.4931865828092246,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029493847321800038,
      "loss": 0.4671,
      "num_input_tokens_seen": 4337416,
      "step": 6665
    },
    {
      "epoch": 3.4958071278825997,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.029492078647465145,
      "loss": 0.3716,
      "num_input_tokens_seen": 4340328,
      "step": 6670
    },
    {
      "epoch": 3.498427672955975,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02949030694155273,
      "loss": 0.441,
      "num_input_tokens_seen": 4344456,
      "step": 6675
    },
    {
      "epoch": 3.50104821802935,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02948853220443342,
      "loss": 0.4369,
      "num_input_tokens_seen": 4347944,
      "step": 6680
    },
    {
      "epoch": 3.5036687631027252,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029486754436478465,
      "loss": 0.503,
      "num_input_tokens_seen": 4350760,
      "step": 6685
    },
    {
      "epoch": 3.5062893081761004,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02948497363805976,
      "loss": 0.506,
      "num_input_tokens_seen": 4354760,
      "step": 6690
    },
    {
      "epoch": 3.508909853249476,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.029483189809549816,
      "loss": 0.41,
      "num_input_tokens_seen": 4358120,
      "step": 6695
    },
    {
      "epoch": 3.511530398322851,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.0294814029513218,
      "loss": 0.42,
      "num_input_tokens_seen": 4361352,
      "step": 6700
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0294796130637495,
      "loss": 0.5321,
      "num_input_tokens_seen": 4363784,
      "step": 6705
    },
    {
      "epoch": 3.5167714884696015,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02947782014720734,
      "loss": 0.444,
      "num_input_tokens_seen": 4367400,
      "step": 6710
    },
    {
      "epoch": 3.519392033542977,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029476024202070374,
      "loss": 0.5454,
      "num_input_tokens_seen": 4372232,
      "step": 6715
    },
    {
      "epoch": 3.5220125786163523,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029474225228714297,
      "loss": 0.4331,
      "num_input_tokens_seen": 4375560,
      "step": 6720
    },
    {
      "epoch": 3.5246331236897275,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.029472423227515435,
      "loss": 0.4705,
      "num_input_tokens_seen": 4378216,
      "step": 6725
    },
    {
      "epoch": 3.5272536687631026,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.029470618198850742,
      "loss": 0.5844,
      "num_input_tokens_seen": 4380744,
      "step": 6730
    },
    {
      "epoch": 3.529874213836478,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02946881014309781,
      "loss": 0.4215,
      "num_input_tokens_seen": 4384264,
      "step": 6735
    },
    {
      "epoch": 3.532494758909853,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029466999060634866,
      "loss": 0.5949,
      "num_input_tokens_seen": 4388200,
      "step": 6740
    },
    {
      "epoch": 3.5351153039832286,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.029465184951840768,
      "loss": 0.4477,
      "num_input_tokens_seen": 4390760,
      "step": 6745
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.029463367817095007,
      "loss": 0.5299,
      "num_input_tokens_seen": 4393416,
      "step": 6750
    },
    {
      "epoch": 3.540356394129979,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029461547656777703,
      "loss": 0.4713,
      "num_input_tokens_seen": 4396392,
      "step": 6755
    },
    {
      "epoch": 3.5429769392033545,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029459724471269613,
      "loss": 0.4301,
      "num_input_tokens_seen": 4399496,
      "step": 6760
    },
    {
      "epoch": 3.5455974842767297,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02945789826095213,
      "loss": 0.4872,
      "num_input_tokens_seen": 4402440,
      "step": 6765
    },
    {
      "epoch": 3.548218029350105,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02945606902620728,
      "loss": 0.561,
      "num_input_tokens_seen": 4405672,
      "step": 6770
    },
    {
      "epoch": 3.55083857442348,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.02945423676741771,
      "loss": 0.5495,
      "num_input_tokens_seen": 4408872,
      "step": 6775
    },
    {
      "epoch": 3.5534591194968552,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029452401484966712,
      "loss": 0.4284,
      "num_input_tokens_seen": 4411112,
      "step": 6780
    },
    {
      "epoch": 3.5560796645702304,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029450563179238203,
      "loss": 0.3972,
      "num_input_tokens_seen": 4414248,
      "step": 6785
    },
    {
      "epoch": 3.558700209643606,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02944872185061674,
      "loss": 0.57,
      "num_input_tokens_seen": 4417160,
      "step": 6790
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029446877499487507,
      "loss": 0.4179,
      "num_input_tokens_seen": 4421032,
      "step": 6795
    },
    {
      "epoch": 3.5639412997903563,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029445030126236314,
      "loss": 0.6233,
      "num_input_tokens_seen": 4423848,
      "step": 6800
    },
    {
      "epoch": 3.5665618448637315,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02944317973124962,
      "loss": 0.419,
      "num_input_tokens_seen": 4427144,
      "step": 6805
    },
    {
      "epoch": 3.569182389937107,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029441326314914504,
      "loss": 0.5367,
      "num_input_tokens_seen": 4429896,
      "step": 6810
    },
    {
      "epoch": 3.5718029350104823,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029439469877618675,
      "loss": 0.498,
      "num_input_tokens_seen": 4435720,
      "step": 6815
    },
    {
      "epoch": 3.5744234800838575,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02943761041975048,
      "loss": 0.4304,
      "num_input_tokens_seen": 4439048,
      "step": 6820
    },
    {
      "epoch": 3.5770440251572326,
      "grad_norm": 0.03125,
      "learning_rate": 0.029435747941698905,
      "loss": 0.4938,
      "num_input_tokens_seen": 4441928,
      "step": 6825
    },
    {
      "epoch": 3.579664570230608,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029433882443853546,
      "loss": 0.4143,
      "num_input_tokens_seen": 4445160,
      "step": 6830
    },
    {
      "epoch": 3.582285115303983,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.029432013926604654,
      "loss": 0.4882,
      "num_input_tokens_seen": 4448328,
      "step": 6835
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029430142390343098,
      "loss": 0.3864,
      "num_input_tokens_seen": 4451816,
      "step": 6840
    },
    {
      "epoch": 3.5875262054507338,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.02942826783546038,
      "loss": 0.4873,
      "num_input_tokens_seen": 4454856,
      "step": 6845
    },
    {
      "epoch": 3.590146750524109,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.029426390262348636,
      "loss": 0.4798,
      "num_input_tokens_seen": 4457544,
      "step": 6850
    },
    {
      "epoch": 3.5927672955974845,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029424509671400634,
      "loss": 0.5322,
      "num_input_tokens_seen": 4461160,
      "step": 6855
    },
    {
      "epoch": 3.5953878406708597,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02942262606300978,
      "loss": 0.3957,
      "num_input_tokens_seen": 4465000,
      "step": 6860
    },
    {
      "epoch": 3.598008385744235,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.029420739437570082,
      "loss": 0.4928,
      "num_input_tokens_seen": 4467944,
      "step": 6865
    },
    {
      "epoch": 3.60062893081761,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029418849795476223,
      "loss": 0.543,
      "num_input_tokens_seen": 4471880,
      "step": 6870
    },
    {
      "epoch": 3.603249475890985,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.029416957137123487,
      "loss": 0.4364,
      "num_input_tokens_seen": 4474184,
      "step": 6875
    },
    {
      "epoch": 3.6058700209643604,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02941506146290779,
      "loss": 0.4143,
      "num_input_tokens_seen": 4476584,
      "step": 6880
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 0.044921875,
      "learning_rate": 0.029413162773225695,
      "loss": 0.4373,
      "num_input_tokens_seen": 4479592,
      "step": 6885
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.02941126106847438,
      "loss": 0.4607,
      "num_input_tokens_seen": 4482280,
      "step": 6890
    },
    {
      "epoch": 3.6137316561844863,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.02940935634905166,
      "loss": 0.4048,
      "num_input_tokens_seen": 4485224,
      "step": 6895
    },
    {
      "epoch": 3.6163522012578615,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029407448615355984,
      "loss": 0.5535,
      "num_input_tokens_seen": 4488680,
      "step": 6900
    },
    {
      "epoch": 3.618972746331237,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02940553786778643,
      "loss": 0.4961,
      "num_input_tokens_seen": 4492104,
      "step": 6905
    },
    {
      "epoch": 3.6215932914046123,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029403624106742698,
      "loss": 0.4523,
      "num_input_tokens_seen": 4495240,
      "step": 6910
    },
    {
      "epoch": 3.6242138364779874,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.029401707332625125,
      "loss": 0.3498,
      "num_input_tokens_seen": 4498088,
      "step": 6915
    },
    {
      "epoch": 3.6268343815513626,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029399787545834683,
      "loss": 0.4439,
      "num_input_tokens_seen": 4501224,
      "step": 6920
    },
    {
      "epoch": 3.629454926624738,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.029397864746772973,
      "loss": 0.5557,
      "num_input_tokens_seen": 4503784,
      "step": 6925
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029395938935842213,
      "loss": 0.456,
      "num_input_tokens_seen": 4506568,
      "step": 6930
    },
    {
      "epoch": 3.6346960167714886,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029394010113445265,
      "loss": 0.3726,
      "num_input_tokens_seen": 4509096,
      "step": 6935
    },
    {
      "epoch": 3.6373165618448637,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02939207827998562,
      "loss": 0.3616,
      "num_input_tokens_seen": 4512200,
      "step": 6940
    },
    {
      "epoch": 3.639937106918239,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02939014343586739,
      "loss": 0.4268,
      "num_input_tokens_seen": 4514984,
      "step": 6945
    },
    {
      "epoch": 3.6425576519916145,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02938820558149532,
      "loss": 0.4445,
      "num_input_tokens_seen": 4518568,
      "step": 6950
    },
    {
      "epoch": 3.6451781970649897,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0293862647172748,
      "loss": 0.7949,
      "num_input_tokens_seen": 4521832,
      "step": 6955
    },
    {
      "epoch": 3.647798742138365,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.029384320843611825,
      "loss": 0.5469,
      "num_input_tokens_seen": 4525640,
      "step": 6960
    },
    {
      "epoch": 3.65041928721174,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.029382373960913033,
      "loss": 0.4393,
      "num_input_tokens_seen": 4529608,
      "step": 6965
    },
    {
      "epoch": 3.653039832285115,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.029380424069585694,
      "loss": 0.5463,
      "num_input_tokens_seen": 4532328,
      "step": 6970
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029378471170037695,
      "loss": 0.4947,
      "num_input_tokens_seen": 4535336,
      "step": 6975
    },
    {
      "epoch": 3.658280922431866,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029376515262677573,
      "loss": 0.3607,
      "num_input_tokens_seen": 4538760,
      "step": 6980
    },
    {
      "epoch": 3.660901467505241,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029374556347914468,
      "loss": 0.4195,
      "num_input_tokens_seen": 4546376,
      "step": 6985
    },
    {
      "epoch": 3.6635220125786163,
      "grad_norm": 0.046875,
      "learning_rate": 0.02937259442615817,
      "loss": 0.4778,
      "num_input_tokens_seen": 4550376,
      "step": 6990
    },
    {
      "epoch": 3.6661425576519915,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029370629497819094,
      "loss": 0.5891,
      "num_input_tokens_seen": 4553736,
      "step": 6995
    },
    {
      "epoch": 3.668763102725367,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02936866156330827,
      "loss": 0.4313,
      "num_input_tokens_seen": 4556936,
      "step": 7000
    },
    {
      "epoch": 3.6713836477987423,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.029366690623037373,
      "loss": 0.4458,
      "num_input_tokens_seen": 4559816,
      "step": 7005
    },
    {
      "epoch": 3.6740041928721174,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.029364716677418707,
      "loss": 0.4486,
      "num_input_tokens_seen": 4564520,
      "step": 7010
    },
    {
      "epoch": 3.6766247379454926,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02936273972686519,
      "loss": 0.4986,
      "num_input_tokens_seen": 4567592,
      "step": 7015
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02936075977179038,
      "loss": 0.4871,
      "num_input_tokens_seen": 4570408,
      "step": 7020
    },
    {
      "epoch": 3.681865828092243,
      "grad_norm": 0.05078125,
      "learning_rate": 0.029358776812608463,
      "loss": 0.4421,
      "num_input_tokens_seen": 4573320,
      "step": 7025
    },
    {
      "epoch": 3.6844863731656186,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02935679084973425,
      "loss": 0.5301,
      "num_input_tokens_seen": 4576680,
      "step": 7030
    },
    {
      "epoch": 3.6871069182389937,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.029354801883583186,
      "loss": 0.466,
      "num_input_tokens_seen": 4579368,
      "step": 7035
    },
    {
      "epoch": 3.689727463312369,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02935280991457133,
      "loss": 0.605,
      "num_input_tokens_seen": 4582440,
      "step": 7040
    },
    {
      "epoch": 3.6923480083857445,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029350814943115387,
      "loss": 0.4001,
      "num_input_tokens_seen": 4585128,
      "step": 7045
    },
    {
      "epoch": 3.6949685534591197,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029348816969632683,
      "loss": 0.45,
      "num_input_tokens_seen": 4588168,
      "step": 7050
    },
    {
      "epoch": 3.697589098532495,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.029346815994541166,
      "loss": 0.4468,
      "num_input_tokens_seen": 4591400,
      "step": 7055
    },
    {
      "epoch": 3.70020964360587,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.029344812018259423,
      "loss": 0.4691,
      "num_input_tokens_seen": 4594280,
      "step": 7060
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02934280504120666,
      "loss": 0.3663,
      "num_input_tokens_seen": 4598536,
      "step": 7065
    },
    {
      "epoch": 3.7054507337526204,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029340795063802712,
      "loss": 0.4938,
      "num_input_tokens_seen": 4602440,
      "step": 7070
    },
    {
      "epoch": 3.708071278825996,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029338782086468045,
      "loss": 0.5052,
      "num_input_tokens_seen": 4605352,
      "step": 7075
    },
    {
      "epoch": 3.710691823899371,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02933676610962375,
      "loss": 0.3629,
      "num_input_tokens_seen": 4610152,
      "step": 7080
    },
    {
      "epoch": 3.7133123689727463,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029334747133691552,
      "loss": 0.3919,
      "num_input_tokens_seen": 4612872,
      "step": 7085
    },
    {
      "epoch": 3.7159329140461215,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029332725159093787,
      "loss": 0.5393,
      "num_input_tokens_seen": 4617064,
      "step": 7090
    },
    {
      "epoch": 3.718553459119497,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02933070018625344,
      "loss": 0.5326,
      "num_input_tokens_seen": 4621608,
      "step": 7095
    },
    {
      "epoch": 3.7211740041928723,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.02932867221559411,
      "loss": 0.4564,
      "num_input_tokens_seen": 4623880,
      "step": 7100
    },
    {
      "epoch": 3.7237945492662474,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029326641247540013,
      "loss": 0.4485,
      "num_input_tokens_seen": 4627080,
      "step": 7105
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02932460728251602,
      "loss": 0.4241,
      "num_input_tokens_seen": 4630536,
      "step": 7110
    },
    {
      "epoch": 3.7290356394129978,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.029322570320947608,
      "loss": 0.5086,
      "num_input_tokens_seen": 4633256,
      "step": 7115
    },
    {
      "epoch": 3.731656184486373,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029320530363260884,
      "loss": 0.4606,
      "num_input_tokens_seen": 4636104,
      "step": 7120
    },
    {
      "epoch": 3.7342767295597485,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029318487409882583,
      "loss": 0.4555,
      "num_input_tokens_seen": 4639272,
      "step": 7125
    },
    {
      "epoch": 3.7368972746331237,
      "grad_norm": 0.046875,
      "learning_rate": 0.029316441461240073,
      "loss": 0.3498,
      "num_input_tokens_seen": 4641800,
      "step": 7130
    },
    {
      "epoch": 3.739517819706499,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029314392517761335,
      "loss": 0.5879,
      "num_input_tokens_seen": 4645160,
      "step": 7135
    },
    {
      "epoch": 3.742138364779874,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.029312340579874998,
      "loss": 0.3609,
      "num_input_tokens_seen": 4648488,
      "step": 7140
    },
    {
      "epoch": 3.7447589098532497,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.029310285648010286,
      "loss": 0.4313,
      "num_input_tokens_seen": 4651944,
      "step": 7145
    },
    {
      "epoch": 3.747379454926625,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029308227722597086,
      "loss": 0.5057,
      "num_input_tokens_seen": 4655816,
      "step": 7150
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.029306166804065874,
      "loss": 0.4251,
      "num_input_tokens_seen": 4658824,
      "step": 7155
    },
    {
      "epoch": 3.752620545073375,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029304102892847787,
      "loss": 0.5091,
      "num_input_tokens_seen": 4662440,
      "step": 7160
    },
    {
      "epoch": 3.7552410901467503,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.029302035989374562,
      "loss": 0.452,
      "num_input_tokens_seen": 4666280,
      "step": 7165
    },
    {
      "epoch": 3.757861635220126,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029299966094078575,
      "loss": 0.3876,
      "num_input_tokens_seen": 4669544,
      "step": 7170
    },
    {
      "epoch": 3.760482180293501,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029297893207392827,
      "loss": 0.4442,
      "num_input_tokens_seen": 4672392,
      "step": 7175
    },
    {
      "epoch": 3.7631027253668763,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029295817329750933,
      "loss": 0.5457,
      "num_input_tokens_seen": 4676872,
      "step": 7180
    },
    {
      "epoch": 3.7657232704402515,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02929373846158715,
      "loss": 0.6006,
      "num_input_tokens_seen": 4680520,
      "step": 7185
    },
    {
      "epoch": 3.768343815513627,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029291656603336356,
      "loss": 0.456,
      "num_input_tokens_seen": 4683464,
      "step": 7190
    },
    {
      "epoch": 3.7709643605870022,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029289571755434046,
      "loss": 0.4289,
      "num_input_tokens_seen": 4686728,
      "step": 7195
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02928748391831635,
      "loss": 0.4782,
      "num_input_tokens_seen": 4690376,
      "step": 7200
    },
    {
      "epoch": 3.7762054507337526,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.029285393092420016,
      "loss": 0.4522,
      "num_input_tokens_seen": 4695304,
      "step": 7205
    },
    {
      "epoch": 3.7788259958071277,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.029283299278182424,
      "loss": 0.5364,
      "num_input_tokens_seen": 4698216,
      "step": 7210
    },
    {
      "epoch": 3.781446540880503,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.029281202476041577,
      "loss": 0.5809,
      "num_input_tokens_seen": 4701512,
      "step": 7215
    },
    {
      "epoch": 3.7840670859538785,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0292791026864361,
      "loss": 0.5334,
      "num_input_tokens_seen": 4704744,
      "step": 7220
    },
    {
      "epoch": 3.7866876310272537,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029276999909805246,
      "loss": 0.6038,
      "num_input_tokens_seen": 4707880,
      "step": 7225
    },
    {
      "epoch": 3.789308176100629,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029274894146588893,
      "loss": 0.413,
      "num_input_tokens_seen": 4711080,
      "step": 7230
    },
    {
      "epoch": 3.791928721174004,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029272785397227538,
      "loss": 0.4018,
      "num_input_tokens_seen": 4714760,
      "step": 7235
    },
    {
      "epoch": 3.7945492662473796,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02927067366216231,
      "loss": 0.4029,
      "num_input_tokens_seen": 4718248,
      "step": 7240
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02926855894183496,
      "loss": 0.7858,
      "num_input_tokens_seen": 4721480,
      "step": 7245
    },
    {
      "epoch": 3.79979035639413,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02926644123668787,
      "loss": 0.4638,
      "num_input_tokens_seen": 4724328,
      "step": 7250
    },
    {
      "epoch": 3.802410901467505,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02926432054716403,
      "loss": 0.398,
      "num_input_tokens_seen": 4727432,
      "step": 7255
    },
    {
      "epoch": 3.8050314465408803,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029262196873707062,
      "loss": 0.5698,
      "num_input_tokens_seen": 4730952,
      "step": 7260
    },
    {
      "epoch": 3.8076519916142555,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.02926007021676123,
      "loss": 0.5218,
      "num_input_tokens_seen": 4733704,
      "step": 7265
    },
    {
      "epoch": 3.810272536687631,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02925794057677139,
      "loss": 0.484,
      "num_input_tokens_seen": 4736424,
      "step": 7270
    },
    {
      "epoch": 3.8128930817610063,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029255807954183047,
      "loss": 0.4435,
      "num_input_tokens_seen": 4739944,
      "step": 7275
    },
    {
      "epoch": 3.8155136268343814,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029253672349442317,
      "loss": 0.3885,
      "num_input_tokens_seen": 4743656,
      "step": 7280
    },
    {
      "epoch": 3.818134171907757,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029251533762995944,
      "loss": 0.4468,
      "num_input_tokens_seen": 4746440,
      "step": 7285
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0292493921952913,
      "loss": 0.4789,
      "num_input_tokens_seen": 4749672,
      "step": 7290
    },
    {
      "epoch": 3.8233752620545074,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029247247646776373,
      "loss": 0.5712,
      "num_input_tokens_seen": 4753224,
      "step": 7295
    },
    {
      "epoch": 3.8259958071278826,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02924510011789978,
      "loss": 0.4003,
      "num_input_tokens_seen": 4756072,
      "step": 7300
    },
    {
      "epoch": 3.8286163522012577,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029242949609110763,
      "loss": 0.4179,
      "num_input_tokens_seen": 4759368,
      "step": 7305
    },
    {
      "epoch": 3.831236897274633,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029240796120859173,
      "loss": 0.572,
      "num_input_tokens_seen": 4762280,
      "step": 7310
    },
    {
      "epoch": 3.8338574423480085,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029238639653595508,
      "loss": 0.4114,
      "num_input_tokens_seen": 4765544,
      "step": 7315
    },
    {
      "epoch": 3.8364779874213837,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.02923648020777087,
      "loss": 0.464,
      "num_input_tokens_seen": 4768424,
      "step": 7320
    },
    {
      "epoch": 3.839098532494759,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.029234317783836988,
      "loss": 0.4493,
      "num_input_tokens_seen": 4771464,
      "step": 7325
    },
    {
      "epoch": 3.841719077568134,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029232152382246218,
      "loss": 0.4281,
      "num_input_tokens_seen": 4775208,
      "step": 7330
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.029229984003451545,
      "loss": 0.3553,
      "num_input_tokens_seen": 4777768,
      "step": 7335
    },
    {
      "epoch": 3.846960167714885,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029227812647906557,
      "loss": 0.4025,
      "num_input_tokens_seen": 4781864,
      "step": 7340
    },
    {
      "epoch": 3.84958071278826,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02922563831606548,
      "loss": 0.4006,
      "num_input_tokens_seen": 4785352,
      "step": 7345
    },
    {
      "epoch": 3.852201257861635,
      "grad_norm": 0.0703125,
      "learning_rate": 0.029223461008383172,
      "loss": 0.47,
      "num_input_tokens_seen": 4787848,
      "step": 7350
    },
    {
      "epoch": 3.8548218029350103,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.029221280725315085,
      "loss": 0.3285,
      "num_input_tokens_seen": 4790952,
      "step": 7355
    },
    {
      "epoch": 3.8574423480083855,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.029219097467317316,
      "loss": 0.2936,
      "num_input_tokens_seen": 4794824,
      "step": 7360
    },
    {
      "epoch": 3.860062893081761,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.029216911234846574,
      "loss": 0.425,
      "num_input_tokens_seen": 4798120,
      "step": 7365
    },
    {
      "epoch": 3.8626834381551363,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029214722028360198,
      "loss": 0.4361,
      "num_input_tokens_seen": 4801704,
      "step": 7370
    },
    {
      "epoch": 3.8653039832285114,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.029212529848316147,
      "loss": 0.6224,
      "num_input_tokens_seen": 4804776,
      "step": 7375
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029210334695172997,
      "loss": 0.6323,
      "num_input_tokens_seen": 4807496,
      "step": 7380
    },
    {
      "epoch": 3.870545073375262,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029208136569389944,
      "loss": 0.4658,
      "num_input_tokens_seen": 4810152,
      "step": 7385
    },
    {
      "epoch": 3.8731656184486374,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029205935471426817,
      "loss": 0.3539,
      "num_input_tokens_seen": 4813736,
      "step": 7390
    },
    {
      "epoch": 3.8757861635220126,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.02920373140174406,
      "loss": 0.5543,
      "num_input_tokens_seen": 4816904,
      "step": 7395
    },
    {
      "epoch": 3.8784067085953877,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029201524360802738,
      "loss": 0.4019,
      "num_input_tokens_seen": 4820168,
      "step": 7400
    },
    {
      "epoch": 3.881027253668763,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02919931434906454,
      "loss": 0.641,
      "num_input_tokens_seen": 4823176,
      "step": 7405
    },
    {
      "epoch": 3.8836477987421385,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029197101366991773,
      "loss": 0.4443,
      "num_input_tokens_seen": 4826344,
      "step": 7410
    },
    {
      "epoch": 3.8862683438155137,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029194885415047375,
      "loss": 0.4426,
      "num_input_tokens_seen": 4830120,
      "step": 7415
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02919266649369489,
      "loss": 0.4292,
      "num_input_tokens_seen": 4833576,
      "step": 7420
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029190444603398487,
      "loss": 0.3821,
      "num_input_tokens_seen": 4836776,
      "step": 7425
    },
    {
      "epoch": 3.8941299790356396,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02918821974462297,
      "loss": 0.5886,
      "num_input_tokens_seen": 4840808,
      "step": 7430
    },
    {
      "epoch": 3.896750524109015,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02918599191783375,
      "loss": 0.5507,
      "num_input_tokens_seen": 4847624,
      "step": 7435
    },
    {
      "epoch": 3.89937106918239,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029183761123496865,
      "loss": 0.5518,
      "num_input_tokens_seen": 4850504,
      "step": 7440
    },
    {
      "epoch": 3.901991614255765,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02918152736207897,
      "loss": 0.3735,
      "num_input_tokens_seen": 4853448,
      "step": 7445
    },
    {
      "epoch": 3.9046121593291403,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.02917929063404734,
      "loss": 0.4873,
      "num_input_tokens_seen": 4856040,
      "step": 7450
    },
    {
      "epoch": 3.9072327044025155,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029177050939869872,
      "loss": 0.4874,
      "num_input_tokens_seen": 4859976,
      "step": 7455
    },
    {
      "epoch": 3.909853249475891,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02917480828001509,
      "loss": 0.467,
      "num_input_tokens_seen": 4863304,
      "step": 7460
    },
    {
      "epoch": 3.9124737945492662,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029172562654952133,
      "loss": 0.6014,
      "num_input_tokens_seen": 4866568,
      "step": 7465
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.029170314065150756,
      "loss": 0.5185,
      "num_input_tokens_seen": 4870312,
      "step": 7470
    },
    {
      "epoch": 3.917714884696017,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029168062511081342,
      "loss": 0.6368,
      "num_input_tokens_seen": 4873864,
      "step": 7475
    },
    {
      "epoch": 3.920335429769392,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029165807993214887,
      "loss": 0.5433,
      "num_input_tokens_seen": 4877032,
      "step": 7480
    },
    {
      "epoch": 3.9229559748427674,
      "grad_norm": 0.111328125,
      "learning_rate": 0.02916355051202301,
      "loss": 0.5022,
      "num_input_tokens_seen": 4880136,
      "step": 7485
    },
    {
      "epoch": 3.9255765199161425,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.029161290067977956,
      "loss": 0.4645,
      "num_input_tokens_seen": 4883720,
      "step": 7490
    },
    {
      "epoch": 3.9281970649895177,
      "grad_norm": 0.060546875,
      "learning_rate": 0.02915902666155258,
      "loss": 0.5713,
      "num_input_tokens_seen": 4886440,
      "step": 7495
    },
    {
      "epoch": 3.930817610062893,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.029156760293220358,
      "loss": 0.4787,
      "num_input_tokens_seen": 4888936,
      "step": 7500
    },
    {
      "epoch": 3.9334381551362685,
      "grad_norm": 0.115234375,
      "learning_rate": 0.029154490963455397,
      "loss": 0.4911,
      "num_input_tokens_seen": 4890952,
      "step": 7505
    },
    {
      "epoch": 3.9360587002096437,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.029152218672732405,
      "loss": 0.5257,
      "num_input_tokens_seen": 4893736,
      "step": 7510
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029149943421526724,
      "loss": 0.5497,
      "num_input_tokens_seen": 4897384,
      "step": 7515
    },
    {
      "epoch": 3.941299790356394,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02914766521031431,
      "loss": 0.4869,
      "num_input_tokens_seen": 4900808,
      "step": 7520
    },
    {
      "epoch": 3.9439203354297696,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.02914538403957174,
      "loss": 0.7633,
      "num_input_tokens_seen": 4903592,
      "step": 7525
    },
    {
      "epoch": 3.9465408805031448,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029143099909776204,
      "loss": 0.4657,
      "num_input_tokens_seen": 4907048,
      "step": 7530
    },
    {
      "epoch": 3.94916142557652,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02914081282140553,
      "loss": 0.4661,
      "num_input_tokens_seen": 4909960,
      "step": 7535
    },
    {
      "epoch": 3.951781970649895,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02913852277493813,
      "loss": 0.3581,
      "num_input_tokens_seen": 4913768,
      "step": 7540
    },
    {
      "epoch": 3.9544025157232703,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.029136229770853072,
      "loss": 0.5229,
      "num_input_tokens_seen": 4917416,
      "step": 7545
    },
    {
      "epoch": 3.9570230607966455,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02913393380963002,
      "loss": 0.3632,
      "num_input_tokens_seen": 4922056,
      "step": 7550
    },
    {
      "epoch": 3.959643605870021,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.02913163489174926,
      "loss": 0.4999,
      "num_input_tokens_seen": 4925320,
      "step": 7555
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.029129333017691706,
      "loss": 0.6662,
      "num_input_tokens_seen": 4928168,
      "step": 7560
    },
    {
      "epoch": 3.9648846960167714,
      "grad_norm": 0.046875,
      "learning_rate": 0.029127028187938882,
      "loss": 0.4403,
      "num_input_tokens_seen": 4931176,
      "step": 7565
    },
    {
      "epoch": 3.967505241090147,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.029124720402972933,
      "loss": 0.4107,
      "num_input_tokens_seen": 4935304,
      "step": 7570
    },
    {
      "epoch": 3.970125786163522,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02912240966327661,
      "loss": 0.5247,
      "num_input_tokens_seen": 4938600,
      "step": 7575
    },
    {
      "epoch": 3.9727463312368974,
      "grad_norm": 0.056640625,
      "learning_rate": 0.02912009596933331,
      "loss": 0.3878,
      "num_input_tokens_seen": 4941512,
      "step": 7580
    },
    {
      "epoch": 3.9753668763102725,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02911777932162702,
      "loss": 0.4556,
      "num_input_tokens_seen": 4944744,
      "step": 7585
    },
    {
      "epoch": 3.9779874213836477,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.02911545972064236,
      "loss": 0.4805,
      "num_input_tokens_seen": 4948072,
      "step": 7590
    },
    {
      "epoch": 3.980607966457023,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.029113137166864568,
      "loss": 0.552,
      "num_input_tokens_seen": 4951656,
      "step": 7595
    },
    {
      "epoch": 3.9832285115303985,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.029110811660779487,
      "loss": 0.3018,
      "num_input_tokens_seen": 4955400,
      "step": 7600
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029108483202873594,
      "loss": 0.4269,
      "num_input_tokens_seen": 4958568,
      "step": 7605
    },
    {
      "epoch": 3.988469601677149,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029106151793633975,
      "loss": 0.447,
      "num_input_tokens_seen": 4961608,
      "step": 7610
    },
    {
      "epoch": 3.991090146750524,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.029103817433548326,
      "loss": 0.4612,
      "num_input_tokens_seen": 4965288,
      "step": 7615
    },
    {
      "epoch": 3.9937106918238996,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029101480123104977,
      "loss": 0.4219,
      "num_input_tokens_seen": 4968424,
      "step": 7620
    },
    {
      "epoch": 3.9963312368972748,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029099139862792868,
      "loss": 0.5012,
      "num_input_tokens_seen": 4972200,
      "step": 7625
    },
    {
      "epoch": 3.99895178197065,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029096796653101553,
      "loss": 0.5105,
      "num_input_tokens_seen": 4975560,
      "step": 7630
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.4566926062107086,
      "eval_runtime": 13.5954,
      "eval_samples_per_second": 62.374,
      "eval_steps_per_second": 15.593,
      "num_input_tokens_seen": 4976128,
      "step": 7632
    },
    {
      "epoch": 4.001572327044025,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029094450494521196,
      "loss": 0.456,
      "num_input_tokens_seen": 4977984,
      "step": 7635
    },
    {
      "epoch": 4.0041928721174,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029092101387542594,
      "loss": 0.3607,
      "num_input_tokens_seen": 4980896,
      "step": 7640
    },
    {
      "epoch": 4.006813417190775,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.029089749332657157,
      "loss": 0.3969,
      "num_input_tokens_seen": 4984512,
      "step": 7645
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.029087394330356904,
      "loss": 0.4577,
      "num_input_tokens_seen": 4988320,
      "step": 7650
    },
    {
      "epoch": 4.012054507337526,
      "grad_norm": 0.03125,
      "learning_rate": 0.029085036381134473,
      "loss": 0.6228,
      "num_input_tokens_seen": 4991392,
      "step": 7655
    },
    {
      "epoch": 4.014675052410902,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02908267548548312,
      "loss": 0.4706,
      "num_input_tokens_seen": 4995680,
      "step": 7660
    },
    {
      "epoch": 4.017295597484277,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02908031164389672,
      "loss": 0.5539,
      "num_input_tokens_seen": 4999424,
      "step": 7665
    },
    {
      "epoch": 4.019916142557652,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.029077944856869764,
      "loss": 0.5055,
      "num_input_tokens_seen": 5002272,
      "step": 7670
    },
    {
      "epoch": 4.022536687631027,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02907557512489735,
      "loss": 0.3972,
      "num_input_tokens_seen": 5005888,
      "step": 7675
    },
    {
      "epoch": 4.0251572327044025,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029073202448475204,
      "loss": 0.3418,
      "num_input_tokens_seen": 5009600,
      "step": 7680
    },
    {
      "epoch": 4.027777777777778,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029070826828099663,
      "loss": 0.4176,
      "num_input_tokens_seen": 5013568,
      "step": 7685
    },
    {
      "epoch": 4.030398322851153,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.029068448264267674,
      "loss": 0.466,
      "num_input_tokens_seen": 5016800,
      "step": 7690
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.029066066757476812,
      "loss": 0.513,
      "num_input_tokens_seen": 5020032,
      "step": 7695
    },
    {
      "epoch": 4.035639412997903,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029063682308225258,
      "loss": 0.4682,
      "num_input_tokens_seen": 5023168,
      "step": 7700
    },
    {
      "epoch": 4.038259958071279,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029061294917011815,
      "loss": 0.3579,
      "num_input_tokens_seen": 5027040,
      "step": 7705
    },
    {
      "epoch": 4.040880503144654,
      "grad_norm": 0.05078125,
      "learning_rate": 0.02905890458433589,
      "loss": 0.609,
      "num_input_tokens_seen": 5030112,
      "step": 7710
    },
    {
      "epoch": 4.04350104821803,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.029056511310697516,
      "loss": 0.3672,
      "num_input_tokens_seen": 5033600,
      "step": 7715
    },
    {
      "epoch": 4.046121593291405,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02905411509659735,
      "loss": 0.2782,
      "num_input_tokens_seen": 5036416,
      "step": 7720
    },
    {
      "epoch": 4.04874213836478,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.029051715942536638,
      "loss": 0.5437,
      "num_input_tokens_seen": 5039552,
      "step": 7725
    },
    {
      "epoch": 4.051362683438155,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029049313849017266,
      "loss": 0.4813,
      "num_input_tokens_seen": 5044032,
      "step": 7730
    },
    {
      "epoch": 4.05398322851153,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029046908816541712,
      "loss": 0.3524,
      "num_input_tokens_seen": 5047232,
      "step": 7735
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029044500845613094,
      "loss": 0.5011,
      "num_input_tokens_seen": 5050624,
      "step": 7740
    },
    {
      "epoch": 4.059224318658281,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.02904208993673513,
      "loss": 0.4548,
      "num_input_tokens_seen": 5055008,
      "step": 7745
    },
    {
      "epoch": 4.061844863731656,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.029039676090412147,
      "loss": 0.4434,
      "num_input_tokens_seen": 5058240,
      "step": 7750
    },
    {
      "epoch": 4.064465408805032,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0290372593071491,
      "loss": 0.3228,
      "num_input_tokens_seen": 5060864,
      "step": 7755
    },
    {
      "epoch": 4.067085953878407,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029034839587451552,
      "loss": 0.5435,
      "num_input_tokens_seen": 5064384,
      "step": 7760
    },
    {
      "epoch": 4.069706498951782,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029032416931825684,
      "loss": 0.4241,
      "num_input_tokens_seen": 5068096,
      "step": 7765
    },
    {
      "epoch": 4.072327044025157,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.02902999134077828,
      "loss": 0.5054,
      "num_input_tokens_seen": 5070944,
      "step": 7770
    },
    {
      "epoch": 4.0749475890985325,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02902756281481675,
      "loss": 0.4754,
      "num_input_tokens_seen": 5074624,
      "step": 7775
    },
    {
      "epoch": 4.077568134171908,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.02902513135444911,
      "loss": 0.5441,
      "num_input_tokens_seen": 5077216,
      "step": 7780
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029022696960184006,
      "loss": 0.5104,
      "num_input_tokens_seen": 5079872,
      "step": 7785
    },
    {
      "epoch": 4.082809224318658,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.029020259632530673,
      "loss": 0.5136,
      "num_input_tokens_seen": 5082592,
      "step": 7790
    },
    {
      "epoch": 4.085429769392033,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029017819371998977,
      "loss": 0.3994,
      "num_input_tokens_seen": 5086144,
      "step": 7795
    },
    {
      "epoch": 4.088050314465409,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02901537617909939,
      "loss": 0.357,
      "num_input_tokens_seen": 5089696,
      "step": 7800
    },
    {
      "epoch": 4.090670859538784,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02901293005434301,
      "loss": 0.4424,
      "num_input_tokens_seen": 5092672,
      "step": 7805
    },
    {
      "epoch": 4.09329140461216,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029010480998241522,
      "loss": 0.4722,
      "num_input_tokens_seen": 5094848,
      "step": 7810
    },
    {
      "epoch": 4.095911949685535,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02900802901130725,
      "loss": 0.3475,
      "num_input_tokens_seen": 5097408,
      "step": 7815
    },
    {
      "epoch": 4.09853249475891,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029005574094053126,
      "loss": 0.4737,
      "num_input_tokens_seen": 5099872,
      "step": 7820
    },
    {
      "epoch": 4.101153039832285,
      "grad_norm": 0.04296875,
      "learning_rate": 0.029003116246992682,
      "loss": 0.537,
      "num_input_tokens_seen": 5103392,
      "step": 7825
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.029000655470640076,
      "loss": 0.4197,
      "num_input_tokens_seen": 5105696,
      "step": 7830
    },
    {
      "epoch": 4.106394129979035,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.028998191765510075,
      "loss": 0.5391,
      "num_input_tokens_seen": 5109312,
      "step": 7835
    },
    {
      "epoch": 4.109014675052411,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.028995725132118053,
      "loss": 0.5505,
      "num_input_tokens_seen": 5111840,
      "step": 7840
    },
    {
      "epoch": 4.111635220125786,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02899325557098001,
      "loss": 0.563,
      "num_input_tokens_seen": 5114752,
      "step": 7845
    },
    {
      "epoch": 4.114255765199162,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.028990783082612537,
      "loss": 0.5248,
      "num_input_tokens_seen": 5117760,
      "step": 7850
    },
    {
      "epoch": 4.116876310272537,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028988307667532864,
      "loss": 0.52,
      "num_input_tokens_seen": 5122176,
      "step": 7855
    },
    {
      "epoch": 4.119496855345912,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.028985829326258813,
      "loss": 0.7985,
      "num_input_tokens_seen": 5124256,
      "step": 7860
    },
    {
      "epoch": 4.122117400419287,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.028983348059308824,
      "loss": 0.5421,
      "num_input_tokens_seen": 5126912,
      "step": 7865
    },
    {
      "epoch": 4.1247379454926625,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02898086386720196,
      "loss": 0.5406,
      "num_input_tokens_seen": 5130144,
      "step": 7870
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.028978376750457866,
      "loss": 0.4573,
      "num_input_tokens_seen": 5133440,
      "step": 7875
    },
    {
      "epoch": 4.129979035639413,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.028975886709596832,
      "loss": 0.5539,
      "num_input_tokens_seen": 5140064,
      "step": 7880
    },
    {
      "epoch": 4.132599580712788,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02897339374513975,
      "loss": 0.4859,
      "num_input_tokens_seen": 5142720,
      "step": 7885
    },
    {
      "epoch": 4.135220125786163,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02897089785760811,
      "loss": 0.4445,
      "num_input_tokens_seen": 5145632,
      "step": 7890
    },
    {
      "epoch": 4.137840670859539,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.028968399047524025,
      "loss": 0.4146,
      "num_input_tokens_seen": 5148224,
      "step": 7895
    },
    {
      "epoch": 4.140461215932914,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.02896589731541022,
      "loss": 0.5474,
      "num_input_tokens_seen": 5152096,
      "step": 7900
    },
    {
      "epoch": 4.1430817610062896,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.028963392661790027,
      "loss": 0.5662,
      "num_input_tokens_seen": 5155968,
      "step": 7905
    },
    {
      "epoch": 4.145702306079665,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028960885087187395,
      "loss": 0.4406,
      "num_input_tokens_seen": 5159168,
      "step": 7910
    },
    {
      "epoch": 4.14832285115304,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02895837459212687,
      "loss": 0.37,
      "num_input_tokens_seen": 5162816,
      "step": 7915
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02895586117713363,
      "loss": 0.54,
      "num_input_tokens_seen": 5167232,
      "step": 7920
    },
    {
      "epoch": 4.15356394129979,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02895334484273345,
      "loss": 0.5029,
      "num_input_tokens_seen": 5170560,
      "step": 7925
    },
    {
      "epoch": 4.156184486373165,
      "grad_norm": 0.05078125,
      "learning_rate": 0.028950825589452715,
      "loss": 0.3911,
      "num_input_tokens_seen": 5173184,
      "step": 7930
    },
    {
      "epoch": 4.158805031446541,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02894830341781843,
      "loss": 0.4272,
      "num_input_tokens_seen": 5176288,
      "step": 7935
    },
    {
      "epoch": 4.161425576519916,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0289457783283582,
      "loss": 0.4316,
      "num_input_tokens_seen": 5181216,
      "step": 7940
    },
    {
      "epoch": 4.164046121593292,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.028943250321600243,
      "loss": 0.3646,
      "num_input_tokens_seen": 5184256,
      "step": 7945
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.028940719398073394,
      "loss": 0.3497,
      "num_input_tokens_seen": 5187616,
      "step": 7950
    },
    {
      "epoch": 4.169287211740042,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02893818555830709,
      "loss": 0.5931,
      "num_input_tokens_seen": 5189952,
      "step": 7955
    },
    {
      "epoch": 4.171907756813417,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.028935648802831383,
      "loss": 0.51,
      "num_input_tokens_seen": 5192608,
      "step": 7960
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02893310913217693,
      "loss": 0.4125,
      "num_input_tokens_seen": 5196256,
      "step": 7965
    },
    {
      "epoch": 4.177148846960168,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.028930566546875005,
      "loss": 0.4425,
      "num_input_tokens_seen": 5199840,
      "step": 7970
    },
    {
      "epoch": 4.179769392033543,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.028928021047457488,
      "loss": 0.3069,
      "num_input_tokens_seen": 5202688,
      "step": 7975
    },
    {
      "epoch": 4.182389937106918,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.028925472634456866,
      "loss": 0.6256,
      "num_input_tokens_seen": 5205248,
      "step": 7980
    },
    {
      "epoch": 4.185010482180293,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02892292130840624,
      "loss": 0.3902,
      "num_input_tokens_seen": 5207648,
      "step": 7985
    },
    {
      "epoch": 4.187631027253669,
      "grad_norm": 0.041015625,
      "learning_rate": 0.028920367069839317,
      "loss": 0.4088,
      "num_input_tokens_seen": 5210272,
      "step": 7990
    },
    {
      "epoch": 4.190251572327044,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.028917809919290414,
      "loss": 0.5804,
      "num_input_tokens_seen": 5214848,
      "step": 7995
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 0.03515625,
      "learning_rate": 0.028915249857294458,
      "loss": 0.4196,
      "num_input_tokens_seen": 5217888,
      "step": 8000
    },
    {
      "epoch": 4.195492662473795,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.028912686884386987,
      "loss": 0.5047,
      "num_input_tokens_seen": 5221952,
      "step": 8005
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.028910121001104145,
      "loss": 0.5335,
      "num_input_tokens_seen": 5225120,
      "step": 8010
    },
    {
      "epoch": 4.200733752620545,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.028907552207982686,
      "loss": 0.5494,
      "num_input_tokens_seen": 5228448,
      "step": 8015
    },
    {
      "epoch": 4.20335429769392,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.028904980505559966,
      "loss": 0.5421,
      "num_input_tokens_seen": 5230976,
      "step": 8020
    },
    {
      "epoch": 4.205974842767295,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.028902405894373966,
      "loss": 0.5184,
      "num_input_tokens_seen": 5234112,
      "step": 8025
    },
    {
      "epoch": 4.2085953878406706,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.028899828374963254,
      "loss": 0.5929,
      "num_input_tokens_seen": 5238592,
      "step": 8030
    },
    {
      "epoch": 4.211215932914046,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.028897247947867025,
      "loss": 0.4419,
      "num_input_tokens_seen": 5241568,
      "step": 8035
    },
    {
      "epoch": 4.213836477987422,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.028894664613625075,
      "loss": 0.3747,
      "num_input_tokens_seen": 5244704,
      "step": 8040
    },
    {
      "epoch": 4.216457023060797,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02889207837277781,
      "loss": 0.3648,
      "num_input_tokens_seen": 5248000,
      "step": 8045
    },
    {
      "epoch": 4.219077568134172,
      "grad_norm": 0.03125,
      "learning_rate": 0.028889489225866233,
      "loss": 0.3749,
      "num_input_tokens_seen": 5252096,
      "step": 8050
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028886897173431975,
      "loss": 0.5465,
      "num_input_tokens_seen": 5255360,
      "step": 8055
    },
    {
      "epoch": 4.2243186582809225,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028884302216017253,
      "loss": 0.5951,
      "num_input_tokens_seen": 5259712,
      "step": 8060
    },
    {
      "epoch": 4.226939203354298,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.02888170435416491,
      "loss": 0.4413,
      "num_input_tokens_seen": 5262912,
      "step": 8065
    },
    {
      "epoch": 4.229559748427673,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.028879103588418387,
      "loss": 0.4705,
      "num_input_tokens_seen": 5265664,
      "step": 8070
    },
    {
      "epoch": 4.232180293501048,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.028876499919321737,
      "loss": 0.4091,
      "num_input_tokens_seen": 5268928,
      "step": 8075
    },
    {
      "epoch": 4.234800838574423,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.028873893347419608,
      "loss": 0.5014,
      "num_input_tokens_seen": 5272032,
      "step": 8080
    },
    {
      "epoch": 4.237421383647799,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02887128387325728,
      "loss": 0.4048,
      "num_input_tokens_seen": 5275520,
      "step": 8085
    },
    {
      "epoch": 4.240041928721174,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.028868671497380618,
      "loss": 0.4868,
      "num_input_tokens_seen": 5278528,
      "step": 8090
    },
    {
      "epoch": 4.2426624737945495,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.028866056220336098,
      "loss": 0.6158,
      "num_input_tokens_seen": 5281536,
      "step": 8095
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.028863438042670806,
      "loss": 0.4627,
      "num_input_tokens_seen": 5287168,
      "step": 8100
    },
    {
      "epoch": 4.2479035639413,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02886081696493244,
      "loss": 0.4037,
      "num_input_tokens_seen": 5291008,
      "step": 8105
    },
    {
      "epoch": 4.250524109014675,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0288581929876693,
      "loss": 0.375,
      "num_input_tokens_seen": 5294144,
      "step": 8110
    },
    {
      "epoch": 4.25314465408805,
      "grad_norm": 0.052734375,
      "learning_rate": 0.028855566111430288,
      "loss": 0.575,
      "num_input_tokens_seen": 5297184,
      "step": 8115
    },
    {
      "epoch": 4.255765199161425,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.028852936336764917,
      "loss": 0.4419,
      "num_input_tokens_seen": 5300480,
      "step": 8120
    },
    {
      "epoch": 4.2583857442348005,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.028850303664223313,
      "loss": 0.484,
      "num_input_tokens_seen": 5303936,
      "step": 8125
    },
    {
      "epoch": 4.261006289308176,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.028847668094356186,
      "loss": 0.4019,
      "num_input_tokens_seen": 5306816,
      "step": 8130
    },
    {
      "epoch": 4.263626834381552,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.02884502962771488,
      "loss": 0.4456,
      "num_input_tokens_seen": 5309792,
      "step": 8135
    },
    {
      "epoch": 4.266247379454927,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02884238826485133,
      "loss": 0.5686,
      "num_input_tokens_seen": 5313056,
      "step": 8140
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.02883974400631808,
      "loss": 0.3491,
      "num_input_tokens_seen": 5315488,
      "step": 8145
    },
    {
      "epoch": 4.271488469601677,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.02883709685266827,
      "loss": 0.3989,
      "num_input_tokens_seen": 5318272,
      "step": 8150
    },
    {
      "epoch": 4.274109014675052,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02883444680445566,
      "loss": 0.4937,
      "num_input_tokens_seen": 5321408,
      "step": 8155
    },
    {
      "epoch": 4.276729559748428,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.028831793862234615,
      "loss": 0.4082,
      "num_input_tokens_seen": 5325696,
      "step": 8160
    },
    {
      "epoch": 4.279350104821803,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.02882913802656009,
      "loss": 0.5442,
      "num_input_tokens_seen": 5328832,
      "step": 8165
    },
    {
      "epoch": 4.281970649895178,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.028826479297987668,
      "loss": 0.4852,
      "num_input_tokens_seen": 5332448,
      "step": 8170
    },
    {
      "epoch": 4.284591194968553,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.028823817677073513,
      "loss": 0.5141,
      "num_input_tokens_seen": 5335520,
      "step": 8175
    },
    {
      "epoch": 4.287211740041929,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.028821153164374412,
      "loss": 0.3508,
      "num_input_tokens_seen": 5339328,
      "step": 8180
    },
    {
      "epoch": 4.289832285115304,
      "grad_norm": 0.03515625,
      "learning_rate": 0.028818485760447748,
      "loss": 0.4821,
      "num_input_tokens_seen": 5342368,
      "step": 8185
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.028815815465851515,
      "loss": 0.5861,
      "num_input_tokens_seen": 5345632,
      "step": 8190
    },
    {
      "epoch": 4.295073375262055,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.028813142281144302,
      "loss": 0.5776,
      "num_input_tokens_seen": 5348608,
      "step": 8195
    },
    {
      "epoch": 4.29769392033543,
      "grad_norm": 0.046875,
      "learning_rate": 0.028810466206885314,
      "loss": 0.3929,
      "num_input_tokens_seen": 5351776,
      "step": 8200
    },
    {
      "epoch": 4.300314465408805,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02880778724363435,
      "loss": 0.5811,
      "num_input_tokens_seen": 5354880,
      "step": 8205
    },
    {
      "epoch": 4.30293501048218,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.02880510539195183,
      "loss": 0.4119,
      "num_input_tokens_seen": 5357824,
      "step": 8210
    },
    {
      "epoch": 4.305555555555555,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.028802420652398752,
      "loss": 0.4189,
      "num_input_tokens_seen": 5360512,
      "step": 8215
    },
    {
      "epoch": 4.3081761006289305,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.028799733025536744,
      "loss": 0.3841,
      "num_input_tokens_seen": 5364064,
      "step": 8220
    },
    {
      "epoch": 4.310796645702306,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.028797042511928023,
      "loss": 0.5384,
      "num_input_tokens_seen": 5367200,
      "step": 8225
    },
    {
      "epoch": 4.313417190775682,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.028794349112135412,
      "loss": 0.5093,
      "num_input_tokens_seen": 5370272,
      "step": 8230
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02879165282672234,
      "loss": 0.4678,
      "num_input_tokens_seen": 5372896,
      "step": 8235
    },
    {
      "epoch": 4.318658280922432,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02878895365625284,
      "loss": 0.552,
      "num_input_tokens_seen": 5376800,
      "step": 8240
    },
    {
      "epoch": 4.321278825995807,
      "grad_norm": 0.03125,
      "learning_rate": 0.028786251601291547,
      "loss": 0.6142,
      "num_input_tokens_seen": 5380288,
      "step": 8245
    },
    {
      "epoch": 4.323899371069182,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.028783546662403705,
      "loss": 0.405,
      "num_input_tokens_seen": 5383072,
      "step": 8250
    },
    {
      "epoch": 4.326519916142558,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02878083884015515,
      "loss": 0.6014,
      "num_input_tokens_seen": 5386400,
      "step": 8255
    },
    {
      "epoch": 4.329140461215933,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02877812813511233,
      "loss": 0.5648,
      "num_input_tokens_seen": 5388832,
      "step": 8260
    },
    {
      "epoch": 4.331761006289308,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.028775414547842295,
      "loss": 0.4606,
      "num_input_tokens_seen": 5391968,
      "step": 8265
    },
    {
      "epoch": 4.334381551362683,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.028772698078912688,
      "loss": 0.3914,
      "num_input_tokens_seen": 5398624,
      "step": 8270
    },
    {
      "epoch": 4.337002096436059,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02876997872889178,
      "loss": 0.4716,
      "num_input_tokens_seen": 5401824,
      "step": 8275
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02876725649834841,
      "loss": 0.4081,
      "num_input_tokens_seen": 5404928,
      "step": 8280
    },
    {
      "epoch": 4.3422431865828095,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02876453138785205,
      "loss": 0.4223,
      "num_input_tokens_seen": 5408320,
      "step": 8285
    },
    {
      "epoch": 4.344863731656185,
      "grad_norm": 0.0390625,
      "learning_rate": 0.028761803397972758,
      "loss": 0.5286,
      "num_input_tokens_seen": 5410752,
      "step": 8290
    },
    {
      "epoch": 4.34748427672956,
      "grad_norm": 0.033203125,
      "learning_rate": 0.028759072529281195,
      "loss": 0.3854,
      "num_input_tokens_seen": 5413312,
      "step": 8295
    },
    {
      "epoch": 4.350104821802935,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028756338782348633,
      "loss": 0.5209,
      "num_input_tokens_seen": 5416992,
      "step": 8300
    },
    {
      "epoch": 4.35272536687631,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.028753602157746937,
      "loss": 0.353,
      "num_input_tokens_seen": 5420672,
      "step": 8305
    },
    {
      "epoch": 4.355345911949685,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.028750862656048582,
      "loss": 0.387,
      "num_input_tokens_seen": 5423360,
      "step": 8310
    },
    {
      "epoch": 4.3579664570230605,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.028748120277826635,
      "loss": 0.4444,
      "num_input_tokens_seen": 5426336,
      "step": 8315
    },
    {
      "epoch": 4.360587002096436,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.028745375023654776,
      "loss": 0.4982,
      "num_input_tokens_seen": 5429632,
      "step": 8320
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.028742626894107268,
      "loss": 0.4364,
      "num_input_tokens_seen": 5433312,
      "step": 8325
    },
    {
      "epoch": 4.365828092243187,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.028739875889759008,
      "loss": 0.511,
      "num_input_tokens_seen": 5437248,
      "step": 8330
    },
    {
      "epoch": 4.368448637316562,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02873712201118546,
      "loss": 0.3859,
      "num_input_tokens_seen": 5440384,
      "step": 8335
    },
    {
      "epoch": 4.371069182389937,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.028734365258962707,
      "loss": 0.5377,
      "num_input_tokens_seen": 5444000,
      "step": 8340
    },
    {
      "epoch": 4.373689727463312,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.028731605633667435,
      "loss": 0.4514,
      "num_input_tokens_seen": 5447936,
      "step": 8345
    },
    {
      "epoch": 4.376310272536688,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.028728843135876925,
      "loss": 0.5255,
      "num_input_tokens_seen": 5451488,
      "step": 8350
    },
    {
      "epoch": 4.378930817610063,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.028726077766169058,
      "loss": 0.4554,
      "num_input_tokens_seen": 5455520,
      "step": 8355
    },
    {
      "epoch": 4.381551362683438,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.028723309525122313,
      "loss": 0.613,
      "num_input_tokens_seen": 5458560,
      "step": 8360
    },
    {
      "epoch": 4.384171907756813,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.028720538413315778,
      "loss": 0.4401,
      "num_input_tokens_seen": 5461600,
      "step": 8365
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.028717764431329142,
      "loss": 0.395,
      "num_input_tokens_seen": 5465536,
      "step": 8370
    },
    {
      "epoch": 4.389412997903564,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.028714987579742692,
      "loss": 0.4284,
      "num_input_tokens_seen": 5468416,
      "step": 8375
    },
    {
      "epoch": 4.3920335429769395,
      "grad_norm": 0.08203125,
      "learning_rate": 0.028712207859137303,
      "loss": 0.5355,
      "num_input_tokens_seen": 5472000,
      "step": 8380
    },
    {
      "epoch": 4.394654088050315,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.028709425270094468,
      "loss": 0.3759,
      "num_input_tokens_seen": 5475168,
      "step": 8385
    },
    {
      "epoch": 4.39727463312369,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.028706639813196273,
      "loss": 0.4041,
      "num_input_tokens_seen": 5477760,
      "step": 8390
    },
    {
      "epoch": 4.399895178197065,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.028703851489025402,
      "loss": 0.4029,
      "num_input_tokens_seen": 5480544,
      "step": 8395
    },
    {
      "epoch": 4.40251572327044,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.028701060298165142,
      "loss": 0.3635,
      "num_input_tokens_seen": 5483776,
      "step": 8400
    },
    {
      "epoch": 4.405136268343815,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.028698266241199376,
      "loss": 0.4247,
      "num_input_tokens_seen": 5485888,
      "step": 8405
    },
    {
      "epoch": 4.4077568134171905,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.028695469318712596,
      "loss": 0.5954,
      "num_input_tokens_seen": 5489056,
      "step": 8410
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028692669531289874,
      "loss": 0.5501,
      "num_input_tokens_seen": 5492224,
      "step": 8415
    },
    {
      "epoch": 4.412997903563941,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0286898668795169,
      "loss": 0.4022,
      "num_input_tokens_seen": 5495872,
      "step": 8420
    },
    {
      "epoch": 4.415618448637317,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02868706136397996,
      "loss": 0.3558,
      "num_input_tokens_seen": 5499488,
      "step": 8425
    },
    {
      "epoch": 4.418238993710692,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02868425298526593,
      "loss": 0.5075,
      "num_input_tokens_seen": 5503104,
      "step": 8430
    },
    {
      "epoch": 4.420859538784067,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0286814417439623,
      "loss": 0.4672,
      "num_input_tokens_seen": 5506912,
      "step": 8435
    },
    {
      "epoch": 4.423480083857442,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02867862764065714,
      "loss": 0.5384,
      "num_input_tokens_seen": 5509312,
      "step": 8440
    },
    {
      "epoch": 4.426100628930818,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.028675810675939134,
      "loss": 0.371,
      "num_input_tokens_seen": 5512096,
      "step": 8445
    },
    {
      "epoch": 4.428721174004193,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.028672990850397553,
      "loss": 0.5103,
      "num_input_tokens_seen": 5515168,
      "step": 8450
    },
    {
      "epoch": 4.431341719077568,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.028670168164622277,
      "loss": 0.4877,
      "num_input_tokens_seen": 5518464,
      "step": 8455
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028667342619203777,
      "loss": 0.4466,
      "num_input_tokens_seen": 5521344,
      "step": 8460
    },
    {
      "epoch": 4.436582809224318,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.028664514214733126,
      "loss": 0.4918,
      "num_input_tokens_seen": 5524416,
      "step": 8465
    },
    {
      "epoch": 4.439203354297694,
      "grad_norm": 0.044921875,
      "learning_rate": 0.028661682951802,
      "loss": 0.5551,
      "num_input_tokens_seen": 5527488,
      "step": 8470
    },
    {
      "epoch": 4.4418238993710695,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.028658848831002653,
      "loss": 0.4622,
      "num_input_tokens_seen": 5530144,
      "step": 8475
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.028656011852927966,
      "loss": 0.5947,
      "num_input_tokens_seen": 5533536,
      "step": 8480
    },
    {
      "epoch": 4.44706498951782,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.028653172018171393,
      "loss": 0.4862,
      "num_input_tokens_seen": 5536032,
      "step": 8485
    },
    {
      "epoch": 4.449685534591195,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.028650329327326998,
      "loss": 0.4212,
      "num_input_tokens_seen": 5539552,
      "step": 8490
    },
    {
      "epoch": 4.45230607966457,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.028647483780989434,
      "loss": 0.4965,
      "num_input_tokens_seen": 5542528,
      "step": 8495
    },
    {
      "epoch": 4.454926624737945,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.028644635379753966,
      "loss": 0.5488,
      "num_input_tokens_seen": 5545408,
      "step": 8500
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02864178412421644,
      "loss": 0.3686,
      "num_input_tokens_seen": 5549120,
      "step": 8505
    },
    {
      "epoch": 4.460167714884696,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02863893001497331,
      "loss": 0.5366,
      "num_input_tokens_seen": 5552352,
      "step": 8510
    },
    {
      "epoch": 4.462788259958071,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.028636073052621627,
      "loss": 0.3489,
      "num_input_tokens_seen": 5555296,
      "step": 8515
    },
    {
      "epoch": 4.465408805031447,
      "grad_norm": 0.046875,
      "learning_rate": 0.028633213237759025,
      "loss": 0.3385,
      "num_input_tokens_seen": 5558720,
      "step": 8520
    },
    {
      "epoch": 4.468029350104822,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02863035057098375,
      "loss": 0.4783,
      "num_input_tokens_seen": 5562176,
      "step": 8525
    },
    {
      "epoch": 4.470649895178197,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02862748505289464,
      "loss": 0.6096,
      "num_input_tokens_seen": 5565408,
      "step": 8530
    },
    {
      "epoch": 4.473270440251572,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.028624616684091126,
      "loss": 0.4395,
      "num_input_tokens_seen": 5568288,
      "step": 8535
    },
    {
      "epoch": 4.475890985324948,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02862174546517324,
      "loss": 0.6893,
      "num_input_tokens_seen": 5571744,
      "step": 8540
    },
    {
      "epoch": 4.478511530398323,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02861887139674161,
      "loss": 0.5308,
      "num_input_tokens_seen": 5575008,
      "step": 8545
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.028615994479397455,
      "loss": 0.5936,
      "num_input_tokens_seen": 5577824,
      "step": 8550
    },
    {
      "epoch": 4.483752620545073,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02861311471374259,
      "loss": 0.4471,
      "num_input_tokens_seen": 5581216,
      "step": 8555
    },
    {
      "epoch": 4.486373165618448,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02861023210037944,
      "loss": 0.4769,
      "num_input_tokens_seen": 5586272,
      "step": 8560
    },
    {
      "epoch": 4.488993710691824,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.028607346639911005,
      "loss": 0.4693,
      "num_input_tokens_seen": 5588608,
      "step": 8565
    },
    {
      "epoch": 4.4916142557651995,
      "grad_norm": 0.041015625,
      "learning_rate": 0.028604458332940893,
      "loss": 0.3357,
      "num_input_tokens_seen": 5591296,
      "step": 8570
    },
    {
      "epoch": 4.494234800838575,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.02860156718007331,
      "loss": 0.411,
      "num_input_tokens_seen": 5593568,
      "step": 8575
    },
    {
      "epoch": 4.49685534591195,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.02859867318191304,
      "loss": 0.4501,
      "num_input_tokens_seen": 5596480,
      "step": 8580
    },
    {
      "epoch": 4.499475890985325,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.028595776339065487,
      "loss": 0.5803,
      "num_input_tokens_seen": 5599968,
      "step": 8585
    },
    {
      "epoch": 4.5020964360587,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.028592876652136633,
      "loss": 0.4401,
      "num_input_tokens_seen": 5603168,
      "step": 8590
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02858997412173306,
      "loss": 0.6054,
      "num_input_tokens_seen": 5606464,
      "step": 8595
    },
    {
      "epoch": 4.5073375262054505,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.028587068748461945,
      "loss": 0.5172,
      "num_input_tokens_seen": 5609216,
      "step": 8600
    },
    {
      "epoch": 4.509958071278826,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02858416053293105,
      "loss": 0.4173,
      "num_input_tokens_seen": 5612800,
      "step": 8605
    },
    {
      "epoch": 4.512578616352201,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.028581249475748756,
      "loss": 0.4237,
      "num_input_tokens_seen": 5615488,
      "step": 8610
    },
    {
      "epoch": 4.515199161425577,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02857833557752401,
      "loss": 0.3961,
      "num_input_tokens_seen": 5619264,
      "step": 8615
    },
    {
      "epoch": 4.517819706498952,
      "grad_norm": 0.048828125,
      "learning_rate": 0.02857541883886637,
      "loss": 0.3981,
      "num_input_tokens_seen": 5623104,
      "step": 8620
    },
    {
      "epoch": 4.520440251572327,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02857249926038599,
      "loss": 0.4591,
      "num_input_tokens_seen": 5626144,
      "step": 8625
    },
    {
      "epoch": 4.523060796645702,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0285695768426936,
      "loss": 0.4659,
      "num_input_tokens_seen": 5629216,
      "step": 8630
    },
    {
      "epoch": 4.5256813417190775,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02856665158640055,
      "loss": 0.5309,
      "num_input_tokens_seen": 5632448,
      "step": 8635
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.028563723492118765,
      "loss": 0.528,
      "num_input_tokens_seen": 5636704,
      "step": 8640
    },
    {
      "epoch": 4.530922431865828,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028560792560460773,
      "loss": 0.4278,
      "num_input_tokens_seen": 5641408,
      "step": 8645
    },
    {
      "epoch": 4.533542976939203,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02855785879203968,
      "loss": 0.4415,
      "num_input_tokens_seen": 5644096,
      "step": 8650
    },
    {
      "epoch": 4.536163522012579,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.028554922187469207,
      "loss": 0.3909,
      "num_input_tokens_seen": 5646720,
      "step": 8655
    },
    {
      "epoch": 4.538784067085954,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.028551982747363654,
      "loss": 0.5477,
      "num_input_tokens_seen": 5649696,
      "step": 8660
    },
    {
      "epoch": 4.5414046121593294,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02854904047233792,
      "loss": 0.3931,
      "num_input_tokens_seen": 5653504,
      "step": 8665
    },
    {
      "epoch": 4.544025157232705,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.028546095363007493,
      "loss": 0.4815,
      "num_input_tokens_seen": 5656896,
      "step": 8670
    },
    {
      "epoch": 4.54664570230608,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02854314741998846,
      "loss": 0.4732,
      "num_input_tokens_seen": 5660928,
      "step": 8675
    },
    {
      "epoch": 4.549266247379455,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02854019664389749,
      "loss": 0.479,
      "num_input_tokens_seen": 5664064,
      "step": 8680
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.028537243035351868,
      "loss": 0.5432,
      "num_input_tokens_seen": 5667232,
      "step": 8685
    },
    {
      "epoch": 4.554507337526205,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.028534286594969437,
      "loss": 0.489,
      "num_input_tokens_seen": 5670176,
      "step": 8690
    },
    {
      "epoch": 4.5571278825995805,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.028531327323368657,
      "loss": 0.3659,
      "num_input_tokens_seen": 5673120,
      "step": 8695
    },
    {
      "epoch": 4.559748427672956,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.028528365221168576,
      "loss": 0.4893,
      "num_input_tokens_seen": 5675936,
      "step": 8700
    },
    {
      "epoch": 4.562368972746331,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02852540028898883,
      "loss": 0.359,
      "num_input_tokens_seen": 5679072,
      "step": 8705
    },
    {
      "epoch": 4.564989517819707,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.028522432527449647,
      "loss": 0.4617,
      "num_input_tokens_seen": 5682112,
      "step": 8710
    },
    {
      "epoch": 4.567610062893082,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028519461937171857,
      "loss": 0.5754,
      "num_input_tokens_seen": 5685696,
      "step": 8715
    },
    {
      "epoch": 4.570230607966457,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.028516488518776865,
      "loss": 0.4613,
      "num_input_tokens_seen": 5688096,
      "step": 8720
    },
    {
      "epoch": 4.572851153039832,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.028513512272886676,
      "loss": 0.4325,
      "num_input_tokens_seen": 5691072,
      "step": 8725
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 0.03515625,
      "learning_rate": 0.028510533200123898,
      "loss": 0.4849,
      "num_input_tokens_seen": 5694592,
      "step": 8730
    },
    {
      "epoch": 4.578092243186583,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.0285075513011117,
      "loss": 0.378,
      "num_input_tokens_seen": 5697312,
      "step": 8735
    },
    {
      "epoch": 4.580712788259958,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02850456657647388,
      "loss": 0.3377,
      "num_input_tokens_seen": 5701088,
      "step": 8740
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 0.02734375,
      "learning_rate": 0.028501579026834797,
      "loss": 0.5249,
      "num_input_tokens_seen": 5703904,
      "step": 8745
    },
    {
      "epoch": 4.585953878406709,
      "grad_norm": 0.03125,
      "learning_rate": 0.02849858865281942,
      "loss": 0.4746,
      "num_input_tokens_seen": 5706816,
      "step": 8750
    },
    {
      "epoch": 4.588574423480084,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02849559545505329,
      "loss": 0.4616,
      "num_input_tokens_seen": 5709632,
      "step": 8755
    },
    {
      "epoch": 4.591194968553459,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02849259943416256,
      "loss": 0.4901,
      "num_input_tokens_seen": 5713472,
      "step": 8760
    },
    {
      "epoch": 4.593815513626835,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.02848960059077396,
      "loss": 0.6212,
      "num_input_tokens_seen": 5716736,
      "step": 8765
    },
    {
      "epoch": 4.59643605870021,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.028486598925514816,
      "loss": 0.3977,
      "num_input_tokens_seen": 5720672,
      "step": 8770
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02848359443901304,
      "loss": 0.6678,
      "num_input_tokens_seen": 5724000,
      "step": 8775
    },
    {
      "epoch": 4.60167714884696,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.028480587131897134,
      "loss": 0.6307,
      "num_input_tokens_seen": 5726784,
      "step": 8780
    },
    {
      "epoch": 4.604297693920335,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.028477577004796195,
      "loss": 0.4552,
      "num_input_tokens_seen": 5730272,
      "step": 8785
    },
    {
      "epoch": 4.6069182389937104,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02847456405833991,
      "loss": 0.4669,
      "num_input_tokens_seen": 5733120,
      "step": 8790
    },
    {
      "epoch": 4.609538784067086,
      "grad_norm": 0.046875,
      "learning_rate": 0.028471548293158542,
      "loss": 0.2989,
      "num_input_tokens_seen": 5737824,
      "step": 8795
    },
    {
      "epoch": 4.612159329140461,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02846852970988297,
      "loss": 0.4007,
      "num_input_tokens_seen": 5740384,
      "step": 8800
    },
    {
      "epoch": 4.614779874213837,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.028465508309144637,
      "loss": 0.4477,
      "num_input_tokens_seen": 5743232,
      "step": 8805
    },
    {
      "epoch": 4.617400419287212,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02846248409157559,
      "loss": 0.5367,
      "num_input_tokens_seen": 5746432,
      "step": 8810
    },
    {
      "epoch": 4.620020964360587,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02845945705780846,
      "loss": 0.5287,
      "num_input_tokens_seen": 5749696,
      "step": 8815
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.028456427208476465,
      "loss": 0.3972,
      "num_input_tokens_seen": 5753152,
      "step": 8820
    },
    {
      "epoch": 4.6252620545073375,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028453394544213417,
      "loss": 0.4064,
      "num_input_tokens_seen": 5755712,
      "step": 8825
    },
    {
      "epoch": 4.627882599580713,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.028450359065653713,
      "loss": 0.4841,
      "num_input_tokens_seen": 5759040,
      "step": 8830
    },
    {
      "epoch": 4.630503144654088,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.028447320773432348,
      "loss": 0.5354,
      "num_input_tokens_seen": 5761408,
      "step": 8835
    },
    {
      "epoch": 4.633123689727463,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02844427966818489,
      "loss": 0.5219,
      "num_input_tokens_seen": 5764672,
      "step": 8840
    },
    {
      "epoch": 4.635744234800838,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.0284412357505475,
      "loss": 0.4308,
      "num_input_tokens_seen": 5768160,
      "step": 8845
    },
    {
      "epoch": 4.638364779874214,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.028438189021156943,
      "loss": 0.4381,
      "num_input_tokens_seen": 5770400,
      "step": 8850
    },
    {
      "epoch": 4.640985324947589,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.028435139480650556,
      "loss": 0.3768,
      "num_input_tokens_seen": 5774080,
      "step": 8855
    },
    {
      "epoch": 4.643605870020965,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.02843208712966626,
      "loss": 0.5143,
      "num_input_tokens_seen": 5777184,
      "step": 8860
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.02842903196884258,
      "loss": 0.5113,
      "num_input_tokens_seen": 5780736,
      "step": 8865
    },
    {
      "epoch": 4.648846960167715,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02842597399881862,
      "loss": 0.5346,
      "num_input_tokens_seen": 5784320,
      "step": 8870
    },
    {
      "epoch": 4.65146750524109,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.028422913220234065,
      "loss": 0.4432,
      "num_input_tokens_seen": 5787840,
      "step": 8875
    },
    {
      "epoch": 4.654088050314465,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.028419849633729205,
      "loss": 0.6148,
      "num_input_tokens_seen": 5791040,
      "step": 8880
    },
    {
      "epoch": 4.65670859538784,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.02841678323994491,
      "loss": 0.732,
      "num_input_tokens_seen": 5794560,
      "step": 8885
    },
    {
      "epoch": 4.659329140461216,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.028413714039522617,
      "loss": 0.461,
      "num_input_tokens_seen": 5797504,
      "step": 8890
    },
    {
      "epoch": 4.661949685534591,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02841064203310438,
      "loss": 0.4397,
      "num_input_tokens_seen": 5799808,
      "step": 8895
    },
    {
      "epoch": 4.664570230607967,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.028407567221332836,
      "loss": 0.3381,
      "num_input_tokens_seen": 5802752,
      "step": 8900
    },
    {
      "epoch": 4.667190775681342,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.028404489604851185,
      "loss": 0.4664,
      "num_input_tokens_seen": 5806144,
      "step": 8905
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02840140918430323,
      "loss": 0.5768,
      "num_input_tokens_seen": 5809536,
      "step": 8910
    },
    {
      "epoch": 4.672431865828092,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.028398325960333374,
      "loss": 0.454,
      "num_input_tokens_seen": 5813088,
      "step": 8915
    },
    {
      "epoch": 4.6750524109014675,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02839523993358658,
      "loss": 0.3364,
      "num_input_tokens_seen": 5815488,
      "step": 8920
    },
    {
      "epoch": 4.677672955974843,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02839215110470841,
      "loss": 0.6029,
      "num_input_tokens_seen": 5818528,
      "step": 8925
    },
    {
      "epoch": 4.680293501048218,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02838905947434502,
      "loss": 0.5216,
      "num_input_tokens_seen": 5822016,
      "step": 8930
    },
    {
      "epoch": 4.682914046121593,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02838596504314313,
      "loss": 0.6387,
      "num_input_tokens_seen": 5825952,
      "step": 8935
    },
    {
      "epoch": 4.685534591194968,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.028382867811750076,
      "loss": 0.3649,
      "num_input_tokens_seen": 5828416,
      "step": 8940
    },
    {
      "epoch": 4.688155136268344,
      "grad_norm": 0.03125,
      "learning_rate": 0.02837976778081375,
      "loss": 0.4571,
      "num_input_tokens_seen": 5832352,
      "step": 8945
    },
    {
      "epoch": 4.690775681341719,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.028376664950982653,
      "loss": 0.5384,
      "num_input_tokens_seen": 5835488,
      "step": 8950
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.028373559322905854,
      "loss": 0.5048,
      "num_input_tokens_seen": 5838624,
      "step": 8955
    },
    {
      "epoch": 4.69601677148847,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.028370450897233022,
      "loss": 0.4596,
      "num_input_tokens_seen": 5842592,
      "step": 8960
    },
    {
      "epoch": 4.698637316561845,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0283673396746144,
      "loss": 0.4903,
      "num_input_tokens_seen": 5846080,
      "step": 8965
    },
    {
      "epoch": 4.70125786163522,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.028364225655700816,
      "loss": 0.5223,
      "num_input_tokens_seen": 5849536,
      "step": 8970
    },
    {
      "epoch": 4.703878406708595,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.028361108841143696,
      "loss": 0.4109,
      "num_input_tokens_seen": 5852032,
      "step": 8975
    },
    {
      "epoch": 4.70649895178197,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.02835798923159504,
      "loss": 0.5527,
      "num_input_tokens_seen": 5855072,
      "step": 8980
    },
    {
      "epoch": 4.709119496855346,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02835486682770743,
      "loss": 0.5093,
      "num_input_tokens_seen": 5858272,
      "step": 8985
    },
    {
      "epoch": 4.711740041928721,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02835174163013404,
      "loss": 0.4632,
      "num_input_tokens_seen": 5861856,
      "step": 8990
    },
    {
      "epoch": 4.714360587002097,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.028348613639528626,
      "loss": 0.462,
      "num_input_tokens_seen": 5864448,
      "step": 8995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.028345482856545527,
      "loss": 0.4299,
      "num_input_tokens_seen": 5867744,
      "step": 9000
    },
    {
      "epoch": 4.719601677148847,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.028342349281839666,
      "loss": 0.3871,
      "num_input_tokens_seen": 5870336,
      "step": 9005
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.028339212916066555,
      "loss": 0.4687,
      "num_input_tokens_seen": 5873728,
      "step": 9010
    },
    {
      "epoch": 4.7248427672955975,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.028336073759882285,
      "loss": 0.4101,
      "num_input_tokens_seen": 5876320,
      "step": 9015
    },
    {
      "epoch": 4.727463312368973,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.02833293181394353,
      "loss": 0.4299,
      "num_input_tokens_seen": 5879552,
      "step": 9020
    },
    {
      "epoch": 4.730083857442348,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.02832978707890755,
      "loss": 0.4065,
      "num_input_tokens_seen": 5882432,
      "step": 9025
    },
    {
      "epoch": 4.732704402515723,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.028326639555432184,
      "loss": 0.4875,
      "num_input_tokens_seen": 5885248,
      "step": 9030
    },
    {
      "epoch": 4.735324947589098,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.028323489244175864,
      "loss": 0.4467,
      "num_input_tokens_seen": 5887680,
      "step": 9035
    },
    {
      "epoch": 4.737945492662474,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.028320336145797598,
      "loss": 0.4658,
      "num_input_tokens_seen": 5890848,
      "step": 9040
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.028317180260956972,
      "loss": 0.4442,
      "num_input_tokens_seen": 5893664,
      "step": 9045
    },
    {
      "epoch": 4.743186582809225,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.028314021590314167,
      "loss": 0.5245,
      "num_input_tokens_seen": 5897216,
      "step": 9050
    },
    {
      "epoch": 4.7458071278826,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.028310860134529946,
      "loss": 0.4694,
      "num_input_tokens_seen": 5900992,
      "step": 9055
    },
    {
      "epoch": 4.748427672955975,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.028307695894265638,
      "loss": 0.5634,
      "num_input_tokens_seen": 5903904,
      "step": 9060
    },
    {
      "epoch": 4.75104821802935,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02830452887018317,
      "loss": 0.5489,
      "num_input_tokens_seen": 5907008,
      "step": 9065
    },
    {
      "epoch": 4.753668763102725,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02830135906294505,
      "loss": 0.3978,
      "num_input_tokens_seen": 5910112,
      "step": 9070
    },
    {
      "epoch": 4.7562893081761,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028298186473214363,
      "loss": 0.4237,
      "num_input_tokens_seen": 5913216,
      "step": 9075
    },
    {
      "epoch": 4.758909853249476,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.028295011101654784,
      "loss": 0.4573,
      "num_input_tokens_seen": 5916256,
      "step": 9080
    },
    {
      "epoch": 4.761530398322851,
      "grad_norm": 0.033203125,
      "learning_rate": 0.028291832948930555,
      "loss": 0.4617,
      "num_input_tokens_seen": 5919040,
      "step": 9085
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.028288652015706513,
      "loss": 0.4251,
      "num_input_tokens_seen": 5922944,
      "step": 9090
    },
    {
      "epoch": 4.766771488469602,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.02828546830264808,
      "loss": 0.3557,
      "num_input_tokens_seen": 5926368,
      "step": 9095
    },
    {
      "epoch": 4.769392033542977,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.028282281810421243,
      "loss": 0.4183,
      "num_input_tokens_seen": 5929632,
      "step": 9100
    },
    {
      "epoch": 4.772012578616352,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.028279092539692584,
      "loss": 0.4726,
      "num_input_tokens_seen": 5932832,
      "step": 9105
    },
    {
      "epoch": 4.7746331236897275,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.028275900491129265,
      "loss": 0.3963,
      "num_input_tokens_seen": 5936608,
      "step": 9110
    },
    {
      "epoch": 4.777253668763103,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.028272705665399016,
      "loss": 0.5621,
      "num_input_tokens_seen": 5940640,
      "step": 9115
    },
    {
      "epoch": 4.779874213836478,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.028269508063170173,
      "loss": 0.4325,
      "num_input_tokens_seen": 5944416,
      "step": 9120
    },
    {
      "epoch": 4.782494758909853,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.02826630768511162,
      "loss": 0.4634,
      "num_input_tokens_seen": 5948352,
      "step": 9125
    },
    {
      "epoch": 4.785115303983228,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02826310453189286,
      "loss": 0.3976,
      "num_input_tokens_seen": 5952000,
      "step": 9130
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.028259898604183938,
      "loss": 0.4552,
      "num_input_tokens_seen": 5954528,
      "step": 9135
    },
    {
      "epoch": 4.790356394129979,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.028256689902655507,
      "loss": 0.4357,
      "num_input_tokens_seen": 5957664,
      "step": 9140
    },
    {
      "epoch": 4.7929769392033545,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.028253478427978786,
      "loss": 0.3641,
      "num_input_tokens_seen": 5960448,
      "step": 9145
    },
    {
      "epoch": 4.79559748427673,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028250264180825588,
      "loss": 0.4103,
      "num_input_tokens_seen": 5963072,
      "step": 9150
    },
    {
      "epoch": 4.798218029350105,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.028247047161868283,
      "loss": 0.3719,
      "num_input_tokens_seen": 5966240,
      "step": 9155
    },
    {
      "epoch": 4.80083857442348,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028243827371779846,
      "loss": 0.392,
      "num_input_tokens_seen": 5969664,
      "step": 9160
    },
    {
      "epoch": 4.803459119496855,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02824060481123382,
      "loss": 0.4462,
      "num_input_tokens_seen": 5972320,
      "step": 9165
    },
    {
      "epoch": 4.80607966457023,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02823737948090432,
      "loss": 0.5222,
      "num_input_tokens_seen": 5975264,
      "step": 9170
    },
    {
      "epoch": 4.808700209643606,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.02823415138146605,
      "loss": 0.5211,
      "num_input_tokens_seen": 5978336,
      "step": 9175
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.028230920513594295,
      "loss": 0.4874,
      "num_input_tokens_seen": 5981536,
      "step": 9180
    },
    {
      "epoch": 4.813941299790356,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02822768687796492,
      "loss": 0.5821,
      "num_input_tokens_seen": 5984640,
      "step": 9185
    },
    {
      "epoch": 4.816561844863732,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.02822445047525436,
      "loss": 0.4875,
      "num_input_tokens_seen": 5987168,
      "step": 9190
    },
    {
      "epoch": 4.819182389937107,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028221211306139627,
      "loss": 0.5148,
      "num_input_tokens_seen": 5990880,
      "step": 9195
    },
    {
      "epoch": 4.821802935010482,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.028217969371298328,
      "loss": 0.4921,
      "num_input_tokens_seen": 5993664,
      "step": 9200
    },
    {
      "epoch": 4.8244234800838575,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02821472467140863,
      "loss": 0.5466,
      "num_input_tokens_seen": 5997216,
      "step": 9205
    },
    {
      "epoch": 4.827044025157233,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0282114772071493,
      "loss": 0.3223,
      "num_input_tokens_seen": 5999840,
      "step": 9210
    },
    {
      "epoch": 4.829664570230608,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.02820822697919966,
      "loss": 0.4409,
      "num_input_tokens_seen": 6003008,
      "step": 9215
    },
    {
      "epoch": 4.832285115303983,
      "grad_norm": 0.03515625,
      "learning_rate": 0.028204973988239616,
      "loss": 0.4884,
      "num_input_tokens_seen": 6007552,
      "step": 9220
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02820171823494967,
      "loss": 0.5006,
      "num_input_tokens_seen": 6011040,
      "step": 9225
    },
    {
      "epoch": 4.837526205450734,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.028198459720010886,
      "loss": 0.3845,
      "num_input_tokens_seen": 6013536,
      "step": 9230
    },
    {
      "epoch": 4.840146750524109,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0281951984441049,
      "loss": 0.387,
      "num_input_tokens_seen": 6015808,
      "step": 9235
    },
    {
      "epoch": 4.8427672955974845,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02819193440791394,
      "loss": 0.3447,
      "num_input_tokens_seen": 6019328,
      "step": 9240
    },
    {
      "epoch": 4.84538784067086,
      "grad_norm": 0.0234375,
      "learning_rate": 0.028188667612120806,
      "loss": 0.4109,
      "num_input_tokens_seen": 6022496,
      "step": 9245
    },
    {
      "epoch": 4.848008385744235,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.028185398057408866,
      "loss": 0.4242,
      "num_input_tokens_seen": 6025408,
      "step": 9250
    },
    {
      "epoch": 4.85062893081761,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02818212574446208,
      "loss": 0.5435,
      "num_input_tokens_seen": 6029056,
      "step": 9255
    },
    {
      "epoch": 4.853249475890985,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.028178850673964983,
      "loss": 0.3331,
      "num_input_tokens_seen": 6031584,
      "step": 9260
    },
    {
      "epoch": 4.85587002096436,
      "grad_norm": 0.0390625,
      "learning_rate": 0.028175572846602673,
      "loss": 0.51,
      "num_input_tokens_seen": 6034432,
      "step": 9265
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.028172292263060842,
      "loss": 0.395,
      "num_input_tokens_seen": 6037792,
      "step": 9270
    },
    {
      "epoch": 4.861111111111111,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.028169008924025747,
      "loss": 0.4774,
      "num_input_tokens_seen": 6040832,
      "step": 9275
    },
    {
      "epoch": 4.863731656184486,
      "grad_norm": 0.03125,
      "learning_rate": 0.02816572283018422,
      "loss": 0.5848,
      "num_input_tokens_seen": 6044480,
      "step": 9280
    },
    {
      "epoch": 4.866352201257862,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.028162433982223685,
      "loss": 0.4188,
      "num_input_tokens_seen": 6047520,
      "step": 9285
    },
    {
      "epoch": 4.868972746331237,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02815914238083212,
      "loss": 0.4278,
      "num_input_tokens_seen": 6051008,
      "step": 9290
    },
    {
      "epoch": 4.871593291404612,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.0281558480266981,
      "loss": 0.5057,
      "num_input_tokens_seen": 6053216,
      "step": 9295
    },
    {
      "epoch": 4.8742138364779874,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.028152550920510763,
      "loss": 0.5527,
      "num_input_tokens_seen": 6056864,
      "step": 9300
    },
    {
      "epoch": 4.876834381551363,
      "grad_norm": 0.03125,
      "learning_rate": 0.028149251062959825,
      "loss": 0.4068,
      "num_input_tokens_seen": 6059744,
      "step": 9305
    },
    {
      "epoch": 4.879454926624738,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02814594845473558,
      "loss": 0.3816,
      "num_input_tokens_seen": 6062624,
      "step": 9310
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.028142643096528894,
      "loss": 0.4439,
      "num_input_tokens_seen": 6066080,
      "step": 9315
    },
    {
      "epoch": 4.884696016771488,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.028139334989031213,
      "loss": 0.5473,
      "num_input_tokens_seen": 6069696,
      "step": 9320
    },
    {
      "epoch": 4.887316561844864,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02813602413293455,
      "loss": 0.5678,
      "num_input_tokens_seen": 6073408,
      "step": 9325
    },
    {
      "epoch": 4.889937106918239,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.028132710528931505,
      "loss": 0.3113,
      "num_input_tokens_seen": 6078592,
      "step": 9330
    },
    {
      "epoch": 4.8925576519916145,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.028129394177715244,
      "loss": 0.3417,
      "num_input_tokens_seen": 6088128,
      "step": 9335
    },
    {
      "epoch": 4.89517819706499,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.028126075079979506,
      "loss": 0.5021,
      "num_input_tokens_seen": 6090816,
      "step": 9340
    },
    {
      "epoch": 4.897798742138365,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.028122753236418618,
      "loss": 0.4746,
      "num_input_tokens_seen": 6093856,
      "step": 9345
    },
    {
      "epoch": 4.90041928721174,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02811942864772746,
      "loss": 0.441,
      "num_input_tokens_seen": 6097792,
      "step": 9350
    },
    {
      "epoch": 4.903039832285115,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.028116101314601502,
      "loss": 0.3144,
      "num_input_tokens_seen": 6100160,
      "step": 9355
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.028112771237736787,
      "loss": 0.5599,
      "num_input_tokens_seen": 6103264,
      "step": 9360
    },
    {
      "epoch": 4.9082809224318655,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.028109438417829927,
      "loss": 0.5659,
      "num_input_tokens_seen": 6105856,
      "step": 9365
    },
    {
      "epoch": 4.910901467505241,
      "grad_norm": 0.046875,
      "learning_rate": 0.02810610285557811,
      "loss": 0.5421,
      "num_input_tokens_seen": 6108800,
      "step": 9370
    },
    {
      "epoch": 4.913522012578616,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.0281027645516791,
      "loss": 0.4094,
      "num_input_tokens_seen": 6111552,
      "step": 9375
    },
    {
      "epoch": 4.916142557651992,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.028099423506831225,
      "loss": 0.5293,
      "num_input_tokens_seen": 6115488,
      "step": 9380
    },
    {
      "epoch": 4.918763102725367,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.028096079721733405,
      "loss": 0.3804,
      "num_input_tokens_seen": 6119136,
      "step": 9385
    },
    {
      "epoch": 4.921383647798742,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.028092733197085114,
      "loss": 0.5793,
      "num_input_tokens_seen": 6121696,
      "step": 9390
    },
    {
      "epoch": 4.924004192872117,
      "grad_norm": 0.0390625,
      "learning_rate": 0.028089383933586405,
      "loss": 0.4008,
      "num_input_tokens_seen": 6124480,
      "step": 9395
    },
    {
      "epoch": 4.926624737945493,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.02808603193193791,
      "loss": 0.4446,
      "num_input_tokens_seen": 6127008,
      "step": 9400
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 0.052734375,
      "learning_rate": 0.02808267719284083,
      "loss": 0.2967,
      "num_input_tokens_seen": 6130080,
      "step": 9405
    },
    {
      "epoch": 4.931865828092243,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.028079319716996936,
      "loss": 0.4467,
      "num_input_tokens_seen": 6133632,
      "step": 9410
    },
    {
      "epoch": 4.934486373165618,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02807595950510857,
      "loss": 0.4462,
      "num_input_tokens_seen": 6136992,
      "step": 9415
    },
    {
      "epoch": 4.937106918238994,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.028072596557878658,
      "loss": 0.3744,
      "num_input_tokens_seen": 6139776,
      "step": 9420
    },
    {
      "epoch": 4.939727463312369,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.028069230876010685,
      "loss": 0.4969,
      "num_input_tokens_seen": 6142880,
      "step": 9425
    },
    {
      "epoch": 4.9423480083857445,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.028065862460208715,
      "loss": 0.4486,
      "num_input_tokens_seen": 6147200,
      "step": 9430
    },
    {
      "epoch": 4.94496855345912,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02806249131117738,
      "loss": 0.4676,
      "num_input_tokens_seen": 6150016,
      "step": 9435
    },
    {
      "epoch": 4.947589098532495,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.028059117429621885,
      "loss": 0.396,
      "num_input_tokens_seen": 6152960,
      "step": 9440
    },
    {
      "epoch": 4.95020964360587,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02805574081624802,
      "loss": 0.3698,
      "num_input_tokens_seen": 6156544,
      "step": 9445
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.028052361471762117,
      "loss": 0.4561,
      "num_input_tokens_seen": 6159584,
      "step": 9450
    },
    {
      "epoch": 4.95545073375262,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0280489793968711,
      "loss": 0.4681,
      "num_input_tokens_seen": 6162752,
      "step": 9455
    },
    {
      "epoch": 4.9580712788259955,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02804559459228247,
      "loss": 0.3206,
      "num_input_tokens_seen": 6166016,
      "step": 9460
    },
    {
      "epoch": 4.960691823899371,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.028042207058704277,
      "loss": 0.4364,
      "num_input_tokens_seen": 6169184,
      "step": 9465
    },
    {
      "epoch": 4.963312368972746,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.028038816796845167,
      "loss": 0.4751,
      "num_input_tokens_seen": 6172960,
      "step": 9470
    },
    {
      "epoch": 4.965932914046122,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028035423807414334,
      "loss": 0.4357,
      "num_input_tokens_seen": 6176832,
      "step": 9475
    },
    {
      "epoch": 4.968553459119497,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02803202809112156,
      "loss": 0.4577,
      "num_input_tokens_seen": 6179392,
      "step": 9480
    },
    {
      "epoch": 4.971174004192872,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02802862964867718,
      "loss": 0.4076,
      "num_input_tokens_seen": 6182208,
      "step": 9485
    },
    {
      "epoch": 4.973794549266247,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.028025228480792124,
      "loss": 0.5545,
      "num_input_tokens_seen": 6185376,
      "step": 9490
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028021824588177865,
      "loss": 0.3668,
      "num_input_tokens_seen": 6188896,
      "step": 9495
    },
    {
      "epoch": 4.979035639412998,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.028018417971546465,
      "loss": 0.4781,
      "num_input_tokens_seen": 6191840,
      "step": 9500
    },
    {
      "epoch": 4.981656184486373,
      "grad_norm": 0.037109375,
      "learning_rate": 0.028015008631610543,
      "loss": 0.41,
      "num_input_tokens_seen": 6194784,
      "step": 9505
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.046875,
      "learning_rate": 0.028011596569083305,
      "loss": 0.3489,
      "num_input_tokens_seen": 6198048,
      "step": 9510
    },
    {
      "epoch": 4.986897274633124,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.028008181784678503,
      "loss": 0.4463,
      "num_input_tokens_seen": 6202688,
      "step": 9515
    },
    {
      "epoch": 4.989517819706499,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02800476427911048,
      "loss": 0.6269,
      "num_input_tokens_seen": 6205952,
      "step": 9520
    },
    {
      "epoch": 4.9921383647798745,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.028001344053094134,
      "loss": 0.3705,
      "num_input_tokens_seen": 6209504,
      "step": 9525
    },
    {
      "epoch": 4.99475890985325,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02799792110734494,
      "loss": 0.5143,
      "num_input_tokens_seen": 6212928,
      "step": 9530
    },
    {
      "epoch": 4.997379454926625,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02799449544257894,
      "loss": 0.4139,
      "num_input_tokens_seen": 6216672,
      "step": 9535
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.02799106705951274,
      "loss": 0.5727,
      "num_input_tokens_seen": 6219592,
      "step": 9540
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.4546113610267639,
      "eval_runtime": 13.6942,
      "eval_samples_per_second": 61.924,
      "eval_steps_per_second": 15.481,
      "num_input_tokens_seen": 6219592,
      "step": 9540
    },
    {
      "epoch": 5.002620545073375,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.027987635958863524,
      "loss": 0.4378,
      "num_input_tokens_seen": 6222216,
      "step": 9545
    },
    {
      "epoch": 5.00524109014675,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.027984202141349033,
      "loss": 0.4536,
      "num_input_tokens_seen": 6224872,
      "step": 9550
    },
    {
      "epoch": 5.0078616352201255,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02798076560768759,
      "loss": 0.4418,
      "num_input_tokens_seen": 6228072,
      "step": 9555
    },
    {
      "epoch": 5.010482180293501,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.027977326358598068,
      "loss": 0.5203,
      "num_input_tokens_seen": 6230984,
      "step": 9560
    },
    {
      "epoch": 5.013102725366877,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02797388439479993,
      "loss": 0.5261,
      "num_input_tokens_seen": 6234056,
      "step": 9565
    },
    {
      "epoch": 5.015723270440252,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02797043971701319,
      "loss": 0.3073,
      "num_input_tokens_seen": 6237800,
      "step": 9570
    },
    {
      "epoch": 5.018343815513627,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.027966992325958433,
      "loss": 0.6247,
      "num_input_tokens_seen": 6240808,
      "step": 9575
    },
    {
      "epoch": 5.020964360587002,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.027963542222356818,
      "loss": 0.5147,
      "num_input_tokens_seen": 6244040,
      "step": 9580
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.027960089406930067,
      "loss": 0.5209,
      "num_input_tokens_seen": 6246952,
      "step": 9585
    },
    {
      "epoch": 5.026205450733753,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.027956633880400466,
      "loss": 0.4817,
      "num_input_tokens_seen": 6250472,
      "step": 9590
    },
    {
      "epoch": 5.028825995807128,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.027953175643490875,
      "loss": 0.4021,
      "num_input_tokens_seen": 6253640,
      "step": 9595
    },
    {
      "epoch": 5.031446540880503,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02794971469692472,
      "loss": 0.4475,
      "num_input_tokens_seen": 6257160,
      "step": 9600
    },
    {
      "epoch": 5.034067085953878,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.027946251041425986,
      "loss": 0.414,
      "num_input_tokens_seen": 6259688,
      "step": 9605
    },
    {
      "epoch": 5.036687631027253,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.027942784677719233,
      "loss": 0.423,
      "num_input_tokens_seen": 6263080,
      "step": 9610
    },
    {
      "epoch": 5.039308176100629,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.027939315606529586,
      "loss": 0.2812,
      "num_input_tokens_seen": 6269320,
      "step": 9615
    },
    {
      "epoch": 5.0419287211740045,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.027935843828582736,
      "loss": 0.4339,
      "num_input_tokens_seen": 6272392,
      "step": 9620
    },
    {
      "epoch": 5.04454926624738,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.027932369344604933,
      "loss": 0.5577,
      "num_input_tokens_seen": 6275496,
      "step": 9625
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 0.041015625,
      "learning_rate": 0.027928892155323005,
      "loss": 0.4021,
      "num_input_tokens_seen": 6277960,
      "step": 9630
    },
    {
      "epoch": 5.04979035639413,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.027925412261464333,
      "loss": 0.8359,
      "num_input_tokens_seen": 6280520,
      "step": 9635
    },
    {
      "epoch": 5.052410901467505,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.027921929663756884,
      "loss": 0.4535,
      "num_input_tokens_seen": 6283304,
      "step": 9640
    },
    {
      "epoch": 5.05503144654088,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.027918444362929166,
      "loss": 0.5667,
      "num_input_tokens_seen": 6286536,
      "step": 9645
    },
    {
      "epoch": 5.0576519916142555,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.027914956359710273,
      "loss": 0.4541,
      "num_input_tokens_seen": 6289928,
      "step": 9650
    },
    {
      "epoch": 5.060272536687631,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02791146565482985,
      "loss": 0.5057,
      "num_input_tokens_seen": 6293416,
      "step": 9655
    },
    {
      "epoch": 5.062893081761007,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02790797224901811,
      "loss": 0.5019,
      "num_input_tokens_seen": 6297320,
      "step": 9660
    },
    {
      "epoch": 5.065513626834382,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.027904476143005844,
      "loss": 0.4054,
      "num_input_tokens_seen": 6300008,
      "step": 9665
    },
    {
      "epoch": 5.068134171907757,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027900977337524387,
      "loss": 0.4074,
      "num_input_tokens_seen": 6304136,
      "step": 9670
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02789747583330565,
      "loss": 0.5621,
      "num_input_tokens_seen": 6306792,
      "step": 9675
    },
    {
      "epoch": 5.073375262054507,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.027893971631082117,
      "loss": 0.5004,
      "num_input_tokens_seen": 6309800,
      "step": 9680
    },
    {
      "epoch": 5.075995807127883,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02789046473158682,
      "loss": 0.5244,
      "num_input_tokens_seen": 6312360,
      "step": 9685
    },
    {
      "epoch": 5.078616352201258,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.027886955135553358,
      "loss": 0.489,
      "num_input_tokens_seen": 6316008,
      "step": 9690
    },
    {
      "epoch": 5.081236897274633,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02788344284371591,
      "loss": 0.4422,
      "num_input_tokens_seen": 6318952,
      "step": 9695
    },
    {
      "epoch": 5.083857442348008,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0278799278568092,
      "loss": 0.473,
      "num_input_tokens_seen": 6322056,
      "step": 9700
    },
    {
      "epoch": 5.086477987421383,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.027876410175568515,
      "loss": 0.44,
      "num_input_tokens_seen": 6325000,
      "step": 9705
    },
    {
      "epoch": 5.089098532494759,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027872889800729734,
      "loss": 0.3876,
      "num_input_tokens_seen": 6329960,
      "step": 9710
    },
    {
      "epoch": 5.0917190775681345,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.027869366733029265,
      "loss": 0.3498,
      "num_input_tokens_seen": 6333640,
      "step": 9715
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.027865840973204097,
      "loss": 0.3729,
      "num_input_tokens_seen": 6336488,
      "step": 9720
    },
    {
      "epoch": 5.096960167714885,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.02786231252199178,
      "loss": 0.4727,
      "num_input_tokens_seen": 6340328,
      "step": 9725
    },
    {
      "epoch": 5.09958071278826,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02785878138013042,
      "loss": 0.5309,
      "num_input_tokens_seen": 6343464,
      "step": 9730
    },
    {
      "epoch": 5.102201257861635,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.027855247548358702,
      "loss": 0.6473,
      "num_input_tokens_seen": 6349352,
      "step": 9735
    },
    {
      "epoch": 5.10482180293501,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.027851711027415856,
      "loss": 0.3684,
      "num_input_tokens_seen": 6351784,
      "step": 9740
    },
    {
      "epoch": 5.1074423480083855,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02784817181804168,
      "loss": 0.5886,
      "num_input_tokens_seen": 6355368,
      "step": 9745
    },
    {
      "epoch": 5.110062893081761,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02784462992097654,
      "loss": 0.4969,
      "num_input_tokens_seen": 6358312,
      "step": 9750
    },
    {
      "epoch": 5.112683438155136,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.02784108533696136,
      "loss": 0.4774,
      "num_input_tokens_seen": 6360776,
      "step": 9755
    },
    {
      "epoch": 5.115303983228512,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.027837538066737628,
      "loss": 0.5347,
      "num_input_tokens_seen": 6364424,
      "step": 9760
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02783398811104739,
      "loss": 0.4313,
      "num_input_tokens_seen": 6367496,
      "step": 9765
    },
    {
      "epoch": 5.120545073375262,
      "grad_norm": 0.06640625,
      "learning_rate": 0.02783043547063326,
      "loss": 0.4084,
      "num_input_tokens_seen": 6370472,
      "step": 9770
    },
    {
      "epoch": 5.123165618448637,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0278268801462384,
      "loss": 0.3581,
      "num_input_tokens_seen": 6373832,
      "step": 9775
    },
    {
      "epoch": 5.1257861635220126,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.027823322138606554,
      "loss": 0.4282,
      "num_input_tokens_seen": 6377672,
      "step": 9780
    },
    {
      "epoch": 5.128406708595388,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.027819761448482017,
      "loss": 0.4531,
      "num_input_tokens_seen": 6380712,
      "step": 9785
    },
    {
      "epoch": 5.131027253668763,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.027816198076609636,
      "loss": 0.4821,
      "num_input_tokens_seen": 6384776,
      "step": 9790
    },
    {
      "epoch": 5.133647798742138,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.027812632023734834,
      "loss": 0.4753,
      "num_input_tokens_seen": 6387560,
      "step": 9795
    },
    {
      "epoch": 5.136268343815513,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.027809063290603585,
      "loss": 0.5805,
      "num_input_tokens_seen": 6390952,
      "step": 9800
    },
    {
      "epoch": 5.138888888888889,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02780549187796243,
      "loss": 0.4868,
      "num_input_tokens_seen": 6394088,
      "step": 9805
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027801917786558467,
      "loss": 0.3696,
      "num_input_tokens_seen": 6397064,
      "step": 9810
    },
    {
      "epoch": 5.14412997903564,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.027798341017139357,
      "loss": 0.6429,
      "num_input_tokens_seen": 6400584,
      "step": 9815
    },
    {
      "epoch": 5.146750524109015,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02779476157045331,
      "loss": 0.3529,
      "num_input_tokens_seen": 6403976,
      "step": 9820
    },
    {
      "epoch": 5.14937106918239,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.02779117944724912,
      "loss": 0.4635,
      "num_input_tokens_seen": 6406536,
      "step": 9825
    },
    {
      "epoch": 5.151991614255765,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.027787594648276116,
      "loss": 0.4677,
      "num_input_tokens_seen": 6410312,
      "step": 9830
    },
    {
      "epoch": 5.15461215932914,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.027784007174284205,
      "loss": 0.4091,
      "num_input_tokens_seen": 6414184,
      "step": 9835
    },
    {
      "epoch": 5.1572327044025155,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.027780417026023834,
      "loss": 0.6939,
      "num_input_tokens_seen": 6417224,
      "step": 9840
    },
    {
      "epoch": 5.159853249475891,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.027776824204246032,
      "loss": 0.5343,
      "num_input_tokens_seen": 6421128,
      "step": 9845
    },
    {
      "epoch": 5.162473794549266,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.027773228709702367,
      "loss": 0.4175,
      "num_input_tokens_seen": 6424232,
      "step": 9850
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.027769630543144987,
      "loss": 0.4404,
      "num_input_tokens_seen": 6427464,
      "step": 9855
    },
    {
      "epoch": 5.167714884696017,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.027766029705326576,
      "loss": 0.3894,
      "num_input_tokens_seen": 6430504,
      "step": 9860
    },
    {
      "epoch": 5.170335429769392,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.027762426197000402,
      "loss": 0.4775,
      "num_input_tokens_seen": 6434056,
      "step": 9865
    },
    {
      "epoch": 5.172955974842767,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.02775882001892026,
      "loss": 0.6242,
      "num_input_tokens_seen": 6437256,
      "step": 9870
    },
    {
      "epoch": 5.1755765199161425,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.027755211171840533,
      "loss": 0.4553,
      "num_input_tokens_seen": 6441064,
      "step": 9875
    },
    {
      "epoch": 5.178197064989518,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02775159965651615,
      "loss": 0.3978,
      "num_input_tokens_seen": 6444648,
      "step": 9880
    },
    {
      "epoch": 5.180817610062893,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.027747985473702597,
      "loss": 0.4898,
      "num_input_tokens_seen": 6447880,
      "step": 9885
    },
    {
      "epoch": 5.183438155136268,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.027744368624155913,
      "loss": 0.4476,
      "num_input_tokens_seen": 6450728,
      "step": 9890
    },
    {
      "epoch": 5.186058700209643,
      "grad_norm": 0.03515625,
      "learning_rate": 0.027740749108632713,
      "loss": 0.3593,
      "num_input_tokens_seen": 6453640,
      "step": 9895
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.027737126927890158,
      "loss": 0.3546,
      "num_input_tokens_seen": 6456840,
      "step": 9900
    },
    {
      "epoch": 5.191299790356394,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.027733502082685955,
      "loss": 0.4072,
      "num_input_tokens_seen": 6459848,
      "step": 9905
    },
    {
      "epoch": 5.19392033542977,
      "grad_norm": 0.04296875,
      "learning_rate": 0.027729874573778394,
      "loss": 0.5264,
      "num_input_tokens_seen": 6463624,
      "step": 9910
    },
    {
      "epoch": 5.196540880503145,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.0277262444019263,
      "loss": 0.3966,
      "num_input_tokens_seen": 6467176,
      "step": 9915
    },
    {
      "epoch": 5.19916142557652,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.02772261156788906,
      "loss": 0.4661,
      "num_input_tokens_seen": 6469896,
      "step": 9920
    },
    {
      "epoch": 5.201781970649895,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.027718976072426633,
      "loss": 0.5569,
      "num_input_tokens_seen": 6472904,
      "step": 9925
    },
    {
      "epoch": 5.20440251572327,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02771533791629952,
      "loss": 0.4916,
      "num_input_tokens_seen": 6477288,
      "step": 9930
    },
    {
      "epoch": 5.2070230607966455,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027711697100268773,
      "loss": 0.4349,
      "num_input_tokens_seen": 6480488,
      "step": 9935
    },
    {
      "epoch": 5.209643605870021,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.02770805362509602,
      "loss": 0.3602,
      "num_input_tokens_seen": 6483368,
      "step": 9940
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.027704407491543428,
      "loss": 0.4969,
      "num_input_tokens_seen": 6485928,
      "step": 9945
    },
    {
      "epoch": 5.214884696016772,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.027700758700373727,
      "loss": 0.4784,
      "num_input_tokens_seen": 6489768,
      "step": 9950
    },
    {
      "epoch": 5.217505241090147,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.02769710725235021,
      "loss": 0.4753,
      "num_input_tokens_seen": 6493352,
      "step": 9955
    },
    {
      "epoch": 5.220125786163522,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0276934531482367,
      "loss": 0.4778,
      "num_input_tokens_seen": 6495944,
      "step": 9960
    },
    {
      "epoch": 5.222746331236897,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.027689796388797615,
      "loss": 0.4369,
      "num_input_tokens_seen": 6498472,
      "step": 9965
    },
    {
      "epoch": 5.2253668763102725,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0276861369747979,
      "loss": 0.5076,
      "num_input_tokens_seen": 6502088,
      "step": 9970
    },
    {
      "epoch": 5.227987421383648,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.027682474907003055,
      "loss": 0.3473,
      "num_input_tokens_seen": 6505032,
      "step": 9975
    },
    {
      "epoch": 5.230607966457023,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02767881018617915,
      "loss": 0.3356,
      "num_input_tokens_seen": 6508488,
      "step": 9980
    },
    {
      "epoch": 5.233228511530398,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0276751428130928,
      "loss": 0.4371,
      "num_input_tokens_seen": 6510984,
      "step": 9985
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.027671472788511182,
      "loss": 0.536,
      "num_input_tokens_seen": 6514376,
      "step": 9990
    },
    {
      "epoch": 5.238469601677149,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.027667800113202023,
      "loss": 0.5752,
      "num_input_tokens_seen": 6517192,
      "step": 9995
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0276641247879336,
      "loss": 0.3827,
      "num_input_tokens_seen": 6520424,
      "step": 10000
    },
    {
      "epoch": 5.2437106918239,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02766044681347475,
      "loss": 0.249,
      "num_input_tokens_seen": 6523976,
      "step": 10005
    },
    {
      "epoch": 5.246331236897275,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.027656766190594866,
      "loss": 0.3713,
      "num_input_tokens_seen": 6526408,
      "step": 10010
    },
    {
      "epoch": 5.24895178197065,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.02765308292006389,
      "loss": 0.4761,
      "num_input_tokens_seen": 6529288,
      "step": 10015
    },
    {
      "epoch": 5.251572327044025,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.027649397002652322,
      "loss": 0.469,
      "num_input_tokens_seen": 6532072,
      "step": 10020
    },
    {
      "epoch": 5.2541928721174,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.027645708439131212,
      "loss": 0.415,
      "num_input_tokens_seen": 6536520,
      "step": 10025
    },
    {
      "epoch": 5.256813417190775,
      "grad_norm": 0.0625,
      "learning_rate": 0.027642017230272167,
      "loss": 0.495,
      "num_input_tokens_seen": 6539880,
      "step": 10030
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.027638323376847344,
      "loss": 0.4619,
      "num_input_tokens_seen": 6542632,
      "step": 10035
    },
    {
      "epoch": 5.262054507337526,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.02763462687962946,
      "loss": 0.3965,
      "num_input_tokens_seen": 6545416,
      "step": 10040
    },
    {
      "epoch": 5.264675052410902,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02763092773939177,
      "loss": 0.3865,
      "num_input_tokens_seen": 6548680,
      "step": 10045
    },
    {
      "epoch": 5.267295597484277,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.027627225956908103,
      "loss": 0.4218,
      "num_input_tokens_seen": 6552744,
      "step": 10050
    },
    {
      "epoch": 5.269916142557652,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.027623521532952826,
      "loss": 0.3744,
      "num_input_tokens_seen": 6557128,
      "step": 10055
    },
    {
      "epoch": 5.272536687631027,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.027619814468300856,
      "loss": 0.4652,
      "num_input_tokens_seen": 6565352,
      "step": 10060
    },
    {
      "epoch": 5.2751572327044025,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.027616104763727674,
      "loss": 0.4909,
      "num_input_tokens_seen": 6568680,
      "step": 10065
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.027612392420009308,
      "loss": 0.4169,
      "num_input_tokens_seen": 6572520,
      "step": 10070
    },
    {
      "epoch": 5.280398322851153,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.027608677437922342,
      "loss": 0.4961,
      "num_input_tokens_seen": 6576232,
      "step": 10075
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0276049598182439,
      "loss": 0.4306,
      "num_input_tokens_seen": 6579016,
      "step": 10080
    },
    {
      "epoch": 5.285639412997903,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.027601239561751673,
      "loss": 0.4401,
      "num_input_tokens_seen": 6581864,
      "step": 10085
    },
    {
      "epoch": 5.288259958071279,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.027597516669223887,
      "loss": 0.5478,
      "num_input_tokens_seen": 6585256,
      "step": 10090
    },
    {
      "epoch": 5.290880503144654,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02759379114143934,
      "loss": 0.5321,
      "num_input_tokens_seen": 6587976,
      "step": 10095
    },
    {
      "epoch": 5.29350104821803,
      "grad_norm": 0.03125,
      "learning_rate": 0.02759006297917736,
      "loss": 0.603,
      "num_input_tokens_seen": 6591080,
      "step": 10100
    },
    {
      "epoch": 5.296121593291405,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.027586332183217835,
      "loss": 0.4934,
      "num_input_tokens_seen": 6594088,
      "step": 10105
    },
    {
      "epoch": 5.29874213836478,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.027582598754341214,
      "loss": 0.3762,
      "num_input_tokens_seen": 6597544,
      "step": 10110
    },
    {
      "epoch": 5.301362683438155,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.027578862693328487,
      "loss": 0.4468,
      "num_input_tokens_seen": 6600552,
      "step": 10115
    },
    {
      "epoch": 5.30398322851153,
      "grad_norm": 0.078125,
      "learning_rate": 0.027575124000961185,
      "loss": 0.4739,
      "num_input_tokens_seen": 6602888,
      "step": 10120
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.027571382678021412,
      "loss": 0.4259,
      "num_input_tokens_seen": 6606344,
      "step": 10125
    },
    {
      "epoch": 5.309224318658281,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0275676387252918,
      "loss": 0.5278,
      "num_input_tokens_seen": 6609864,
      "step": 10130
    },
    {
      "epoch": 5.311844863731656,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.027563892143555552,
      "loss": 0.3834,
      "num_input_tokens_seen": 6612616,
      "step": 10135
    },
    {
      "epoch": 5.314465408805032,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0275601429335964,
      "loss": 0.386,
      "num_input_tokens_seen": 6616072,
      "step": 10140
    },
    {
      "epoch": 5.317085953878407,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02755639109619864,
      "loss": 0.621,
      "num_input_tokens_seen": 6619528,
      "step": 10145
    },
    {
      "epoch": 5.319706498951782,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02755263663214711,
      "loss": 0.4517,
      "num_input_tokens_seen": 6622824,
      "step": 10150
    },
    {
      "epoch": 5.322327044025157,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.027548879542227213,
      "loss": 0.487,
      "num_input_tokens_seen": 6626152,
      "step": 10155
    },
    {
      "epoch": 5.3249475890985325,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.027545119827224875,
      "loss": 0.4637,
      "num_input_tokens_seen": 6628808,
      "step": 10160
    },
    {
      "epoch": 5.327568134171908,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02754135748792659,
      "loss": 0.4575,
      "num_input_tokens_seen": 6632296,
      "step": 10165
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.027537592525119402,
      "loss": 0.5015,
      "num_input_tokens_seen": 6635400,
      "step": 10170
    },
    {
      "epoch": 5.332809224318658,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.027533824939590897,
      "loss": 0.5027,
      "num_input_tokens_seen": 6639432,
      "step": 10175
    },
    {
      "epoch": 5.335429769392033,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0275300547321292,
      "loss": 0.4968,
      "num_input_tokens_seen": 6643336,
      "step": 10180
    },
    {
      "epoch": 5.338050314465409,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02752628190352301,
      "loss": 0.4932,
      "num_input_tokens_seen": 6646088,
      "step": 10185
    },
    {
      "epoch": 5.340670859538784,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.027522506454561547,
      "loss": 0.3208,
      "num_input_tokens_seen": 6649384,
      "step": 10190
    },
    {
      "epoch": 5.34329140461216,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.027518728386034606,
      "loss": 0.4285,
      "num_input_tokens_seen": 6652424,
      "step": 10195
    },
    {
      "epoch": 5.345911949685535,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.027514947698732504,
      "loss": 0.4447,
      "num_input_tokens_seen": 6655464,
      "step": 10200
    },
    {
      "epoch": 5.34853249475891,
      "grad_norm": 0.044921875,
      "learning_rate": 0.02751116439344612,
      "loss": 0.4898,
      "num_input_tokens_seen": 6659112,
      "step": 10205
    },
    {
      "epoch": 5.351153039832285,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.02750737847096688,
      "loss": 0.508,
      "num_input_tokens_seen": 6663304,
      "step": 10210
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027503589932086753,
      "loss": 0.3284,
      "num_input_tokens_seen": 6666888,
      "step": 10215
    },
    {
      "epoch": 5.356394129979035,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.027499798777598266,
      "loss": 0.4217,
      "num_input_tokens_seen": 6670504,
      "step": 10220
    },
    {
      "epoch": 5.359014675052411,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02749600500829448,
      "loss": 0.3951,
      "num_input_tokens_seen": 6673832,
      "step": 10225
    },
    {
      "epoch": 5.361635220125786,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.027492208624969008,
      "loss": 0.4481,
      "num_input_tokens_seen": 6677160,
      "step": 10230
    },
    {
      "epoch": 5.364255765199162,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02748840962841601,
      "loss": 0.4412,
      "num_input_tokens_seen": 6681160,
      "step": 10235
    },
    {
      "epoch": 5.366876310272537,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02748460801943019,
      "loss": 0.4228,
      "num_input_tokens_seen": 6684424,
      "step": 10240
    },
    {
      "epoch": 5.369496855345912,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.027480803798806804,
      "loss": 0.4233,
      "num_input_tokens_seen": 6687528,
      "step": 10245
    },
    {
      "epoch": 5.372117400419287,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.027476996967341655,
      "loss": 0.5048,
      "num_input_tokens_seen": 6690120,
      "step": 10250
    },
    {
      "epoch": 5.3747379454926625,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.027473187525831083,
      "loss": 0.5066,
      "num_input_tokens_seen": 6693992,
      "step": 10255
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.027469375475071986,
      "loss": 0.4436,
      "num_input_tokens_seen": 6698568,
      "step": 10260
    },
    {
      "epoch": 5.379979035639413,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.027465560815861795,
      "loss": 0.4564,
      "num_input_tokens_seen": 6701320,
      "step": 10265
    },
    {
      "epoch": 5.382599580712788,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.027461743548998496,
      "loss": 0.4286,
      "num_input_tokens_seen": 6704168,
      "step": 10270
    },
    {
      "epoch": 5.385220125786163,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02745792367528062,
      "loss": 0.4841,
      "num_input_tokens_seen": 6707688,
      "step": 10275
    },
    {
      "epoch": 5.387840670859539,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.027454101195507234,
      "loss": 0.4087,
      "num_input_tokens_seen": 6711048,
      "step": 10280
    },
    {
      "epoch": 5.390461215932914,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.02745027611047797,
      "loss": 0.4685,
      "num_input_tokens_seen": 6715016,
      "step": 10285
    },
    {
      "epoch": 5.3930817610062896,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.02744644842099297,
      "loss": 0.4391,
      "num_input_tokens_seen": 6717448,
      "step": 10290
    },
    {
      "epoch": 5.395702306079665,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02744261812785297,
      "loss": 0.322,
      "num_input_tokens_seen": 6721032,
      "step": 10295
    },
    {
      "epoch": 5.39832285115304,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.027438785231859202,
      "loss": 0.4357,
      "num_input_tokens_seen": 6723304,
      "step": 10300
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02743494973381348,
      "loss": 0.419,
      "num_input_tokens_seen": 6726344,
      "step": 10305
    },
    {
      "epoch": 5.40356394129979,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.027431111634518134,
      "loss": 0.4595,
      "num_input_tokens_seen": 6730248,
      "step": 10310
    },
    {
      "epoch": 5.406184486373165,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027427270934776055,
      "loss": 0.4853,
      "num_input_tokens_seen": 6733352,
      "step": 10315
    },
    {
      "epoch": 5.408805031446541,
      "grad_norm": 0.064453125,
      "learning_rate": 0.02742342763539068,
      "loss": 0.3409,
      "num_input_tokens_seen": 6736104,
      "step": 10320
    },
    {
      "epoch": 5.411425576519916,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.027419581737165976,
      "loss": 0.4393,
      "num_input_tokens_seen": 6739336,
      "step": 10325
    },
    {
      "epoch": 5.414046121593292,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02741573324090646,
      "loss": 0.4822,
      "num_input_tokens_seen": 6742216,
      "step": 10330
    },
    {
      "epoch": 5.416666666666667,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0274118821474172,
      "loss": 0.3739,
      "num_input_tokens_seen": 6745320,
      "step": 10335
    },
    {
      "epoch": 5.419287211740042,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027408028457503793,
      "loss": 0.3411,
      "num_input_tokens_seen": 6748328,
      "step": 10340
    },
    {
      "epoch": 5.421907756813417,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.027404172171972392,
      "loss": 0.4215,
      "num_input_tokens_seen": 6751240,
      "step": 10345
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.027400313291629685,
      "loss": 0.3706,
      "num_input_tokens_seen": 6754344,
      "step": 10350
    },
    {
      "epoch": 5.427148846960168,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.027396451817282907,
      "loss": 0.4367,
      "num_input_tokens_seen": 6757320,
      "step": 10355
    },
    {
      "epoch": 5.429769392033543,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02739258774973984,
      "loss": 0.376,
      "num_input_tokens_seen": 6760744,
      "step": 10360
    },
    {
      "epoch": 5.432389937106918,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.027388721089808794,
      "loss": 0.5196,
      "num_input_tokens_seen": 6763848,
      "step": 10365
    },
    {
      "epoch": 5.435010482180293,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.027384851838298637,
      "loss": 0.5286,
      "num_input_tokens_seen": 6766792,
      "step": 10370
    },
    {
      "epoch": 5.437631027253669,
      "grad_norm": 0.0390625,
      "learning_rate": 0.027380979996018767,
      "loss": 0.4071,
      "num_input_tokens_seen": 6770024,
      "step": 10375
    },
    {
      "epoch": 5.440251572327044,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.027377105563779133,
      "loss": 0.4696,
      "num_input_tokens_seen": 6772584,
      "step": 10380
    },
    {
      "epoch": 5.4428721174004195,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02737322854239022,
      "loss": 0.4189,
      "num_input_tokens_seen": 6776040,
      "step": 10385
    },
    {
      "epoch": 5.445492662473795,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.027369348932663056,
      "loss": 0.4421,
      "num_input_tokens_seen": 6780616,
      "step": 10390
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02736546673540921,
      "loss": 0.5424,
      "num_input_tokens_seen": 6783720,
      "step": 10395
    },
    {
      "epoch": 5.450733752620545,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0273615819514408,
      "loss": 0.4454,
      "num_input_tokens_seen": 6786824,
      "step": 10400
    },
    {
      "epoch": 5.45335429769392,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.027357694581570474,
      "loss": 0.5141,
      "num_input_tokens_seen": 6790696,
      "step": 10405
    },
    {
      "epoch": 5.455974842767295,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.027353804626611425,
      "loss": 0.428,
      "num_input_tokens_seen": 6794408,
      "step": 10410
    },
    {
      "epoch": 5.4585953878406706,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.027349912087377387,
      "loss": 0.4713,
      "num_input_tokens_seen": 6798184,
      "step": 10415
    },
    {
      "epoch": 5.461215932914046,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.027346016964682637,
      "loss": 0.4006,
      "num_input_tokens_seen": 6801928,
      "step": 10420
    },
    {
      "epoch": 5.463836477987422,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.027342119259341987,
      "loss": 0.4057,
      "num_input_tokens_seen": 6805640,
      "step": 10425
    },
    {
      "epoch": 5.466457023060797,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.027338218972170796,
      "loss": 0.3509,
      "num_input_tokens_seen": 6808680,
      "step": 10430
    },
    {
      "epoch": 5.469077568134172,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.027334316103984953,
      "loss": 0.459,
      "num_input_tokens_seen": 6811752,
      "step": 10435
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.027330410655600903,
      "loss": 0.5021,
      "num_input_tokens_seen": 6814888,
      "step": 10440
    },
    {
      "epoch": 5.4743186582809225,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02732650262783562,
      "loss": 0.4433,
      "num_input_tokens_seen": 6819048,
      "step": 10445
    },
    {
      "epoch": 5.476939203354298,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.02732259202150661,
      "loss": 0.3295,
      "num_input_tokens_seen": 6822792,
      "step": 10450
    },
    {
      "epoch": 5.479559748427673,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.027318678837431935,
      "loss": 0.5481,
      "num_input_tokens_seen": 6826248,
      "step": 10455
    },
    {
      "epoch": 5.482180293501048,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02731476307643018,
      "loss": 0.4011,
      "num_input_tokens_seen": 6830248,
      "step": 10460
    },
    {
      "epoch": 5.484800838574423,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02731084473932049,
      "loss": 0.3798,
      "num_input_tokens_seen": 6833160,
      "step": 10465
    },
    {
      "epoch": 5.487421383647799,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02730692382692253,
      "loss": 0.5971,
      "num_input_tokens_seen": 6836392,
      "step": 10470
    },
    {
      "epoch": 5.490041928721174,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.027303000340056508,
      "loss": 0.4784,
      "num_input_tokens_seen": 6840872,
      "step": 10475
    },
    {
      "epoch": 5.4926624737945495,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.027299074279543176,
      "loss": 0.4389,
      "num_input_tokens_seen": 6843432,
      "step": 10480
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.027295145646203823,
      "loss": 0.3552,
      "num_input_tokens_seen": 6846760,
      "step": 10485
    },
    {
      "epoch": 5.4979035639413,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.027291214440860266,
      "loss": 0.6028,
      "num_input_tokens_seen": 6849256,
      "step": 10490
    },
    {
      "epoch": 5.500524109014675,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02728728066433488,
      "loss": 0.3586,
      "num_input_tokens_seen": 6853256,
      "step": 10495
    },
    {
      "epoch": 5.50314465408805,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02728334431745055,
      "loss": 0.5495,
      "num_input_tokens_seen": 6859688,
      "step": 10500
    },
    {
      "epoch": 5.505765199161425,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.027279405401030733,
      "loss": 0.4095,
      "num_input_tokens_seen": 6863080,
      "step": 10505
    },
    {
      "epoch": 5.5083857442348005,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02727546391589939,
      "loss": 0.4086,
      "num_input_tokens_seen": 6865704,
      "step": 10510
    },
    {
      "epoch": 5.511006289308176,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02727151986288105,
      "loss": 0.516,
      "num_input_tokens_seen": 6868392,
      "step": 10515
    },
    {
      "epoch": 5.513626834381551,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.027267573242800748,
      "loss": 0.4252,
      "num_input_tokens_seen": 6871624,
      "step": 10520
    },
    {
      "epoch": 5.516247379454927,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02726362405648408,
      "loss": 0.4015,
      "num_input_tokens_seen": 6875752,
      "step": 10525
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.027259672304757177,
      "loss": 0.605,
      "num_input_tokens_seen": 6878728,
      "step": 10530
    },
    {
      "epoch": 5.521488469601677,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02725571798844669,
      "loss": 0.5338,
      "num_input_tokens_seen": 6882408,
      "step": 10535
    },
    {
      "epoch": 5.524109014675052,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02725176110837982,
      "loss": 0.5778,
      "num_input_tokens_seen": 6885448,
      "step": 10540
    },
    {
      "epoch": 5.526729559748428,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0272478016653843,
      "loss": 0.4809,
      "num_input_tokens_seen": 6889192,
      "step": 10545
    },
    {
      "epoch": 5.529350104821803,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.027243839660288406,
      "loss": 0.5236,
      "num_input_tokens_seen": 6892424,
      "step": 10550
    },
    {
      "epoch": 5.531970649895178,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.027239875093920937,
      "loss": 0.4317,
      "num_input_tokens_seen": 6895368,
      "step": 10555
    },
    {
      "epoch": 5.534591194968553,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.027235907967111242,
      "loss": 0.4516,
      "num_input_tokens_seen": 6899784,
      "step": 10560
    },
    {
      "epoch": 5.537211740041929,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.027231938280689188,
      "loss": 0.4792,
      "num_input_tokens_seen": 6902504,
      "step": 10565
    },
    {
      "epoch": 5.539832285115304,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.027227966035485197,
      "loss": 0.4699,
      "num_input_tokens_seen": 6906184,
      "step": 10570
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.027223991232330217,
      "loss": 0.4698,
      "num_input_tokens_seen": 6909320,
      "step": 10575
    },
    {
      "epoch": 5.545073375262055,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.027220013872055734,
      "loss": 0.5064,
      "num_input_tokens_seen": 6912168,
      "step": 10580
    },
    {
      "epoch": 5.54769392033543,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.027216033955493758,
      "loss": 0.4753,
      "num_input_tokens_seen": 6915784,
      "step": 10585
    },
    {
      "epoch": 5.550314465408805,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.027212051483476846,
      "loss": 0.4978,
      "num_input_tokens_seen": 6918920,
      "step": 10590
    },
    {
      "epoch": 5.55293501048218,
      "grad_norm": 0.037109375,
      "learning_rate": 0.027208066456838088,
      "loss": 0.4311,
      "num_input_tokens_seen": 6921992,
      "step": 10595
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.056640625,
      "learning_rate": 0.027204078876411103,
      "loss": 0.4134,
      "num_input_tokens_seen": 6924840,
      "step": 10600
    },
    {
      "epoch": 5.5581761006289305,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027200088743030044,
      "loss": 0.4002,
      "num_input_tokens_seen": 6927848,
      "step": 10605
    },
    {
      "epoch": 5.560796645702306,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.027196096057529612,
      "loss": 0.5756,
      "num_input_tokens_seen": 6930568,
      "step": 10610
    },
    {
      "epoch": 5.563417190775681,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.027192100820745024,
      "loss": 0.3591,
      "num_input_tokens_seen": 6933992,
      "step": 10615
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02718810303351204,
      "loss": 0.4302,
      "num_input_tokens_seen": 6936680,
      "step": 10620
    },
    {
      "epoch": 5.568658280922432,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.027184102696666948,
      "loss": 0.4849,
      "num_input_tokens_seen": 6938792,
      "step": 10625
    },
    {
      "epoch": 5.571278825995807,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.027180099811046576,
      "loss": 0.4133,
      "num_input_tokens_seen": 6942664,
      "step": 10630
    },
    {
      "epoch": 5.573899371069182,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.02717609437748828,
      "loss": 0.4762,
      "num_input_tokens_seen": 6944968,
      "step": 10635
    },
    {
      "epoch": 5.576519916142558,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02717208639682996,
      "loss": 0.4689,
      "num_input_tokens_seen": 6948264,
      "step": 10640
    },
    {
      "epoch": 5.579140461215933,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.027168075869910026,
      "loss": 0.3909,
      "num_input_tokens_seen": 6951304,
      "step": 10645
    },
    {
      "epoch": 5.581761006289308,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.027164062797567442,
      "loss": 0.3938,
      "num_input_tokens_seen": 6954184,
      "step": 10650
    },
    {
      "epoch": 5.584381551362683,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.027160047180641696,
      "loss": 0.5885,
      "num_input_tokens_seen": 6956552,
      "step": 10655
    },
    {
      "epoch": 5.587002096436059,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.027156029019972806,
      "loss": 0.5153,
      "num_input_tokens_seen": 6959496,
      "step": 10660
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.027152008316401334,
      "loss": 0.631,
      "num_input_tokens_seen": 6962600,
      "step": 10665
    },
    {
      "epoch": 5.5922431865828095,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.027147985070768356,
      "loss": 0.4538,
      "num_input_tokens_seen": 6965160,
      "step": 10670
    },
    {
      "epoch": 5.594863731656185,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02714395928391549,
      "loss": 0.4345,
      "num_input_tokens_seen": 6967880,
      "step": 10675
    },
    {
      "epoch": 5.59748427672956,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.027139930956684894,
      "loss": 0.4473,
      "num_input_tokens_seen": 6971624,
      "step": 10680
    },
    {
      "epoch": 5.600104821802935,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.027135900089919236,
      "loss": 0.5008,
      "num_input_tokens_seen": 6974696,
      "step": 10685
    },
    {
      "epoch": 5.60272536687631,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.02713186668446173,
      "loss": 0.6292,
      "num_input_tokens_seen": 6977544,
      "step": 10690
    },
    {
      "epoch": 5.605345911949685,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02712783074115613,
      "loss": 0.2849,
      "num_input_tokens_seen": 6983688,
      "step": 10695
    },
    {
      "epoch": 5.6079664570230605,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.027123792260846696,
      "loss": 0.4471,
      "num_input_tokens_seen": 6987112,
      "step": 10700
    },
    {
      "epoch": 5.610587002096436,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02711975124437824,
      "loss": 0.5876,
      "num_input_tokens_seen": 6990312,
      "step": 10705
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.027115707692596087,
      "loss": 0.3994,
      "num_input_tokens_seen": 6992456,
      "step": 10710
    },
    {
      "epoch": 5.615828092243187,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02711166160634611,
      "loss": 0.411,
      "num_input_tokens_seen": 6994792,
      "step": 10715
    },
    {
      "epoch": 5.618448637316562,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02710761298647471,
      "loss": 0.4783,
      "num_input_tokens_seen": 6997960,
      "step": 10720
    },
    {
      "epoch": 5.621069182389937,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.027103561833828796,
      "loss": 0.5508,
      "num_input_tokens_seen": 7001640,
      "step": 10725
    },
    {
      "epoch": 5.623689727463312,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02709950814925583,
      "loss": 0.4142,
      "num_input_tokens_seen": 7004456,
      "step": 10730
    },
    {
      "epoch": 5.626310272536688,
      "grad_norm": 0.0625,
      "learning_rate": 0.0270954519336038,
      "loss": 0.507,
      "num_input_tokens_seen": 7006888,
      "step": 10735
    },
    {
      "epoch": 5.628930817610063,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.027091393187721222,
      "loss": 0.4138,
      "num_input_tokens_seen": 7010600,
      "step": 10740
    },
    {
      "epoch": 5.631551362683438,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.027087331912457133,
      "loss": 0.4483,
      "num_input_tokens_seen": 7014312,
      "step": 10745
    },
    {
      "epoch": 5.634171907756813,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.0270832681086611,
      "loss": 0.3827,
      "num_input_tokens_seen": 7017896,
      "step": 10750
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.027079201777183242,
      "loss": 0.5,
      "num_input_tokens_seen": 7021256,
      "step": 10755
    },
    {
      "epoch": 5.639412997903564,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027075132918874176,
      "loss": 0.4554,
      "num_input_tokens_seen": 7024968,
      "step": 10760
    },
    {
      "epoch": 5.6420335429769395,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.027071061534585064,
      "loss": 0.5197,
      "num_input_tokens_seen": 7027528,
      "step": 10765
    },
    {
      "epoch": 5.644654088050315,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.02706698762516759,
      "loss": 0.4357,
      "num_input_tokens_seen": 7030632,
      "step": 10770
    },
    {
      "epoch": 5.64727463312369,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02706291119147397,
      "loss": 0.3983,
      "num_input_tokens_seen": 7034568,
      "step": 10775
    },
    {
      "epoch": 5.649895178197065,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.027058832234356952,
      "loss": 0.4994,
      "num_input_tokens_seen": 7037544,
      "step": 10780
    },
    {
      "epoch": 5.65251572327044,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.027054750754669808,
      "loss": 0.4393,
      "num_input_tokens_seen": 7041256,
      "step": 10785
    },
    {
      "epoch": 5.655136268343815,
      "grad_norm": 0.044921875,
      "learning_rate": 0.027050666753266332,
      "loss": 0.3654,
      "num_input_tokens_seen": 7043528,
      "step": 10790
    },
    {
      "epoch": 5.6577568134171905,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02704658023100085,
      "loss": 0.4437,
      "num_input_tokens_seen": 7046696,
      "step": 10795
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.027042491188728222,
      "loss": 0.5743,
      "num_input_tokens_seen": 7050216,
      "step": 10800
    },
    {
      "epoch": 5.662997903563941,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.027038399627303828,
      "loss": 0.6381,
      "num_input_tokens_seen": 7053832,
      "step": 10805
    },
    {
      "epoch": 5.665618448637317,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.027034305547583567,
      "loss": 0.4523,
      "num_input_tokens_seen": 7056584,
      "step": 10810
    },
    {
      "epoch": 5.668238993710692,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.027030208950423882,
      "loss": 0.3415,
      "num_input_tokens_seen": 7059240,
      "step": 10815
    },
    {
      "epoch": 5.670859538784067,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.027026109836681737,
      "loss": 0.5214,
      "num_input_tokens_seen": 7062280,
      "step": 10820
    },
    {
      "epoch": 5.673480083857442,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02702200820721461,
      "loss": 0.5104,
      "num_input_tokens_seen": 7065288,
      "step": 10825
    },
    {
      "epoch": 5.676100628930818,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.027017904062880523,
      "loss": 0.5028,
      "num_input_tokens_seen": 7069160,
      "step": 10830
    },
    {
      "epoch": 5.678721174004193,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02701379740453801,
      "loss": 0.4825,
      "num_input_tokens_seen": 7072552,
      "step": 10835
    },
    {
      "epoch": 5.681341719077568,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.027009688233046145,
      "loss": 0.3487,
      "num_input_tokens_seen": 7076040,
      "step": 10840
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.027005576549264512,
      "loss": 0.5418,
      "num_input_tokens_seen": 7079368,
      "step": 10845
    },
    {
      "epoch": 5.686582809224319,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02700146235405323,
      "loss": 0.3747,
      "num_input_tokens_seen": 7082632,
      "step": 10850
    },
    {
      "epoch": 5.689203354297694,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026997345648272955,
      "loss": 0.4945,
      "num_input_tokens_seen": 7085192,
      "step": 10855
    },
    {
      "epoch": 5.6918238993710695,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.026993226432784834,
      "loss": 0.3833,
      "num_input_tokens_seen": 7088040,
      "step": 10860
    },
    {
      "epoch": 5.694444444444445,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.02698910470845057,
      "loss": 0.4919,
      "num_input_tokens_seen": 7090664,
      "step": 10865
    },
    {
      "epoch": 5.69706498951782,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02698498047613238,
      "loss": 0.4364,
      "num_input_tokens_seen": 7093224,
      "step": 10870
    },
    {
      "epoch": 5.699685534591195,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.026980853736693004,
      "loss": 0.4568,
      "num_input_tokens_seen": 7096008,
      "step": 10875
    },
    {
      "epoch": 5.70230607966457,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.026976724490995713,
      "loss": 0.426,
      "num_input_tokens_seen": 7098760,
      "step": 10880
    },
    {
      "epoch": 5.704926624737945,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.026972592739904298,
      "loss": 0.5181,
      "num_input_tokens_seen": 7102120,
      "step": 10885
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.026968458484283073,
      "loss": 0.442,
      "num_input_tokens_seen": 7105672,
      "step": 10890
    },
    {
      "epoch": 5.710167714884696,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.02696432172499687,
      "loss": 0.4391,
      "num_input_tokens_seen": 7109288,
      "step": 10895
    },
    {
      "epoch": 5.712788259958071,
      "grad_norm": 0.04296875,
      "learning_rate": 0.026960182462911063,
      "loss": 0.3725,
      "num_input_tokens_seen": 7112264,
      "step": 10900
    },
    {
      "epoch": 5.715408805031447,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.02695604069889154,
      "loss": 0.4011,
      "num_input_tokens_seen": 7114984,
      "step": 10905
    },
    {
      "epoch": 5.718029350104822,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.026951896433804697,
      "loss": 0.4527,
      "num_input_tokens_seen": 7117896,
      "step": 10910
    },
    {
      "epoch": 5.720649895178197,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.026947749668517474,
      "loss": 0.3758,
      "num_input_tokens_seen": 7121192,
      "step": 10915
    },
    {
      "epoch": 5.723270440251572,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02694360040389733,
      "loss": 0.4332,
      "num_input_tokens_seen": 7124680,
      "step": 10920
    },
    {
      "epoch": 5.725890985324948,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.026939448640812238,
      "loss": 0.4369,
      "num_input_tokens_seen": 7127624,
      "step": 10925
    },
    {
      "epoch": 5.728511530398323,
      "grad_norm": 0.03125,
      "learning_rate": 0.026935294380130707,
      "loss": 0.429,
      "num_input_tokens_seen": 7134792,
      "step": 10930
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.026931137622721757,
      "loss": 0.3554,
      "num_input_tokens_seen": 7137448,
      "step": 10935
    },
    {
      "epoch": 5.733752620545073,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.02692697836945493,
      "loss": 0.4935,
      "num_input_tokens_seen": 7141384,
      "step": 10940
    },
    {
      "epoch": 5.736373165618449,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.026922816621200298,
      "loss": 0.5071,
      "num_input_tokens_seen": 7144296,
      "step": 10945
    },
    {
      "epoch": 5.738993710691824,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.026918652378828452,
      "loss": 0.5732,
      "num_input_tokens_seen": 7148328,
      "step": 10950
    },
    {
      "epoch": 5.7416142557651995,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.026914485643210504,
      "loss": 0.5178,
      "num_input_tokens_seen": 7151304,
      "step": 10955
    },
    {
      "epoch": 5.744234800838575,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02691031641521809,
      "loss": 0.3973,
      "num_input_tokens_seen": 7155240,
      "step": 10960
    },
    {
      "epoch": 5.74685534591195,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.026906144695723364,
      "loss": 0.3629,
      "num_input_tokens_seen": 7159496,
      "step": 10965
    },
    {
      "epoch": 5.749475890985325,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.026901970485598992,
      "loss": 0.3703,
      "num_input_tokens_seen": 7163688,
      "step": 10970
    },
    {
      "epoch": 5.7520964360587,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.026897793785718185,
      "loss": 0.3652,
      "num_input_tokens_seen": 7166728,
      "step": 10975
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.026893614596954652,
      "loss": 0.5096,
      "num_input_tokens_seen": 7169928,
      "step": 10980
    },
    {
      "epoch": 5.7573375262054505,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.026889432920182636,
      "loss": 0.5178,
      "num_input_tokens_seen": 7172936,
      "step": 10985
    },
    {
      "epoch": 5.759958071278826,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0268852487562769,
      "loss": 0.3899,
      "num_input_tokens_seen": 7175240,
      "step": 10990
    },
    {
      "epoch": 5.762578616352201,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.026881062106112713,
      "loss": 0.4436,
      "num_input_tokens_seen": 7178216,
      "step": 10995
    },
    {
      "epoch": 5.765199161425577,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.026876872970565877,
      "loss": 0.4953,
      "num_input_tokens_seen": 7182056,
      "step": 11000
    },
    {
      "epoch": 5.767819706498952,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026872681350512722,
      "loss": 0.4428,
      "num_input_tokens_seen": 7186056,
      "step": 11005
    },
    {
      "epoch": 5.770440251572327,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.026868487246830074,
      "loss": 0.5613,
      "num_input_tokens_seen": 7188808,
      "step": 11010
    },
    {
      "epoch": 5.773060796645702,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.026864290660395296,
      "loss": 0.4613,
      "num_input_tokens_seen": 7192328,
      "step": 11015
    },
    {
      "epoch": 5.7756813417190775,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.02686009159208627,
      "loss": 0.545,
      "num_input_tokens_seen": 7195368,
      "step": 11020
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02685589004278139,
      "loss": 0.4642,
      "num_input_tokens_seen": 7199752,
      "step": 11025
    },
    {
      "epoch": 5.780922431865828,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02685168601335957,
      "loss": 0.6198,
      "num_input_tokens_seen": 7203432,
      "step": 11030
    },
    {
      "epoch": 5.783542976939203,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.026847479504700247,
      "loss": 0.3956,
      "num_input_tokens_seen": 7205864,
      "step": 11035
    },
    {
      "epoch": 5.786163522012579,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.026843270517683374,
      "loss": 0.4159,
      "num_input_tokens_seen": 7208424,
      "step": 11040
    },
    {
      "epoch": 5.788784067085954,
      "grad_norm": 0.03125,
      "learning_rate": 0.026839059053189427,
      "loss": 0.7807,
      "num_input_tokens_seen": 7211464,
      "step": 11045
    },
    {
      "epoch": 5.7914046121593294,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.026834845112099392,
      "loss": 0.5209,
      "num_input_tokens_seen": 7214472,
      "step": 11050
    },
    {
      "epoch": 5.794025157232705,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02683062869529478,
      "loss": 0.4899,
      "num_input_tokens_seen": 7218184,
      "step": 11055
    },
    {
      "epoch": 5.79664570230608,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.026826409803657615,
      "loss": 0.4327,
      "num_input_tokens_seen": 7221480,
      "step": 11060
    },
    {
      "epoch": 5.799266247379455,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.026822188438070438,
      "loss": 0.3537,
      "num_input_tokens_seen": 7224424,
      "step": 11065
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02681796459941632,
      "loss": 0.4924,
      "num_input_tokens_seen": 7227144,
      "step": 11070
    },
    {
      "epoch": 5.804507337526205,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02681373828857883,
      "loss": 0.4491,
      "num_input_tokens_seen": 7229768,
      "step": 11075
    },
    {
      "epoch": 5.8071278825995805,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.02680950950644207,
      "loss": 0.3733,
      "num_input_tokens_seen": 7233352,
      "step": 11080
    },
    {
      "epoch": 5.809748427672956,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.026805278253890653,
      "loss": 0.4448,
      "num_input_tokens_seen": 7236328,
      "step": 11085
    },
    {
      "epoch": 5.812368972746331,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.026801044531809703,
      "loss": 0.576,
      "num_input_tokens_seen": 7238952,
      "step": 11090
    },
    {
      "epoch": 5.814989517819707,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026796808341084874,
      "loss": 0.4225,
      "num_input_tokens_seen": 7242568,
      "step": 11095
    },
    {
      "epoch": 5.817610062893082,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.026792569682602324,
      "loss": 0.5555,
      "num_input_tokens_seen": 7245320,
      "step": 11100
    },
    {
      "epoch": 5.820230607966457,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.026788328557248733,
      "loss": 0.393,
      "num_input_tokens_seen": 7248264,
      "step": 11105
    },
    {
      "epoch": 5.822851153039832,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0267840849659113,
      "loss": 0.5673,
      "num_input_tokens_seen": 7251176,
      "step": 11110
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.026779838909477726,
      "loss": 0.4888,
      "num_input_tokens_seen": 7253864,
      "step": 11115
    },
    {
      "epoch": 5.828092243186583,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.026775590388836246,
      "loss": 0.3618,
      "num_input_tokens_seen": 7257704,
      "step": 11120
    },
    {
      "epoch": 5.830712788259958,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0267713394048756,
      "loss": 0.4115,
      "num_input_tokens_seen": 7260936,
      "step": 11125
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026767085958485044,
      "loss": 0.4794,
      "num_input_tokens_seen": 7264456,
      "step": 11130
    },
    {
      "epoch": 5.835953878406709,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.026762830050554354,
      "loss": 0.6296,
      "num_input_tokens_seen": 7267624,
      "step": 11135
    },
    {
      "epoch": 5.838574423480084,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02675857168197381,
      "loss": 0.4262,
      "num_input_tokens_seen": 7270760,
      "step": 11140
    },
    {
      "epoch": 5.841194968553459,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.026754310853634222,
      "loss": 0.4474,
      "num_input_tokens_seen": 7273608,
      "step": 11145
    },
    {
      "epoch": 5.843815513626835,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.026750047566426904,
      "loss": 0.5074,
      "num_input_tokens_seen": 7276264,
      "step": 11150
    },
    {
      "epoch": 5.84643605870021,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02674578182124369,
      "loss": 0.4499,
      "num_input_tokens_seen": 7279336,
      "step": 11155
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.026741513618976916,
      "loss": 0.4709,
      "num_input_tokens_seen": 7282376,
      "step": 11160
    },
    {
      "epoch": 5.85167714884696,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02673724296051945,
      "loss": 0.4277,
      "num_input_tokens_seen": 7285448,
      "step": 11165
    },
    {
      "epoch": 5.854297693920335,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.026732969846764664,
      "loss": 0.5079,
      "num_input_tokens_seen": 7288648,
      "step": 11170
    },
    {
      "epoch": 5.8569182389937104,
      "grad_norm": 0.0546875,
      "learning_rate": 0.026728694278606437,
      "loss": 0.3998,
      "num_input_tokens_seen": 7291432,
      "step": 11175
    },
    {
      "epoch": 5.859538784067086,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.026724416256939183,
      "loss": 0.3084,
      "num_input_tokens_seen": 7296744,
      "step": 11180
    },
    {
      "epoch": 5.862159329140461,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.026720135782657802,
      "loss": 0.3728,
      "num_input_tokens_seen": 7299496,
      "step": 11185
    },
    {
      "epoch": 5.864779874213837,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026715852856657722,
      "loss": 0.4178,
      "num_input_tokens_seen": 7302440,
      "step": 11190
    },
    {
      "epoch": 5.867400419287212,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.026711567479834893,
      "loss": 0.3387,
      "num_input_tokens_seen": 7305864,
      "step": 11195
    },
    {
      "epoch": 5.870020964360587,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.026707279653085756,
      "loss": 0.4792,
      "num_input_tokens_seen": 7309096,
      "step": 11200
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.026702989377307283,
      "loss": 0.6363,
      "num_input_tokens_seen": 7312648,
      "step": 11205
    },
    {
      "epoch": 5.8752620545073375,
      "grad_norm": 0.03515625,
      "learning_rate": 0.026698696653396945,
      "loss": 0.439,
      "num_input_tokens_seen": 7316264,
      "step": 11210
    },
    {
      "epoch": 5.877882599580713,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.026694401482252734,
      "loss": 0.4938,
      "num_input_tokens_seen": 7320584,
      "step": 11215
    },
    {
      "epoch": 5.880503144654088,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.026690103864773145,
      "loss": 0.6643,
      "num_input_tokens_seen": 7323880,
      "step": 11220
    },
    {
      "epoch": 5.883123689727463,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0266858038018572,
      "loss": 0.4608,
      "num_input_tokens_seen": 7326984,
      "step": 11225
    },
    {
      "epoch": 5.885744234800838,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.026681501294404417,
      "loss": 0.4392,
      "num_input_tokens_seen": 7330504,
      "step": 11230
    },
    {
      "epoch": 5.888364779874214,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02667719634331483,
      "loss": 0.4746,
      "num_input_tokens_seen": 7333480,
      "step": 11235
    },
    {
      "epoch": 5.890985324947589,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.026672888949488993,
      "loss": 0.3606,
      "num_input_tokens_seen": 7336232,
      "step": 11240
    },
    {
      "epoch": 5.893605870020965,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02666857911382796,
      "loss": 0.3542,
      "num_input_tokens_seen": 7338920,
      "step": 11245
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026664266837233296,
      "loss": 0.4028,
      "num_input_tokens_seen": 7341512,
      "step": 11250
    },
    {
      "epoch": 5.898846960167715,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.02665995212060708,
      "loss": 0.5838,
      "num_input_tokens_seen": 7345128,
      "step": 11255
    },
    {
      "epoch": 5.90146750524109,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026655634964851904,
      "loss": 0.4232,
      "num_input_tokens_seen": 7348328,
      "step": 11260
    },
    {
      "epoch": 5.904088050314465,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.026651315370870874,
      "loss": 0.6125,
      "num_input_tokens_seen": 7351336,
      "step": 11265
    },
    {
      "epoch": 5.90670859538784,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.02664699333956759,
      "loss": 0.5033,
      "num_input_tokens_seen": 7355208,
      "step": 11270
    },
    {
      "epoch": 5.909329140461216,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.026642668871846173,
      "loss": 0.3656,
      "num_input_tokens_seen": 7357768,
      "step": 11275
    },
    {
      "epoch": 5.911949685534591,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.026638341968611257,
      "loss": 0.4709,
      "num_input_tokens_seen": 7360808,
      "step": 11280
    },
    {
      "epoch": 5.914570230607967,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02663401263076798,
      "loss": 0.4411,
      "num_input_tokens_seen": 7363208,
      "step": 11285
    },
    {
      "epoch": 5.917190775681342,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026629680859221985,
      "loss": 0.3695,
      "num_input_tokens_seen": 7366504,
      "step": 11290
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.026625346654879437,
      "loss": 0.4099,
      "num_input_tokens_seen": 7369800,
      "step": 11295
    },
    {
      "epoch": 5.922431865828092,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.026621010018646998,
      "loss": 0.4576,
      "num_input_tokens_seen": 7372680,
      "step": 11300
    },
    {
      "epoch": 5.9250524109014675,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.026616670951431838,
      "loss": 0.4768,
      "num_input_tokens_seen": 7377480,
      "step": 11305
    },
    {
      "epoch": 5.927672955974843,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.02661232945414165,
      "loss": 0.4049,
      "num_input_tokens_seen": 7380456,
      "step": 11310
    },
    {
      "epoch": 5.930293501048218,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.026607985527684612,
      "loss": 0.5164,
      "num_input_tokens_seen": 7383208,
      "step": 11315
    },
    {
      "epoch": 5.932914046121593,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02660363917296944,
      "loss": 0.4009,
      "num_input_tokens_seen": 7385576,
      "step": 11320
    },
    {
      "epoch": 5.935534591194968,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02659929039090533,
      "loss": 0.3872,
      "num_input_tokens_seen": 7388296,
      "step": 11325
    },
    {
      "epoch": 5.938155136268344,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.026594939182402005,
      "loss": 0.5473,
      "num_input_tokens_seen": 7391784,
      "step": 11330
    },
    {
      "epoch": 5.940775681341719,
      "grad_norm": 0.044921875,
      "learning_rate": 0.026590585548369686,
      "loss": 0.4251,
      "num_input_tokens_seen": 7395016,
      "step": 11335
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.0265862294897191,
      "loss": 0.3611,
      "num_input_tokens_seen": 7398856,
      "step": 11340
    },
    {
      "epoch": 5.94601677148847,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02658187100736149,
      "loss": 0.499,
      "num_input_tokens_seen": 7401864,
      "step": 11345
    },
    {
      "epoch": 5.948637316561845,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.026577510102208594,
      "loss": 0.6496,
      "num_input_tokens_seen": 7404360,
      "step": 11350
    },
    {
      "epoch": 5.95125786163522,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02657314677517267,
      "loss": 0.4906,
      "num_input_tokens_seen": 7407464,
      "step": 11355
    },
    {
      "epoch": 5.953878406708595,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.026568781027166477,
      "loss": 0.5111,
      "num_input_tokens_seen": 7411144,
      "step": 11360
    },
    {
      "epoch": 5.95649895178197,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.026564412859103274,
      "loss": 0.4656,
      "num_input_tokens_seen": 7414376,
      "step": 11365
    },
    {
      "epoch": 5.959119496855346,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.026560042271896834,
      "loss": 0.549,
      "num_input_tokens_seen": 7417672,
      "step": 11370
    },
    {
      "epoch": 5.961740041928721,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.026555669266461437,
      "loss": 0.4319,
      "num_input_tokens_seen": 7420392,
      "step": 11375
    },
    {
      "epoch": 5.964360587002097,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02655129384371186,
      "loss": 0.4121,
      "num_input_tokens_seen": 7423880,
      "step": 11380
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0265469160045634,
      "loss": 0.5248,
      "num_input_tokens_seen": 7427624,
      "step": 11385
    },
    {
      "epoch": 5.969601677148847,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.02654253574993184,
      "loss": 0.3972,
      "num_input_tokens_seen": 7430536,
      "step": 11390
    },
    {
      "epoch": 5.972222222222222,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02653815308073349,
      "loss": 0.5764,
      "num_input_tokens_seen": 7433192,
      "step": 11395
    },
    {
      "epoch": 5.9748427672955975,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02653376799788515,
      "loss": 0.4266,
      "num_input_tokens_seen": 7436104,
      "step": 11400
    },
    {
      "epoch": 5.977463312368973,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.026529380502304127,
      "loss": 0.3883,
      "num_input_tokens_seen": 7439432,
      "step": 11405
    },
    {
      "epoch": 5.980083857442348,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.026524990594908236,
      "loss": 0.4212,
      "num_input_tokens_seen": 7441960,
      "step": 11410
    },
    {
      "epoch": 5.982704402515723,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.0265205982766158,
      "loss": 0.3752,
      "num_input_tokens_seen": 7445576,
      "step": 11415
    },
    {
      "epoch": 5.985324947589098,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.026516203548345635,
      "loss": 0.4502,
      "num_input_tokens_seen": 7448104,
      "step": 11420
    },
    {
      "epoch": 5.987945492662474,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026511806411017073,
      "loss": 0.4992,
      "num_input_tokens_seen": 7451976,
      "step": 11425
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02650740686554994,
      "loss": 0.4366,
      "num_input_tokens_seen": 7456744,
      "step": 11430
    },
    {
      "epoch": 5.993186582809225,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02650300491286458,
      "loss": 0.5281,
      "num_input_tokens_seen": 7460392,
      "step": 11435
    },
    {
      "epoch": 5.9958071278826,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.026498600553881823,
      "loss": 0.4818,
      "num_input_tokens_seen": 7463496,
      "step": 11440
    },
    {
      "epoch": 5.998427672955975,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.026494193789523016,
      "loss": 0.5876,
      "num_input_tokens_seen": 7466376,
      "step": 11445
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.45464035868644714,
      "eval_runtime": 13.6246,
      "eval_samples_per_second": 62.24,
      "eval_steps_per_second": 15.56,
      "num_input_tokens_seen": 7467968,
      "step": 11448
    },
    {
      "epoch": 6.00104821802935,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.026489784620709996,
      "loss": 0.5223,
      "num_input_tokens_seen": 7468896,
      "step": 11450
    },
    {
      "epoch": 6.003668763102725,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.026485373048365117,
      "loss": 0.4762,
      "num_input_tokens_seen": 7472640,
      "step": 11455
    },
    {
      "epoch": 6.0062893081761,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02648095907341123,
      "loss": 0.4275,
      "num_input_tokens_seen": 7476192,
      "step": 11460
    },
    {
      "epoch": 6.008909853249476,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02647654269677168,
      "loss": 0.3858,
      "num_input_tokens_seen": 7479712,
      "step": 11465
    },
    {
      "epoch": 6.011530398322851,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02647212391937034,
      "loss": 0.4317,
      "num_input_tokens_seen": 7482816,
      "step": 11470
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02646770274213155,
      "loss": 0.5379,
      "num_input_tokens_seen": 7485760,
      "step": 11475
    },
    {
      "epoch": 6.016771488469602,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02646327916598018,
      "loss": 0.5099,
      "num_input_tokens_seen": 7489440,
      "step": 11480
    },
    {
      "epoch": 6.019392033542977,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.026458853191841585,
      "loss": 0.286,
      "num_input_tokens_seen": 7492704,
      "step": 11485
    },
    {
      "epoch": 6.022012578616352,
      "grad_norm": 0.0390625,
      "learning_rate": 0.026454424820641635,
      "loss": 0.4653,
      "num_input_tokens_seen": 7495776,
      "step": 11490
    },
    {
      "epoch": 6.0246331236897275,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.026449994053306693,
      "loss": 0.4129,
      "num_input_tokens_seen": 7498560,
      "step": 11495
    },
    {
      "epoch": 6.027253668763103,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02644556089076362,
      "loss": 0.4257,
      "num_input_tokens_seen": 7500992,
      "step": 11500
    },
    {
      "epoch": 6.029874213836478,
      "grad_norm": 0.09375,
      "learning_rate": 0.026441125333939785,
      "loss": 0.5498,
      "num_input_tokens_seen": 7504192,
      "step": 11505
    },
    {
      "epoch": 6.032494758909853,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02643668738376306,
      "loss": 0.4345,
      "num_input_tokens_seen": 7507552,
      "step": 11510
    },
    {
      "epoch": 6.035115303983228,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.026432247041161813,
      "loss": 0.5171,
      "num_input_tokens_seen": 7510592,
      "step": 11515
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.026427804307064907,
      "loss": 0.5156,
      "num_input_tokens_seen": 7513664,
      "step": 11520
    },
    {
      "epoch": 6.040356394129979,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.026423359182401716,
      "loss": 0.486,
      "num_input_tokens_seen": 7516320,
      "step": 11525
    },
    {
      "epoch": 6.0429769392033545,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.02641891166810211,
      "loss": 0.3658,
      "num_input_tokens_seen": 7518944,
      "step": 11530
    },
    {
      "epoch": 6.04559748427673,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.026414461765096455,
      "loss": 0.4289,
      "num_input_tokens_seen": 7521376,
      "step": 11535
    },
    {
      "epoch": 6.048218029350105,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.026410009474315627,
      "loss": 0.4465,
      "num_input_tokens_seen": 7526432,
      "step": 11540
    },
    {
      "epoch": 6.05083857442348,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.026405554796690982,
      "loss": 0.361,
      "num_input_tokens_seen": 7532064,
      "step": 11545
    },
    {
      "epoch": 6.053459119496855,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0264010977331544,
      "loss": 0.3971,
      "num_input_tokens_seen": 7535360,
      "step": 11550
    },
    {
      "epoch": 6.05607966457023,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02639663828463824,
      "loss": 0.3576,
      "num_input_tokens_seen": 7539232,
      "step": 11555
    },
    {
      "epoch": 6.058700209643606,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.026392176452075373,
      "loss": 0.5422,
      "num_input_tokens_seen": 7542464,
      "step": 11560
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.026387712236399154,
      "loss": 0.3718,
      "num_input_tokens_seen": 7546176,
      "step": 11565
    },
    {
      "epoch": 6.063941299790357,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.026383245638543464,
      "loss": 0.5637,
      "num_input_tokens_seen": 7549696,
      "step": 11570
    },
    {
      "epoch": 6.066561844863732,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.026378776659442646,
      "loss": 0.3652,
      "num_input_tokens_seen": 7552544,
      "step": 11575
    },
    {
      "epoch": 6.069182389937107,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.02637430530003157,
      "loss": 0.3724,
      "num_input_tokens_seen": 7555168,
      "step": 11580
    },
    {
      "epoch": 6.071802935010482,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.026369831561245588,
      "loss": 0.6208,
      "num_input_tokens_seen": 7558528,
      "step": 11585
    },
    {
      "epoch": 6.0744234800838575,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02636535544402056,
      "loss": 0.4884,
      "num_input_tokens_seen": 7561760,
      "step": 11590
    },
    {
      "epoch": 6.077044025157233,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.026360876949292837,
      "loss": 0.5175,
      "num_input_tokens_seen": 7565312,
      "step": 11595
    },
    {
      "epoch": 6.079664570230608,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.026356396077999264,
      "loss": 0.4843,
      "num_input_tokens_seen": 7568640,
      "step": 11600
    },
    {
      "epoch": 6.082285115303983,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02635191283107719,
      "loss": 0.4912,
      "num_input_tokens_seen": 7572096,
      "step": 11605
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.026347427209464466,
      "loss": 0.4639,
      "num_input_tokens_seen": 7574944,
      "step": 11610
    },
    {
      "epoch": 6.087526205450734,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.026342939214099425,
      "loss": 0.6123,
      "num_input_tokens_seen": 7578080,
      "step": 11615
    },
    {
      "epoch": 6.090146750524109,
      "grad_norm": 0.09375,
      "learning_rate": 0.026338448845920914,
      "loss": 0.4194,
      "num_input_tokens_seen": 7580128,
      "step": 11620
    },
    {
      "epoch": 6.0927672955974845,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.026333956105868254,
      "loss": 0.4898,
      "num_input_tokens_seen": 7583968,
      "step": 11625
    },
    {
      "epoch": 6.09538784067086,
      "grad_norm": 0.130859375,
      "learning_rate": 0.02632946099488128,
      "loss": 0.3691,
      "num_input_tokens_seen": 7586368,
      "step": 11630
    },
    {
      "epoch": 6.098008385744235,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.02632496351390032,
      "loss": 0.4565,
      "num_input_tokens_seen": 7589728,
      "step": 11635
    },
    {
      "epoch": 6.10062893081761,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0263204636638662,
      "loss": 0.5515,
      "num_input_tokens_seen": 7592224,
      "step": 11640
    },
    {
      "epoch": 6.103249475890985,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02631596144572023,
      "loss": 0.6225,
      "num_input_tokens_seen": 7595648,
      "step": 11645
    },
    {
      "epoch": 6.10587002096436,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02631145686040422,
      "loss": 0.352,
      "num_input_tokens_seen": 7598784,
      "step": 11650
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02630694990886049,
      "loss": 0.3841,
      "num_input_tokens_seen": 7602112,
      "step": 11655
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.026302440592031833,
      "loss": 0.4148,
      "num_input_tokens_seen": 7608768,
      "step": 11660
    },
    {
      "epoch": 6.113731656184487,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.026297928910861544,
      "loss": 0.5628,
      "num_input_tokens_seen": 7612224,
      "step": 11665
    },
    {
      "epoch": 6.116352201257862,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.026293414866293427,
      "loss": 0.3642,
      "num_input_tokens_seen": 7616768,
      "step": 11670
    },
    {
      "epoch": 6.118972746331237,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02628889845927176,
      "loss": 0.445,
      "num_input_tokens_seen": 7619840,
      "step": 11675
    },
    {
      "epoch": 6.121593291404612,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.026284379690741325,
      "loss": 0.4739,
      "num_input_tokens_seen": 7623104,
      "step": 11680
    },
    {
      "epoch": 6.1242138364779874,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0262798585616474,
      "loss": 0.4964,
      "num_input_tokens_seen": 7625856,
      "step": 11685
    },
    {
      "epoch": 6.126834381551363,
      "grad_norm": 0.037109375,
      "learning_rate": 0.026275335072935748,
      "loss": 0.3233,
      "num_input_tokens_seen": 7629600,
      "step": 11690
    },
    {
      "epoch": 6.129454926624738,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.026270809225552638,
      "loss": 0.5148,
      "num_input_tokens_seen": 7632960,
      "step": 11695
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.02626628102044482,
      "loss": 0.4677,
      "num_input_tokens_seen": 7635840,
      "step": 11700
    },
    {
      "epoch": 6.134696016771488,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026261750458559545,
      "loss": 0.5708,
      "num_input_tokens_seen": 7639232,
      "step": 11705
    },
    {
      "epoch": 6.137316561844864,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.026257217540844556,
      "loss": 0.4238,
      "num_input_tokens_seen": 7645888,
      "step": 11710
    },
    {
      "epoch": 6.139937106918239,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.026252682268248086,
      "loss": 0.3709,
      "num_input_tokens_seen": 7648608,
      "step": 11715
    },
    {
      "epoch": 6.1425576519916145,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.026248144641718863,
      "loss": 0.5396,
      "num_input_tokens_seen": 7651072,
      "step": 11720
    },
    {
      "epoch": 6.14517819706499,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02624360466220611,
      "loss": 0.6342,
      "num_input_tokens_seen": 7654304,
      "step": 11725
    },
    {
      "epoch": 6.147798742138365,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.026239062330659528,
      "loss": 0.4554,
      "num_input_tokens_seen": 7657568,
      "step": 11730
    },
    {
      "epoch": 6.15041928721174,
      "grad_norm": 0.033203125,
      "learning_rate": 0.026234517648029334,
      "loss": 0.3684,
      "num_input_tokens_seen": 7661056,
      "step": 11735
    },
    {
      "epoch": 6.153039832285115,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.026229970615266215,
      "loss": 0.3947,
      "num_input_tokens_seen": 7667360,
      "step": 11740
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 0.044921875,
      "learning_rate": 0.026225421233321364,
      "loss": 0.4007,
      "num_input_tokens_seen": 7671200,
      "step": 11745
    },
    {
      "epoch": 6.1582809224318655,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02622086950314646,
      "loss": 0.62,
      "num_input_tokens_seen": 7674144,
      "step": 11750
    },
    {
      "epoch": 6.160901467505241,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.026216315425693665,
      "loss": 0.4088,
      "num_input_tokens_seen": 7677408,
      "step": 11755
    },
    {
      "epoch": 6.163522012578617,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02621175900191565,
      "loss": 0.568,
      "num_input_tokens_seen": 7680160,
      "step": 11760
    },
    {
      "epoch": 6.166142557651992,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02620720023276556,
      "loss": 0.4819,
      "num_input_tokens_seen": 7682848,
      "step": 11765
    },
    {
      "epoch": 6.168763102725367,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.026202639119197037,
      "loss": 0.4636,
      "num_input_tokens_seen": 7685728,
      "step": 11770
    },
    {
      "epoch": 6.171383647798742,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.026198075662164217,
      "loss": 0.3997,
      "num_input_tokens_seen": 7688960,
      "step": 11775
    },
    {
      "epoch": 6.174004192872117,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.026193509862621728,
      "loss": 0.4991,
      "num_input_tokens_seen": 7691872,
      "step": 11780
    },
    {
      "epoch": 6.176624737945493,
      "grad_norm": 0.0546875,
      "learning_rate": 0.026188941721524674,
      "loss": 0.4871,
      "num_input_tokens_seen": 7694336,
      "step": 11785
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.026184371239828665,
      "loss": 0.4564,
      "num_input_tokens_seen": 7697280,
      "step": 11790
    },
    {
      "epoch": 6.181865828092243,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.026179798418489785,
      "loss": 0.4891,
      "num_input_tokens_seen": 7700640,
      "step": 11795
    },
    {
      "epoch": 6.184486373165618,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.026175223258464626,
      "loss": 0.4501,
      "num_input_tokens_seen": 7704064,
      "step": 11800
    },
    {
      "epoch": 6.187106918238993,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.026170645760710255,
      "loss": 0.4776,
      "num_input_tokens_seen": 7707712,
      "step": 11805
    },
    {
      "epoch": 6.189727463312369,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.026166065926184232,
      "loss": 0.4918,
      "num_input_tokens_seen": 7710656,
      "step": 11810
    },
    {
      "epoch": 6.1923480083857445,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.026161483755844605,
      "loss": 0.5381,
      "num_input_tokens_seen": 7713664,
      "step": 11815
    },
    {
      "epoch": 6.19496855345912,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02615689925064992,
      "loss": 0.473,
      "num_input_tokens_seen": 7716000,
      "step": 11820
    },
    {
      "epoch": 6.197589098532495,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.026152312411559187,
      "loss": 0.4754,
      "num_input_tokens_seen": 7718432,
      "step": 11825
    },
    {
      "epoch": 6.20020964360587,
      "grad_norm": 0.068359375,
      "learning_rate": 0.026147723239531934,
      "loss": 0.4078,
      "num_input_tokens_seen": 7721216,
      "step": 11830
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02614313173552816,
      "loss": 0.3982,
      "num_input_tokens_seen": 7724128,
      "step": 11835
    },
    {
      "epoch": 6.20545073375262,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.026138537900508352,
      "loss": 0.3538,
      "num_input_tokens_seen": 7727744,
      "step": 11840
    },
    {
      "epoch": 6.2080712788259955,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.026133941735433493,
      "loss": 0.5352,
      "num_input_tokens_seen": 7730880,
      "step": 11845
    },
    {
      "epoch": 6.210691823899371,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.026129343241265042,
      "loss": 0.4887,
      "num_input_tokens_seen": 7734048,
      "step": 11850
    },
    {
      "epoch": 6.213312368972747,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.02612474241896496,
      "loss": 0.3785,
      "num_input_tokens_seen": 7736832,
      "step": 11855
    },
    {
      "epoch": 6.215932914046122,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02612013926949568,
      "loss": 0.441,
      "num_input_tokens_seen": 7739808,
      "step": 11860
    },
    {
      "epoch": 6.218553459119497,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.02611553379382013,
      "loss": 0.4096,
      "num_input_tokens_seen": 7743072,
      "step": 11865
    },
    {
      "epoch": 6.221174004192872,
      "grad_norm": 0.037109375,
      "learning_rate": 0.026110925992901723,
      "loss": 0.4812,
      "num_input_tokens_seen": 7746336,
      "step": 11870
    },
    {
      "epoch": 6.223794549266247,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.026106315867704358,
      "loss": 0.5238,
      "num_input_tokens_seen": 7748864,
      "step": 11875
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02610170341919242,
      "loss": 0.3746,
      "num_input_tokens_seen": 7751424,
      "step": 11880
    },
    {
      "epoch": 6.229035639412998,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.026097088648330782,
      "loss": 0.5114,
      "num_input_tokens_seen": 7754272,
      "step": 11885
    },
    {
      "epoch": 6.231656184486373,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.026092471556084802,
      "loss": 0.4009,
      "num_input_tokens_seen": 7757920,
      "step": 11890
    },
    {
      "epoch": 6.234276729559748,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.026087852143420322,
      "loss": 0.4019,
      "num_input_tokens_seen": 7761792,
      "step": 11895
    },
    {
      "epoch": 6.236897274633123,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02608323041130367,
      "loss": 0.3501,
      "num_input_tokens_seen": 7764832,
      "step": 11900
    },
    {
      "epoch": 6.239517819706499,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02607860636070166,
      "loss": 0.4368,
      "num_input_tokens_seen": 7768640,
      "step": 11905
    },
    {
      "epoch": 6.2421383647798745,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026073979992581594,
      "loss": 0.3393,
      "num_input_tokens_seen": 7771360,
      "step": 11910
    },
    {
      "epoch": 6.24475890985325,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.026069351307911246,
      "loss": 0.3369,
      "num_input_tokens_seen": 7775104,
      "step": 11915
    },
    {
      "epoch": 6.247379454926625,
      "grad_norm": 0.046875,
      "learning_rate": 0.026064720307658894,
      "loss": 0.428,
      "num_input_tokens_seen": 7782720,
      "step": 11920
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.026060086992793288,
      "loss": 0.3916,
      "num_input_tokens_seen": 7785568,
      "step": 11925
    },
    {
      "epoch": 6.252620545073375,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026055451364283663,
      "loss": 0.4774,
      "num_input_tokens_seen": 7789248,
      "step": 11930
    },
    {
      "epoch": 6.25524109014675,
      "grad_norm": 0.056640625,
      "learning_rate": 0.02605081342309974,
      "loss": 0.3789,
      "num_input_tokens_seen": 7791744,
      "step": 11935
    },
    {
      "epoch": 6.2578616352201255,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02604617317021172,
      "loss": 0.5932,
      "num_input_tokens_seen": 7795008,
      "step": 11940
    },
    {
      "epoch": 6.260482180293501,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.026041530606590296,
      "loss": 0.4886,
      "num_input_tokens_seen": 7798144,
      "step": 11945
    },
    {
      "epoch": 6.263102725366876,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.026036885733206633,
      "loss": 0.378,
      "num_input_tokens_seen": 7800608,
      "step": 11950
    },
    {
      "epoch": 6.265723270440252,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.026032238551032396,
      "loss": 0.5751,
      "num_input_tokens_seen": 7804000,
      "step": 11955
    },
    {
      "epoch": 6.268343815513627,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.026027589061039715,
      "loss": 0.411,
      "num_input_tokens_seen": 7807296,
      "step": 11960
    },
    {
      "epoch": 6.270964360587002,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.026022937264201216,
      "loss": 0.4139,
      "num_input_tokens_seen": 7810816,
      "step": 11965
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02601828316148999,
      "loss": 0.4956,
      "num_input_tokens_seen": 7814976,
      "step": 11970
    },
    {
      "epoch": 6.276205450733753,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.026013626753879634,
      "loss": 0.7052,
      "num_input_tokens_seen": 7818656,
      "step": 11975
    },
    {
      "epoch": 6.278825995807128,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.026008968042344212,
      "loss": 0.4561,
      "num_input_tokens_seen": 7821664,
      "step": 11980
    },
    {
      "epoch": 6.281446540880503,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02600430702785827,
      "loss": 0.358,
      "num_input_tokens_seen": 7824800,
      "step": 11985
    },
    {
      "epoch": 6.284067085953878,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02599964371139684,
      "loss": 0.488,
      "num_input_tokens_seen": 7827456,
      "step": 11990
    },
    {
      "epoch": 6.286687631027253,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.025994978093935438,
      "loss": 0.4375,
      "num_input_tokens_seen": 7832064,
      "step": 11995
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.02599031017645006,
      "loss": 0.4706,
      "num_input_tokens_seen": 7834976,
      "step": 12000
    },
    {
      "epoch": 6.2919287211740045,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.025985639959917175,
      "loss": 0.3798,
      "num_input_tokens_seen": 7837920,
      "step": 12005
    },
    {
      "epoch": 6.29454926624738,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.025980967445313734,
      "loss": 0.449,
      "num_input_tokens_seen": 7840448,
      "step": 12010
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02597629263361719,
      "loss": 0.4302,
      "num_input_tokens_seen": 7843680,
      "step": 12015
    },
    {
      "epoch": 6.29979035639413,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.025971615525805448,
      "loss": 0.4323,
      "num_input_tokens_seen": 7847264,
      "step": 12020
    },
    {
      "epoch": 6.302410901467505,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.025966936122856908,
      "loss": 0.3978,
      "num_input_tokens_seen": 7850144,
      "step": 12025
    },
    {
      "epoch": 6.30503144654088,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.025962254425750453,
      "loss": 0.3997,
      "num_input_tokens_seen": 7853696,
      "step": 12030
    },
    {
      "epoch": 6.3076519916142555,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.02595757043546543,
      "loss": 0.4293,
      "num_input_tokens_seen": 7856736,
      "step": 12035
    },
    {
      "epoch": 6.310272536687631,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.025952884152981685,
      "loss": 0.4481,
      "num_input_tokens_seen": 7860640,
      "step": 12040
    },
    {
      "epoch": 6.312893081761006,
      "grad_norm": 0.033203125,
      "learning_rate": 0.025948195579279534,
      "loss": 0.3617,
      "num_input_tokens_seen": 7863680,
      "step": 12045
    },
    {
      "epoch": 6.315513626834382,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02594350471533977,
      "loss": 0.5545,
      "num_input_tokens_seen": 7866688,
      "step": 12050
    },
    {
      "epoch": 6.318134171907757,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.025938811562143668,
      "loss": 0.3842,
      "num_input_tokens_seen": 7869824,
      "step": 12055
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02593411612067299,
      "loss": 0.412,
      "num_input_tokens_seen": 7872640,
      "step": 12060
    },
    {
      "epoch": 6.323375262054507,
      "grad_norm": 0.048828125,
      "learning_rate": 0.025929418391909955,
      "loss": 0.4886,
      "num_input_tokens_seen": 7875424,
      "step": 12065
    },
    {
      "epoch": 6.325995807127883,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025924718376837286,
      "loss": 0.4492,
      "num_input_tokens_seen": 7879360,
      "step": 12070
    },
    {
      "epoch": 6.328616352201258,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.025920016076438165,
      "loss": 0.4119,
      "num_input_tokens_seen": 7882144,
      "step": 12075
    },
    {
      "epoch": 6.331236897274633,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.025915311491696267,
      "loss": 0.4135,
      "num_input_tokens_seen": 7886176,
      "step": 12080
    },
    {
      "epoch": 6.333857442348008,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.02591060462359573,
      "loss": 0.4099,
      "num_input_tokens_seen": 7889760,
      "step": 12085
    },
    {
      "epoch": 6.336477987421383,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.025905895473121183,
      "loss": 0.5137,
      "num_input_tokens_seen": 7892448,
      "step": 12090
    },
    {
      "epoch": 6.339098532494759,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02590118404125772,
      "loss": 0.4211,
      "num_input_tokens_seen": 7896608,
      "step": 12095
    },
    {
      "epoch": 6.3417190775681345,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.02589647032899093,
      "loss": 0.4729,
      "num_input_tokens_seen": 7899584,
      "step": 12100
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.025891754337306856,
      "loss": 0.4952,
      "num_input_tokens_seen": 7903008,
      "step": 12105
    },
    {
      "epoch": 6.346960167714885,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.025887036067192032,
      "loss": 0.4884,
      "num_input_tokens_seen": 7905696,
      "step": 12110
    },
    {
      "epoch": 6.34958071278826,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.025882315519633475,
      "loss": 0.5206,
      "num_input_tokens_seen": 7908832,
      "step": 12115
    },
    {
      "epoch": 6.352201257861635,
      "grad_norm": 0.078125,
      "learning_rate": 0.025877592695618658,
      "loss": 0.469,
      "num_input_tokens_seen": 7911520,
      "step": 12120
    },
    {
      "epoch": 6.35482180293501,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.025872867596135548,
      "loss": 0.5153,
      "num_input_tokens_seen": 7914848,
      "step": 12125
    },
    {
      "epoch": 6.3574423480083855,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02586814022217258,
      "loss": 0.4703,
      "num_input_tokens_seen": 7918400,
      "step": 12130
    },
    {
      "epoch": 6.360062893081761,
      "grad_norm": 0.04296875,
      "learning_rate": 0.025863410574718672,
      "loss": 0.7049,
      "num_input_tokens_seen": 7921248,
      "step": 12135
    },
    {
      "epoch": 6.362683438155136,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.025858678654763205,
      "loss": 0.5167,
      "num_input_tokens_seen": 7926240,
      "step": 12140
    },
    {
      "epoch": 6.365303983228512,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.025853944463296048,
      "loss": 0.4554,
      "num_input_tokens_seen": 7930176,
      "step": 12145
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.025849208001307536,
      "loss": 0.4361,
      "num_input_tokens_seen": 7933120,
      "step": 12150
    },
    {
      "epoch": 6.370545073375262,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.025844469269788485,
      "loss": 0.5548,
      "num_input_tokens_seen": 7935552,
      "step": 12155
    },
    {
      "epoch": 6.373165618448637,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.025839728269730182,
      "loss": 0.3237,
      "num_input_tokens_seen": 7937824,
      "step": 12160
    },
    {
      "epoch": 6.3757861635220126,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.025834985002124392,
      "loss": 0.5536,
      "num_input_tokens_seen": 7940128,
      "step": 12165
    },
    {
      "epoch": 6.378406708595388,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.02583023946796335,
      "loss": 0.2801,
      "num_input_tokens_seen": 7943008,
      "step": 12170
    },
    {
      "epoch": 6.381027253668763,
      "grad_norm": 0.10546875,
      "learning_rate": 0.025825491668239775,
      "loss": 0.3974,
      "num_input_tokens_seen": 7945472,
      "step": 12175
    },
    {
      "epoch": 6.383647798742138,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.025820741603946838,
      "loss": 0.4008,
      "num_input_tokens_seen": 7948608,
      "step": 12180
    },
    {
      "epoch": 6.386268343815513,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.025815989276078217,
      "loss": 0.3689,
      "num_input_tokens_seen": 7951296,
      "step": 12185
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.025811234685628025,
      "loss": 0.3631,
      "num_input_tokens_seen": 7955488,
      "step": 12190
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.02580647783359088,
      "loss": 0.4388,
      "num_input_tokens_seen": 7958880,
      "step": 12195
    },
    {
      "epoch": 6.39412997903564,
      "grad_norm": 0.0703125,
      "learning_rate": 0.025801718720961864,
      "loss": 0.4559,
      "num_input_tokens_seen": 7961888,
      "step": 12200
    },
    {
      "epoch": 6.396750524109015,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02579695734873652,
      "loss": 0.4651,
      "num_input_tokens_seen": 7964896,
      "step": 12205
    },
    {
      "epoch": 6.39937106918239,
      "grad_norm": 0.037109375,
      "learning_rate": 0.025792193717910878,
      "loss": 0.4628,
      "num_input_tokens_seen": 7968032,
      "step": 12210
    },
    {
      "epoch": 6.401991614255765,
      "grad_norm": 0.041015625,
      "learning_rate": 0.025787427829481437,
      "loss": 0.3651,
      "num_input_tokens_seen": 7971040,
      "step": 12215
    },
    {
      "epoch": 6.40461215932914,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02578265968444516,
      "loss": 0.3893,
      "num_input_tokens_seen": 7974560,
      "step": 12220
    },
    {
      "epoch": 6.4072327044025155,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.02577788928379949,
      "loss": 0.3839,
      "num_input_tokens_seen": 7978144,
      "step": 12225
    },
    {
      "epoch": 6.409853249475891,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02577311662854235,
      "loss": 0.3437,
      "num_input_tokens_seen": 7981024,
      "step": 12230
    },
    {
      "epoch": 6.412473794549266,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02576834171967211,
      "loss": 0.4619,
      "num_input_tokens_seen": 7984448,
      "step": 12235
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.025763564558187638,
      "loss": 0.4793,
      "num_input_tokens_seen": 7987520,
      "step": 12240
    },
    {
      "epoch": 6.417714884696017,
      "grad_norm": 0.048828125,
      "learning_rate": 0.025758785145088257,
      "loss": 0.4592,
      "num_input_tokens_seen": 7990784,
      "step": 12245
    },
    {
      "epoch": 6.420335429769392,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02575400348137376,
      "loss": 0.4705,
      "num_input_tokens_seen": 7993792,
      "step": 12250
    },
    {
      "epoch": 6.422955974842767,
      "grad_norm": 0.025390625,
      "learning_rate": 0.025749219568044435,
      "loss": 0.4581,
      "num_input_tokens_seen": 7996768,
      "step": 12255
    },
    {
      "epoch": 6.4255765199161425,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.025744433406101004,
      "loss": 0.344,
      "num_input_tokens_seen": 7999424,
      "step": 12260
    },
    {
      "epoch": 6.428197064989518,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.025739644996544683,
      "loss": 0.3612,
      "num_input_tokens_seen": 8003456,
      "step": 12265
    },
    {
      "epoch": 6.430817610062893,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02573485434037715,
      "loss": 0.3737,
      "num_input_tokens_seen": 8007360,
      "step": 12270
    },
    {
      "epoch": 6.433438155136268,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.025730061438600564,
      "loss": 0.3113,
      "num_input_tokens_seen": 8010112,
      "step": 12275
    },
    {
      "epoch": 6.436058700209643,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.025725266292217535,
      "loss": 0.4186,
      "num_input_tokens_seen": 8013024,
      "step": 12280
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02572046890223116,
      "loss": 0.57,
      "num_input_tokens_seen": 8015584,
      "step": 12285
    },
    {
      "epoch": 6.441299790356394,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.025715669269644997,
      "loss": 0.4416,
      "num_input_tokens_seen": 8018720,
      "step": 12290
    },
    {
      "epoch": 6.44392033542977,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.025710867395463074,
      "loss": 0.382,
      "num_input_tokens_seen": 8021184,
      "step": 12295
    },
    {
      "epoch": 6.446540880503145,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.025706063280689886,
      "loss": 0.5035,
      "num_input_tokens_seen": 8024416,
      "step": 12300
    },
    {
      "epoch": 6.44916142557652,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.025701256926330403,
      "loss": 0.3544,
      "num_input_tokens_seen": 8028160,
      "step": 12305
    },
    {
      "epoch": 6.451781970649895,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02569644833339006,
      "loss": 0.304,
      "num_input_tokens_seen": 8031296,
      "step": 12310
    },
    {
      "epoch": 6.45440251572327,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.025691637502874752,
      "loss": 0.5358,
      "num_input_tokens_seen": 8034624,
      "step": 12315
    },
    {
      "epoch": 6.4570230607966455,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02568682443579086,
      "loss": 0.4154,
      "num_input_tokens_seen": 8038176,
      "step": 12320
    },
    {
      "epoch": 6.459643605870021,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.025682009133145216,
      "loss": 0.4397,
      "num_input_tokens_seen": 8041920,
      "step": 12325
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.025677191595945125,
      "loss": 0.4314,
      "num_input_tokens_seen": 8044992,
      "step": 12330
    },
    {
      "epoch": 6.464884696016772,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.025672371825198367,
      "loss": 0.4559,
      "num_input_tokens_seen": 8048352,
      "step": 12335
    },
    {
      "epoch": 6.467505241090147,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.025667549821913174,
      "loss": 0.5346,
      "num_input_tokens_seen": 8051168,
      "step": 12340
    },
    {
      "epoch": 6.470125786163522,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02566272558709827,
      "loss": 0.4657,
      "num_input_tokens_seen": 8054848,
      "step": 12345
    },
    {
      "epoch": 6.472746331236897,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.025657899121762813,
      "loss": 0.3991,
      "num_input_tokens_seen": 8057952,
      "step": 12350
    },
    {
      "epoch": 6.4753668763102725,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.025653070426916457,
      "loss": 0.4422,
      "num_input_tokens_seen": 8061344,
      "step": 12355
    },
    {
      "epoch": 6.477987421383648,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.025648239503569297,
      "loss": 0.4674,
      "num_input_tokens_seen": 8064704,
      "step": 12360
    },
    {
      "epoch": 6.480607966457023,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02564340635273192,
      "loss": 0.5146,
      "num_input_tokens_seen": 8067360,
      "step": 12365
    },
    {
      "epoch": 6.483228511530398,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.025638570975415362,
      "loss": 0.5227,
      "num_input_tokens_seen": 8071520,
      "step": 12370
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02563373337263112,
      "loss": 0.4429,
      "num_input_tokens_seen": 8075680,
      "step": 12375
    },
    {
      "epoch": 6.488469601677149,
      "grad_norm": 0.06640625,
      "learning_rate": 0.025628893545391176,
      "loss": 0.8509,
      "num_input_tokens_seen": 8077600,
      "step": 12380
    },
    {
      "epoch": 6.491090146750524,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.025624051494707967,
      "loss": 0.4476,
      "num_input_tokens_seen": 8079872,
      "step": 12385
    },
    {
      "epoch": 6.4937106918239,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.025619207221594387,
      "loss": 0.4226,
      "num_input_tokens_seen": 8082720,
      "step": 12390
    },
    {
      "epoch": 6.496331236897275,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.025614360727063804,
      "loss": 0.4989,
      "num_input_tokens_seen": 8085728,
      "step": 12395
    },
    {
      "epoch": 6.49895178197065,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02560951201213006,
      "loss": 0.442,
      "num_input_tokens_seen": 8089280,
      "step": 12400
    },
    {
      "epoch": 6.501572327044025,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02560466107780743,
      "loss": 0.4487,
      "num_input_tokens_seen": 8092576,
      "step": 12405
    },
    {
      "epoch": 6.5041928721174,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0255998079251107,
      "loss": 0.4896,
      "num_input_tokens_seen": 8096032,
      "step": 12410
    },
    {
      "epoch": 6.506813417190775,
      "grad_norm": 0.060546875,
      "learning_rate": 0.025594952555055072,
      "loss": 0.5138,
      "num_input_tokens_seen": 8099424,
      "step": 12415
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.025590094968656246,
      "loss": 0.4071,
      "num_input_tokens_seen": 8101952,
      "step": 12420
    },
    {
      "epoch": 6.512054507337526,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.025585235166930374,
      "loss": 0.4295,
      "num_input_tokens_seen": 8105472,
      "step": 12425
    },
    {
      "epoch": 6.514675052410902,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.025580373150894065,
      "loss": 0.5862,
      "num_input_tokens_seen": 8108256,
      "step": 12430
    },
    {
      "epoch": 6.517295597484277,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0255755089215644,
      "loss": 0.3827,
      "num_input_tokens_seen": 8111008,
      "step": 12435
    },
    {
      "epoch": 6.519916142557652,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.02557064247995892,
      "loss": 0.5192,
      "num_input_tokens_seen": 8114208,
      "step": 12440
    },
    {
      "epoch": 6.522536687631027,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.025565773827095628,
      "loss": 0.3502,
      "num_input_tokens_seen": 8117376,
      "step": 12445
    },
    {
      "epoch": 6.5251572327044025,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.025560902963993,
      "loss": 0.5486,
      "num_input_tokens_seen": 8120256,
      "step": 12450
    },
    {
      "epoch": 6.527777777777778,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.025556029891669947,
      "loss": 0.3089,
      "num_input_tokens_seen": 8122848,
      "step": 12455
    },
    {
      "epoch": 6.530398322851153,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.025551154611145876,
      "loss": 0.3895,
      "num_input_tokens_seen": 8126336,
      "step": 12460
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.025546277123440632,
      "loss": 0.2891,
      "num_input_tokens_seen": 8129760,
      "step": 12465
    },
    {
      "epoch": 6.535639412997903,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.025541397429574533,
      "loss": 0.3998,
      "num_input_tokens_seen": 8133248,
      "step": 12470
    },
    {
      "epoch": 6.538259958071279,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.025536515530568356,
      "loss": 0.468,
      "num_input_tokens_seen": 8136032,
      "step": 12475
    },
    {
      "epoch": 6.540880503144654,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.025531631427443337,
      "loss": 0.3693,
      "num_input_tokens_seen": 8140224,
      "step": 12480
    },
    {
      "epoch": 6.54350104821803,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02552674512122117,
      "loss": 0.4821,
      "num_input_tokens_seen": 8143424,
      "step": 12485
    },
    {
      "epoch": 6.546121593291405,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.025521856612924023,
      "loss": 0.4111,
      "num_input_tokens_seen": 8146336,
      "step": 12490
    },
    {
      "epoch": 6.54874213836478,
      "grad_norm": 0.048828125,
      "learning_rate": 0.02551696590357451,
      "loss": 0.5037,
      "num_input_tokens_seen": 8148768,
      "step": 12495
    },
    {
      "epoch": 6.551362683438155,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.025512072994195716,
      "loss": 0.3148,
      "num_input_tokens_seen": 8151840,
      "step": 12500
    },
    {
      "epoch": 6.55398322851153,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02550717788581118,
      "loss": 0.4011,
      "num_input_tokens_seen": 8154560,
      "step": 12505
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0255022805794449,
      "loss": 0.4472,
      "num_input_tokens_seen": 8158144,
      "step": 12510
    },
    {
      "epoch": 6.559224318658281,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.02549738107612134,
      "loss": 0.5713,
      "num_input_tokens_seen": 8160672,
      "step": 12515
    },
    {
      "epoch": 6.561844863731656,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02549247937686542,
      "loss": 0.4226,
      "num_input_tokens_seen": 8163744,
      "step": 12520
    },
    {
      "epoch": 6.564465408805032,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02548757548270252,
      "loss": 0.6201,
      "num_input_tokens_seen": 8167104,
      "step": 12525
    },
    {
      "epoch": 6.567085953878407,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.025482669394658475,
      "loss": 0.3909,
      "num_input_tokens_seen": 8170304,
      "step": 12530
    },
    {
      "epoch": 6.569706498951782,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02547776111375959,
      "loss": 0.4188,
      "num_input_tokens_seen": 8173184,
      "step": 12535
    },
    {
      "epoch": 6.572327044025157,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02547285064103261,
      "loss": 0.484,
      "num_input_tokens_seen": 8177472,
      "step": 12540
    },
    {
      "epoch": 6.5749475890985325,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.025467937977504757,
      "loss": 0.4047,
      "num_input_tokens_seen": 8180480,
      "step": 12545
    },
    {
      "epoch": 6.577568134171908,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025463023124203708,
      "loss": 0.3814,
      "num_input_tokens_seen": 8185088,
      "step": 12550
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.025458106082157587,
      "loss": 0.5378,
      "num_input_tokens_seen": 8188384,
      "step": 12555
    },
    {
      "epoch": 6.582809224318658,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.02545318685239499,
      "loss": 0.5137,
      "num_input_tokens_seen": 8191360,
      "step": 12560
    },
    {
      "epoch": 6.585429769392033,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.025448265435944955,
      "loss": 0.3732,
      "num_input_tokens_seen": 8194816,
      "step": 12565
    },
    {
      "epoch": 6.588050314465409,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02544334183383699,
      "loss": 0.3355,
      "num_input_tokens_seen": 8198016,
      "step": 12570
    },
    {
      "epoch": 6.590670859538784,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.025438416047101065,
      "loss": 0.4544,
      "num_input_tokens_seen": 8200448,
      "step": 12575
    },
    {
      "epoch": 6.59329140461216,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.025433488076767586,
      "loss": 0.6091,
      "num_input_tokens_seen": 8203488,
      "step": 12580
    },
    {
      "epoch": 6.595911949685535,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.025428557923867428,
      "loss": 0.4605,
      "num_input_tokens_seen": 8206464,
      "step": 12585
    },
    {
      "epoch": 6.59853249475891,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.025423625589431936,
      "loss": 0.3866,
      "num_input_tokens_seen": 8212928,
      "step": 12590
    },
    {
      "epoch": 6.601153039832285,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02541869107449288,
      "loss": 0.4552,
      "num_input_tokens_seen": 8216128,
      "step": 12595
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.109375,
      "learning_rate": 0.02541375438008252,
      "loss": 0.5393,
      "num_input_tokens_seen": 8218496,
      "step": 12600
    },
    {
      "epoch": 6.606394129979035,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.025408815507233555,
      "loss": 0.2886,
      "num_input_tokens_seen": 8221824,
      "step": 12605
    },
    {
      "epoch": 6.609014675052411,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.02540387445697913,
      "loss": 0.3724,
      "num_input_tokens_seen": 8225376,
      "step": 12610
    },
    {
      "epoch": 6.611635220125786,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02539893123035286,
      "loss": 0.5038,
      "num_input_tokens_seen": 8228224,
      "step": 12615
    },
    {
      "epoch": 6.614255765199162,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.025393985828388824,
      "loss": 0.4998,
      "num_input_tokens_seen": 8230656,
      "step": 12620
    },
    {
      "epoch": 6.616876310272537,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.025389038252121524,
      "loss": 0.3766,
      "num_input_tokens_seen": 8233408,
      "step": 12625
    },
    {
      "epoch": 6.619496855345912,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.025384088502585948,
      "loss": 0.4885,
      "num_input_tokens_seen": 8237216,
      "step": 12630
    },
    {
      "epoch": 6.622117400419287,
      "grad_norm": 0.044921875,
      "learning_rate": 0.025379136580817523,
      "loss": 0.5123,
      "num_input_tokens_seen": 8240000,
      "step": 12635
    },
    {
      "epoch": 6.6247379454926625,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.025374182487852143,
      "loss": 0.4569,
      "num_input_tokens_seen": 8242848,
      "step": 12640
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.025369226224726135,
      "loss": 0.3993,
      "num_input_tokens_seen": 8246304,
      "step": 12645
    },
    {
      "epoch": 6.629979035639413,
      "grad_norm": 0.025390625,
      "learning_rate": 0.025364267792476303,
      "loss": 0.423,
      "num_input_tokens_seen": 8249984,
      "step": 12650
    },
    {
      "epoch": 6.632599580712788,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02535930719213989,
      "loss": 0.5215,
      "num_input_tokens_seen": 8254240,
      "step": 12655
    },
    {
      "epoch": 6.635220125786163,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0253543444247546,
      "loss": 0.4005,
      "num_input_tokens_seen": 8257312,
      "step": 12660
    },
    {
      "epoch": 6.637840670859539,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.025349379491358584,
      "loss": 0.4469,
      "num_input_tokens_seen": 8261216,
      "step": 12665
    },
    {
      "epoch": 6.640461215932914,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.025344412392990446,
      "loss": 0.4289,
      "num_input_tokens_seen": 8264608,
      "step": 12670
    },
    {
      "epoch": 6.6430817610062896,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.025339443130689254,
      "loss": 0.4886,
      "num_input_tokens_seen": 8268384,
      "step": 12675
    },
    {
      "epoch": 6.645702306079665,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.025334471705494518,
      "loss": 0.4346,
      "num_input_tokens_seen": 8272000,
      "step": 12680
    },
    {
      "epoch": 6.64832285115304,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.025329498118446196,
      "loss": 0.3528,
      "num_input_tokens_seen": 8275776,
      "step": 12685
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.02532452237058472,
      "loss": 0.4005,
      "num_input_tokens_seen": 8279008,
      "step": 12690
    },
    {
      "epoch": 6.65356394129979,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02531954446295095,
      "loss": 0.4544,
      "num_input_tokens_seen": 8281952,
      "step": 12695
    },
    {
      "epoch": 6.656184486373165,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025314564396586216,
      "loss": 0.3787,
      "num_input_tokens_seen": 8285472,
      "step": 12700
    },
    {
      "epoch": 6.658805031446541,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02530958217253228,
      "loss": 0.5329,
      "num_input_tokens_seen": 8288416,
      "step": 12705
    },
    {
      "epoch": 6.661425576519916,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.025304597791831373,
      "loss": 0.4622,
      "num_input_tokens_seen": 8292064,
      "step": 12710
    },
    {
      "epoch": 6.664046121593291,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02529961125552617,
      "loss": 0.4637,
      "num_input_tokens_seen": 8295008,
      "step": 12715
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.025294622564659794,
      "loss": 0.4164,
      "num_input_tokens_seen": 8298432,
      "step": 12720
    },
    {
      "epoch": 6.669287211740042,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02528963172027582,
      "loss": 0.4252,
      "num_input_tokens_seen": 8302240,
      "step": 12725
    },
    {
      "epoch": 6.671907756813417,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02528463872341829,
      "loss": 0.5028,
      "num_input_tokens_seen": 8304960,
      "step": 12730
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.025279643575131665,
      "loss": 0.357,
      "num_input_tokens_seen": 8308320,
      "step": 12735
    },
    {
      "epoch": 6.677148846960168,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.025274646276460883,
      "loss": 0.4383,
      "num_input_tokens_seen": 8311360,
      "step": 12740
    },
    {
      "epoch": 6.679769392033543,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.02526964682845132,
      "loss": 0.4881,
      "num_input_tokens_seen": 8315680,
      "step": 12745
    },
    {
      "epoch": 6.682389937106918,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.025264645232148806,
      "loss": 0.5523,
      "num_input_tokens_seen": 8319200,
      "step": 12750
    },
    {
      "epoch": 6.685010482180293,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.025259641488599608,
      "loss": 0.4414,
      "num_input_tokens_seen": 8322816,
      "step": 12755
    },
    {
      "epoch": 6.687631027253669,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.025254635598850463,
      "loss": 0.4717,
      "num_input_tokens_seen": 8325408,
      "step": 12760
    },
    {
      "epoch": 6.690251572327044,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.025249627563948544,
      "loss": 0.3751,
      "num_input_tokens_seen": 8329408,
      "step": 12765
    },
    {
      "epoch": 6.6928721174004195,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.025244617384941468,
      "loss": 0.5942,
      "num_input_tokens_seen": 8332736,
      "step": 12770
    },
    {
      "epoch": 6.695492662473795,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.025239605062877313,
      "loss": 0.2994,
      "num_input_tokens_seen": 8338272,
      "step": 12775
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0252345905988046,
      "loss": 0.6201,
      "num_input_tokens_seen": 8341472,
      "step": 12780
    },
    {
      "epoch": 6.700733752620545,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.025229573993772292,
      "loss": 0.4174,
      "num_input_tokens_seen": 8344288,
      "step": 12785
    },
    {
      "epoch": 6.70335429769392,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.025224555248829814,
      "loss": 0.4267,
      "num_input_tokens_seen": 8347392,
      "step": 12790
    },
    {
      "epoch": 6.705974842767295,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02521953436502702,
      "loss": 0.4128,
      "num_input_tokens_seen": 8350656,
      "step": 12795
    },
    {
      "epoch": 6.7085953878406706,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02521451134341423,
      "loss": 0.4279,
      "num_input_tokens_seen": 8353888,
      "step": 12800
    },
    {
      "epoch": 6.711215932914046,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.025209486185042207,
      "loss": 0.449,
      "num_input_tokens_seen": 8356736,
      "step": 12805
    },
    {
      "epoch": 6.713836477987421,
      "grad_norm": 0.0390625,
      "learning_rate": 0.025204458890962143,
      "loss": 0.4406,
      "num_input_tokens_seen": 8360992,
      "step": 12810
    },
    {
      "epoch": 6.716457023060797,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.025199429462225705,
      "loss": 0.4484,
      "num_input_tokens_seen": 8364000,
      "step": 12815
    },
    {
      "epoch": 6.719077568134172,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.025194397899884976,
      "loss": 0.3836,
      "num_input_tokens_seen": 8366304,
      "step": 12820
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.025189364204992515,
      "loss": 0.3471,
      "num_input_tokens_seen": 8369792,
      "step": 12825
    },
    {
      "epoch": 6.7243186582809225,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02518432837860131,
      "loss": 0.402,
      "num_input_tokens_seen": 8372384,
      "step": 12830
    },
    {
      "epoch": 6.726939203354298,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02517929042176479,
      "loss": 0.4536,
      "num_input_tokens_seen": 8374720,
      "step": 12835
    },
    {
      "epoch": 6.729559748427673,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02517425033553685,
      "loss": 0.5476,
      "num_input_tokens_seen": 8378144,
      "step": 12840
    },
    {
      "epoch": 6.732180293501048,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.025169208120971814,
      "loss": 0.4141,
      "num_input_tokens_seen": 8381536,
      "step": 12845
    },
    {
      "epoch": 6.734800838574423,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.025164163779124456,
      "loss": 0.4445,
      "num_input_tokens_seen": 8384416,
      "step": 12850
    },
    {
      "epoch": 6.737421383647799,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.025159117311049993,
      "loss": 0.3899,
      "num_input_tokens_seen": 8387616,
      "step": 12855
    },
    {
      "epoch": 6.740041928721174,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.02515406871780409,
      "loss": 0.4432,
      "num_input_tokens_seen": 8390208,
      "step": 12860
    },
    {
      "epoch": 6.7426624737945495,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.025149018000442856,
      "loss": 0.3924,
      "num_input_tokens_seen": 8393792,
      "step": 12865
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.025143965160022836,
      "loss": 0.534,
      "num_input_tokens_seen": 8397120,
      "step": 12870
    },
    {
      "epoch": 6.7479035639413,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.025138910197601044,
      "loss": 0.5483,
      "num_input_tokens_seen": 8400160,
      "step": 12875
    },
    {
      "epoch": 6.750524109014675,
      "grad_norm": 0.0390625,
      "learning_rate": 0.025133853114234904,
      "loss": 0.6769,
      "num_input_tokens_seen": 8404032,
      "step": 12880
    },
    {
      "epoch": 6.75314465408805,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.025128793910982303,
      "loss": 0.4717,
      "num_input_tokens_seen": 8407392,
      "step": 12885
    },
    {
      "epoch": 6.755765199161425,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.02512373258890158,
      "loss": 0.4931,
      "num_input_tokens_seen": 8410656,
      "step": 12890
    },
    {
      "epoch": 6.7583857442348005,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.025118669149051486,
      "loss": 0.5126,
      "num_input_tokens_seen": 8413344,
      "step": 12895
    },
    {
      "epoch": 6.761006289308176,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.025113603592491252,
      "loss": 0.4696,
      "num_input_tokens_seen": 8416960,
      "step": 12900
    },
    {
      "epoch": 6.763626834381551,
      "grad_norm": 0.037109375,
      "learning_rate": 0.025108535920280523,
      "loss": 0.5613,
      "num_input_tokens_seen": 8420352,
      "step": 12905
    },
    {
      "epoch": 6.766247379454927,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.025103466133479413,
      "loss": 0.4972,
      "num_input_tokens_seen": 8423360,
      "step": 12910
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.025098394233148448,
      "loss": 0.2954,
      "num_input_tokens_seen": 8426816,
      "step": 12915
    },
    {
      "epoch": 6.771488469601677,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.025093320220348617,
      "loss": 0.571,
      "num_input_tokens_seen": 8429856,
      "step": 12920
    },
    {
      "epoch": 6.774109014675052,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.025088244096141354,
      "loss": 0.5613,
      "num_input_tokens_seen": 8433056,
      "step": 12925
    },
    {
      "epoch": 6.776729559748428,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02508316586158851,
      "loss": 0.5112,
      "num_input_tokens_seen": 8436448,
      "step": 12930
    },
    {
      "epoch": 6.779350104821803,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.025078085517752407,
      "loss": 0.3092,
      "num_input_tokens_seen": 8439200,
      "step": 12935
    },
    {
      "epoch": 6.781970649895178,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.025073003065695793,
      "loss": 0.4747,
      "num_input_tokens_seen": 8443072,
      "step": 12940
    },
    {
      "epoch": 6.784591194968553,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02506791850648185,
      "loss": 0.3867,
      "num_input_tokens_seen": 8445952,
      "step": 12945
    },
    {
      "epoch": 6.787211740041929,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02506283184117422,
      "loss": 0.5567,
      "num_input_tokens_seen": 8448768,
      "step": 12950
    },
    {
      "epoch": 6.789832285115304,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.025057743070836972,
      "loss": 0.3282,
      "num_input_tokens_seen": 8451232,
      "step": 12955
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02505265219653462,
      "loss": 0.3993,
      "num_input_tokens_seen": 8454688,
      "step": 12960
    },
    {
      "epoch": 6.795073375262055,
      "grad_norm": 0.0625,
      "learning_rate": 0.025047559219332112,
      "loss": 0.3293,
      "num_input_tokens_seen": 8458048,
      "step": 12965
    },
    {
      "epoch": 6.79769392033543,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.025042464140294842,
      "loss": 0.4787,
      "num_input_tokens_seen": 8460960,
      "step": 12970
    },
    {
      "epoch": 6.800314465408805,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.025037366960488652,
      "loss": 0.4808,
      "num_input_tokens_seen": 8464448,
      "step": 12975
    },
    {
      "epoch": 6.80293501048218,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.0250322676809798,
      "loss": 0.3381,
      "num_input_tokens_seen": 8468000,
      "step": 12980
    },
    {
      "epoch": 6.805555555555555,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.025027166302835007,
      "loss": 0.4896,
      "num_input_tokens_seen": 8472288,
      "step": 12985
    },
    {
      "epoch": 6.8081761006289305,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02502206282712142,
      "loss": 0.3824,
      "num_input_tokens_seen": 8475616,
      "step": 12990
    },
    {
      "epoch": 6.810796645702306,
      "grad_norm": 0.078125,
      "learning_rate": 0.025016957254906623,
      "loss": 0.5532,
      "num_input_tokens_seen": 8477984,
      "step": 12995
    },
    {
      "epoch": 6.813417190775681,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.025011849587258647,
      "loss": 0.4579,
      "num_input_tokens_seen": 8480864,
      "step": 13000
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.025006739825245962,
      "loss": 0.4856,
      "num_input_tokens_seen": 8483680,
      "step": 13005
    },
    {
      "epoch": 6.818658280922432,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.025001627969937466,
      "loss": 0.4171,
      "num_input_tokens_seen": 8487296,
      "step": 13010
    },
    {
      "epoch": 6.821278825995807,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.024996514022402508,
      "loss": 0.3814,
      "num_input_tokens_seen": 8490528,
      "step": 13015
    },
    {
      "epoch": 6.823899371069182,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.024991397983710858,
      "loss": 0.4784,
      "num_input_tokens_seen": 8493536,
      "step": 13020
    },
    {
      "epoch": 6.826519916142558,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.02498627985493274,
      "loss": 0.552,
      "num_input_tokens_seen": 8497472,
      "step": 13025
    },
    {
      "epoch": 6.829140461215933,
      "grad_norm": 0.05078125,
      "learning_rate": 0.024981159637138806,
      "loss": 0.5307,
      "num_input_tokens_seen": 8500288,
      "step": 13030
    },
    {
      "epoch": 6.831761006289308,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.024976037331400143,
      "loss": 0.4534,
      "num_input_tokens_seen": 8503136,
      "step": 13035
    },
    {
      "epoch": 6.834381551362683,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.024970912938788288,
      "loss": 0.3078,
      "num_input_tokens_seen": 8507168,
      "step": 13040
    },
    {
      "epoch": 6.837002096436059,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.024965786460375197,
      "loss": 0.4705,
      "num_input_tokens_seen": 8509760,
      "step": 13045
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.024960657897233274,
      "loss": 0.3497,
      "num_input_tokens_seen": 8512672,
      "step": 13050
    },
    {
      "epoch": 6.8422431865828095,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02495552725043536,
      "loss": 0.4254,
      "num_input_tokens_seen": 8516320,
      "step": 13055
    },
    {
      "epoch": 6.844863731656185,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02495039452105472,
      "loss": 0.4527,
      "num_input_tokens_seen": 8519456,
      "step": 13060
    },
    {
      "epoch": 6.84748427672956,
      "grad_norm": 0.107421875,
      "learning_rate": 0.02494525971016507,
      "loss": 0.5521,
      "num_input_tokens_seen": 8522560,
      "step": 13065
    },
    {
      "epoch": 6.850104821802935,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02494012281884055,
      "loss": 0.4242,
      "num_input_tokens_seen": 8525792,
      "step": 13070
    },
    {
      "epoch": 6.85272536687631,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.024934983848155734,
      "loss": 0.4179,
      "num_input_tokens_seen": 8528704,
      "step": 13075
    },
    {
      "epoch": 6.855345911949685,
      "grad_norm": 0.05078125,
      "learning_rate": 0.024929842799185645,
      "loss": 0.4251,
      "num_input_tokens_seen": 8532000,
      "step": 13080
    },
    {
      "epoch": 6.8579664570230605,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.02492469967300573,
      "loss": 0.4991,
      "num_input_tokens_seen": 8535136,
      "step": 13085
    },
    {
      "epoch": 6.860587002096436,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.024919554470691867,
      "loss": 0.5304,
      "num_input_tokens_seen": 8538592,
      "step": 13090
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.024914407193320377,
      "loss": 0.4791,
      "num_input_tokens_seen": 8541344,
      "step": 13095
    },
    {
      "epoch": 6.865828092243187,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.024909257841968013,
      "loss": 0.4654,
      "num_input_tokens_seen": 8544576,
      "step": 13100
    },
    {
      "epoch": 6.868448637316562,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02490410641771196,
      "loss": 0.4306,
      "num_input_tokens_seen": 8547680,
      "step": 13105
    },
    {
      "epoch": 6.871069182389937,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.024898952921629833,
      "loss": 0.5201,
      "num_input_tokens_seen": 8550720,
      "step": 13110
    },
    {
      "epoch": 6.873689727463312,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.024893797354799692,
      "loss": 0.4002,
      "num_input_tokens_seen": 8554112,
      "step": 13115
    },
    {
      "epoch": 6.876310272536688,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.024888639718300014,
      "loss": 0.3351,
      "num_input_tokens_seen": 8557856,
      "step": 13120
    },
    {
      "epoch": 6.878930817610063,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.02488348001320973,
      "loss": 0.4764,
      "num_input_tokens_seen": 8560960,
      "step": 13125
    },
    {
      "epoch": 6.881551362683438,
      "grad_norm": 0.048828125,
      "learning_rate": 0.024878318240608176,
      "loss": 0.4357,
      "num_input_tokens_seen": 8564288,
      "step": 13130
    },
    {
      "epoch": 6.884171907756813,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02487315440157515,
      "loss": 0.4289,
      "num_input_tokens_seen": 8568608,
      "step": 13135
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02486798849719086,
      "loss": 0.4159,
      "num_input_tokens_seen": 8571872,
      "step": 13140
    },
    {
      "epoch": 6.889412997903564,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.024862820528535955,
      "loss": 0.5597,
      "num_input_tokens_seen": 8574336,
      "step": 13145
    },
    {
      "epoch": 6.8920335429769395,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02485765049669152,
      "loss": 0.4383,
      "num_input_tokens_seen": 8577248,
      "step": 13150
    },
    {
      "epoch": 6.894654088050315,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.024852478402739058,
      "loss": 0.4319,
      "num_input_tokens_seen": 8581120,
      "step": 13155
    },
    {
      "epoch": 6.89727463312369,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.024847304247760518,
      "loss": 0.3717,
      "num_input_tokens_seen": 8585280,
      "step": 13160
    },
    {
      "epoch": 6.899895178197065,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.024842128032838275,
      "loss": 0.4237,
      "num_input_tokens_seen": 8590464,
      "step": 13165
    },
    {
      "epoch": 6.90251572327044,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.024836949759055132,
      "loss": 0.41,
      "num_input_tokens_seen": 8593408,
      "step": 13170
    },
    {
      "epoch": 6.905136268343815,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.024831769427494328,
      "loss": 0.4208,
      "num_input_tokens_seen": 8596832,
      "step": 13175
    },
    {
      "epoch": 6.9077568134171905,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.024826587039239522,
      "loss": 0.3296,
      "num_input_tokens_seen": 8600000,
      "step": 13180
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 0.06640625,
      "learning_rate": 0.02482140259537482,
      "loss": 0.4717,
      "num_input_tokens_seen": 8602560,
      "step": 13185
    },
    {
      "epoch": 6.912997903563941,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.024816216096984738,
      "loss": 0.5684,
      "num_input_tokens_seen": 8605088,
      "step": 13190
    },
    {
      "epoch": 6.915618448637317,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02481102754515424,
      "loss": 0.512,
      "num_input_tokens_seen": 8608704,
      "step": 13195
    },
    {
      "epoch": 6.918238993710692,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.024805836940968714,
      "loss": 0.4372,
      "num_input_tokens_seen": 8612320,
      "step": 13200
    },
    {
      "epoch": 6.920859538784067,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.024800644285513967,
      "loss": 0.5017,
      "num_input_tokens_seen": 8615360,
      "step": 13205
    },
    {
      "epoch": 6.923480083857442,
      "grad_norm": 0.033203125,
      "learning_rate": 0.024795449579876245,
      "loss": 0.3656,
      "num_input_tokens_seen": 8618464,
      "step": 13210
    },
    {
      "epoch": 6.926100628930818,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.024790252825142224,
      "loss": 0.2893,
      "num_input_tokens_seen": 8621664,
      "step": 13215
    },
    {
      "epoch": 6.928721174004193,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.024785054022399007,
      "loss": 0.3915,
      "num_input_tokens_seen": 8624608,
      "step": 13220
    },
    {
      "epoch": 6.931341719077568,
      "grad_norm": 0.064453125,
      "learning_rate": 0.024779853172734123,
      "loss": 0.3643,
      "num_input_tokens_seen": 8626880,
      "step": 13225
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 0.04296875,
      "learning_rate": 0.02477465027723553,
      "loss": 0.5322,
      "num_input_tokens_seen": 8630880,
      "step": 13230
    },
    {
      "epoch": 6.936582809224319,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.024769445336991613,
      "loss": 0.5001,
      "num_input_tokens_seen": 8633856,
      "step": 13235
    },
    {
      "epoch": 6.939203354297694,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.024764238353091182,
      "loss": 0.5096,
      "num_input_tokens_seen": 8636288,
      "step": 13240
    },
    {
      "epoch": 6.9418238993710695,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.024759029326623494,
      "loss": 0.6296,
      "num_input_tokens_seen": 8638880,
      "step": 13245
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.024753818258678197,
      "loss": 0.5018,
      "num_input_tokens_seen": 8642080,
      "step": 13250
    },
    {
      "epoch": 6.94706498951782,
      "grad_norm": 0.03515625,
      "learning_rate": 0.024748605150345406,
      "loss": 0.3954,
      "num_input_tokens_seen": 8644672,
      "step": 13255
    },
    {
      "epoch": 6.949685534591195,
      "grad_norm": 0.0390625,
      "learning_rate": 0.024743390002715634,
      "loss": 0.3982,
      "num_input_tokens_seen": 8648160,
      "step": 13260
    },
    {
      "epoch": 6.95230607966457,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02473817281687983,
      "loss": 0.3999,
      "num_input_tokens_seen": 8651168,
      "step": 13265
    },
    {
      "epoch": 6.954926624737945,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.024732953593929373,
      "loss": 0.3911,
      "num_input_tokens_seen": 8653984,
      "step": 13270
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.024727732334956065,
      "loss": 0.3613,
      "num_input_tokens_seen": 8656544,
      "step": 13275
    },
    {
      "epoch": 6.960167714884696,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.024722509041052124,
      "loss": 0.3834,
      "num_input_tokens_seen": 8659456,
      "step": 13280
    },
    {
      "epoch": 6.962788259958071,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.024717283713310223,
      "loss": 0.3871,
      "num_input_tokens_seen": 8662112,
      "step": 13285
    },
    {
      "epoch": 6.965408805031447,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.024712056352823423,
      "loss": 0.5129,
      "num_input_tokens_seen": 8665760,
      "step": 13290
    },
    {
      "epoch": 6.968029350104822,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.024706826960685237,
      "loss": 0.4854,
      "num_input_tokens_seen": 8669152,
      "step": 13295
    },
    {
      "epoch": 6.970649895178197,
      "grad_norm": 0.06640625,
      "learning_rate": 0.024701595537989594,
      "loss": 0.453,
      "num_input_tokens_seen": 8671648,
      "step": 13300
    },
    {
      "epoch": 6.973270440251572,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.024696362085830842,
      "loss": 0.5198,
      "num_input_tokens_seen": 8676192,
      "step": 13305
    },
    {
      "epoch": 6.975890985324948,
      "grad_norm": 0.046875,
      "learning_rate": 0.024691126605303766,
      "loss": 0.538,
      "num_input_tokens_seen": 8679008,
      "step": 13310
    },
    {
      "epoch": 6.978511530398323,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.024685889097503567,
      "loss": 0.4224,
      "num_input_tokens_seen": 8682848,
      "step": 13315
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.02468064956352587,
      "loss": 0.3495,
      "num_input_tokens_seen": 8685376,
      "step": 13320
    },
    {
      "epoch": 6.983752620545073,
      "grad_norm": 0.04296875,
      "learning_rate": 0.024675408004466725,
      "loss": 0.593,
      "num_input_tokens_seen": 8688672,
      "step": 13325
    },
    {
      "epoch": 6.986373165618449,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.024670164421422613,
      "loss": 0.49,
      "num_input_tokens_seen": 8692256,
      "step": 13330
    },
    {
      "epoch": 6.988993710691824,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.024664918815490426,
      "loss": 0.5637,
      "num_input_tokens_seen": 8695968,
      "step": 13335
    },
    {
      "epoch": 6.9916142557651995,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.024659671187767483,
      "loss": 0.4225,
      "num_input_tokens_seen": 8699808,
      "step": 13340
    },
    {
      "epoch": 6.994234800838575,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.024654421539351535,
      "loss": 0.5154,
      "num_input_tokens_seen": 8702432,
      "step": 13345
    },
    {
      "epoch": 6.99685534591195,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.02464916987134074,
      "loss": 0.5443,
      "num_input_tokens_seen": 8705792,
      "step": 13350
    },
    {
      "epoch": 6.999475890985325,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.024643916184833695,
      "loss": 0.4743,
      "num_input_tokens_seen": 8709056,
      "step": 13355
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.45409154891967773,
      "eval_runtime": 13.6299,
      "eval_samples_per_second": 62.216,
      "eval_steps_per_second": 15.554,
      "num_input_tokens_seen": 8709168,
      "step": 13356
    },
    {
      "epoch": 7.0020964360587,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.024638660480929404,
      "loss": 0.4821,
      "num_input_tokens_seen": 8711536,
      "step": 13360
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.024633402760727306,
      "loss": 0.5228,
      "num_input_tokens_seen": 8715728,
      "step": 13365
    },
    {
      "epoch": 7.0073375262054505,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02462814302532726,
      "loss": 0.4619,
      "num_input_tokens_seen": 8719216,
      "step": 13370
    },
    {
      "epoch": 7.009958071278826,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02462288127582953,
      "loss": 0.3496,
      "num_input_tokens_seen": 8722160,
      "step": 13375
    },
    {
      "epoch": 7.012578616352202,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.024617617513334817,
      "loss": 0.3301,
      "num_input_tokens_seen": 8725584,
      "step": 13380
    },
    {
      "epoch": 7.015199161425577,
      "grad_norm": 0.041015625,
      "learning_rate": 0.024612351738944254,
      "loss": 0.3919,
      "num_input_tokens_seen": 8728272,
      "step": 13385
    },
    {
      "epoch": 7.017819706498952,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.024607083953759365,
      "loss": 0.4377,
      "num_input_tokens_seen": 8731568,
      "step": 13390
    },
    {
      "epoch": 7.020440251572327,
      "grad_norm": 0.076171875,
      "learning_rate": 0.024601814158882115,
      "loss": 0.4601,
      "num_input_tokens_seen": 8735120,
      "step": 13395
    },
    {
      "epoch": 7.023060796645702,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02459654235541489,
      "loss": 0.33,
      "num_input_tokens_seen": 8737552,
      "step": 13400
    },
    {
      "epoch": 7.0256813417190775,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.024591268544460483,
      "loss": 0.3954,
      "num_input_tokens_seen": 8741712,
      "step": 13405
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.024585992727122124,
      "loss": 0.464,
      "num_input_tokens_seen": 8744464,
      "step": 13410
    },
    {
      "epoch": 7.030922431865828,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.024580714904503444,
      "loss": 0.3976,
      "num_input_tokens_seen": 8748048,
      "step": 13415
    },
    {
      "epoch": 7.033542976939203,
      "grad_norm": 0.03515625,
      "learning_rate": 0.024575435077708513,
      "loss": 0.3571,
      "num_input_tokens_seen": 8751024,
      "step": 13420
    },
    {
      "epoch": 7.036163522012578,
      "grad_norm": 0.0625,
      "learning_rate": 0.024570153247841807,
      "loss": 0.4275,
      "num_input_tokens_seen": 8753264,
      "step": 13425
    },
    {
      "epoch": 7.038784067085954,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02456486941600822,
      "loss": 0.4317,
      "num_input_tokens_seen": 8756240,
      "step": 13430
    },
    {
      "epoch": 7.0414046121593294,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.024559583583313078,
      "loss": 0.3598,
      "num_input_tokens_seen": 8759184,
      "step": 13435
    },
    {
      "epoch": 7.044025157232705,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.024554295750862105,
      "loss": 0.5532,
      "num_input_tokens_seen": 8762864,
      "step": 13440
    },
    {
      "epoch": 7.04664570230608,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.024549005919761467,
      "loss": 0.3263,
      "num_input_tokens_seen": 8767856,
      "step": 13445
    },
    {
      "epoch": 7.049266247379455,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02454371409111773,
      "loss": 0.4781,
      "num_input_tokens_seen": 8771344,
      "step": 13450
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.024538420266037884,
      "loss": 0.4252,
      "num_input_tokens_seen": 8775216,
      "step": 13455
    },
    {
      "epoch": 7.054507337526205,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02453312444562934,
      "loss": 0.495,
      "num_input_tokens_seen": 8779536,
      "step": 13460
    },
    {
      "epoch": 7.0571278825995805,
      "grad_norm": 0.044921875,
      "learning_rate": 0.024527826630999918,
      "loss": 0.5639,
      "num_input_tokens_seen": 8782640,
      "step": 13465
    },
    {
      "epoch": 7.059748427672956,
      "grad_norm": 0.03515625,
      "learning_rate": 0.024522526823257874,
      "loss": 0.6155,
      "num_input_tokens_seen": 8786384,
      "step": 13470
    },
    {
      "epoch": 7.062368972746331,
      "grad_norm": 0.04296875,
      "learning_rate": 0.02451722502351185,
      "loss": 0.3999,
      "num_input_tokens_seen": 8789616,
      "step": 13475
    },
    {
      "epoch": 7.064989517819707,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.024511921232870926,
      "loss": 0.4714,
      "num_input_tokens_seen": 8792816,
      "step": 13480
    },
    {
      "epoch": 7.067610062893082,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.024506615452444605,
      "loss": 0.3916,
      "num_input_tokens_seen": 8796784,
      "step": 13485
    },
    {
      "epoch": 7.070230607966457,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.02450130768334278,
      "loss": 0.4612,
      "num_input_tokens_seen": 8800016,
      "step": 13490
    },
    {
      "epoch": 7.072851153039832,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.024495997926675794,
      "loss": 0.4619,
      "num_input_tokens_seen": 8803088,
      "step": 13495
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.024490686183554374,
      "loss": 0.3896,
      "num_input_tokens_seen": 8806096,
      "step": 13500
    },
    {
      "epoch": 7.078092243186583,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.024485372455089676,
      "loss": 0.4976,
      "num_input_tokens_seen": 8808624,
      "step": 13505
    },
    {
      "epoch": 7.080712788259958,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02448005674239328,
      "loss": 0.5555,
      "num_input_tokens_seen": 8811920,
      "step": 13510
    },
    {
      "epoch": 7.083333333333333,
      "grad_norm": 0.00885009765625,
      "learning_rate": 0.02447473904657717,
      "loss": 0.2871,
      "num_input_tokens_seen": 8817904,
      "step": 13515
    },
    {
      "epoch": 7.085953878406708,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02446941936875374,
      "loss": 0.4046,
      "num_input_tokens_seen": 8820496,
      "step": 13520
    },
    {
      "epoch": 7.088574423480084,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.024464097710035812,
      "loss": 0.4882,
      "num_input_tokens_seen": 8823120,
      "step": 13525
    },
    {
      "epoch": 7.091194968553459,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.024458774071536616,
      "loss": 0.4113,
      "num_input_tokens_seen": 8826192,
      "step": 13530
    },
    {
      "epoch": 7.093815513626835,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.024453448454369794,
      "loss": 0.4645,
      "num_input_tokens_seen": 8829456,
      "step": 13535
    },
    {
      "epoch": 7.09643605870021,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.024448120859649407,
      "loss": 0.5448,
      "num_input_tokens_seen": 8832112,
      "step": 13540
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 0.056640625,
      "learning_rate": 0.024442791288489925,
      "loss": 0.529,
      "num_input_tokens_seen": 8834832,
      "step": 13545
    },
    {
      "epoch": 7.10167714884696,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.024437459742006236,
      "loss": 0.4872,
      "num_input_tokens_seen": 8838480,
      "step": 13550
    },
    {
      "epoch": 7.104297693920335,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.02443212622131364,
      "loss": 0.3671,
      "num_input_tokens_seen": 8842128,
      "step": 13555
    },
    {
      "epoch": 7.1069182389937104,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02442679072752784,
      "loss": 0.3687,
      "num_input_tokens_seen": 8845872,
      "step": 13560
    },
    {
      "epoch": 7.109538784067086,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02442145326176497,
      "loss": 0.407,
      "num_input_tokens_seen": 8848720,
      "step": 13565
    },
    {
      "epoch": 7.112159329140461,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.02441611382514156,
      "loss": 0.3701,
      "num_input_tokens_seen": 8853072,
      "step": 13570
    },
    {
      "epoch": 7.114779874213837,
      "grad_norm": 0.033203125,
      "learning_rate": 0.024410772418774564,
      "loss": 0.4149,
      "num_input_tokens_seen": 8856208,
      "step": 13575
    },
    {
      "epoch": 7.117400419287212,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.024405429043781346,
      "loss": 0.4929,
      "num_input_tokens_seen": 8859600,
      "step": 13580
    },
    {
      "epoch": 7.120020964360587,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02440008370127967,
      "loss": 0.4545,
      "num_input_tokens_seen": 8862576,
      "step": 13585
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.024394736392387723,
      "loss": 0.3956,
      "num_input_tokens_seen": 8866704,
      "step": 13590
    },
    {
      "epoch": 7.1252620545073375,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.024389387118224106,
      "loss": 0.3667,
      "num_input_tokens_seen": 8869584,
      "step": 13595
    },
    {
      "epoch": 7.127882599580713,
      "grad_norm": 0.03125,
      "learning_rate": 0.02438403587990783,
      "loss": 0.4059,
      "num_input_tokens_seen": 8873040,
      "step": 13600
    },
    {
      "epoch": 7.130503144654088,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.024378682678558302,
      "loss": 0.481,
      "num_input_tokens_seen": 8876208,
      "step": 13605
    },
    {
      "epoch": 7.133123689727463,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.024373327515295355,
      "loss": 0.3894,
      "num_input_tokens_seen": 8880240,
      "step": 13610
    },
    {
      "epoch": 7.135744234800838,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.02436797039123923,
      "loss": 0.3988,
      "num_input_tokens_seen": 8883280,
      "step": 13615
    },
    {
      "epoch": 7.138364779874214,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.024362611307510572,
      "loss": 0.4484,
      "num_input_tokens_seen": 8886192,
      "step": 13620
    },
    {
      "epoch": 7.140985324947589,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02435725026523044,
      "loss": 0.3853,
      "num_input_tokens_seen": 8889456,
      "step": 13625
    },
    {
      "epoch": 7.143605870020965,
      "grad_norm": 0.078125,
      "learning_rate": 0.02435188726552031,
      "loss": 0.3781,
      "num_input_tokens_seen": 8892528,
      "step": 13630
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.02434652230950206,
      "loss": 0.4441,
      "num_input_tokens_seen": 8895824,
      "step": 13635
    },
    {
      "epoch": 7.148846960167715,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.024341155398297967,
      "loss": 0.4699,
      "num_input_tokens_seen": 8898544,
      "step": 13640
    },
    {
      "epoch": 7.15146750524109,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.024335786533030736,
      "loss": 0.5474,
      "num_input_tokens_seen": 8902480,
      "step": 13645
    },
    {
      "epoch": 7.154088050314465,
      "grad_norm": 0.0546875,
      "learning_rate": 0.02433041571482347,
      "loss": 0.5628,
      "num_input_tokens_seen": 8905968,
      "step": 13650
    },
    {
      "epoch": 7.15670859538784,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.024325042944799688,
      "loss": 0.4094,
      "num_input_tokens_seen": 8909296,
      "step": 13655
    },
    {
      "epoch": 7.159329140461216,
      "grad_norm": 0.044921875,
      "learning_rate": 0.024319668224083303,
      "loss": 0.3751,
      "num_input_tokens_seen": 8911792,
      "step": 13660
    },
    {
      "epoch": 7.161949685534591,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.02431429155379865,
      "loss": 0.4667,
      "num_input_tokens_seen": 8915280,
      "step": 13665
    },
    {
      "epoch": 7.164570230607967,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.024308912935070467,
      "loss": 0.3715,
      "num_input_tokens_seen": 8918736,
      "step": 13670
    },
    {
      "epoch": 7.167190775681342,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.024303532369023897,
      "loss": 0.4885,
      "num_input_tokens_seen": 8921776,
      "step": 13675
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.024298149856784503,
      "loss": 0.5119,
      "num_input_tokens_seen": 8925264,
      "step": 13680
    },
    {
      "epoch": 7.172431865828092,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.024292765399478236,
      "loss": 0.4613,
      "num_input_tokens_seen": 8928400,
      "step": 13685
    },
    {
      "epoch": 7.1750524109014675,
      "grad_norm": 0.03125,
      "learning_rate": 0.02428737899823146,
      "loss": 0.5281,
      "num_input_tokens_seen": 8931792,
      "step": 13690
    },
    {
      "epoch": 7.177672955974843,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02428199065417096,
      "loss": 0.3517,
      "num_input_tokens_seen": 8934768,
      "step": 13695
    },
    {
      "epoch": 7.180293501048218,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.024276600368423905,
      "loss": 0.2594,
      "num_input_tokens_seen": 8937552,
      "step": 13700
    },
    {
      "epoch": 7.182914046121593,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.024271208142117886,
      "loss": 0.484,
      "num_input_tokens_seen": 8941520,
      "step": 13705
    },
    {
      "epoch": 7.185534591194968,
      "grad_norm": 0.03125,
      "learning_rate": 0.024265813976380897,
      "loss": 0.3027,
      "num_input_tokens_seen": 8945616,
      "step": 13710
    },
    {
      "epoch": 7.188155136268344,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.024260417872341334,
      "loss": 0.5133,
      "num_input_tokens_seen": 8948592,
      "step": 13715
    },
    {
      "epoch": 7.190775681341719,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.024255019831128,
      "loss": 0.4154,
      "num_input_tokens_seen": 8951344,
      "step": 13720
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.024249619853870107,
      "loss": 0.4339,
      "num_input_tokens_seen": 8955056,
      "step": 13725
    },
    {
      "epoch": 7.19601677148847,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.024244217941697264,
      "loss": 0.3586,
      "num_input_tokens_seen": 8958128,
      "step": 13730
    },
    {
      "epoch": 7.198637316561845,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.024238814095739487,
      "loss": 0.3872,
      "num_input_tokens_seen": 8961840,
      "step": 13735
    },
    {
      "epoch": 7.20125786163522,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.024233408317127208,
      "loss": 0.4886,
      "num_input_tokens_seen": 8965040,
      "step": 13740
    },
    {
      "epoch": 7.203878406708595,
      "grad_norm": 0.056640625,
      "learning_rate": 0.024228000606991245,
      "loss": 0.4506,
      "num_input_tokens_seen": 8969200,
      "step": 13745
    },
    {
      "epoch": 7.20649895178197,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.024222590966462836,
      "loss": 0.6212,
      "num_input_tokens_seen": 8972624,
      "step": 13750
    },
    {
      "epoch": 7.209119496855346,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.024217179396673616,
      "loss": 0.4389,
      "num_input_tokens_seen": 8975728,
      "step": 13755
    },
    {
      "epoch": 7.211740041928721,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02421176589875562,
      "loss": 0.5472,
      "num_input_tokens_seen": 8979312,
      "step": 13760
    },
    {
      "epoch": 7.214360587002097,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02420635047384129,
      "loss": 0.4192,
      "num_input_tokens_seen": 8982032,
      "step": 13765
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.024200933123063477,
      "loss": 0.4372,
      "num_input_tokens_seen": 8984432,
      "step": 13770
    },
    {
      "epoch": 7.219601677148847,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.024195513847555417,
      "loss": 0.4948,
      "num_input_tokens_seen": 8988624,
      "step": 13775
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.024190092648450772,
      "loss": 0.6381,
      "num_input_tokens_seen": 8991664,
      "step": 13780
    },
    {
      "epoch": 7.2248427672955975,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.024184669526883595,
      "loss": 0.419,
      "num_input_tokens_seen": 8994928,
      "step": 13785
    },
    {
      "epoch": 7.227463312368973,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.024179244483988328,
      "loss": 0.6058,
      "num_input_tokens_seen": 8997872,
      "step": 13790
    },
    {
      "epoch": 7.230083857442348,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.024173817520899844,
      "loss": 0.4652,
      "num_input_tokens_seen": 9002384,
      "step": 13795
    },
    {
      "epoch": 7.232704402515723,
      "grad_norm": 0.07421875,
      "learning_rate": 0.024168388638753393,
      "loss": 0.3742,
      "num_input_tokens_seen": 9005296,
      "step": 13800
    },
    {
      "epoch": 7.235324947589098,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.024162957838684638,
      "loss": 0.507,
      "num_input_tokens_seen": 9009968,
      "step": 13805
    },
    {
      "epoch": 7.237945492662474,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.02415752512182964,
      "loss": 0.4717,
      "num_input_tokens_seen": 9013776,
      "step": 13810
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 0.05078125,
      "learning_rate": 0.02415209048932486,
      "loss": 0.367,
      "num_input_tokens_seen": 9017072,
      "step": 13815
    },
    {
      "epoch": 7.243186582809225,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.024146653942307163,
      "loss": 0.687,
      "num_input_tokens_seen": 9020272,
      "step": 13820
    },
    {
      "epoch": 7.2458071278826,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02414121548191381,
      "loss": 0.4037,
      "num_input_tokens_seen": 9024528,
      "step": 13825
    },
    {
      "epoch": 7.248427672955975,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02413577510928247,
      "loss": 0.3871,
      "num_input_tokens_seen": 9027376,
      "step": 13830
    },
    {
      "epoch": 7.25104821802935,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0241303328255512,
      "loss": 0.4879,
      "num_input_tokens_seen": 9030832,
      "step": 13835
    },
    {
      "epoch": 7.253668763102725,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.024124888631858472,
      "loss": 0.5794,
      "num_input_tokens_seen": 9034064,
      "step": 13840
    },
    {
      "epoch": 7.2562893081761,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.02411944252934314,
      "loss": 0.4341,
      "num_input_tokens_seen": 9037072,
      "step": 13845
    },
    {
      "epoch": 7.258909853249476,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02411399451914447,
      "loss": 0.3479,
      "num_input_tokens_seen": 9040624,
      "step": 13850
    },
    {
      "epoch": 7.261530398322851,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.02410854460240213,
      "loss": 0.4125,
      "num_input_tokens_seen": 9043792,
      "step": 13855
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.024103092780256175,
      "loss": 0.428,
      "num_input_tokens_seen": 9046640,
      "step": 13860
    },
    {
      "epoch": 7.266771488469602,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.024097639053847056,
      "loss": 0.4338,
      "num_input_tokens_seen": 9049328,
      "step": 13865
    },
    {
      "epoch": 7.269392033542977,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.024092183424315648,
      "loss": 0.4591,
      "num_input_tokens_seen": 9052944,
      "step": 13870
    },
    {
      "epoch": 7.272012578616352,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.02408672589280319,
      "loss": 0.2951,
      "num_input_tokens_seen": 9055344,
      "step": 13875
    },
    {
      "epoch": 7.2746331236897275,
      "grad_norm": 0.03125,
      "learning_rate": 0.02408126646045134,
      "loss": 0.3945,
      "num_input_tokens_seen": 9058928,
      "step": 13880
    },
    {
      "epoch": 7.277253668763103,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.02407580512840216,
      "loss": 0.5334,
      "num_input_tokens_seen": 9062384,
      "step": 13885
    },
    {
      "epoch": 7.279874213836478,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02407034189779808,
      "loss": 0.3737,
      "num_input_tokens_seen": 9066288,
      "step": 13890
    },
    {
      "epoch": 7.282494758909853,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.024064876769781963,
      "loss": 0.4669,
      "num_input_tokens_seen": 9070000,
      "step": 13895
    },
    {
      "epoch": 7.285115303983228,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.02405940974549704,
      "loss": 0.3851,
      "num_input_tokens_seen": 9072784,
      "step": 13900
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02405394082608695,
      "loss": 0.3312,
      "num_input_tokens_seen": 9075632,
      "step": 13905
    },
    {
      "epoch": 7.290356394129979,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.02404847001269574,
      "loss": 0.564,
      "num_input_tokens_seen": 9078448,
      "step": 13910
    },
    {
      "epoch": 7.2929769392033545,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.024042997306467827,
      "loss": 0.4594,
      "num_input_tokens_seen": 9082352,
      "step": 13915
    },
    {
      "epoch": 7.29559748427673,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.02403752270854805,
      "loss": 0.5079,
      "num_input_tokens_seen": 9085104,
      "step": 13920
    },
    {
      "epoch": 7.298218029350105,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02403204622008162,
      "loss": 0.4276,
      "num_input_tokens_seen": 9087856,
      "step": 13925
    },
    {
      "epoch": 7.30083857442348,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.024026567842214172,
      "loss": 0.3335,
      "num_input_tokens_seen": 9091216,
      "step": 13930
    },
    {
      "epoch": 7.303459119496855,
      "grad_norm": 0.064453125,
      "learning_rate": 0.02402108757609171,
      "loss": 0.4168,
      "num_input_tokens_seen": 9094256,
      "step": 13935
    },
    {
      "epoch": 7.30607966457023,
      "grad_norm": 0.0234375,
      "learning_rate": 0.024015605422860645,
      "loss": 0.3529,
      "num_input_tokens_seen": 9098192,
      "step": 13940
    },
    {
      "epoch": 7.308700209643606,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.024010121383667777,
      "loss": 0.4783,
      "num_input_tokens_seen": 9100432,
      "step": 13945
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.024004635459660313,
      "loss": 0.437,
      "num_input_tokens_seen": 9104016,
      "step": 13950
    },
    {
      "epoch": 7.313941299790357,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.02399914765198584,
      "loss": 0.7878,
      "num_input_tokens_seen": 9107088,
      "step": 13955
    },
    {
      "epoch": 7.316561844863732,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.023993657961792345,
      "loss": 0.5131,
      "num_input_tokens_seen": 9110064,
      "step": 13960
    },
    {
      "epoch": 7.319182389937107,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02398816639022821,
      "loss": 0.5052,
      "num_input_tokens_seen": 9113040,
      "step": 13965
    },
    {
      "epoch": 7.321802935010482,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02398267293844221,
      "loss": 0.4374,
      "num_input_tokens_seen": 9116208,
      "step": 13970
    },
    {
      "epoch": 7.3244234800838575,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02397717760758351,
      "loss": 0.4196,
      "num_input_tokens_seen": 9119280,
      "step": 13975
    },
    {
      "epoch": 7.327044025157233,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02397168039880167,
      "loss": 0.4763,
      "num_input_tokens_seen": 9124048,
      "step": 13980
    },
    {
      "epoch": 7.329664570230608,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.023966181313246647,
      "loss": 0.4508,
      "num_input_tokens_seen": 9128080,
      "step": 13985
    },
    {
      "epoch": 7.332285115303983,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.023960680352068782,
      "loss": 0.4568,
      "num_input_tokens_seen": 9131120,
      "step": 13990
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 0.029296875,
      "learning_rate": 0.023955177516418815,
      "loss": 0.4976,
      "num_input_tokens_seen": 9133776,
      "step": 13995
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.023949672807447888,
      "loss": 0.5139,
      "num_input_tokens_seen": 9137680,
      "step": 14000
    },
    {
      "epoch": 7.340146750524109,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.0239441662263075,
      "loss": 0.3792,
      "num_input_tokens_seen": 9140912,
      "step": 14005
    },
    {
      "epoch": 7.3427672955974845,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.023938657774149583,
      "loss": 0.3959,
      "num_input_tokens_seen": 9144432,
      "step": 14010
    },
    {
      "epoch": 7.34538784067086,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.02393314745212644,
      "loss": 0.5058,
      "num_input_tokens_seen": 9147504,
      "step": 14015
    },
    {
      "epoch": 7.348008385744235,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.023927635261390762,
      "loss": 0.4058,
      "num_input_tokens_seen": 9150256,
      "step": 14020
    },
    {
      "epoch": 7.35062893081761,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.023922121203095642,
      "loss": 0.4331,
      "num_input_tokens_seen": 9153552,
      "step": 14025
    },
    {
      "epoch": 7.353249475890985,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.02391660527839455,
      "loss": 0.3329,
      "num_input_tokens_seen": 9156624,
      "step": 14030
    },
    {
      "epoch": 7.35587002096436,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02391108748844136,
      "loss": 0.415,
      "num_input_tokens_seen": 9159792,
      "step": 14035
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02390556783439034,
      "loss": 0.4091,
      "num_input_tokens_seen": 9162960,
      "step": 14040
    },
    {
      "epoch": 7.361111111111111,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02390004631739612,
      "loss": 0.4568,
      "num_input_tokens_seen": 9166416,
      "step": 14045
    },
    {
      "epoch": 7.363731656184487,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.02389452293861375,
      "loss": 0.5468,
      "num_input_tokens_seen": 9169712,
      "step": 14050
    },
    {
      "epoch": 7.366352201257862,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.023888997699198656,
      "loss": 0.4218,
      "num_input_tokens_seen": 9172208,
      "step": 14055
    },
    {
      "epoch": 7.368972746331237,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.023883470600306652,
      "loss": 0.439,
      "num_input_tokens_seen": 9174640,
      "step": 14060
    },
    {
      "epoch": 7.371593291404612,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02387794164309395,
      "loss": 0.4455,
      "num_input_tokens_seen": 9178416,
      "step": 14065
    },
    {
      "epoch": 7.3742138364779874,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02387241082871714,
      "loss": 0.4217,
      "num_input_tokens_seen": 9181040,
      "step": 14070
    },
    {
      "epoch": 7.376834381551363,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.023866878158333205,
      "loss": 0.4443,
      "num_input_tokens_seen": 9184048,
      "step": 14075
    },
    {
      "epoch": 7.379454926624738,
      "grad_norm": 0.046875,
      "learning_rate": 0.023861343633099517,
      "loss": 0.3587,
      "num_input_tokens_seen": 9187344,
      "step": 14080
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.023855807254173838,
      "loss": 0.3408,
      "num_input_tokens_seen": 9190608,
      "step": 14085
    },
    {
      "epoch": 7.384696016771488,
      "grad_norm": 0.056640625,
      "learning_rate": 0.023850269022714317,
      "loss": 0.4075,
      "num_input_tokens_seen": 9194160,
      "step": 14090
    },
    {
      "epoch": 7.387316561844864,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.023844728939879482,
      "loss": 0.4054,
      "num_input_tokens_seen": 9196688,
      "step": 14095
    },
    {
      "epoch": 7.389937106918239,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02383918700682826,
      "loss": 0.4563,
      "num_input_tokens_seen": 9200880,
      "step": 14100
    },
    {
      "epoch": 7.3925576519916145,
      "grad_norm": 0.044921875,
      "learning_rate": 0.023833643224719956,
      "loss": 0.3996,
      "num_input_tokens_seen": 9204720,
      "step": 14105
    },
    {
      "epoch": 7.39517819706499,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.023828097594714274,
      "loss": 0.3979,
      "num_input_tokens_seen": 9207184,
      "step": 14110
    },
    {
      "epoch": 7.397798742138365,
      "grad_norm": 0.04296875,
      "learning_rate": 0.023822550117971287,
      "loss": 0.4468,
      "num_input_tokens_seen": 9210224,
      "step": 14115
    },
    {
      "epoch": 7.40041928721174,
      "grad_norm": 0.064453125,
      "learning_rate": 0.02381700079565147,
      "loss": 0.366,
      "num_input_tokens_seen": 9213776,
      "step": 14120
    },
    {
      "epoch": 7.403039832285115,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.023811449628915675,
      "loss": 0.4636,
      "num_input_tokens_seen": 9216368,
      "step": 14125
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.023805896618925147,
      "loss": 0.6639,
      "num_input_tokens_seen": 9220112,
      "step": 14130
    },
    {
      "epoch": 7.4082809224318655,
      "grad_norm": 0.05859375,
      "learning_rate": 0.023800341766841503,
      "loss": 0.455,
      "num_input_tokens_seen": 9223120,
      "step": 14135
    },
    {
      "epoch": 7.410901467505241,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.02379478507382677,
      "loss": 0.548,
      "num_input_tokens_seen": 9225744,
      "step": 14140
    },
    {
      "epoch": 7.413522012578617,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02378922654104333,
      "loss": 0.3974,
      "num_input_tokens_seen": 9229200,
      "step": 14145
    },
    {
      "epoch": 7.416142557651992,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02378366616965397,
      "loss": 0.4696,
      "num_input_tokens_seen": 9232752,
      "step": 14150
    },
    {
      "epoch": 7.418763102725367,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02377810396082186,
      "loss": 0.3654,
      "num_input_tokens_seen": 9238704,
      "step": 14155
    },
    {
      "epoch": 7.421383647798742,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.023772539915710544,
      "loss": 0.6792,
      "num_input_tokens_seen": 9241648,
      "step": 14160
    },
    {
      "epoch": 7.424004192872117,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.023766974035483963,
      "loss": 0.4561,
      "num_input_tokens_seen": 9244944,
      "step": 14165
    },
    {
      "epoch": 7.426624737945493,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.02376140632130643,
      "loss": 0.4159,
      "num_input_tokens_seen": 9248144,
      "step": 14170
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02375583677434265,
      "loss": 0.4958,
      "num_input_tokens_seen": 9251792,
      "step": 14175
    },
    {
      "epoch": 7.431865828092243,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02375026539575771,
      "loss": 0.517,
      "num_input_tokens_seen": 9255248,
      "step": 14180
    },
    {
      "epoch": 7.434486373165618,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02374469218671708,
      "loss": 0.4326,
      "num_input_tokens_seen": 9258096,
      "step": 14185
    },
    {
      "epoch": 7.437106918238994,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.023739117148386606,
      "loss": 0.4453,
      "num_input_tokens_seen": 9260816,
      "step": 14190
    },
    {
      "epoch": 7.439727463312369,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.02373354028193253,
      "loss": 0.3947,
      "num_input_tokens_seen": 9263472,
      "step": 14195
    },
    {
      "epoch": 7.4423480083857445,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02372796158852146,
      "loss": 0.3904,
      "num_input_tokens_seen": 9266832,
      "step": 14200
    },
    {
      "epoch": 7.44496855345912,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0237223810693204,
      "loss": 0.4537,
      "num_input_tokens_seen": 9270512,
      "step": 14205
    },
    {
      "epoch": 7.447589098532495,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.02371679872549673,
      "loss": 0.4686,
      "num_input_tokens_seen": 9274032,
      "step": 14210
    },
    {
      "epoch": 7.45020964360587,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.023711214558218216,
      "loss": 0.415,
      "num_input_tokens_seen": 9276720,
      "step": 14215
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.023705628568652992,
      "loss": 0.3886,
      "num_input_tokens_seen": 9279504,
      "step": 14220
    },
    {
      "epoch": 7.45545073375262,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0237000407579696,
      "loss": 0.3953,
      "num_input_tokens_seen": 9282864,
      "step": 14225
    },
    {
      "epoch": 7.4580712788259955,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.023694451127336934,
      "loss": 0.5204,
      "num_input_tokens_seen": 9285520,
      "step": 14230
    },
    {
      "epoch": 7.460691823899371,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.023688859677924282,
      "loss": 0.4689,
      "num_input_tokens_seen": 9289840,
      "step": 14235
    },
    {
      "epoch": 7.463312368972747,
      "grad_norm": 0.0546875,
      "learning_rate": 0.023683266410901314,
      "loss": 0.3633,
      "num_input_tokens_seen": 9292016,
      "step": 14240
    },
    {
      "epoch": 7.465932914046122,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.023677671327438077,
      "loss": 0.3986,
      "num_input_tokens_seen": 9295312,
      "step": 14245
    },
    {
      "epoch": 7.468553459119497,
      "grad_norm": 0.0859375,
      "learning_rate": 0.023672074428704996,
      "loss": 0.5646,
      "num_input_tokens_seen": 9298032,
      "step": 14250
    },
    {
      "epoch": 7.471174004192872,
      "grad_norm": 0.05078125,
      "learning_rate": 0.02366647571587288,
      "loss": 0.5107,
      "num_input_tokens_seen": 9301552,
      "step": 14255
    },
    {
      "epoch": 7.473794549266247,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.02366087519011292,
      "loss": 0.3661,
      "num_input_tokens_seen": 9304688,
      "step": 14260
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.023655272852596675,
      "loss": 0.522,
      "num_input_tokens_seen": 9307664,
      "step": 14265
    },
    {
      "epoch": 7.479035639412998,
      "grad_norm": 0.048828125,
      "learning_rate": 0.023649668704496103,
      "loss": 0.423,
      "num_input_tokens_seen": 9310736,
      "step": 14270
    },
    {
      "epoch": 7.481656184486373,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.02364406274698351,
      "loss": 0.4424,
      "num_input_tokens_seen": 9315056,
      "step": 14275
    },
    {
      "epoch": 7.484276729559748,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.023638454981231606,
      "loss": 0.4091,
      "num_input_tokens_seen": 9318672,
      "step": 14280
    },
    {
      "epoch": 7.486897274633124,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.02363284540841348,
      "loss": 0.5596,
      "num_input_tokens_seen": 9325040,
      "step": 14285
    },
    {
      "epoch": 7.489517819706499,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.023627234029702578,
      "loss": 0.3545,
      "num_input_tokens_seen": 9328496,
      "step": 14290
    },
    {
      "epoch": 7.4921383647798745,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.023621620846272742,
      "loss": 0.4653,
      "num_input_tokens_seen": 9331824,
      "step": 14295
    },
    {
      "epoch": 7.49475890985325,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.023616005859298187,
      "loss": 0.3248,
      "num_input_tokens_seen": 9334768,
      "step": 14300
    },
    {
      "epoch": 7.497379454926625,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0236103890699535,
      "loss": 0.4983,
      "num_input_tokens_seen": 9337616,
      "step": 14305
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02360477047941366,
      "loss": 0.4907,
      "num_input_tokens_seen": 9340304,
      "step": 14310
    },
    {
      "epoch": 7.502620545073375,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.023599150088853996,
      "loss": 0.4898,
      "num_input_tokens_seen": 9343440,
      "step": 14315
    },
    {
      "epoch": 7.50524109014675,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02359352789945024,
      "loss": 0.4912,
      "num_input_tokens_seen": 9346864,
      "step": 14320
    },
    {
      "epoch": 7.5078616352201255,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.023587903912378486,
      "loss": 0.5252,
      "num_input_tokens_seen": 9350384,
      "step": 14325
    },
    {
      "epoch": 7.510482180293501,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.023582278128815214,
      "loss": 0.3948,
      "num_input_tokens_seen": 9354256,
      "step": 14330
    },
    {
      "epoch": 7.513102725366876,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.023576650549937263,
      "loss": 0.4092,
      "num_input_tokens_seen": 9357456,
      "step": 14335
    },
    {
      "epoch": 7.515723270440252,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02357102117692187,
      "loss": 0.3994,
      "num_input_tokens_seen": 9360560,
      "step": 14340
    },
    {
      "epoch": 7.518343815513627,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.023565390010946628,
      "loss": 0.451,
      "num_input_tokens_seen": 9363248,
      "step": 14345
    },
    {
      "epoch": 7.520964360587002,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.023559757053189512,
      "loss": 0.3738,
      "num_input_tokens_seen": 9367088,
      "step": 14350
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02355412230482888,
      "loss": 0.392,
      "num_input_tokens_seen": 9370960,
      "step": 14355
    },
    {
      "epoch": 7.526205450733753,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02354848576704345,
      "loss": 0.421,
      "num_input_tokens_seen": 9373936,
      "step": 14360
    },
    {
      "epoch": 7.528825995807128,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.023542847441012326,
      "loss": 0.3468,
      "num_input_tokens_seen": 9377264,
      "step": 14365
    },
    {
      "epoch": 7.531446540880503,
      "grad_norm": 0.046875,
      "learning_rate": 0.023537207327914977,
      "loss": 0.5058,
      "num_input_tokens_seen": 9380240,
      "step": 14370
    },
    {
      "epoch": 7.534067085953878,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.023531565428931256,
      "loss": 0.3012,
      "num_input_tokens_seen": 9385136,
      "step": 14375
    },
    {
      "epoch": 7.536687631027254,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.023525921745241377,
      "loss": 0.5055,
      "num_input_tokens_seen": 9387760,
      "step": 14380
    },
    {
      "epoch": 7.539308176100629,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.023520276278025935,
      "loss": 0.5396,
      "num_input_tokens_seen": 9391184,
      "step": 14385
    },
    {
      "epoch": 7.5419287211740045,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.023514629028465905,
      "loss": 0.4266,
      "num_input_tokens_seen": 9393776,
      "step": 14390
    },
    {
      "epoch": 7.54454926624738,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.023508979997742624,
      "loss": 0.4982,
      "num_input_tokens_seen": 9397008,
      "step": 14395
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0235033291870378,
      "loss": 0.5529,
      "num_input_tokens_seen": 9400144,
      "step": 14400
    },
    {
      "epoch": 7.54979035639413,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.023497676597533518,
      "loss": 0.3286,
      "num_input_tokens_seen": 9403728,
      "step": 14405
    },
    {
      "epoch": 7.552410901467505,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.023492022230412246,
      "loss": 0.5678,
      "num_input_tokens_seen": 9408112,
      "step": 14410
    },
    {
      "epoch": 7.55503144654088,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0234863660868568,
      "loss": 0.5343,
      "num_input_tokens_seen": 9411152,
      "step": 14415
    },
    {
      "epoch": 7.5576519916142555,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.023480708168050386,
      "loss": 0.3196,
      "num_input_tokens_seen": 9414416,
      "step": 14420
    },
    {
      "epoch": 7.560272536687631,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.023475048475176577,
      "loss": 0.411,
      "num_input_tokens_seen": 9417168,
      "step": 14425
    },
    {
      "epoch": 7.562893081761006,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.023469387009419317,
      "loss": 0.3699,
      "num_input_tokens_seen": 9420880,
      "step": 14430
    },
    {
      "epoch": 7.565513626834382,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.023463723771962916,
      "loss": 0.4192,
      "num_input_tokens_seen": 9424144,
      "step": 14435
    },
    {
      "epoch": 7.568134171907757,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.023458058763992064,
      "loss": 0.4278,
      "num_input_tokens_seen": 9426448,
      "step": 14440
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.023452391986691815,
      "loss": 0.5017,
      "num_input_tokens_seen": 9429456,
      "step": 14445
    },
    {
      "epoch": 7.573375262054507,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.02344672344124759,
      "loss": 0.497,
      "num_input_tokens_seen": 9432720,
      "step": 14450
    },
    {
      "epoch": 7.575995807127883,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.023441053128845184,
      "loss": 0.5539,
      "num_input_tokens_seen": 9435344,
      "step": 14455
    },
    {
      "epoch": 7.578616352201258,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.023435381050670768,
      "loss": 0.5496,
      "num_input_tokens_seen": 9438416,
      "step": 14460
    },
    {
      "epoch": 7.581236897274633,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.02342970720791087,
      "loss": 0.4145,
      "num_input_tokens_seen": 9440912,
      "step": 14465
    },
    {
      "epoch": 7.583857442348008,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.023424031601752392,
      "loss": 0.2637,
      "num_input_tokens_seen": 9444272,
      "step": 14470
    },
    {
      "epoch": 7.586477987421384,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02341835423338261,
      "loss": 0.5609,
      "num_input_tokens_seen": 9446704,
      "step": 14475
    },
    {
      "epoch": 7.589098532494759,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.02341267510398917,
      "loss": 0.4311,
      "num_input_tokens_seen": 9450128,
      "step": 14480
    },
    {
      "epoch": 7.5917190775681345,
      "grad_norm": 0.04296875,
      "learning_rate": 0.02340699421476007,
      "loss": 0.5464,
      "num_input_tokens_seen": 9453072,
      "step": 14485
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.023401311566883697,
      "loss": 0.3562,
      "num_input_tokens_seen": 9455856,
      "step": 14490
    },
    {
      "epoch": 7.596960167714885,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.023395627161548787,
      "loss": 0.5428,
      "num_input_tokens_seen": 9458736,
      "step": 14495
    },
    {
      "epoch": 7.59958071278826,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02338994099994446,
      "loss": 0.3244,
      "num_input_tokens_seen": 9461840,
      "step": 14500
    },
    {
      "epoch": 7.602201257861635,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02338425308326019,
      "loss": 0.4137,
      "num_input_tokens_seen": 9468368,
      "step": 14505
    },
    {
      "epoch": 7.60482180293501,
      "grad_norm": 0.0625,
      "learning_rate": 0.023378563412685835,
      "loss": 0.567,
      "num_input_tokens_seen": 9470928,
      "step": 14510
    },
    {
      "epoch": 7.6074423480083855,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.023372871989411594,
      "loss": 0.6699,
      "num_input_tokens_seen": 9475024,
      "step": 14515
    },
    {
      "epoch": 7.610062893081761,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.023367178814628055,
      "loss": 0.4873,
      "num_input_tokens_seen": 9479408,
      "step": 14520
    },
    {
      "epoch": 7.612683438155136,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.023361483889526174,
      "loss": 0.3814,
      "num_input_tokens_seen": 9482480,
      "step": 14525
    },
    {
      "epoch": 7.615303983228512,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.023355787215297248,
      "loss": 0.3525,
      "num_input_tokens_seen": 9484816,
      "step": 14530
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.023350088793132964,
      "loss": 0.4767,
      "num_input_tokens_seen": 9487728,
      "step": 14535
    },
    {
      "epoch": 7.620545073375262,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.023344388624225377,
      "loss": 0.6166,
      "num_input_tokens_seen": 9490448,
      "step": 14540
    },
    {
      "epoch": 7.623165618448637,
      "grad_norm": 0.033203125,
      "learning_rate": 0.023338686709766878,
      "loss": 0.4721,
      "num_input_tokens_seen": 9493360,
      "step": 14545
    },
    {
      "epoch": 7.6257861635220126,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.023332983050950256,
      "loss": 0.3632,
      "num_input_tokens_seen": 9496560,
      "step": 14550
    },
    {
      "epoch": 7.628406708595388,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.023327277648968647,
      "loss": 0.333,
      "num_input_tokens_seen": 9499632,
      "step": 14555
    },
    {
      "epoch": 7.631027253668763,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.023321570505015557,
      "loss": 0.4622,
      "num_input_tokens_seen": 9503696,
      "step": 14560
    },
    {
      "epoch": 7.633647798742138,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.023315861620284855,
      "loss": 0.5127,
      "num_input_tokens_seen": 9506768,
      "step": 14565
    },
    {
      "epoch": 7.636268343815514,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.023310150995970773,
      "loss": 0.3864,
      "num_input_tokens_seen": 9509488,
      "step": 14570
    },
    {
      "epoch": 7.638888888888889,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02330443863326791,
      "loss": 0.397,
      "num_input_tokens_seen": 9512336,
      "step": 14575
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02329872453337123,
      "loss": 0.5461,
      "num_input_tokens_seen": 9515376,
      "step": 14580
    },
    {
      "epoch": 7.64412997903564,
      "grad_norm": 0.044921875,
      "learning_rate": 0.02329300869747605,
      "loss": 0.3685,
      "num_input_tokens_seen": 9518160,
      "step": 14585
    },
    {
      "epoch": 7.646750524109015,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.023287291126778065,
      "loss": 0.4105,
      "num_input_tokens_seen": 9520912,
      "step": 14590
    },
    {
      "epoch": 7.64937106918239,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023281571822473323,
      "loss": 0.385,
      "num_input_tokens_seen": 9523728,
      "step": 14595
    },
    {
      "epoch": 7.651991614255765,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.02327585078575823,
      "loss": 0.5635,
      "num_input_tokens_seen": 9526416,
      "step": 14600
    },
    {
      "epoch": 7.65461215932914,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.02327012801782958,
      "loss": 0.458,
      "num_input_tokens_seen": 9529936,
      "step": 14605
    },
    {
      "epoch": 7.6572327044025155,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.023264403519884493,
      "loss": 0.4724,
      "num_input_tokens_seen": 9533232,
      "step": 14610
    },
    {
      "epoch": 7.659853249475891,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.023258677293120474,
      "loss": 0.3986,
      "num_input_tokens_seen": 9536656,
      "step": 14615
    },
    {
      "epoch": 7.662473794549266,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.023252949338735385,
      "loss": 0.4884,
      "num_input_tokens_seen": 9539344,
      "step": 14620
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02324721965792745,
      "loss": 0.4908,
      "num_input_tokens_seen": 9542832,
      "step": 14625
    },
    {
      "epoch": 7.667714884696017,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.023241488251895252,
      "loss": 0.4796,
      "num_input_tokens_seen": 9546640,
      "step": 14630
    },
    {
      "epoch": 7.670335429769392,
      "grad_norm": 0.037109375,
      "learning_rate": 0.023235755121837738,
      "loss": 0.3359,
      "num_input_tokens_seen": 9550000,
      "step": 14635
    },
    {
      "epoch": 7.672955974842767,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.023230020268954207,
      "loss": 0.4679,
      "num_input_tokens_seen": 9553424,
      "step": 14640
    },
    {
      "epoch": 7.6755765199161425,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.023224283694444334,
      "loss": 0.4639,
      "num_input_tokens_seen": 9556144,
      "step": 14645
    },
    {
      "epoch": 7.678197064989518,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.023218545399508135,
      "loss": 0.4011,
      "num_input_tokens_seen": 9559024,
      "step": 14650
    },
    {
      "epoch": 7.680817610062893,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.023212805385346,
      "loss": 0.5185,
      "num_input_tokens_seen": 9562640,
      "step": 14655
    },
    {
      "epoch": 7.683438155136268,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.02320706365315868,
      "loss": 0.4051,
      "num_input_tokens_seen": 9565360,
      "step": 14660
    },
    {
      "epoch": 7.686058700209644,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.023201320204147273,
      "loss": 0.463,
      "num_input_tokens_seen": 9568688,
      "step": 14665
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.023195575039513243,
      "loss": 0.3674,
      "num_input_tokens_seen": 9571568,
      "step": 14670
    },
    {
      "epoch": 7.691299790356394,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.023189828160458413,
      "loss": 0.4432,
      "num_input_tokens_seen": 9574128,
      "step": 14675
    },
    {
      "epoch": 7.69392033542977,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02318407956818497,
      "loss": 0.4458,
      "num_input_tokens_seen": 9577680,
      "step": 14680
    },
    {
      "epoch": 7.696540880503145,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.02317832926389545,
      "loss": 0.4082,
      "num_input_tokens_seen": 9580656,
      "step": 14685
    },
    {
      "epoch": 7.69916142557652,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02317257724879275,
      "loss": 0.4744,
      "num_input_tokens_seen": 9584496,
      "step": 14690
    },
    {
      "epoch": 7.701781970649895,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02316682352408013,
      "loss": 0.4909,
      "num_input_tokens_seen": 9591056,
      "step": 14695
    },
    {
      "epoch": 7.70440251572327,
      "grad_norm": 0.048828125,
      "learning_rate": 0.023161068090961197,
      "loss": 0.4038,
      "num_input_tokens_seen": 9594320,
      "step": 14700
    },
    {
      "epoch": 7.7070230607966455,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023155310950639923,
      "loss": 0.4432,
      "num_input_tokens_seen": 9597616,
      "step": 14705
    },
    {
      "epoch": 7.709643605870021,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.023149552104320646,
      "loss": 0.4736,
      "num_input_tokens_seen": 9601072,
      "step": 14710
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.023143791553208042,
      "loss": 0.4189,
      "num_input_tokens_seen": 9604752,
      "step": 14715
    },
    {
      "epoch": 7.714884696016772,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02313802929850715,
      "loss": 0.4123,
      "num_input_tokens_seen": 9607600,
      "step": 14720
    },
    {
      "epoch": 7.717505241090147,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.023132265341423378,
      "loss": 0.4548,
      "num_input_tokens_seen": 9611184,
      "step": 14725
    },
    {
      "epoch": 7.720125786163522,
      "grad_norm": 0.046875,
      "learning_rate": 0.02312649968316247,
      "loss": 0.4324,
      "num_input_tokens_seen": 9614288,
      "step": 14730
    },
    {
      "epoch": 7.722746331236897,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.023120732324930546,
      "loss": 0.4524,
      "num_input_tokens_seen": 9617680,
      "step": 14735
    },
    {
      "epoch": 7.7253668763102725,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.023114963267934063,
      "loss": 0.3633,
      "num_input_tokens_seen": 9620944,
      "step": 14740
    },
    {
      "epoch": 7.727987421383648,
      "grad_norm": 0.068359375,
      "learning_rate": 0.023109192513379843,
      "loss": 0.4315,
      "num_input_tokens_seen": 9627536,
      "step": 14745
    },
    {
      "epoch": 7.730607966457023,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.023103420062475063,
      "loss": 0.5327,
      "num_input_tokens_seen": 9631216,
      "step": 14750
    },
    {
      "epoch": 7.733228511530398,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.023097645916427257,
      "loss": 0.3571,
      "num_input_tokens_seen": 9634160,
      "step": 14755
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.023091870076444306,
      "loss": 0.3979,
      "num_input_tokens_seen": 9637456,
      "step": 14760
    },
    {
      "epoch": 7.738469601677149,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.023086092543734456,
      "loss": 0.4394,
      "num_input_tokens_seen": 9641520,
      "step": 14765
    },
    {
      "epoch": 7.741090146750524,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.023080313319506288,
      "loss": 0.4415,
      "num_input_tokens_seen": 9644912,
      "step": 14770
    },
    {
      "epoch": 7.7437106918239,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02307453240496877,
      "loss": 0.3811,
      "num_input_tokens_seen": 9647568,
      "step": 14775
    },
    {
      "epoch": 7.746331236897275,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02306874980133119,
      "loss": 0.3906,
      "num_input_tokens_seen": 9651600,
      "step": 14780
    },
    {
      "epoch": 7.74895178197065,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.023062965509803194,
      "loss": 0.3267,
      "num_input_tokens_seen": 9654576,
      "step": 14785
    },
    {
      "epoch": 7.751572327044025,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.023057179531594807,
      "loss": 0.4309,
      "num_input_tokens_seen": 9658000,
      "step": 14790
    },
    {
      "epoch": 7.7541928721174,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02305139186791639,
      "loss": 0.4733,
      "num_input_tokens_seen": 9660880,
      "step": 14795
    },
    {
      "epoch": 7.756813417190775,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.023045602519978642,
      "loss": 0.3605,
      "num_input_tokens_seen": 9663920,
      "step": 14800
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.023039811488992635,
      "loss": 0.4097,
      "num_input_tokens_seen": 9667792,
      "step": 14805
    },
    {
      "epoch": 7.762054507337526,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02303401877616979,
      "loss": 0.3863,
      "num_input_tokens_seen": 9671120,
      "step": 14810
    },
    {
      "epoch": 7.764675052410902,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02302822438272188,
      "loss": 0.3618,
      "num_input_tokens_seen": 9674608,
      "step": 14815
    },
    {
      "epoch": 7.767295597484277,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.023022428309861006,
      "loss": 0.422,
      "num_input_tokens_seen": 9677904,
      "step": 14820
    },
    {
      "epoch": 7.769916142557652,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.023016630558799663,
      "loss": 0.4299,
      "num_input_tokens_seen": 9681648,
      "step": 14825
    },
    {
      "epoch": 7.772536687631027,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.023010831130750666,
      "loss": 0.5283,
      "num_input_tokens_seen": 9684208,
      "step": 14830
    },
    {
      "epoch": 7.7751572327044025,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02300503002692719,
      "loss": 0.4965,
      "num_input_tokens_seen": 9686800,
      "step": 14835
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.06640625,
      "learning_rate": 0.022999227248542753,
      "loss": 0.4088,
      "num_input_tokens_seen": 9689424,
      "step": 14840
    },
    {
      "epoch": 7.780398322851153,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.022993422796811238,
      "loss": 0.3774,
      "num_input_tokens_seen": 9692368,
      "step": 14845
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.022987616672946874,
      "loss": 0.3786,
      "num_input_tokens_seen": 9695856,
      "step": 14850
    },
    {
      "epoch": 7.785639412997903,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02298180887816422,
      "loss": 0.3674,
      "num_input_tokens_seen": 9699280,
      "step": 14855
    },
    {
      "epoch": 7.788259958071279,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.022975999413678217,
      "loss": 0.5466,
      "num_input_tokens_seen": 9701840,
      "step": 14860
    },
    {
      "epoch": 7.790880503144654,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02297018828070413,
      "loss": 0.4551,
      "num_input_tokens_seen": 9704400,
      "step": 14865
    },
    {
      "epoch": 7.79350104821803,
      "grad_norm": 0.033203125,
      "learning_rate": 0.022964375480457582,
      "loss": 0.3942,
      "num_input_tokens_seen": 9707664,
      "step": 14870
    },
    {
      "epoch": 7.796121593291405,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02295856101415454,
      "loss": 0.4789,
      "num_input_tokens_seen": 9710672,
      "step": 14875
    },
    {
      "epoch": 7.79874213836478,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.022952744883011342,
      "loss": 0.5233,
      "num_input_tokens_seen": 9713872,
      "step": 14880
    },
    {
      "epoch": 7.801362683438155,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.022946927088244636,
      "loss": 0.3785,
      "num_input_tokens_seen": 9717008,
      "step": 14885
    },
    {
      "epoch": 7.80398322851153,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.022941107631071447,
      "loss": 0.579,
      "num_input_tokens_seen": 9720560,
      "step": 14890
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02293528651270914,
      "loss": 0.3803,
      "num_input_tokens_seen": 9723792,
      "step": 14895
    },
    {
      "epoch": 7.809224318658281,
      "grad_norm": 0.044921875,
      "learning_rate": 0.022929463734375426,
      "loss": 0.3977,
      "num_input_tokens_seen": 9727184,
      "step": 14900
    },
    {
      "epoch": 7.811844863731656,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.02292363929728836,
      "loss": 0.4302,
      "num_input_tokens_seen": 9730832,
      "step": 14905
    },
    {
      "epoch": 7.814465408805032,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.022917813202666355,
      "loss": 0.447,
      "num_input_tokens_seen": 9734640,
      "step": 14910
    },
    {
      "epoch": 7.817085953878407,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.022911985451728154,
      "loss": 0.3057,
      "num_input_tokens_seen": 9737520,
      "step": 14915
    },
    {
      "epoch": 7.819706498951782,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.02290615604569287,
      "loss": 0.4193,
      "num_input_tokens_seen": 9740880,
      "step": 14920
    },
    {
      "epoch": 7.822327044025157,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.02290032498577993,
      "loss": 0.3305,
      "num_input_tokens_seen": 9744080,
      "step": 14925
    },
    {
      "epoch": 7.8249475890985325,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02289449227320914,
      "loss": 0.4285,
      "num_input_tokens_seen": 9747184,
      "step": 14930
    },
    {
      "epoch": 7.827568134171908,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.022888657909200628,
      "loss": 0.5828,
      "num_input_tokens_seen": 9749648,
      "step": 14935
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.022882821894974884,
      "loss": 0.4824,
      "num_input_tokens_seen": 9751984,
      "step": 14940
    },
    {
      "epoch": 7.832809224318658,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02287698423175273,
      "loss": 0.4177,
      "num_input_tokens_seen": 9755152,
      "step": 14945
    },
    {
      "epoch": 7.835429769392033,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.022871144920755336,
      "loss": 0.4803,
      "num_input_tokens_seen": 9758448,
      "step": 14950
    },
    {
      "epoch": 7.838050314465409,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022865303963204225,
      "loss": 0.4106,
      "num_input_tokens_seen": 9761264,
      "step": 14955
    },
    {
      "epoch": 7.840670859538784,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.02285946136032126,
      "loss": 0.5485,
      "num_input_tokens_seen": 9764976,
      "step": 14960
    },
    {
      "epoch": 7.84329140461216,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.022853617113328645,
      "loss": 0.5341,
      "num_input_tokens_seen": 9768208,
      "step": 14965
    },
    {
      "epoch": 7.845911949685535,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02284777122344893,
      "loss": 0.4048,
      "num_input_tokens_seen": 9771248,
      "step": 14970
    },
    {
      "epoch": 7.84853249475891,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.022841923691905,
      "loss": 0.4333,
      "num_input_tokens_seen": 9773776,
      "step": 14975
    },
    {
      "epoch": 7.851153039832285,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.022836074519920105,
      "loss": 0.5092,
      "num_input_tokens_seen": 9777360,
      "step": 14980
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.022830223708717814,
      "loss": 0.4157,
      "num_input_tokens_seen": 9779984,
      "step": 14985
    },
    {
      "epoch": 7.856394129979035,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.022824371259522056,
      "loss": 0.4397,
      "num_input_tokens_seen": 9782480,
      "step": 14990
    },
    {
      "epoch": 7.859014675052411,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.022818517173557096,
      "loss": 0.5497,
      "num_input_tokens_seen": 9786608,
      "step": 14995
    },
    {
      "epoch": 7.861635220125786,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.022812661452047547,
      "loss": 0.4953,
      "num_input_tokens_seen": 9790096,
      "step": 15000
    },
    {
      "epoch": 7.864255765199162,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.02280680409621835,
      "loss": 0.5162,
      "num_input_tokens_seen": 9793072,
      "step": 15005
    },
    {
      "epoch": 7.866876310272537,
      "grad_norm": 0.107421875,
      "learning_rate": 0.022800945107294798,
      "loss": 0.4231,
      "num_input_tokens_seen": 9796528,
      "step": 15010
    },
    {
      "epoch": 7.869496855345912,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.02279508448650253,
      "loss": 0.4547,
      "num_input_tokens_seen": 9800048,
      "step": 15015
    },
    {
      "epoch": 7.872117400419287,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.022789222235067515,
      "loss": 0.404,
      "num_input_tokens_seen": 9802672,
      "step": 15020
    },
    {
      "epoch": 7.8747379454926625,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.022783358354216084,
      "loss": 0.5557,
      "num_input_tokens_seen": 9806352,
      "step": 15025
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.02277749284517487,
      "loss": 0.3843,
      "num_input_tokens_seen": 9809200,
      "step": 15030
    },
    {
      "epoch": 7.879979035639413,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.02277162570917089,
      "loss": 0.5559,
      "num_input_tokens_seen": 9811568,
      "step": 15035
    },
    {
      "epoch": 7.882599580712788,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02276575694743147,
      "loss": 0.5091,
      "num_input_tokens_seen": 9814480,
      "step": 15040
    },
    {
      "epoch": 7.885220125786163,
      "grad_norm": 0.04296875,
      "learning_rate": 0.022759886561184297,
      "loss": 0.4236,
      "num_input_tokens_seen": 9817200,
      "step": 15045
    },
    {
      "epoch": 7.887840670859539,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022754014551657382,
      "loss": 0.3379,
      "num_input_tokens_seen": 9820336,
      "step": 15050
    },
    {
      "epoch": 7.890461215932914,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02274814092007909,
      "loss": 0.5035,
      "num_input_tokens_seen": 9823664,
      "step": 15055
    },
    {
      "epoch": 7.8930817610062896,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02274226566767811,
      "loss": 0.455,
      "num_input_tokens_seen": 9827280,
      "step": 15060
    },
    {
      "epoch": 7.895702306079665,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.02273638879568348,
      "loss": 0.5037,
      "num_input_tokens_seen": 9830192,
      "step": 15065
    },
    {
      "epoch": 7.89832285115304,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.022730510305324574,
      "loss": 0.4598,
      "num_input_tokens_seen": 9832816,
      "step": 15070
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02272463019783111,
      "loss": 0.3816,
      "num_input_tokens_seen": 9836080,
      "step": 15075
    },
    {
      "epoch": 7.90356394129979,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.022718748474433136,
      "loss": 0.3887,
      "num_input_tokens_seen": 9839152,
      "step": 15080
    },
    {
      "epoch": 7.906184486373165,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.022712865136361036,
      "loss": 0.3477,
      "num_input_tokens_seen": 9843120,
      "step": 15085
    },
    {
      "epoch": 7.908805031446541,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.022706980184845547,
      "loss": 0.4673,
      "num_input_tokens_seen": 9846608,
      "step": 15090
    },
    {
      "epoch": 7.911425576519916,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.02270109362111773,
      "loss": 0.456,
      "num_input_tokens_seen": 9849552,
      "step": 15095
    },
    {
      "epoch": 7.914046121593291,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.022695205446408982,
      "loss": 0.4633,
      "num_input_tokens_seen": 9852336,
      "step": 15100
    },
    {
      "epoch": 7.916666666666667,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02268931566195105,
      "loss": 0.4448,
      "num_input_tokens_seen": 9855376,
      "step": 15105
    },
    {
      "epoch": 7.919287211740042,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.022683424268976005,
      "loss": 0.4901,
      "num_input_tokens_seen": 9858608,
      "step": 15110
    },
    {
      "epoch": 7.921907756813417,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.02267753126871626,
      "loss": 0.4385,
      "num_input_tokens_seen": 9861776,
      "step": 15115
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.05859375,
      "learning_rate": 0.022671636662404564,
      "loss": 0.4659,
      "num_input_tokens_seen": 9864144,
      "step": 15120
    },
    {
      "epoch": 7.927148846960168,
      "grad_norm": 0.0546875,
      "learning_rate": 0.022665740451273998,
      "loss": 0.4606,
      "num_input_tokens_seen": 9866512,
      "step": 15125
    },
    {
      "epoch": 7.929769392033543,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.022659842636557986,
      "loss": 0.5142,
      "num_input_tokens_seen": 9869936,
      "step": 15130
    },
    {
      "epoch": 7.932389937106918,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.022653943219490286,
      "loss": 0.3482,
      "num_input_tokens_seen": 9873168,
      "step": 15135
    },
    {
      "epoch": 7.935010482180293,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02264804220130498,
      "loss": 0.5469,
      "num_input_tokens_seen": 9875888,
      "step": 15140
    },
    {
      "epoch": 7.937631027253669,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.022642139583236503,
      "loss": 0.5457,
      "num_input_tokens_seen": 9880432,
      "step": 15145
    },
    {
      "epoch": 7.940251572327044,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.02263623536651961,
      "loss": 0.4252,
      "num_input_tokens_seen": 9885136,
      "step": 15150
    },
    {
      "epoch": 7.9428721174004195,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.022630329552389396,
      "loss": 0.4839,
      "num_input_tokens_seen": 9887920,
      "step": 15155
    },
    {
      "epoch": 7.945492662473795,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02262442214208129,
      "loss": 0.4002,
      "num_input_tokens_seen": 9892464,
      "step": 15160
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 0.0546875,
      "learning_rate": 0.02261851313683106,
      "loss": 0.4423,
      "num_input_tokens_seen": 9895344,
      "step": 15165
    },
    {
      "epoch": 7.950733752620545,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0226126025378748,
      "loss": 0.4637,
      "num_input_tokens_seen": 9898032,
      "step": 15170
    },
    {
      "epoch": 7.95335429769392,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.02260669034644893,
      "loss": 0.3222,
      "num_input_tokens_seen": 9902160,
      "step": 15175
    },
    {
      "epoch": 7.955974842767295,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.022600776563790226,
      "loss": 0.3994,
      "num_input_tokens_seen": 9906160,
      "step": 15180
    },
    {
      "epoch": 7.9585953878406706,
      "grad_norm": 0.037109375,
      "learning_rate": 0.022594861191135778,
      "loss": 0.4157,
      "num_input_tokens_seen": 9909296,
      "step": 15185
    },
    {
      "epoch": 7.961215932914046,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.022588944229723018,
      "loss": 0.464,
      "num_input_tokens_seen": 9911952,
      "step": 15190
    },
    {
      "epoch": 7.963836477987421,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.022583025680789705,
      "loss": 0.5244,
      "num_input_tokens_seen": 9914992,
      "step": 15195
    },
    {
      "epoch": 7.966457023060797,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.022577105545573926,
      "loss": 0.319,
      "num_input_tokens_seen": 9919184,
      "step": 15200
    },
    {
      "epoch": 7.969077568134172,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.022571183825314114,
      "loss": 0.5604,
      "num_input_tokens_seen": 9921744,
      "step": 15205
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.022565260521249027,
      "loss": 0.4584,
      "num_input_tokens_seen": 9924912,
      "step": 15210
    },
    {
      "epoch": 7.9743186582809225,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02255933563461774,
      "loss": 0.3727,
      "num_input_tokens_seen": 9927728,
      "step": 15215
    },
    {
      "epoch": 7.976939203354298,
      "grad_norm": 0.044921875,
      "learning_rate": 0.022553409166659686,
      "loss": 0.3915,
      "num_input_tokens_seen": 9930448,
      "step": 15220
    },
    {
      "epoch": 7.979559748427673,
      "grad_norm": 0.0703125,
      "learning_rate": 0.022547481118614605,
      "loss": 0.4474,
      "num_input_tokens_seen": 9934608,
      "step": 15225
    },
    {
      "epoch": 7.982180293501048,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.02254155149172258,
      "loss": 0.3975,
      "num_input_tokens_seen": 9937456,
      "step": 15230
    },
    {
      "epoch": 7.984800838574423,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.022535620287224025,
      "loss": 0.49,
      "num_input_tokens_seen": 9940816,
      "step": 15235
    },
    {
      "epoch": 7.987421383647799,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.022529687506359673,
      "loss": 0.3818,
      "num_input_tokens_seen": 9944080,
      "step": 15240
    },
    {
      "epoch": 7.990041928721174,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0225237531503706,
      "loss": 0.4162,
      "num_input_tokens_seen": 9946480,
      "step": 15245
    },
    {
      "epoch": 7.9926624737945495,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.022517817220498202,
      "loss": 0.4629,
      "num_input_tokens_seen": 9949872,
      "step": 15250
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02251187971798421,
      "loss": 0.4102,
      "num_input_tokens_seen": 9953520,
      "step": 15255
    },
    {
      "epoch": 7.9979035639413,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.02250594064407068,
      "loss": 0.4622,
      "num_input_tokens_seen": 9956624,
      "step": 15260
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.4542999267578125,
      "eval_runtime": 13.6315,
      "eval_samples_per_second": 62.209,
      "eval_steps_per_second": 15.552,
      "num_input_tokens_seen": 9958360,
      "step": 15264
    },
    {
      "epoch": 8.000524109014675,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0225,
      "loss": 0.4359,
      "num_input_tokens_seen": 9958968,
      "step": 15265
    },
    {
      "epoch": 8.00314465408805,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02249405778701488,
      "loss": 0.3697,
      "num_input_tokens_seen": 9961624,
      "step": 15270
    },
    {
      "epoch": 8.005765199161425,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.02248811400635838,
      "loss": 0.4829,
      "num_input_tokens_seen": 9964600,
      "step": 15275
    },
    {
      "epoch": 8.0083857442348,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.022482168659273848,
      "loss": 0.4075,
      "num_input_tokens_seen": 9967096,
      "step": 15280
    },
    {
      "epoch": 8.011006289308176,
      "grad_norm": 0.037109375,
      "learning_rate": 0.022476221747005,
      "loss": 0.3752,
      "num_input_tokens_seen": 9969848,
      "step": 15285
    },
    {
      "epoch": 8.01362683438155,
      "grad_norm": 0.03125,
      "learning_rate": 0.022470273270795846,
      "loss": 0.3977,
      "num_input_tokens_seen": 9973272,
      "step": 15290
    },
    {
      "epoch": 8.016247379454926,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.022464323231890758,
      "loss": 0.38,
      "num_input_tokens_seen": 9975992,
      "step": 15295
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.022458371631534406,
      "loss": 0.5053,
      "num_input_tokens_seen": 9979704,
      "step": 15300
    },
    {
      "epoch": 8.021488469601676,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.022452418470971795,
      "loss": 0.4483,
      "num_input_tokens_seen": 9982712,
      "step": 15305
    },
    {
      "epoch": 8.024109014675052,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.022446463751448268,
      "loss": 0.3088,
      "num_input_tokens_seen": 9985560,
      "step": 15310
    },
    {
      "epoch": 8.026729559748428,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.022440507474209472,
      "loss": 0.3766,
      "num_input_tokens_seen": 9989272,
      "step": 15315
    },
    {
      "epoch": 8.029350104821804,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.022434549640501403,
      "loss": 0.3599,
      "num_input_tokens_seen": 9992088,
      "step": 15320
    },
    {
      "epoch": 8.031970649895179,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.022428590251570363,
      "loss": 0.431,
      "num_input_tokens_seen": 9995032,
      "step": 15325
    },
    {
      "epoch": 8.034591194968554,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.02242262930866299,
      "loss": 0.3089,
      "num_input_tokens_seen": 9998008,
      "step": 15330
    },
    {
      "epoch": 8.03721174004193,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.022416666813026252,
      "loss": 0.5663,
      "num_input_tokens_seen": 10001464,
      "step": 15335
    },
    {
      "epoch": 8.039832285115304,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.022410702765907432,
      "loss": 0.3907,
      "num_input_tokens_seen": 10004472,
      "step": 15340
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.022404737168554137,
      "loss": 0.398,
      "num_input_tokens_seen": 10007160,
      "step": 15345
    },
    {
      "epoch": 8.045073375262055,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.0223987700222143,
      "loss": 0.4986,
      "num_input_tokens_seen": 10010136,
      "step": 15350
    },
    {
      "epoch": 8.04769392033543,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.022392801328136192,
      "loss": 0.4658,
      "num_input_tokens_seen": 10013560,
      "step": 15355
    },
    {
      "epoch": 8.050314465408805,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02238683108756838,
      "loss": 0.4038,
      "num_input_tokens_seen": 10016280,
      "step": 15360
    },
    {
      "epoch": 8.05293501048218,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.022380859301759783,
      "loss": 0.5699,
      "num_input_tokens_seen": 10019096,
      "step": 15365
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.022374885971959625,
      "loss": 0.4763,
      "num_input_tokens_seen": 10022200,
      "step": 15370
    },
    {
      "epoch": 8.05817610062893,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.02236891109941745,
      "loss": 0.4122,
      "num_input_tokens_seen": 10025528,
      "step": 15375
    },
    {
      "epoch": 8.060796645702306,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.022362934685383157,
      "loss": 0.5009,
      "num_input_tokens_seen": 10030104,
      "step": 15380
    },
    {
      "epoch": 8.06341719077568,
      "grad_norm": 0.056640625,
      "learning_rate": 0.022356956731106915,
      "loss": 0.5067,
      "num_input_tokens_seen": 10033112,
      "step": 15385
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 0.05859375,
      "learning_rate": 0.022350977237839267,
      "loss": 0.3925,
      "num_input_tokens_seen": 10036280,
      "step": 15390
    },
    {
      "epoch": 8.068658280922431,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.022344996206831044,
      "loss": 0.3438,
      "num_input_tokens_seen": 10039160,
      "step": 15395
    },
    {
      "epoch": 8.071278825995806,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.022339013639333407,
      "loss": 0.3871,
      "num_input_tokens_seen": 10043224,
      "step": 15400
    },
    {
      "epoch": 8.073899371069182,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.022333029536597847,
      "loss": 0.378,
      "num_input_tokens_seen": 10045976,
      "step": 15405
    },
    {
      "epoch": 8.076519916142558,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022327043899876164,
      "loss": 0.3611,
      "num_input_tokens_seen": 10048824,
      "step": 15410
    },
    {
      "epoch": 8.079140461215934,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.022321056730420487,
      "loss": 0.4063,
      "num_input_tokens_seen": 10051512,
      "step": 15415
    },
    {
      "epoch": 8.081761006289309,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.022315068029483274,
      "loss": 0.5056,
      "num_input_tokens_seen": 10054712,
      "step": 15420
    },
    {
      "epoch": 8.084381551362684,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.022309077798317277,
      "loss": 0.5116,
      "num_input_tokens_seen": 10058232,
      "step": 15425
    },
    {
      "epoch": 8.08700209643606,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02230308603817559,
      "loss": 0.4512,
      "num_input_tokens_seen": 10061880,
      "step": 15430
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02229709275031163,
      "loss": 0.4384,
      "num_input_tokens_seen": 10064984,
      "step": 15435
    },
    {
      "epoch": 8.09224318658281,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02229109793597911,
      "loss": 0.4258,
      "num_input_tokens_seen": 10067384,
      "step": 15440
    },
    {
      "epoch": 8.094863731656185,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.022285101596432082,
      "loss": 0.4189,
      "num_input_tokens_seen": 10070776,
      "step": 15445
    },
    {
      "epoch": 8.09748427672956,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.022279103732924917,
      "loss": 0.3367,
      "num_input_tokens_seen": 10074104,
      "step": 15450
    },
    {
      "epoch": 8.100104821802935,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.02227310434671229,
      "loss": 0.3804,
      "num_input_tokens_seen": 10077784,
      "step": 15455
    },
    {
      "epoch": 8.10272536687631,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.022267103439049214,
      "loss": 0.4817,
      "num_input_tokens_seen": 10082392,
      "step": 15460
    },
    {
      "epoch": 8.105345911949685,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.022261101011191005,
      "loss": 0.3707,
      "num_input_tokens_seen": 10085656,
      "step": 15465
    },
    {
      "epoch": 8.10796645702306,
      "grad_norm": 0.0546875,
      "learning_rate": 0.022255097064393305,
      "loss": 0.4451,
      "num_input_tokens_seen": 10089688,
      "step": 15470
    },
    {
      "epoch": 8.110587002096436,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.022249091599912067,
      "loss": 0.502,
      "num_input_tokens_seen": 10092632,
      "step": 15475
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.060546875,
      "learning_rate": 0.022243084619003567,
      "loss": 0.4236,
      "num_input_tokens_seen": 10095384,
      "step": 15480
    },
    {
      "epoch": 8.115828092243186,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0222370761229244,
      "loss": 0.4942,
      "num_input_tokens_seen": 10101272,
      "step": 15485
    },
    {
      "epoch": 8.118448637316561,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.022231066112931474,
      "loss": 0.4754,
      "num_input_tokens_seen": 10104920,
      "step": 15490
    },
    {
      "epoch": 8.121069182389936,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.022225054590282015,
      "loss": 0.4563,
      "num_input_tokens_seen": 10108440,
      "step": 15495
    },
    {
      "epoch": 8.123689727463312,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.02221904155623356,
      "loss": 0.3947,
      "num_input_tokens_seen": 10111672,
      "step": 15500
    },
    {
      "epoch": 8.126310272536688,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.022213027012043977,
      "loss": 0.5023,
      "num_input_tokens_seen": 10115448,
      "step": 15505
    },
    {
      "epoch": 8.128930817610064,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02220701095897143,
      "loss": 0.3428,
      "num_input_tokens_seen": 10118328,
      "step": 15510
    },
    {
      "epoch": 8.131551362683439,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.022200993398274405,
      "loss": 0.4379,
      "num_input_tokens_seen": 10122328,
      "step": 15515
    },
    {
      "epoch": 8.134171907756814,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.022194974331211724,
      "loss": 0.463,
      "num_input_tokens_seen": 10125112,
      "step": 15520
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.022188953759042493,
      "loss": 0.3436,
      "num_input_tokens_seen": 10128280,
      "step": 15525
    },
    {
      "epoch": 8.139412997903564,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.02218293168302615,
      "loss": 0.3393,
      "num_input_tokens_seen": 10131320,
      "step": 15530
    },
    {
      "epoch": 8.14203354297694,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.022176908104422452,
      "loss": 0.4274,
      "num_input_tokens_seen": 10135384,
      "step": 15535
    },
    {
      "epoch": 8.144654088050315,
      "grad_norm": 0.05078125,
      "learning_rate": 0.022170883024491453,
      "loss": 0.5445,
      "num_input_tokens_seen": 10138520,
      "step": 15540
    },
    {
      "epoch": 8.14727463312369,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.022164856444493532,
      "loss": 0.4264,
      "num_input_tokens_seen": 10141528,
      "step": 15545
    },
    {
      "epoch": 8.149895178197065,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.022158828365689394,
      "loss": 0.4468,
      "num_input_tokens_seen": 10144568,
      "step": 15550
    },
    {
      "epoch": 8.15251572327044,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.022152798789340025,
      "loss": 0.4354,
      "num_input_tokens_seen": 10148504,
      "step": 15555
    },
    {
      "epoch": 8.155136268343815,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.022146767716706754,
      "loss": 0.4685,
      "num_input_tokens_seen": 10151288,
      "step": 15560
    },
    {
      "epoch": 8.15775681341719,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.022140735149051217,
      "loss": 0.4266,
      "num_input_tokens_seen": 10154744,
      "step": 15565
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 0.0625,
      "learning_rate": 0.022134701087635347,
      "loss": 0.3412,
      "num_input_tokens_seen": 10158136,
      "step": 15570
    },
    {
      "epoch": 8.16299790356394,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.022128665533721406,
      "loss": 0.3972,
      "num_input_tokens_seen": 10161560,
      "step": 15575
    },
    {
      "epoch": 8.165618448637316,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.022122628488571966,
      "loss": 0.4618,
      "num_input_tokens_seen": 10163960,
      "step": 15580
    },
    {
      "epoch": 8.168238993710691,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.022116589953449905,
      "loss": 0.3239,
      "num_input_tokens_seen": 10166776,
      "step": 15585
    },
    {
      "epoch": 8.170859538784066,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.022110549929618414,
      "loss": 0.3087,
      "num_input_tokens_seen": 10169592,
      "step": 15590
    },
    {
      "epoch": 8.173480083857442,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.022104508418341003,
      "loss": 0.3956,
      "num_input_tokens_seen": 10172760,
      "step": 15595
    },
    {
      "epoch": 8.176100628930818,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02209846542088148,
      "loss": 0.4481,
      "num_input_tokens_seen": 10177496,
      "step": 15600
    },
    {
      "epoch": 8.178721174004194,
      "grad_norm": 0.080078125,
      "learning_rate": 0.022092420938503977,
      "loss": 0.4318,
      "num_input_tokens_seen": 10181112,
      "step": 15605
    },
    {
      "epoch": 8.181341719077569,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.022086374972472927,
      "loss": 0.4423,
      "num_input_tokens_seen": 10184632,
      "step": 15610
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02208032752405308,
      "loss": 0.4235,
      "num_input_tokens_seen": 10187576,
      "step": 15615
    },
    {
      "epoch": 8.18658280922432,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.02207427859450949,
      "loss": 0.5455,
      "num_input_tokens_seen": 10190648,
      "step": 15620
    },
    {
      "epoch": 8.189203354297694,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.022068228185107522,
      "loss": 0.3086,
      "num_input_tokens_seen": 10193944,
      "step": 15625
    },
    {
      "epoch": 8.19182389937107,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02206217629711286,
      "loss": 0.3962,
      "num_input_tokens_seen": 10196824,
      "step": 15630
    },
    {
      "epoch": 8.194444444444445,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02205612293179149,
      "loss": 0.499,
      "num_input_tokens_seen": 10200312,
      "step": 15635
    },
    {
      "epoch": 8.19706498951782,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0220500680904097,
      "loss": 0.5511,
      "num_input_tokens_seen": 10203992,
      "step": 15640
    },
    {
      "epoch": 8.199685534591195,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.022044011774234092,
      "loss": 0.3891,
      "num_input_tokens_seen": 10206904,
      "step": 15645
    },
    {
      "epoch": 8.20230607966457,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.022037953984531587,
      "loss": 0.4317,
      "num_input_tokens_seen": 10209880,
      "step": 15650
    },
    {
      "epoch": 8.204926624737945,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.022031894722569403,
      "loss": 0.4936,
      "num_input_tokens_seen": 10215864,
      "step": 15655
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 0.0703125,
      "learning_rate": 0.022025833989615065,
      "loss": 0.4351,
      "num_input_tokens_seen": 10218904,
      "step": 15660
    },
    {
      "epoch": 8.210167714884696,
      "grad_norm": 0.0546875,
      "learning_rate": 0.022019771786936416,
      "loss": 0.3892,
      "num_input_tokens_seen": 10225048,
      "step": 15665
    },
    {
      "epoch": 8.21278825995807,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02201370811580159,
      "loss": 0.4342,
      "num_input_tokens_seen": 10227576,
      "step": 15670
    },
    {
      "epoch": 8.215408805031446,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.022007642977479046,
      "loss": 0.5925,
      "num_input_tokens_seen": 10230712,
      "step": 15675
    },
    {
      "epoch": 8.218029350104821,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.02200157637323754,
      "loss": 0.5005,
      "num_input_tokens_seen": 10233848,
      "step": 15680
    },
    {
      "epoch": 8.220649895178196,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02199550830434613,
      "loss": 0.4601,
      "num_input_tokens_seen": 10237816,
      "step": 15685
    },
    {
      "epoch": 8.223270440251572,
      "grad_norm": 0.037109375,
      "learning_rate": 0.021989438772074193,
      "loss": 0.4045,
      "num_input_tokens_seen": 10241272,
      "step": 15690
    },
    {
      "epoch": 8.225890985324948,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.021983367777691405,
      "loss": 0.43,
      "num_input_tokens_seen": 10243864,
      "step": 15695
    },
    {
      "epoch": 8.228511530398324,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.021977295322467748,
      "loss": 0.3079,
      "num_input_tokens_seen": 10246712,
      "step": 15700
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 0.0234375,
      "learning_rate": 0.021971221407673512,
      "loss": 0.4672,
      "num_input_tokens_seen": 10250648,
      "step": 15705
    },
    {
      "epoch": 8.233752620545074,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.02196514603457929,
      "loss": 0.4501,
      "num_input_tokens_seen": 10253624,
      "step": 15710
    },
    {
      "epoch": 8.23637316561845,
      "grad_norm": 0.06640625,
      "learning_rate": 0.021959069204455978,
      "loss": 0.3948,
      "num_input_tokens_seen": 10256536,
      "step": 15715
    },
    {
      "epoch": 8.238993710691824,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.021952990918574784,
      "loss": 0.3909,
      "num_input_tokens_seen": 10259448,
      "step": 15720
    },
    {
      "epoch": 8.2416142557652,
      "grad_norm": 0.068359375,
      "learning_rate": 0.02194691117820721,
      "loss": 0.4196,
      "num_input_tokens_seen": 10263768,
      "step": 15725
    },
    {
      "epoch": 8.244234800838575,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.02194082998462507,
      "loss": 0.4554,
      "num_input_tokens_seen": 10266712,
      "step": 15730
    },
    {
      "epoch": 8.24685534591195,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02193474733910049,
      "loss": 0.6529,
      "num_input_tokens_seen": 10270712,
      "step": 15735
    },
    {
      "epoch": 8.249475890985325,
      "grad_norm": 0.076171875,
      "learning_rate": 0.021928663242905877,
      "loss": 0.5595,
      "num_input_tokens_seen": 10274456,
      "step": 15740
    },
    {
      "epoch": 8.2520964360587,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.021922577697313966,
      "loss": 0.4236,
      "num_input_tokens_seen": 10278840,
      "step": 15745
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.021916490703597774,
      "loss": 0.3181,
      "num_input_tokens_seen": 10282104,
      "step": 15750
    },
    {
      "epoch": 8.25733752620545,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.021910402263030634,
      "loss": 0.5088,
      "num_input_tokens_seen": 10285272,
      "step": 15755
    },
    {
      "epoch": 8.259958071278826,
      "grad_norm": 0.068359375,
      "learning_rate": 0.02190431237688618,
      "loss": 0.4231,
      "num_input_tokens_seen": 10288472,
      "step": 15760
    },
    {
      "epoch": 8.2625786163522,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02189822104643834,
      "loss": 0.4726,
      "num_input_tokens_seen": 10291128,
      "step": 15765
    },
    {
      "epoch": 8.265199161425576,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.021892128272961362,
      "loss": 0.4386,
      "num_input_tokens_seen": 10295288,
      "step": 15770
    },
    {
      "epoch": 8.267819706498951,
      "grad_norm": 0.052734375,
      "learning_rate": 0.021886034057729778,
      "loss": 0.4079,
      "num_input_tokens_seen": 10298680,
      "step": 15775
    },
    {
      "epoch": 8.270440251572326,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.021879938402018427,
      "loss": 0.4584,
      "num_input_tokens_seen": 10301496,
      "step": 15780
    },
    {
      "epoch": 8.273060796645701,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.021873841307102456,
      "loss": 0.3298,
      "num_input_tokens_seen": 10303992,
      "step": 15785
    },
    {
      "epoch": 8.275681341719078,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.021867742774257303,
      "loss": 0.3907,
      "num_input_tokens_seen": 10306808,
      "step": 15790
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02186164280475871,
      "loss": 0.3814,
      "num_input_tokens_seen": 10310584,
      "step": 15795
    },
    {
      "epoch": 8.280922431865829,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02185554139988273,
      "loss": 0.3333,
      "num_input_tokens_seen": 10314872,
      "step": 15800
    },
    {
      "epoch": 8.283542976939204,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.021849438560905696,
      "loss": 0.3642,
      "num_input_tokens_seen": 10317624,
      "step": 15805
    },
    {
      "epoch": 8.286163522012579,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.021843334289104257,
      "loss": 0.3588,
      "num_input_tokens_seen": 10320408,
      "step": 15810
    },
    {
      "epoch": 8.288784067085954,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.021837228585755365,
      "loss": 0.4826,
      "num_input_tokens_seen": 10323320,
      "step": 15815
    },
    {
      "epoch": 8.29140461215933,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.021831121452136247,
      "loss": 0.3367,
      "num_input_tokens_seen": 10326776,
      "step": 15820
    },
    {
      "epoch": 8.294025157232705,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.021825012889524462,
      "loss": 0.2702,
      "num_input_tokens_seen": 10330584,
      "step": 15825
    },
    {
      "epoch": 8.29664570230608,
      "grad_norm": 0.05859375,
      "learning_rate": 0.02181890289919784,
      "loss": 0.4119,
      "num_input_tokens_seen": 10332824,
      "step": 15830
    },
    {
      "epoch": 8.299266247379455,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.021812791482434528,
      "loss": 0.5227,
      "num_input_tokens_seen": 10335480,
      "step": 15835
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02180667864051296,
      "loss": 0.4763,
      "num_input_tokens_seen": 10338424,
      "step": 15840
    },
    {
      "epoch": 8.304507337526205,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.021800564374711874,
      "loss": 0.4272,
      "num_input_tokens_seen": 10341784,
      "step": 15845
    },
    {
      "epoch": 8.30712788259958,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.021794448686310308,
      "loss": 0.4339,
      "num_input_tokens_seen": 10345144,
      "step": 15850
    },
    {
      "epoch": 8.309748427672956,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.021788331576587595,
      "loss": 0.3803,
      "num_input_tokens_seen": 10348984,
      "step": 15855
    },
    {
      "epoch": 8.31236897274633,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.021782213046823356,
      "loss": 0.4483,
      "num_input_tokens_seen": 10352152,
      "step": 15860
    },
    {
      "epoch": 8.314989517819706,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.02177609309829753,
      "loss": 0.4786,
      "num_input_tokens_seen": 10354392,
      "step": 15865
    },
    {
      "epoch": 8.317610062893081,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.021769971732290328,
      "loss": 0.705,
      "num_input_tokens_seen": 10357368,
      "step": 15870
    },
    {
      "epoch": 8.320230607966456,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.021763848950082272,
      "loss": 0.4013,
      "num_input_tokens_seen": 10359864,
      "step": 15875
    },
    {
      "epoch": 8.322851153039831,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.02175772475295419,
      "loss": 0.5658,
      "num_input_tokens_seen": 10364536,
      "step": 15880
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.021751599142187182,
      "loss": 0.4644,
      "num_input_tokens_seen": 10367864,
      "step": 15885
    },
    {
      "epoch": 8.328092243186584,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02174547211906266,
      "loss": 0.4967,
      "num_input_tokens_seen": 10371736,
      "step": 15890
    },
    {
      "epoch": 8.330712788259959,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.021739343684862324,
      "loss": 0.3201,
      "num_input_tokens_seen": 10375192,
      "step": 15895
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02173321384086818,
      "loss": 0.4067,
      "num_input_tokens_seen": 10383288,
      "step": 15900
    },
    {
      "epoch": 8.335953878406709,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02172708258836252,
      "loss": 0.5286,
      "num_input_tokens_seen": 10387192,
      "step": 15905
    },
    {
      "epoch": 8.338574423480084,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.021720949928627924,
      "loss": 0.4018,
      "num_input_tokens_seen": 10389400,
      "step": 15910
    },
    {
      "epoch": 8.34119496855346,
      "grad_norm": 0.02734375,
      "learning_rate": 0.021714815862947278,
      "loss": 0.4772,
      "num_input_tokens_seen": 10392920,
      "step": 15915
    },
    {
      "epoch": 8.343815513626835,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.021708680392603763,
      "loss": 0.4437,
      "num_input_tokens_seen": 10396280,
      "step": 15920
    },
    {
      "epoch": 8.34643605870021,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.02170254351888085,
      "loss": 0.4714,
      "num_input_tokens_seen": 10399960,
      "step": 15925
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.021696405243062294,
      "loss": 0.428,
      "num_input_tokens_seen": 10403128,
      "step": 15930
    },
    {
      "epoch": 8.35167714884696,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.02169026556643216,
      "loss": 0.402,
      "num_input_tokens_seen": 10405976,
      "step": 15935
    },
    {
      "epoch": 8.354297693920335,
      "grad_norm": 0.095703125,
      "learning_rate": 0.021684124490274796,
      "loss": 0.4202,
      "num_input_tokens_seen": 10408632,
      "step": 15940
    },
    {
      "epoch": 8.35691823899371,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.021677982015874845,
      "loss": 0.5065,
      "num_input_tokens_seen": 10412216,
      "step": 15945
    },
    {
      "epoch": 8.359538784067086,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.02167183814451724,
      "loss": 0.4093,
      "num_input_tokens_seen": 10414680,
      "step": 15950
    },
    {
      "epoch": 8.36215932914046,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.021665692877487215,
      "loss": 0.4864,
      "num_input_tokens_seen": 10417560,
      "step": 15955
    },
    {
      "epoch": 8.364779874213836,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.021659546216070285,
      "loss": 0.3329,
      "num_input_tokens_seen": 10420056,
      "step": 15960
    },
    {
      "epoch": 8.367400419287211,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.021653398161552265,
      "loss": 0.3524,
      "num_input_tokens_seen": 10423288,
      "step": 15965
    },
    {
      "epoch": 8.370020964360586,
      "grad_norm": 0.0546875,
      "learning_rate": 0.021647248715219253,
      "loss": 0.4643,
      "num_input_tokens_seen": 10425752,
      "step": 15970
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02164109787835765,
      "loss": 0.3799,
      "num_input_tokens_seen": 10429432,
      "step": 15975
    },
    {
      "epoch": 8.375262054507338,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.021634945652254132,
      "loss": 0.421,
      "num_input_tokens_seen": 10432856,
      "step": 15980
    },
    {
      "epoch": 8.377882599580714,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.021628792038195677,
      "loss": 0.5095,
      "num_input_tokens_seen": 10436600,
      "step": 15985
    },
    {
      "epoch": 8.380503144654089,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02162263703746956,
      "loss": 0.4026,
      "num_input_tokens_seen": 10439352,
      "step": 15990
    },
    {
      "epoch": 8.383123689727464,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.02161648065136332,
      "loss": 0.4427,
      "num_input_tokens_seen": 10441976,
      "step": 15995
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02161032288116482,
      "loss": 0.4711,
      "num_input_tokens_seen": 10445272,
      "step": 16000
    },
    {
      "epoch": 8.388364779874214,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.021604163728162186,
      "loss": 0.51,
      "num_input_tokens_seen": 10447864,
      "step": 16005
    },
    {
      "epoch": 8.39098532494759,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.021598003193643845,
      "loss": 0.3846,
      "num_input_tokens_seen": 10451992,
      "step": 16010
    },
    {
      "epoch": 8.393605870020965,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.02159184127889851,
      "loss": 0.4905,
      "num_input_tokens_seen": 10454968,
      "step": 16015
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.021585677985215178,
      "loss": 0.4474,
      "num_input_tokens_seen": 10458392,
      "step": 16020
    },
    {
      "epoch": 8.398846960167715,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.021579513313883154,
      "loss": 0.5425,
      "num_input_tokens_seen": 10461144,
      "step": 16025
    },
    {
      "epoch": 8.40146750524109,
      "grad_norm": 0.064453125,
      "learning_rate": 0.021573347266192,
      "loss": 0.388,
      "num_input_tokens_seen": 10463768,
      "step": 16030
    },
    {
      "epoch": 8.404088050314465,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.02156717984343159,
      "loss": 0.3992,
      "num_input_tokens_seen": 10467128,
      "step": 16035
    },
    {
      "epoch": 8.40670859538784,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.021561011046892085,
      "loss": 0.435,
      "num_input_tokens_seen": 10469240,
      "step": 16040
    },
    {
      "epoch": 8.409329140461216,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.021554840877863916,
      "loss": 0.4167,
      "num_input_tokens_seen": 10472728,
      "step": 16045
    },
    {
      "epoch": 8.41194968553459,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.021548669337637827,
      "loss": 0.3738,
      "num_input_tokens_seen": 10474936,
      "step": 16050
    },
    {
      "epoch": 8.414570230607966,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.02154249642750481,
      "loss": 0.4409,
      "num_input_tokens_seen": 10478264,
      "step": 16055
    },
    {
      "epoch": 8.417190775681341,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02153632214875619,
      "loss": 0.4929,
      "num_input_tokens_seen": 10481432,
      "step": 16060
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.021530146502683544,
      "loss": 0.3934,
      "num_input_tokens_seen": 10483864,
      "step": 16065
    },
    {
      "epoch": 8.422431865828091,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02152396949057875,
      "loss": 0.4975,
      "num_input_tokens_seen": 10487480,
      "step": 16070
    },
    {
      "epoch": 8.425052410901468,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.021517791113733968,
      "loss": 0.3407,
      "num_input_tokens_seen": 10490616,
      "step": 16075
    },
    {
      "epoch": 8.427672955974844,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.021511611373441646,
      "loss": 0.3542,
      "num_input_tokens_seen": 10493528,
      "step": 16080
    },
    {
      "epoch": 8.430293501048219,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.021505430270994507,
      "loss": 0.4752,
      "num_input_tokens_seen": 10496568,
      "step": 16085
    },
    {
      "epoch": 8.432914046121594,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02149924780768558,
      "loss": 0.4232,
      "num_input_tokens_seen": 10499992,
      "step": 16090
    },
    {
      "epoch": 8.435534591194969,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02149306398480815,
      "loss": 0.4847,
      "num_input_tokens_seen": 10502840,
      "step": 16095
    },
    {
      "epoch": 8.438155136268344,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02148687880365582,
      "loss": 0.4488,
      "num_input_tokens_seen": 10505752,
      "step": 16100
    },
    {
      "epoch": 8.44077568134172,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.021480692265522447,
      "loss": 0.4455,
      "num_input_tokens_seen": 10508664,
      "step": 16105
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.02147450437170219,
      "loss": 0.4723,
      "num_input_tokens_seen": 10511416,
      "step": 16110
    },
    {
      "epoch": 8.44601677148847,
      "grad_norm": 0.044921875,
      "learning_rate": 0.021468315123489477,
      "loss": 0.4268,
      "num_input_tokens_seen": 10513944,
      "step": 16115
    },
    {
      "epoch": 8.448637316561845,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.021462124522179042,
      "loss": 0.4125,
      "num_input_tokens_seen": 10517208,
      "step": 16120
    },
    {
      "epoch": 8.45125786163522,
      "grad_norm": 0.05078125,
      "learning_rate": 0.021455932569065878,
      "loss": 0.4264,
      "num_input_tokens_seen": 10519704,
      "step": 16125
    },
    {
      "epoch": 8.453878406708595,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.021449739265445276,
      "loss": 0.5607,
      "num_input_tokens_seen": 10523160,
      "step": 16130
    },
    {
      "epoch": 8.45649895178197,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.021443544612612796,
      "loss": 0.3987,
      "num_input_tokens_seen": 10526168,
      "step": 16135
    },
    {
      "epoch": 8.459119496855346,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0214373486118643,
      "loss": 0.3736,
      "num_input_tokens_seen": 10528824,
      "step": 16140
    },
    {
      "epoch": 8.46174004192872,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.021431151264495914,
      "loss": 0.3953,
      "num_input_tokens_seen": 10531640,
      "step": 16145
    },
    {
      "epoch": 8.464360587002096,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.021424952571804053,
      "loss": 0.3583,
      "num_input_tokens_seen": 10534808,
      "step": 16150
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.02141875253508542,
      "loss": 0.452,
      "num_input_tokens_seen": 10537784,
      "step": 16155
    },
    {
      "epoch": 8.469601677148846,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.021412551155636977,
      "loss": 0.4648,
      "num_input_tokens_seen": 10540760,
      "step": 16160
    },
    {
      "epoch": 8.472222222222221,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.021406348434755994,
      "loss": 0.399,
      "num_input_tokens_seen": 10543928,
      "step": 16165
    },
    {
      "epoch": 8.474842767295598,
      "grad_norm": 0.048828125,
      "learning_rate": 0.021400144373740007,
      "loss": 0.5834,
      "num_input_tokens_seen": 10546456,
      "step": 16170
    },
    {
      "epoch": 8.477463312368974,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02139393897388683,
      "loss": 0.439,
      "num_input_tokens_seen": 10549752,
      "step": 16175
    },
    {
      "epoch": 8.480083857442349,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.021387732236494574,
      "loss": 0.4791,
      "num_input_tokens_seen": 10553048,
      "step": 16180
    },
    {
      "epoch": 8.482704402515724,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.021381524162861598,
      "loss": 0.4675,
      "num_input_tokens_seen": 10555768,
      "step": 16185
    },
    {
      "epoch": 8.485324947589099,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.021375314754286577,
      "loss": 0.5104,
      "num_input_tokens_seen": 10559448,
      "step": 16190
    },
    {
      "epoch": 8.487945492662474,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.021369104012068447,
      "loss": 0.6288,
      "num_input_tokens_seen": 10562008,
      "step": 16195
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02136289193750641,
      "loss": 0.4868,
      "num_input_tokens_seen": 10564760,
      "step": 16200
    },
    {
      "epoch": 8.493186582809225,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.02135667853189998,
      "loss": 0.5449,
      "num_input_tokens_seen": 10567512,
      "step": 16205
    },
    {
      "epoch": 8.4958071278826,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.021350463796548917,
      "loss": 0.3307,
      "num_input_tokens_seen": 10570104,
      "step": 16210
    },
    {
      "epoch": 8.498427672955975,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.021344247732753282,
      "loss": 0.5311,
      "num_input_tokens_seen": 10573304,
      "step": 16215
    },
    {
      "epoch": 8.50104821802935,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0213380303418134,
      "loss": 0.3645,
      "num_input_tokens_seen": 10576408,
      "step": 16220
    },
    {
      "epoch": 8.503668763102725,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.021331811625029878,
      "loss": 0.2694,
      "num_input_tokens_seen": 10579864,
      "step": 16225
    },
    {
      "epoch": 8.5062893081761,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.021325591583703597,
      "loss": 0.4841,
      "num_input_tokens_seen": 10582840,
      "step": 16230
    },
    {
      "epoch": 8.508909853249476,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02131937021913573,
      "loss": 0.4481,
      "num_input_tokens_seen": 10586392,
      "step": 16235
    },
    {
      "epoch": 8.51153039832285,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0213131475326277,
      "loss": 0.3596,
      "num_input_tokens_seen": 10591224,
      "step": 16240
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 0.0390625,
      "learning_rate": 0.021306923525481237,
      "loss": 0.4455,
      "num_input_tokens_seen": 10594296,
      "step": 16245
    },
    {
      "epoch": 8.516771488469601,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.021300698198998325,
      "loss": 0.3646,
      "num_input_tokens_seen": 10597912,
      "step": 16250
    },
    {
      "epoch": 8.519392033542976,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.021294471554481227,
      "loss": 0.4333,
      "num_input_tokens_seen": 10600632,
      "step": 16255
    },
    {
      "epoch": 8.522012578616351,
      "grad_norm": 0.04296875,
      "learning_rate": 0.021288243593232496,
      "loss": 0.4531,
      "num_input_tokens_seen": 10603672,
      "step": 16260
    },
    {
      "epoch": 8.524633123689728,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02128201431655494,
      "loss": 0.3425,
      "num_input_tokens_seen": 10606648,
      "step": 16265
    },
    {
      "epoch": 8.527253668763104,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.021275783725751663,
      "loss": 0.4025,
      "num_input_tokens_seen": 10609432,
      "step": 16270
    },
    {
      "epoch": 8.529874213836479,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.021269551822126028,
      "loss": 0.3875,
      "num_input_tokens_seen": 10614424,
      "step": 16275
    },
    {
      "epoch": 8.532494758909854,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.02126331860698168,
      "loss": 0.5487,
      "num_input_tokens_seen": 10617784,
      "step": 16280
    },
    {
      "epoch": 8.535115303983229,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.02125708408162253,
      "loss": 0.4817,
      "num_input_tokens_seen": 10620216,
      "step": 16285
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.021250848247352783,
      "loss": 0.4731,
      "num_input_tokens_seen": 10622808,
      "step": 16290
    },
    {
      "epoch": 8.54035639412998,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.021244611105476888,
      "loss": 0.4523,
      "num_input_tokens_seen": 10625912,
      "step": 16295
    },
    {
      "epoch": 8.542976939203355,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.021238372657299595,
      "loss": 0.4154,
      "num_input_tokens_seen": 10629016,
      "step": 16300
    },
    {
      "epoch": 8.54559748427673,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02123213290412591,
      "loss": 0.5649,
      "num_input_tokens_seen": 10632920,
      "step": 16305
    },
    {
      "epoch": 8.548218029350105,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.021225891847261123,
      "loss": 0.375,
      "num_input_tokens_seen": 10635800,
      "step": 16310
    },
    {
      "epoch": 8.55083857442348,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.021219649488010795,
      "loss": 0.4076,
      "num_input_tokens_seen": 10638712,
      "step": 16315
    },
    {
      "epoch": 8.553459119496855,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.021213405827680748,
      "loss": 0.3512,
      "num_input_tokens_seen": 10641528,
      "step": 16320
    },
    {
      "epoch": 8.55607966457023,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.02120716086757709,
      "loss": 0.396,
      "num_input_tokens_seen": 10644152,
      "step": 16325
    },
    {
      "epoch": 8.558700209643606,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.021200914609006187,
      "loss": 0.4226,
      "num_input_tokens_seen": 10647288,
      "step": 16330
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0211946670532747,
      "loss": 0.368,
      "num_input_tokens_seen": 10650840,
      "step": 16335
    },
    {
      "epoch": 8.563941299790356,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.021188418201689534,
      "loss": 0.3907,
      "num_input_tokens_seen": 10654360,
      "step": 16340
    },
    {
      "epoch": 8.566561844863731,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.021182168055557882,
      "loss": 0.4738,
      "num_input_tokens_seen": 10659448,
      "step": 16345
    },
    {
      "epoch": 8.569182389937106,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0211759166161872,
      "loss": 0.4144,
      "num_input_tokens_seen": 10662520,
      "step": 16350
    },
    {
      "epoch": 8.571802935010481,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.021169663884885226,
      "loss": 0.4393,
      "num_input_tokens_seen": 10665176,
      "step": 16355
    },
    {
      "epoch": 8.574423480083858,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.02116340986295995,
      "loss": 0.3993,
      "num_input_tokens_seen": 10668984,
      "step": 16360
    },
    {
      "epoch": 8.577044025157234,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02115715455171965,
      "loss": 0.3377,
      "num_input_tokens_seen": 10673112,
      "step": 16365
    },
    {
      "epoch": 8.579664570230609,
      "grad_norm": 0.05859375,
      "learning_rate": 0.021150897952472856,
      "loss": 0.3496,
      "num_input_tokens_seen": 10676504,
      "step": 16370
    },
    {
      "epoch": 8.582285115303984,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.021144640066528386,
      "loss": 0.2843,
      "num_input_tokens_seen": 10679480,
      "step": 16375
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 0.123046875,
      "learning_rate": 0.02113838089519532,
      "loss": 0.4244,
      "num_input_tokens_seen": 10682360,
      "step": 16380
    },
    {
      "epoch": 8.587526205450734,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.021132120439782996,
      "loss": 0.4544,
      "num_input_tokens_seen": 10686232,
      "step": 16385
    },
    {
      "epoch": 8.59014675052411,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02112585870160104,
      "loss": 0.3714,
      "num_input_tokens_seen": 10689592,
      "step": 16390
    },
    {
      "epoch": 8.592767295597485,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02111959568195933,
      "loss": 0.4136,
      "num_input_tokens_seen": 10693272,
      "step": 16395
    },
    {
      "epoch": 8.59538784067086,
      "grad_norm": 0.044921875,
      "learning_rate": 0.021113331382168015,
      "loss": 0.3429,
      "num_input_tokens_seen": 10697560,
      "step": 16400
    },
    {
      "epoch": 8.598008385744235,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.02110706580353753,
      "loss": 0.6258,
      "num_input_tokens_seen": 10700664,
      "step": 16405
    },
    {
      "epoch": 8.60062893081761,
      "grad_norm": 0.080078125,
      "learning_rate": 0.021100798947378537,
      "loss": 0.4665,
      "num_input_tokens_seen": 10704696,
      "step": 16410
    },
    {
      "epoch": 8.603249475890985,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.021094530815002017,
      "loss": 0.5169,
      "num_input_tokens_seen": 10707512,
      "step": 16415
    },
    {
      "epoch": 8.60587002096436,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.02108826140771918,
      "loss": 0.319,
      "num_input_tokens_seen": 10711128,
      "step": 16420
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02108199072684151,
      "loss": 0.48,
      "num_input_tokens_seen": 10714392,
      "step": 16425
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.02107571877368077,
      "loss": 0.5057,
      "num_input_tokens_seen": 10718008,
      "step": 16430
    },
    {
      "epoch": 8.613731656184486,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.021069445549548975,
      "loss": 0.449,
      "num_input_tokens_seen": 10721240,
      "step": 16435
    },
    {
      "epoch": 8.616352201257861,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.021063171055758414,
      "loss": 0.4857,
      "num_input_tokens_seen": 10724312,
      "step": 16440
    },
    {
      "epoch": 8.618972746331236,
      "grad_norm": 0.078125,
      "learning_rate": 0.021056895293621647,
      "loss": 0.3744,
      "num_input_tokens_seen": 10726872,
      "step": 16445
    },
    {
      "epoch": 8.621593291404611,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02105061826445148,
      "loss": 0.4721,
      "num_input_tokens_seen": 10730808,
      "step": 16450
    },
    {
      "epoch": 8.624213836477988,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.021044339969561003,
      "loss": 0.3929,
      "num_input_tokens_seen": 10733848,
      "step": 16455
    },
    {
      "epoch": 8.626834381551364,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.021038060410263563,
      "loss": 0.3587,
      "num_input_tokens_seen": 10736952,
      "step": 16460
    },
    {
      "epoch": 8.629454926624739,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02103177958787277,
      "loss": 0.3466,
      "num_input_tokens_seen": 10740024,
      "step": 16465
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0210254975037025,
      "loss": 0.4707,
      "num_input_tokens_seen": 10742680,
      "step": 16470
    },
    {
      "epoch": 8.634696016771489,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.021019214159066896,
      "loss": 0.4217,
      "num_input_tokens_seen": 10746904,
      "step": 16475
    },
    {
      "epoch": 8.637316561844864,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.02101292955528036,
      "loss": 0.4662,
      "num_input_tokens_seen": 10749656,
      "step": 16480
    },
    {
      "epoch": 8.63993710691824,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.02100664369365756,
      "loss": 0.5944,
      "num_input_tokens_seen": 10752280,
      "step": 16485
    },
    {
      "epoch": 8.642557651991615,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.021000356575513426,
      "loss": 0.4035,
      "num_input_tokens_seen": 10755672,
      "step": 16490
    },
    {
      "epoch": 8.64517819706499,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.020994068202163155,
      "loss": 0.4079,
      "num_input_tokens_seen": 10759256,
      "step": 16495
    },
    {
      "epoch": 8.647798742138365,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0209877785749222,
      "loss": 0.5042,
      "num_input_tokens_seen": 10762840,
      "step": 16500
    },
    {
      "epoch": 8.65041928721174,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.020981487695106275,
      "loss": 0.478,
      "num_input_tokens_seen": 10765592,
      "step": 16505
    },
    {
      "epoch": 8.653039832285115,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.02097519556403137,
      "loss": 0.3661,
      "num_input_tokens_seen": 10767800,
      "step": 16510
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.020968902183013716,
      "loss": 0.4583,
      "num_input_tokens_seen": 10775384,
      "step": 16515
    },
    {
      "epoch": 8.658280922431866,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.020962607553369827,
      "loss": 0.495,
      "num_input_tokens_seen": 10777752,
      "step": 16520
    },
    {
      "epoch": 8.66090146750524,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.020956311676416463,
      "loss": 0.2846,
      "num_input_tokens_seen": 10780376,
      "step": 16525
    },
    {
      "epoch": 8.663522012578616,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.020950014553470644,
      "loss": 0.4272,
      "num_input_tokens_seen": 10784344,
      "step": 16530
    },
    {
      "epoch": 8.666142557651991,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.02094371618584967,
      "loss": 0.516,
      "num_input_tokens_seen": 10787128,
      "step": 16535
    },
    {
      "epoch": 8.668763102725366,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.020937416574871075,
      "loss": 0.4016,
      "num_input_tokens_seen": 10790296,
      "step": 16540
    },
    {
      "epoch": 8.671383647798741,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.020931115721852665,
      "loss": 0.4767,
      "num_input_tokens_seen": 10793144,
      "step": 16545
    },
    {
      "epoch": 8.674004192872118,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.02092481362811252,
      "loss": 0.4428,
      "num_input_tokens_seen": 10796280,
      "step": 16550
    },
    {
      "epoch": 8.676624737945493,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020918510294968953,
      "loss": 0.4172,
      "num_input_tokens_seen": 10799000,
      "step": 16555
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.020912205723740557,
      "loss": 0.4964,
      "num_input_tokens_seen": 10802264,
      "step": 16560
    },
    {
      "epoch": 8.681865828092244,
      "grad_norm": 0.03515625,
      "learning_rate": 0.020905899915746172,
      "loss": 0.4253,
      "num_input_tokens_seen": 10805656,
      "step": 16565
    },
    {
      "epoch": 8.684486373165619,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.020899592872304903,
      "loss": 0.402,
      "num_input_tokens_seen": 10809496,
      "step": 16570
    },
    {
      "epoch": 8.687106918238994,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.020893284594736113,
      "loss": 0.4686,
      "num_input_tokens_seen": 10812856,
      "step": 16575
    },
    {
      "epoch": 8.68972746331237,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.020886975084359417,
      "loss": 0.42,
      "num_input_tokens_seen": 10815544,
      "step": 16580
    },
    {
      "epoch": 8.692348008385745,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.020880664342494692,
      "loss": 0.4752,
      "num_input_tokens_seen": 10818072,
      "step": 16585
    },
    {
      "epoch": 8.69496855345912,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.020874352370462087,
      "loss": 0.3936,
      "num_input_tokens_seen": 10821304,
      "step": 16590
    },
    {
      "epoch": 8.697589098532495,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02086803916958198,
      "loss": 0.337,
      "num_input_tokens_seen": 10824984,
      "step": 16595
    },
    {
      "epoch": 8.70020964360587,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02086172474117502,
      "loss": 0.5674,
      "num_input_tokens_seen": 10828024,
      "step": 16600
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.020855409086562124,
      "loss": 0.47,
      "num_input_tokens_seen": 10831640,
      "step": 16605
    },
    {
      "epoch": 8.70545073375262,
      "grad_norm": 0.02734375,
      "learning_rate": 0.020849092207064445,
      "loss": 0.4909,
      "num_input_tokens_seen": 10835160,
      "step": 16610
    },
    {
      "epoch": 8.708071278825996,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.02084277410400341,
      "loss": 0.3378,
      "num_input_tokens_seen": 10838232,
      "step": 16615
    },
    {
      "epoch": 8.71069182389937,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.020836454778700688,
      "loss": 0.3254,
      "num_input_tokens_seen": 10841240,
      "step": 16620
    },
    {
      "epoch": 8.713312368972746,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02083013423247821,
      "loss": 0.3862,
      "num_input_tokens_seen": 10844312,
      "step": 16625
    },
    {
      "epoch": 8.715932914046121,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.020823812466658168,
      "loss": 0.3691,
      "num_input_tokens_seen": 10847320,
      "step": 16630
    },
    {
      "epoch": 8.718553459119496,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.020817489482563,
      "loss": 0.3842,
      "num_input_tokens_seen": 10850648,
      "step": 16635
    },
    {
      "epoch": 8.721174004192871,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0208111652815154,
      "loss": 0.4348,
      "num_input_tokens_seen": 10853912,
      "step": 16640
    },
    {
      "epoch": 8.723794549266248,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02080483986483832,
      "loss": 0.3925,
      "num_input_tokens_seen": 10856824,
      "step": 16645
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.020798513233854964,
      "loss": 0.3671,
      "num_input_tokens_seen": 10860280,
      "step": 16650
    },
    {
      "epoch": 8.729035639412999,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.020792185389888796,
      "loss": 0.5078,
      "num_input_tokens_seen": 10863832,
      "step": 16655
    },
    {
      "epoch": 8.731656184486374,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.020785856334263525,
      "loss": 0.4355,
      "num_input_tokens_seen": 10867576,
      "step": 16660
    },
    {
      "epoch": 8.734276729559749,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.020779526068303113,
      "loss": 0.4326,
      "num_input_tokens_seen": 10871384,
      "step": 16665
    },
    {
      "epoch": 8.736897274633124,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.020773194593331788,
      "loss": 0.5632,
      "num_input_tokens_seen": 10874584,
      "step": 16670
    },
    {
      "epoch": 8.7395178197065,
      "grad_norm": 0.03125,
      "learning_rate": 0.020766861910674016,
      "loss": 0.3996,
      "num_input_tokens_seen": 10878584,
      "step": 16675
    },
    {
      "epoch": 8.742138364779874,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.020760528021654526,
      "loss": 0.6382,
      "num_input_tokens_seen": 10881560,
      "step": 16680
    },
    {
      "epoch": 8.74475890985325,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.020754192927598288,
      "loss": 0.5347,
      "num_input_tokens_seen": 10884600,
      "step": 16685
    },
    {
      "epoch": 8.747379454926625,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.020747856629830544,
      "loss": 0.3441,
      "num_input_tokens_seen": 10888408,
      "step": 16690
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.020741519129676766,
      "loss": 0.4715,
      "num_input_tokens_seen": 10891736,
      "step": 16695
    },
    {
      "epoch": 8.752620545073375,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.020735180428462684,
      "loss": 0.3585,
      "num_input_tokens_seen": 10894648,
      "step": 16700
    },
    {
      "epoch": 8.75524109014675,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.020728840527514295,
      "loss": 0.3555,
      "num_input_tokens_seen": 10897624,
      "step": 16705
    },
    {
      "epoch": 8.757861635220126,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02072249942815782,
      "loss": 0.4642,
      "num_input_tokens_seen": 10900696,
      "step": 16710
    },
    {
      "epoch": 8.7604821802935,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.020716157131719756,
      "loss": 0.5406,
      "num_input_tokens_seen": 10904280,
      "step": 16715
    },
    {
      "epoch": 8.763102725366876,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02070981363952683,
      "loss": 0.4719,
      "num_input_tokens_seen": 10907896,
      "step": 16720
    },
    {
      "epoch": 8.765723270440251,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.020703468952906024,
      "loss": 0.3929,
      "num_input_tokens_seen": 10911192,
      "step": 16725
    },
    {
      "epoch": 8.768343815513626,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.02069712307318459,
      "loss": 0.5662,
      "num_input_tokens_seen": 10914488,
      "step": 16730
    },
    {
      "epoch": 8.770964360587001,
      "grad_norm": 0.060546875,
      "learning_rate": 0.020690776001690007,
      "loss": 0.3934,
      "num_input_tokens_seen": 10918456,
      "step": 16735
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.020684427739750003,
      "loss": 0.4706,
      "num_input_tokens_seen": 10922296,
      "step": 16740
    },
    {
      "epoch": 8.776205450733752,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02067807828869257,
      "loss": 0.4964,
      "num_input_tokens_seen": 10925400,
      "step": 16745
    },
    {
      "epoch": 8.778825995807129,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02067172764984594,
      "loss": 0.4002,
      "num_input_tokens_seen": 10928120,
      "step": 16750
    },
    {
      "epoch": 8.781446540880504,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.020665375824538583,
      "loss": 0.4583,
      "num_input_tokens_seen": 10932088,
      "step": 16755
    },
    {
      "epoch": 8.784067085953879,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.020659022814099252,
      "loss": 0.4636,
      "num_input_tokens_seen": 10935832,
      "step": 16760
    },
    {
      "epoch": 8.786687631027254,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.020652668619856897,
      "loss": 0.558,
      "num_input_tokens_seen": 10938552,
      "step": 16765
    },
    {
      "epoch": 8.78930817610063,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.020646313243140764,
      "loss": 0.4223,
      "num_input_tokens_seen": 10942520,
      "step": 16770
    },
    {
      "epoch": 8.791928721174004,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.020639956685280313,
      "loss": 0.439,
      "num_input_tokens_seen": 10946008,
      "step": 16775
    },
    {
      "epoch": 8.79454926624738,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.020633598947605265,
      "loss": 0.4377,
      "num_input_tokens_seen": 10949144,
      "step": 16780
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 0.064453125,
      "learning_rate": 0.02062724003144559,
      "loss": 0.4259,
      "num_input_tokens_seen": 10953400,
      "step": 16785
    },
    {
      "epoch": 8.79979035639413,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.020620879938131505,
      "loss": 0.4281,
      "num_input_tokens_seen": 10956472,
      "step": 16790
    },
    {
      "epoch": 8.802410901467505,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.02061451866899346,
      "loss": 0.4325,
      "num_input_tokens_seen": 10960312,
      "step": 16795
    },
    {
      "epoch": 8.80503144654088,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.020608156225362165,
      "loss": 0.4212,
      "num_input_tokens_seen": 10962616,
      "step": 16800
    },
    {
      "epoch": 8.807651991614255,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.020601792608568564,
      "loss": 0.4087,
      "num_input_tokens_seen": 10966008,
      "step": 16805
    },
    {
      "epoch": 8.81027253668763,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.02059542781994386,
      "loss": 0.5358,
      "num_input_tokens_seen": 10970040,
      "step": 16810
    },
    {
      "epoch": 8.812893081761006,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.020589061860819498,
      "loss": 0.4716,
      "num_input_tokens_seen": 10972472,
      "step": 16815
    },
    {
      "epoch": 8.815513626834381,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02058269473252715,
      "loss": 0.3893,
      "num_input_tokens_seen": 10976664,
      "step": 16820
    },
    {
      "epoch": 8.818134171907756,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.020576326436398756,
      "loss": 0.4044,
      "num_input_tokens_seen": 10979192,
      "step": 16825
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.02056995697376649,
      "loss": 0.3816,
      "num_input_tokens_seen": 10981688,
      "step": 16830
    },
    {
      "epoch": 8.823375262054507,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02056358634596277,
      "loss": 0.4469,
      "num_input_tokens_seen": 10985176,
      "step": 16835
    },
    {
      "epoch": 8.825995807127882,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.020557214554320264,
      "loss": 0.4459,
      "num_input_tokens_seen": 10988536,
      "step": 16840
    },
    {
      "epoch": 8.828616352201259,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.02055084160017187,
      "loss": 0.3451,
      "num_input_tokens_seen": 10990936,
      "step": 16845
    },
    {
      "epoch": 8.831236897274634,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.020544467484850743,
      "loss": 0.4146,
      "num_input_tokens_seen": 10994200,
      "step": 16850
    },
    {
      "epoch": 8.833857442348009,
      "grad_norm": 0.04296875,
      "learning_rate": 0.020538092209690266,
      "loss": 0.4488,
      "num_input_tokens_seen": 10997240,
      "step": 16855
    },
    {
      "epoch": 8.836477987421384,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.020531715776024084,
      "loss": 0.3863,
      "num_input_tokens_seen": 11000568,
      "step": 16860
    },
    {
      "epoch": 8.83909853249476,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.020525338185186075,
      "loss": 0.3389,
      "num_input_tokens_seen": 11005528,
      "step": 16865
    },
    {
      "epoch": 8.841719077568134,
      "grad_norm": 0.048828125,
      "learning_rate": 0.020518959438510346,
      "loss": 0.4647,
      "num_input_tokens_seen": 11009176,
      "step": 16870
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.020512579537331273,
      "loss": 0.3736,
      "num_input_tokens_seen": 11012504,
      "step": 16875
    },
    {
      "epoch": 8.846960167714885,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02050619848298345,
      "loss": 0.451,
      "num_input_tokens_seen": 11015992,
      "step": 16880
    },
    {
      "epoch": 8.84958071278826,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.02049981627680172,
      "loss": 0.3371,
      "num_input_tokens_seen": 11019928,
      "step": 16885
    },
    {
      "epoch": 8.852201257861635,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02049343292012118,
      "loss": 0.5051,
      "num_input_tokens_seen": 11023096,
      "step": 16890
    },
    {
      "epoch": 8.85482180293501,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.02048704841427713,
      "loss": 0.4827,
      "num_input_tokens_seen": 11026232,
      "step": 16895
    },
    {
      "epoch": 8.857442348008385,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.020480662760605162,
      "loss": 0.625,
      "num_input_tokens_seen": 11029560,
      "step": 16900
    },
    {
      "epoch": 8.86006289308176,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.020474275960441067,
      "loss": 0.3806,
      "num_input_tokens_seen": 11032120,
      "step": 16905
    },
    {
      "epoch": 8.862683438155136,
      "grad_norm": 0.0390625,
      "learning_rate": 0.020467888015120898,
      "loss": 0.4385,
      "num_input_tokens_seen": 11034968,
      "step": 16910
    },
    {
      "epoch": 8.865303983228511,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.020461498925980932,
      "loss": 0.3122,
      "num_input_tokens_seen": 11038456,
      "step": 16915
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.020455108694357706,
      "loss": 0.3844,
      "num_input_tokens_seen": 11042712,
      "step": 16920
    },
    {
      "epoch": 8.870545073375261,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.020448717321587966,
      "loss": 0.5174,
      "num_input_tokens_seen": 11046008,
      "step": 16925
    },
    {
      "epoch": 8.873165618448636,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.02044232480900874,
      "loss": 0.5538,
      "num_input_tokens_seen": 11049336,
      "step": 16930
    },
    {
      "epoch": 8.875786163522012,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.020435931157957242,
      "loss": 0.386,
      "num_input_tokens_seen": 11053272,
      "step": 16935
    },
    {
      "epoch": 8.878406708595389,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.02042953636977096,
      "loss": 0.4752,
      "num_input_tokens_seen": 11057848,
      "step": 16940
    },
    {
      "epoch": 8.881027253668764,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.02042314044578762,
      "loss": 0.4455,
      "num_input_tokens_seen": 11060216,
      "step": 16945
    },
    {
      "epoch": 8.883647798742139,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.020416743387345164,
      "loss": 0.4346,
      "num_input_tokens_seen": 11063640,
      "step": 16950
    },
    {
      "epoch": 8.886268343815514,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.020410345195781793,
      "loss": 0.3548,
      "num_input_tokens_seen": 11066872,
      "step": 16955
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02040394587243593,
      "loss": 0.4562,
      "num_input_tokens_seen": 11069624,
      "step": 16960
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02039754541864624,
      "loss": 0.4258,
      "num_input_tokens_seen": 11073048,
      "step": 16965
    },
    {
      "epoch": 8.89412997903564,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.020391143835751628,
      "loss": 0.3846,
      "num_input_tokens_seen": 11076312,
      "step": 16970
    },
    {
      "epoch": 8.896750524109015,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.020384741125091234,
      "loss": 0.5096,
      "num_input_tokens_seen": 11079416,
      "step": 16975
    },
    {
      "epoch": 8.89937106918239,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.020378337288004426,
      "loss": 0.4564,
      "num_input_tokens_seen": 11082168,
      "step": 16980
    },
    {
      "epoch": 8.901991614255765,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.02037193232583082,
      "loss": 0.3235,
      "num_input_tokens_seen": 11084568,
      "step": 16985
    },
    {
      "epoch": 8.90461215932914,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.020365526239910253,
      "loss": 0.3933,
      "num_input_tokens_seen": 11087736,
      "step": 16990
    },
    {
      "epoch": 8.907232704402515,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02035911903158281,
      "loss": 0.4265,
      "num_input_tokens_seen": 11090648,
      "step": 16995
    },
    {
      "epoch": 8.90985324947589,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02035271070218881,
      "loss": 0.5916,
      "num_input_tokens_seen": 11093240,
      "step": 17000
    },
    {
      "epoch": 8.912473794549266,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.020346301253068796,
      "loss": 0.5238,
      "num_input_tokens_seen": 11096568,
      "step": 17005
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.02033989068556355,
      "loss": 0.4182,
      "num_input_tokens_seen": 11099384,
      "step": 17010
    },
    {
      "epoch": 8.917714884696016,
      "grad_norm": 0.0703125,
      "learning_rate": 0.020333479001014097,
      "loss": 0.261,
      "num_input_tokens_seen": 11101944,
      "step": 17015
    },
    {
      "epoch": 8.920335429769391,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.020327066200761683,
      "loss": 0.5814,
      "num_input_tokens_seen": 11104856,
      "step": 17020
    },
    {
      "epoch": 8.922955974842766,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.020320652286147795,
      "loss": 0.4279,
      "num_input_tokens_seen": 11108440,
      "step": 17025
    },
    {
      "epoch": 8.925576519916142,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02031423725851415,
      "loss": 0.4101,
      "num_input_tokens_seen": 11112504,
      "step": 17030
    },
    {
      "epoch": 8.928197064989519,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0203078211192027,
      "loss": 0.5053,
      "num_input_tokens_seen": 11116024,
      "step": 17035
    },
    {
      "epoch": 8.930817610062894,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.020301403869555625,
      "loss": 0.4325,
      "num_input_tokens_seen": 11119128,
      "step": 17040
    },
    {
      "epoch": 8.933438155136269,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02029498551091534,
      "loss": 0.5251,
      "num_input_tokens_seen": 11122584,
      "step": 17045
    },
    {
      "epoch": 8.936058700209644,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0202885660446245,
      "loss": 0.4376,
      "num_input_tokens_seen": 11125592,
      "step": 17050
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.020282145472025972,
      "loss": 0.4338,
      "num_input_tokens_seen": 11128312,
      "step": 17055
    },
    {
      "epoch": 8.941299790356394,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.020275723794462876,
      "loss": 0.3808,
      "num_input_tokens_seen": 11132120,
      "step": 17060
    },
    {
      "epoch": 8.94392033542977,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.020269301013278555,
      "loss": 0.3412,
      "num_input_tokens_seen": 11135768,
      "step": 17065
    },
    {
      "epoch": 8.946540880503145,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.020262877129816573,
      "loss": 0.4165,
      "num_input_tokens_seen": 11139128,
      "step": 17070
    },
    {
      "epoch": 8.94916142557652,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02025645214542074,
      "loss": 0.6251,
      "num_input_tokens_seen": 11142360,
      "step": 17075
    },
    {
      "epoch": 8.951781970649895,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.020250026061435084,
      "loss": 0.3674,
      "num_input_tokens_seen": 11144984,
      "step": 17080
    },
    {
      "epoch": 8.95440251572327,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.020243598879203874,
      "loss": 0.4864,
      "num_input_tokens_seen": 11147672,
      "step": 17085
    },
    {
      "epoch": 8.957023060796645,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0202371706000716,
      "loss": 0.4816,
      "num_input_tokens_seen": 11150872,
      "step": 17090
    },
    {
      "epoch": 8.95964360587002,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.020230741225382983,
      "loss": 0.4435,
      "num_input_tokens_seen": 11154360,
      "step": 17095
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.020224310756482976,
      "loss": 0.4623,
      "num_input_tokens_seen": 11156984,
      "step": 17100
    },
    {
      "epoch": 8.964884696016771,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.020217879194716768,
      "loss": 0.6931,
      "num_input_tokens_seen": 11160696,
      "step": 17105
    },
    {
      "epoch": 8.967505241090146,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.020211446541429753,
      "loss": 0.4901,
      "num_input_tokens_seen": 11163800,
      "step": 17110
    },
    {
      "epoch": 8.970125786163521,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.020205012797967586,
      "loss": 0.5245,
      "num_input_tokens_seen": 11167480,
      "step": 17115
    },
    {
      "epoch": 8.972746331236896,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.02019857796567611,
      "loss": 0.4663,
      "num_input_tokens_seen": 11170232,
      "step": 17120
    },
    {
      "epoch": 8.975366876310272,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.020192142045901442,
      "loss": 0.6331,
      "num_input_tokens_seen": 11173400,
      "step": 17125
    },
    {
      "epoch": 8.977987421383649,
      "grad_norm": 0.048828125,
      "learning_rate": 0.020185705039989895,
      "loss": 0.4294,
      "num_input_tokens_seen": 11176472,
      "step": 17130
    },
    {
      "epoch": 8.980607966457024,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02017926694928801,
      "loss": 0.3676,
      "num_input_tokens_seen": 11179288,
      "step": 17135
    },
    {
      "epoch": 8.983228511530399,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.020172827775142566,
      "loss": 0.4389,
      "num_input_tokens_seen": 11182136,
      "step": 17140
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.020166387518900573,
      "loss": 0.3878,
      "num_input_tokens_seen": 11186296,
      "step": 17145
    },
    {
      "epoch": 8.98846960167715,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.020159946181909246,
      "loss": 0.3416,
      "num_input_tokens_seen": 11189304,
      "step": 17150
    },
    {
      "epoch": 8.991090146750524,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.020153503765516046,
      "loss": 0.5104,
      "num_input_tokens_seen": 11192024,
      "step": 17155
    },
    {
      "epoch": 8.9937106918239,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02014706027106866,
      "loss": 0.4877,
      "num_input_tokens_seen": 11195160,
      "step": 17160
    },
    {
      "epoch": 8.996331236897275,
      "grad_norm": 0.185546875,
      "learning_rate": 0.020140615699914977,
      "loss": 0.5029,
      "num_input_tokens_seen": 11198264,
      "step": 17165
    },
    {
      "epoch": 8.99895178197065,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.020134170053403146,
      "loss": 0.4449,
      "num_input_tokens_seen": 11202840,
      "step": 17170
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.4556712508201599,
      "eval_runtime": 13.6647,
      "eval_samples_per_second": 62.058,
      "eval_steps_per_second": 15.514,
      "num_input_tokens_seen": 11204000,
      "step": 17172
    },
    {
      "epoch": 9.001572327044025,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.020127723332881504,
      "loss": 0.4626,
      "num_input_tokens_seen": 11206048,
      "step": 17175
    },
    {
      "epoch": 9.0041928721174,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.020121275539698644,
      "loss": 0.5313,
      "num_input_tokens_seen": 11210464,
      "step": 17180
    },
    {
      "epoch": 9.006813417190775,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.02011482667520337,
      "loss": 0.4041,
      "num_input_tokens_seen": 11213312,
      "step": 17185
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.020108376740744702,
      "loss": 0.3784,
      "num_input_tokens_seen": 11216704,
      "step": 17190
    },
    {
      "epoch": 9.012054507337526,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02010192573767191,
      "loss": 0.3994,
      "num_input_tokens_seen": 11221248,
      "step": 17195
    },
    {
      "epoch": 9.014675052410901,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.020095473667334454,
      "loss": 0.3985,
      "num_input_tokens_seen": 11223904,
      "step": 17200
    },
    {
      "epoch": 9.017295597484276,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.020089020531082035,
      "loss": 0.3767,
      "num_input_tokens_seen": 11227648,
      "step": 17205
    },
    {
      "epoch": 9.019916142557651,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.020082566330264586,
      "loss": 0.3815,
      "num_input_tokens_seen": 11230432,
      "step": 17210
    },
    {
      "epoch": 9.022536687631026,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.020076111066232246,
      "loss": 0.4173,
      "num_input_tokens_seen": 11233536,
      "step": 17215
    },
    {
      "epoch": 9.025157232704403,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.02006965474033538,
      "loss": 0.4488,
      "num_input_tokens_seen": 11236512,
      "step": 17220
    },
    {
      "epoch": 9.027777777777779,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02006319735392458,
      "loss": 0.3756,
      "num_input_tokens_seen": 11238656,
      "step": 17225
    },
    {
      "epoch": 9.030398322851154,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020056738908350656,
      "loss": 0.4953,
      "num_input_tokens_seen": 11242400,
      "step": 17230
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.020050279404964642,
      "loss": 0.4251,
      "num_input_tokens_seen": 11245824,
      "step": 17235
    },
    {
      "epoch": 9.035639412997904,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.020043818845117793,
      "loss": 0.3244,
      "num_input_tokens_seen": 11248736,
      "step": 17240
    },
    {
      "epoch": 9.03825995807128,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.020037357230161586,
      "loss": 0.5091,
      "num_input_tokens_seen": 11252096,
      "step": 17245
    },
    {
      "epoch": 9.040880503144654,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.020030894561447715,
      "loss": 0.4028,
      "num_input_tokens_seen": 11258496,
      "step": 17250
    },
    {
      "epoch": 9.04350104821803,
      "grad_norm": 0.07421875,
      "learning_rate": 0.020024430840328098,
      "loss": 0.3845,
      "num_input_tokens_seen": 11260960,
      "step": 17255
    },
    {
      "epoch": 9.046121593291405,
      "grad_norm": 0.06640625,
      "learning_rate": 0.020017966068154876,
      "loss": 0.3891,
      "num_input_tokens_seen": 11264512,
      "step": 17260
    },
    {
      "epoch": 9.04874213836478,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.020011500246280393,
      "loss": 0.4602,
      "num_input_tokens_seen": 11268256,
      "step": 17265
    },
    {
      "epoch": 9.051362683438155,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.020005033376057236,
      "loss": 0.565,
      "num_input_tokens_seen": 11271328,
      "step": 17270
    },
    {
      "epoch": 9.05398322851153,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0199985654588382,
      "loss": 0.3339,
      "num_input_tokens_seen": 11274048,
      "step": 17275
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.01999209649597629,
      "loss": 0.4749,
      "num_input_tokens_seen": 11277440,
      "step": 17280
    },
    {
      "epoch": 9.05922431865828,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.019985626488824747,
      "loss": 0.4173,
      "num_input_tokens_seen": 11281120,
      "step": 17285
    },
    {
      "epoch": 9.061844863731656,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.019979155438737024,
      "loss": 0.4224,
      "num_input_tokens_seen": 11284064,
      "step": 17290
    },
    {
      "epoch": 9.064465408805031,
      "grad_norm": 0.0390625,
      "learning_rate": 0.019972683347066787,
      "loss": 0.367,
      "num_input_tokens_seen": 11288096,
      "step": 17295
    },
    {
      "epoch": 9.067085953878406,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.01996621021516793,
      "loss": 0.309,
      "num_input_tokens_seen": 11292032,
      "step": 17300
    },
    {
      "epoch": 9.069706498951781,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.01995973604439455,
      "loss": 0.3382,
      "num_input_tokens_seen": 11294720,
      "step": 17305
    },
    {
      "epoch": 9.072327044025156,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01995326083610097,
      "loss": 0.4382,
      "num_input_tokens_seen": 11298880,
      "step": 17310
    },
    {
      "epoch": 9.074947589098532,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01994678459164174,
      "loss": 0.5959,
      "num_input_tokens_seen": 11302336,
      "step": 17315
    },
    {
      "epoch": 9.077568134171909,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0199403073123716,
      "loss": 0.3305,
      "num_input_tokens_seen": 11305216,
      "step": 17320
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.01993382899964554,
      "loss": 0.539,
      "num_input_tokens_seen": 11308512,
      "step": 17325
    },
    {
      "epoch": 9.082809224318659,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.01992734965481874,
      "loss": 0.4943,
      "num_input_tokens_seen": 11312416,
      "step": 17330
    },
    {
      "epoch": 9.085429769392034,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019920869279246602,
      "loss": 0.3279,
      "num_input_tokens_seen": 11315904,
      "step": 17335
    },
    {
      "epoch": 9.08805031446541,
      "grad_norm": 0.0086669921875,
      "learning_rate": 0.019914387874284754,
      "loss": 0.3012,
      "num_input_tokens_seen": 11321664,
      "step": 17340
    },
    {
      "epoch": 9.090670859538784,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01990790544128903,
      "loss": 0.3725,
      "num_input_tokens_seen": 11325728,
      "step": 17345
    },
    {
      "epoch": 9.09329140461216,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.019901421981615482,
      "loss": 0.3907,
      "num_input_tokens_seen": 11329344,
      "step": 17350
    },
    {
      "epoch": 9.095911949685535,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.019894937496620368,
      "loss": 0.3833,
      "num_input_tokens_seen": 11331840,
      "step": 17355
    },
    {
      "epoch": 9.09853249475891,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01988845198766018,
      "loss": 0.4535,
      "num_input_tokens_seen": 11334912,
      "step": 17360
    },
    {
      "epoch": 9.101153039832285,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.019881965456091603,
      "loss": 0.4233,
      "num_input_tokens_seen": 11337344,
      "step": 17365
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01987547790327155,
      "loss": 0.5587,
      "num_input_tokens_seen": 11340384,
      "step": 17370
    },
    {
      "epoch": 9.106394129979035,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.01986898933055714,
      "loss": 0.4179,
      "num_input_tokens_seen": 11342976,
      "step": 17375
    },
    {
      "epoch": 9.10901467505241,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.019862499739305713,
      "loss": 0.3859,
      "num_input_tokens_seen": 11345504,
      "step": 17380
    },
    {
      "epoch": 9.111635220125786,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.019856009130874815,
      "loss": 0.3978,
      "num_input_tokens_seen": 11349216,
      "step": 17385
    },
    {
      "epoch": 9.114255765199161,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.019849517506622207,
      "loss": 0.4338,
      "num_input_tokens_seen": 11352576,
      "step": 17390
    },
    {
      "epoch": 9.116876310272536,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01984302486790586,
      "loss": 0.3664,
      "num_input_tokens_seen": 11355104,
      "step": 17395
    },
    {
      "epoch": 9.119496855345911,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.019836531216083968,
      "loss": 0.4792,
      "num_input_tokens_seen": 11361504,
      "step": 17400
    },
    {
      "epoch": 9.122117400419286,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01983003655251492,
      "loss": 0.3879,
      "num_input_tokens_seen": 11365120,
      "step": 17405
    },
    {
      "epoch": 9.124737945492662,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.019823540878557328,
      "loss": 0.4876,
      "num_input_tokens_seen": 11368800,
      "step": 17410
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 0.046875,
      "learning_rate": 0.019817044195570014,
      "loss": 0.6206,
      "num_input_tokens_seen": 11371616,
      "step": 17415
    },
    {
      "epoch": 9.129979035639414,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019810546504912013,
      "loss": 0.5157,
      "num_input_tokens_seen": 11374496,
      "step": 17420
    },
    {
      "epoch": 9.132599580712789,
      "grad_norm": 0.044921875,
      "learning_rate": 0.019804047807942562,
      "loss": 0.3824,
      "num_input_tokens_seen": 11378624,
      "step": 17425
    },
    {
      "epoch": 9.135220125786164,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.019797548106021124,
      "loss": 0.5013,
      "num_input_tokens_seen": 11381312,
      "step": 17430
    },
    {
      "epoch": 9.13784067085954,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.01979104740050735,
      "loss": 0.3462,
      "num_input_tokens_seen": 11383808,
      "step": 17435
    },
    {
      "epoch": 9.140461215932914,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.019784545692761113,
      "loss": 0.4565,
      "num_input_tokens_seen": 11387840,
      "step": 17440
    },
    {
      "epoch": 9.14308176100629,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01977804298414251,
      "loss": 0.343,
      "num_input_tokens_seen": 11392800,
      "step": 17445
    },
    {
      "epoch": 9.145702306079665,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.019771539276011826,
      "loss": 0.3028,
      "num_input_tokens_seen": 11396224,
      "step": 17450
    },
    {
      "epoch": 9.14832285115304,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01976503456972956,
      "loss": 0.435,
      "num_input_tokens_seen": 11399456,
      "step": 17455
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019758528866656425,
      "loss": 0.3901,
      "num_input_tokens_seen": 11402304,
      "step": 17460
    },
    {
      "epoch": 9.15356394129979,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.019752022168153336,
      "loss": 0.4569,
      "num_input_tokens_seen": 11405440,
      "step": 17465
    },
    {
      "epoch": 9.156184486373165,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.01974551447558143,
      "loss": 0.4417,
      "num_input_tokens_seen": 11408256,
      "step": 17470
    },
    {
      "epoch": 9.15880503144654,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01973900579030203,
      "loss": 0.3257,
      "num_input_tokens_seen": 11412288,
      "step": 17475
    },
    {
      "epoch": 9.161425576519916,
      "grad_norm": 0.0859375,
      "learning_rate": 0.01973249611367668,
      "loss": 0.3802,
      "num_input_tokens_seen": 11414816,
      "step": 17480
    },
    {
      "epoch": 9.164046121593291,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01972598544706714,
      "loss": 0.4325,
      "num_input_tokens_seen": 11417632,
      "step": 17485
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.019719473791835362,
      "loss": 0.5466,
      "num_input_tokens_seen": 11420384,
      "step": 17490
    },
    {
      "epoch": 9.169287211740041,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.019712961149343507,
      "loss": 0.484,
      "num_input_tokens_seen": 11423648,
      "step": 17495
    },
    {
      "epoch": 9.171907756813416,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.019706447520953952,
      "loss": 0.3744,
      "num_input_tokens_seen": 11426976,
      "step": 17500
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.019699932908029266,
      "loss": 0.375,
      "num_input_tokens_seen": 11430304,
      "step": 17505
    },
    {
      "epoch": 9.177148846960169,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.019693417311932235,
      "loss": 0.4181,
      "num_input_tokens_seen": 11432704,
      "step": 17510
    },
    {
      "epoch": 9.179769392033544,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.019686900734025847,
      "loss": 0.3854,
      "num_input_tokens_seen": 11436448,
      "step": 17515
    },
    {
      "epoch": 9.182389937106919,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.019680383175673298,
      "loss": 0.3911,
      "num_input_tokens_seen": 11439872,
      "step": 17520
    },
    {
      "epoch": 9.185010482180294,
      "grad_norm": 0.10546875,
      "learning_rate": 0.019673864638237986,
      "loss": 0.5293,
      "num_input_tokens_seen": 11442304,
      "step": 17525
    },
    {
      "epoch": 9.18763102725367,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.019667345123083515,
      "loss": 0.5739,
      "num_input_tokens_seen": 11445856,
      "step": 17530
    },
    {
      "epoch": 9.190251572327044,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.019660824631573693,
      "loss": 0.4204,
      "num_input_tokens_seen": 11452288,
      "step": 17535
    },
    {
      "epoch": 9.19287211740042,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.019654303165072527,
      "loss": 0.5208,
      "num_input_tokens_seen": 11456032,
      "step": 17540
    },
    {
      "epoch": 9.195492662473795,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.019647780724944237,
      "loss": 0.3797,
      "num_input_tokens_seen": 11459392,
      "step": 17545
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.01964125731255325,
      "loss": 0.4508,
      "num_input_tokens_seen": 11462336,
      "step": 17550
    },
    {
      "epoch": 9.200733752620545,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.01963473292926418,
      "loss": 0.4068,
      "num_input_tokens_seen": 11464896,
      "step": 17555
    },
    {
      "epoch": 9.20335429769392,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.019628207576441858,
      "loss": 0.4094,
      "num_input_tokens_seen": 11472224,
      "step": 17560
    },
    {
      "epoch": 9.205974842767295,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.01962168125545131,
      "loss": 0.4516,
      "num_input_tokens_seen": 11474912,
      "step": 17565
    },
    {
      "epoch": 9.20859538784067,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.019615153967657773,
      "loss": 0.6118,
      "num_input_tokens_seen": 11478240,
      "step": 17570
    },
    {
      "epoch": 9.211215932914046,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.01960862571442668,
      "loss": 0.3776,
      "num_input_tokens_seen": 11481216,
      "step": 17575
    },
    {
      "epoch": 9.213836477987421,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01960209649712366,
      "loss": 0.4008,
      "num_input_tokens_seen": 11484384,
      "step": 17580
    },
    {
      "epoch": 9.216457023060796,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.01959556631711456,
      "loss": 0.482,
      "num_input_tokens_seen": 11487104,
      "step": 17585
    },
    {
      "epoch": 9.219077568134171,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.019589035175765414,
      "loss": 0.3419,
      "num_input_tokens_seen": 11489696,
      "step": 17590
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.019582503074442464,
      "loss": 0.5237,
      "num_input_tokens_seen": 11492512,
      "step": 17595
    },
    {
      "epoch": 9.224318658280922,
      "grad_norm": 0.04296875,
      "learning_rate": 0.019575970014512153,
      "loss": 0.3863,
      "num_input_tokens_seen": 11495648,
      "step": 17600
    },
    {
      "epoch": 9.226939203354299,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01956943599734112,
      "loss": 0.4675,
      "num_input_tokens_seen": 11498976,
      "step": 17605
    },
    {
      "epoch": 9.229559748427674,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.019562901024296205,
      "loss": 0.369,
      "num_input_tokens_seen": 11501888,
      "step": 17610
    },
    {
      "epoch": 9.232180293501049,
      "grad_norm": 0.052734375,
      "learning_rate": 0.019556365096744455,
      "loss": 0.5092,
      "num_input_tokens_seen": 11504544,
      "step": 17615
    },
    {
      "epoch": 9.234800838574424,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.019549828216053104,
      "loss": 0.3626,
      "num_input_tokens_seen": 11507168,
      "step": 17620
    },
    {
      "epoch": 9.2374213836478,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0195432903835896,
      "loss": 0.3707,
      "num_input_tokens_seen": 11510272,
      "step": 17625
    },
    {
      "epoch": 9.240041928721174,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.01953675160072158,
      "loss": 0.4447,
      "num_input_tokens_seen": 11513280,
      "step": 17630
    },
    {
      "epoch": 9.24266247379455,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.019530211868816885,
      "loss": 0.3437,
      "num_input_tokens_seen": 11516384,
      "step": 17635
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.019523671189243547,
      "loss": 0.4885,
      "num_input_tokens_seen": 11520672,
      "step": 17640
    },
    {
      "epoch": 9.2479035639413,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.019517129563369807,
      "loss": 0.4644,
      "num_input_tokens_seen": 11523456,
      "step": 17645
    },
    {
      "epoch": 9.250524109014675,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.019510586992564093,
      "loss": 0.3255,
      "num_input_tokens_seen": 11526240,
      "step": 17650
    },
    {
      "epoch": 9.25314465408805,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.019504043478195048,
      "loss": 0.4466,
      "num_input_tokens_seen": 11531040,
      "step": 17655
    },
    {
      "epoch": 9.255765199161425,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.019497499021631486,
      "loss": 0.4311,
      "num_input_tokens_seen": 11533984,
      "step": 17660
    },
    {
      "epoch": 9.2583857442348,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01949095362424244,
      "loss": 0.4321,
      "num_input_tokens_seen": 11537248,
      "step": 17665
    },
    {
      "epoch": 9.261006289308176,
      "grad_norm": 0.0859375,
      "learning_rate": 0.019484407287397133,
      "loss": 0.4337,
      "num_input_tokens_seen": 11540384,
      "step": 17670
    },
    {
      "epoch": 9.26362683438155,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01947786001246498,
      "loss": 0.4996,
      "num_input_tokens_seen": 11543424,
      "step": 17675
    },
    {
      "epoch": 9.266247379454926,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0194713118008156,
      "loss": 0.4172,
      "num_input_tokens_seen": 11547136,
      "step": 17680
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0194647626538188,
      "loss": 0.5174,
      "num_input_tokens_seen": 11550528,
      "step": 17685
    },
    {
      "epoch": 9.271488469601676,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.019458212572844587,
      "loss": 0.3279,
      "num_input_tokens_seen": 11553440,
      "step": 17690
    },
    {
      "epoch": 9.274109014675052,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.019451661559263173,
      "loss": 0.332,
      "num_input_tokens_seen": 11557184,
      "step": 17695
    },
    {
      "epoch": 9.276729559748428,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.01944510961444495,
      "loss": 0.3712,
      "num_input_tokens_seen": 11560416,
      "step": 17700
    },
    {
      "epoch": 9.279350104821804,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0194385567397605,
      "loss": 0.4288,
      "num_input_tokens_seen": 11563712,
      "step": 17705
    },
    {
      "epoch": 9.281970649895179,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01943200293658062,
      "loss": 0.4652,
      "num_input_tokens_seen": 11567776,
      "step": 17710
    },
    {
      "epoch": 9.284591194968554,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.019425448206276293,
      "loss": 0.446,
      "num_input_tokens_seen": 11571456,
      "step": 17715
    },
    {
      "epoch": 9.28721174004193,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.019418892550218687,
      "loss": 0.4401,
      "num_input_tokens_seen": 11575136,
      "step": 17720
    },
    {
      "epoch": 9.289832285115304,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.019412335969779166,
      "loss": 0.4475,
      "num_input_tokens_seen": 11578880,
      "step": 17725
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.019405778466329307,
      "loss": 0.3099,
      "num_input_tokens_seen": 11581792,
      "step": 17730
    },
    {
      "epoch": 9.295073375262055,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.019399220041240856,
      "loss": 0.537,
      "num_input_tokens_seen": 11584704,
      "step": 17735
    },
    {
      "epoch": 9.29769392033543,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.01939266069588576,
      "loss": 0.5318,
      "num_input_tokens_seen": 11587936,
      "step": 17740
    },
    {
      "epoch": 9.300314465408805,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.01938610043163616,
      "loss": 0.4937,
      "num_input_tokens_seen": 11591232,
      "step": 17745
    },
    {
      "epoch": 9.30293501048218,
      "grad_norm": 0.037109375,
      "learning_rate": 0.019379539249864384,
      "loss": 0.6168,
      "num_input_tokens_seen": 11594656,
      "step": 17750
    },
    {
      "epoch": 9.305555555555555,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01937297715194297,
      "loss": 0.3654,
      "num_input_tokens_seen": 11598016,
      "step": 17755
    },
    {
      "epoch": 9.30817610062893,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01936641413924462,
      "loss": 0.3392,
      "num_input_tokens_seen": 11601312,
      "step": 17760
    },
    {
      "epoch": 9.310796645702306,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.01935985021314225,
      "loss": 0.5224,
      "num_input_tokens_seen": 11604896,
      "step": 17765
    },
    {
      "epoch": 9.31341719077568,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.019353285375008956,
      "loss": 0.4061,
      "num_input_tokens_seen": 11607840,
      "step": 17770
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 0.048828125,
      "learning_rate": 0.019346719626218024,
      "loss": 0.3963,
      "num_input_tokens_seen": 11611392,
      "step": 17775
    },
    {
      "epoch": 9.318658280922431,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.019340152968142933,
      "loss": 0.38,
      "num_input_tokens_seen": 11614752,
      "step": 17780
    },
    {
      "epoch": 9.321278825995806,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.019333585402157363,
      "loss": 0.4092,
      "num_input_tokens_seen": 11617728,
      "step": 17785
    },
    {
      "epoch": 9.323899371069182,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.019327016929635163,
      "loss": 0.3892,
      "num_input_tokens_seen": 11620736,
      "step": 17790
    },
    {
      "epoch": 9.326519916142558,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.019320447551950382,
      "loss": 0.5086,
      "num_input_tokens_seen": 11623744,
      "step": 17795
    },
    {
      "epoch": 9.329140461215934,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.019313877270477276,
      "loss": 0.4596,
      "num_input_tokens_seen": 11626496,
      "step": 17800
    },
    {
      "epoch": 9.331761006289309,
      "grad_norm": 0.041015625,
      "learning_rate": 0.019307306086590252,
      "loss": 0.3612,
      "num_input_tokens_seen": 11629504,
      "step": 17805
    },
    {
      "epoch": 9.334381551362684,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.019300734001663933,
      "loss": 0.4653,
      "num_input_tokens_seen": 11632384,
      "step": 17810
    },
    {
      "epoch": 9.33700209643606,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.019294161017073135,
      "loss": 0.3772,
      "num_input_tokens_seen": 11635616,
      "step": 17815
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.019287587134192832,
      "loss": 0.4071,
      "num_input_tokens_seen": 11638912,
      "step": 17820
    },
    {
      "epoch": 9.34224318658281,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.019281012354398226,
      "loss": 0.4741,
      "num_input_tokens_seen": 11642432,
      "step": 17825
    },
    {
      "epoch": 9.344863731656185,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.019274436679064668,
      "loss": 0.3866,
      "num_input_tokens_seen": 11645120,
      "step": 17830
    },
    {
      "epoch": 9.34748427672956,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.01926786010956772,
      "loss": 0.3847,
      "num_input_tokens_seen": 11647456,
      "step": 17835
    },
    {
      "epoch": 9.350104821802935,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.019261282647283138,
      "loss": 0.4517,
      "num_input_tokens_seen": 11650976,
      "step": 17840
    },
    {
      "epoch": 9.35272536687631,
      "grad_norm": 0.05859375,
      "learning_rate": 0.019254704293586833,
      "loss": 0.4031,
      "num_input_tokens_seen": 11654304,
      "step": 17845
    },
    {
      "epoch": 9.355345911949685,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.01924812504985493,
      "loss": 0.4822,
      "num_input_tokens_seen": 11658592,
      "step": 17850
    },
    {
      "epoch": 9.35796645702306,
      "grad_norm": 0.0546875,
      "learning_rate": 0.019241544917463736,
      "loss": 0.5087,
      "num_input_tokens_seen": 11662464,
      "step": 17855
    },
    {
      "epoch": 9.360587002096436,
      "grad_norm": 0.037109375,
      "learning_rate": 0.019234963897789736,
      "loss": 0.3729,
      "num_input_tokens_seen": 11665248,
      "step": 17860
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.019228381992209602,
      "loss": 0.468,
      "num_input_tokens_seen": 11668256,
      "step": 17865
    },
    {
      "epoch": 9.365828092243186,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.019221799202100192,
      "loss": 0.4678,
      "num_input_tokens_seen": 11671264,
      "step": 17870
    },
    {
      "epoch": 9.368448637316561,
      "grad_norm": 0.056640625,
      "learning_rate": 0.019215215528838554,
      "loss": 0.5849,
      "num_input_tokens_seen": 11673728,
      "step": 17875
    },
    {
      "epoch": 9.371069182389936,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.019208630973801922,
      "loss": 0.3793,
      "num_input_tokens_seen": 11676960,
      "step": 17880
    },
    {
      "epoch": 9.373689727463312,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.019202045538367702,
      "loss": 0.4651,
      "num_input_tokens_seen": 11681440,
      "step": 17885
    },
    {
      "epoch": 9.376310272536688,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.019195459223913497,
      "loss": 0.4132,
      "num_input_tokens_seen": 11684544,
      "step": 17890
    },
    {
      "epoch": 9.378930817610064,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01918887203181708,
      "loss": 0.4189,
      "num_input_tokens_seen": 11687584,
      "step": 17895
    },
    {
      "epoch": 9.381551362683439,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01918228396345643,
      "loss": 0.3331,
      "num_input_tokens_seen": 11690944,
      "step": 17900
    },
    {
      "epoch": 9.384171907756814,
      "grad_norm": 0.064453125,
      "learning_rate": 0.019175695020209688,
      "loss": 0.4013,
      "num_input_tokens_seen": 11694112,
      "step": 17905
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.019169105203455182,
      "loss": 0.4431,
      "num_input_tokens_seen": 11697376,
      "step": 17910
    },
    {
      "epoch": 9.389412997903564,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.019162514514571436,
      "loss": 0.3602,
      "num_input_tokens_seen": 11700832,
      "step": 17915
    },
    {
      "epoch": 9.39203354297694,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.019155922954937142,
      "loss": 0.405,
      "num_input_tokens_seen": 11703520,
      "step": 17920
    },
    {
      "epoch": 9.394654088050315,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.019149330525931176,
      "loss": 0.4024,
      "num_input_tokens_seen": 11707008,
      "step": 17925
    },
    {
      "epoch": 9.39727463312369,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.019142737228932605,
      "loss": 0.2949,
      "num_input_tokens_seen": 11709408,
      "step": 17930
    },
    {
      "epoch": 9.399895178197065,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.019136143065320667,
      "loss": 0.3796,
      "num_input_tokens_seen": 11712000,
      "step": 17935
    },
    {
      "epoch": 9.40251572327044,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.019129548036474785,
      "loss": 0.4748,
      "num_input_tokens_seen": 11716064,
      "step": 17940
    },
    {
      "epoch": 9.405136268343815,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.01912295214377457,
      "loss": 0.3662,
      "num_input_tokens_seen": 11719776,
      "step": 17945
    },
    {
      "epoch": 9.40775681341719,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0191163553885998,
      "loss": 0.3621,
      "num_input_tokens_seen": 11722880,
      "step": 17950
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.01910975777233045,
      "loss": 0.5731,
      "num_input_tokens_seen": 11725824,
      "step": 17955
    },
    {
      "epoch": 9.41299790356394,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.019103159296346657,
      "loss": 0.5015,
      "num_input_tokens_seen": 11728480,
      "step": 17960
    },
    {
      "epoch": 9.415618448637316,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.019096559962028747,
      "loss": 0.4053,
      "num_input_tokens_seen": 11731040,
      "step": 17965
    },
    {
      "epoch": 9.418238993710691,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01908995977075724,
      "loss": 0.3254,
      "num_input_tokens_seen": 11733664,
      "step": 17970
    },
    {
      "epoch": 9.420859538784066,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.019083358723912796,
      "loss": 0.3606,
      "num_input_tokens_seen": 11737216,
      "step": 17975
    },
    {
      "epoch": 9.423480083857442,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.019076756822876302,
      "loss": 0.3984,
      "num_input_tokens_seen": 11740864,
      "step": 17980
    },
    {
      "epoch": 9.426100628930818,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01907015406902879,
      "loss": 0.3977,
      "num_input_tokens_seen": 11743776,
      "step": 17985
    },
    {
      "epoch": 9.428721174004194,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.019063550463751477,
      "loss": 0.399,
      "num_input_tokens_seen": 11747552,
      "step": 17990
    },
    {
      "epoch": 9.431341719077569,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.01905694600842577,
      "loss": 0.4525,
      "num_input_tokens_seen": 11750656,
      "step": 17995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.01905034070443324,
      "loss": 0.3503,
      "num_input_tokens_seen": 11753888,
      "step": 18000
    },
    {
      "epoch": 9.43658280922432,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.019043734553155642,
      "loss": 0.4181,
      "num_input_tokens_seen": 11757184,
      "step": 18005
    },
    {
      "epoch": 9.439203354297694,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.019037127555974916,
      "loss": 0.3676,
      "num_input_tokens_seen": 11759936,
      "step": 18010
    },
    {
      "epoch": 9.44182389937107,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.019030519714273157,
      "loss": 0.4262,
      "num_input_tokens_seen": 11762816,
      "step": 18015
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.019023911029432655,
      "loss": 0.3561,
      "num_input_tokens_seen": 11765824,
      "step": 18020
    },
    {
      "epoch": 9.44706498951782,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.019017301502835877,
      "loss": 0.368,
      "num_input_tokens_seen": 11769056,
      "step": 18025
    },
    {
      "epoch": 9.449685534591195,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.019010691135865454,
      "loss": 0.3411,
      "num_input_tokens_seen": 11771808,
      "step": 18030
    },
    {
      "epoch": 9.45230607966457,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.019004079929904202,
      "loss": 0.3127,
      "num_input_tokens_seen": 11774720,
      "step": 18035
    },
    {
      "epoch": 9.454926624737945,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.01899746788633511,
      "loss": 0.3872,
      "num_input_tokens_seen": 11777664,
      "step": 18040
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.018990855006541345,
      "loss": 0.3352,
      "num_input_tokens_seen": 11781696,
      "step": 18045
    },
    {
      "epoch": 9.460167714884696,
      "grad_norm": 0.056640625,
      "learning_rate": 0.018984241291906234,
      "loss": 0.401,
      "num_input_tokens_seen": 11784672,
      "step": 18050
    },
    {
      "epoch": 9.46278825995807,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01897762674381331,
      "loss": 0.4138,
      "num_input_tokens_seen": 11787424,
      "step": 18055
    },
    {
      "epoch": 9.465408805031446,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.018971011363646244,
      "loss": 0.4597,
      "num_input_tokens_seen": 11789824,
      "step": 18060
    },
    {
      "epoch": 9.468029350104821,
      "grad_norm": 0.041015625,
      "learning_rate": 0.018964395152788905,
      "loss": 0.2465,
      "num_input_tokens_seen": 11793120,
      "step": 18065
    },
    {
      "epoch": 9.470649895178196,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.01895777811262533,
      "loss": 0.4135,
      "num_input_tokens_seen": 11798720,
      "step": 18070
    },
    {
      "epoch": 9.473270440251572,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.018951160244539727,
      "loss": 0.4095,
      "num_input_tokens_seen": 11801600,
      "step": 18075
    },
    {
      "epoch": 9.475890985324948,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.018944541549916476,
      "loss": 0.3222,
      "num_input_tokens_seen": 11804640,
      "step": 18080
    },
    {
      "epoch": 9.478511530398324,
      "grad_norm": 0.046875,
      "learning_rate": 0.018937922030140133,
      "loss": 0.4206,
      "num_input_tokens_seen": 11807296,
      "step": 18085
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01893130168659543,
      "loss": 0.2875,
      "num_input_tokens_seen": 11809920,
      "step": 18090
    },
    {
      "epoch": 9.483752620545074,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.018924680520667266,
      "loss": 0.3771,
      "num_input_tokens_seen": 11813440,
      "step": 18095
    },
    {
      "epoch": 9.48637316561845,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.018918058533740713,
      "loss": 0.4669,
      "num_input_tokens_seen": 11817440,
      "step": 18100
    },
    {
      "epoch": 9.488993710691824,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.01891143572720101,
      "loss": 0.4689,
      "num_input_tokens_seen": 11820192,
      "step": 18105
    },
    {
      "epoch": 9.4916142557652,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01890481210243358,
      "loss": 0.5249,
      "num_input_tokens_seen": 11824192,
      "step": 18110
    },
    {
      "epoch": 9.494234800838575,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.018898187660824007,
      "loss": 0.5816,
      "num_input_tokens_seen": 11826944,
      "step": 18115
    },
    {
      "epoch": 9.49685534591195,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01889156240375804,
      "loss": 0.3806,
      "num_input_tokens_seen": 11830784,
      "step": 18120
    },
    {
      "epoch": 9.499475890985325,
      "grad_norm": 0.087890625,
      "learning_rate": 0.018884936332621626,
      "loss": 0.5765,
      "num_input_tokens_seen": 11833632,
      "step": 18125
    },
    {
      "epoch": 9.5020964360587,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.01887830944880085,
      "loss": 0.5452,
      "num_input_tokens_seen": 11837056,
      "step": 18130
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.01887168175368198,
      "loss": 0.3383,
      "num_input_tokens_seen": 11840736,
      "step": 18135
    },
    {
      "epoch": 9.50733752620545,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.018865053248651464,
      "loss": 0.5874,
      "num_input_tokens_seen": 11843328,
      "step": 18140
    },
    {
      "epoch": 9.509958071278826,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.0188584239350959,
      "loss": 0.493,
      "num_input_tokens_seen": 11846784,
      "step": 18145
    },
    {
      "epoch": 9.5125786163522,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.018851793814402062,
      "loss": 0.4669,
      "num_input_tokens_seen": 11849984,
      "step": 18150
    },
    {
      "epoch": 9.515199161425576,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.01884516288795691,
      "loss": 0.4572,
      "num_input_tokens_seen": 11852992,
      "step": 18155
    },
    {
      "epoch": 9.517819706498951,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.018838531157147544,
      "loss": 0.3977,
      "num_input_tokens_seen": 11855776,
      "step": 18160
    },
    {
      "epoch": 9.520440251572326,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.018831898623361258,
      "loss": 0.3327,
      "num_input_tokens_seen": 11859008,
      "step": 18165
    },
    {
      "epoch": 9.523060796645701,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.018825265287985494,
      "loss": 0.3377,
      "num_input_tokens_seen": 11861632,
      "step": 18170
    },
    {
      "epoch": 9.525681341719078,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.01881863115240787,
      "loss": 0.3519,
      "num_input_tokens_seen": 11864864,
      "step": 18175
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.018811996218016176,
      "loss": 0.4925,
      "num_input_tokens_seen": 11868224,
      "step": 18180
    },
    {
      "epoch": 9.530922431865829,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.018805360486198366,
      "loss": 0.4667,
      "num_input_tokens_seen": 11871072,
      "step": 18185
    },
    {
      "epoch": 9.533542976939204,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.018798723958342548,
      "loss": 0.4005,
      "num_input_tokens_seen": 11873888,
      "step": 18190
    },
    {
      "epoch": 9.536163522012579,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.018792086635837025,
      "loss": 0.3623,
      "num_input_tokens_seen": 11876608,
      "step": 18195
    },
    {
      "epoch": 9.538784067085954,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.018785448520070237,
      "loss": 0.388,
      "num_input_tokens_seen": 11880864,
      "step": 18200
    },
    {
      "epoch": 9.54140461215933,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.018778809612430804,
      "loss": 0.4214,
      "num_input_tokens_seen": 11883520,
      "step": 18205
    },
    {
      "epoch": 9.544025157232705,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.018772169914307513,
      "loss": 0.3502,
      "num_input_tokens_seen": 11887168,
      "step": 18210
    },
    {
      "epoch": 9.54664570230608,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.018765529427089308,
      "loss": 0.551,
      "num_input_tokens_seen": 11890976,
      "step": 18215
    },
    {
      "epoch": 9.549266247379455,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01875888815216531,
      "loss": 0.4174,
      "num_input_tokens_seen": 11893888,
      "step": 18220
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.018752246090924792,
      "loss": 0.4135,
      "num_input_tokens_seen": 11896704,
      "step": 18225
    },
    {
      "epoch": 9.554507337526205,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0187456032447572,
      "loss": 0.5222,
      "num_input_tokens_seen": 11900736,
      "step": 18230
    },
    {
      "epoch": 9.55712788259958,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01873895961505214,
      "loss": 0.4945,
      "num_input_tokens_seen": 11904160,
      "step": 18235
    },
    {
      "epoch": 9.559748427672956,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.01873231520319939,
      "loss": 0.3479,
      "num_input_tokens_seen": 11907200,
      "step": 18240
    },
    {
      "epoch": 9.56236897274633,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.018725670010588877,
      "loss": 0.3456,
      "num_input_tokens_seen": 11910368,
      "step": 18245
    },
    {
      "epoch": 9.564989517819706,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0187190240386107,
      "loss": 0.4498,
      "num_input_tokens_seen": 11913472,
      "step": 18250
    },
    {
      "epoch": 9.567610062893081,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.018712377288655126,
      "loss": 0.3428,
      "num_input_tokens_seen": 11917280,
      "step": 18255
    },
    {
      "epoch": 9.570230607966456,
      "grad_norm": 0.033203125,
      "learning_rate": 0.018705729762112572,
      "loss": 0.3859,
      "num_input_tokens_seen": 11919744,
      "step": 18260
    },
    {
      "epoch": 9.572851153039831,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.01869908146037363,
      "loss": 0.5319,
      "num_input_tokens_seen": 11924096,
      "step": 18265
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.01869243238482905,
      "loss": 0.35,
      "num_input_tokens_seen": 11927744,
      "step": 18270
    },
    {
      "epoch": 9.578092243186584,
      "grad_norm": 0.111328125,
      "learning_rate": 0.018685782536869745,
      "loss": 0.3976,
      "num_input_tokens_seen": 11930368,
      "step": 18275
    },
    {
      "epoch": 9.580712788259959,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.018679131917886775,
      "loss": 0.4293,
      "num_input_tokens_seen": 11933344,
      "step": 18280
    },
    {
      "epoch": 9.583333333333334,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.01867248052927139,
      "loss": 0.3308,
      "num_input_tokens_seen": 11936992,
      "step": 18285
    },
    {
      "epoch": 9.585953878406709,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.018665828372414966,
      "loss": 0.4319,
      "num_input_tokens_seen": 11940288,
      "step": 18290
    },
    {
      "epoch": 9.588574423480084,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.018659175448709073,
      "loss": 0.378,
      "num_input_tokens_seen": 11943072,
      "step": 18295
    },
    {
      "epoch": 9.59119496855346,
      "grad_norm": 0.04296875,
      "learning_rate": 0.018652521759545425,
      "loss": 0.3259,
      "num_input_tokens_seen": 11946016,
      "step": 18300
    },
    {
      "epoch": 9.593815513626835,
      "grad_norm": 0.0546875,
      "learning_rate": 0.01864586730631589,
      "loss": 0.4736,
      "num_input_tokens_seen": 11949280,
      "step": 18305
    },
    {
      "epoch": 9.59643605870021,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.018639212090412513,
      "loss": 0.3917,
      "num_input_tokens_seen": 11953216,
      "step": 18310
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.018632556113227484,
      "loss": 0.3938,
      "num_input_tokens_seen": 11956320,
      "step": 18315
    },
    {
      "epoch": 9.60167714884696,
      "grad_norm": 0.0390625,
      "learning_rate": 0.018625899376153155,
      "loss": 0.4261,
      "num_input_tokens_seen": 11959104,
      "step": 18320
    },
    {
      "epoch": 9.604297693920335,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01861924188058205,
      "loss": 0.3862,
      "num_input_tokens_seen": 11962208,
      "step": 18325
    },
    {
      "epoch": 9.60691823899371,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.018612583627906826,
      "loss": 0.5472,
      "num_input_tokens_seen": 11964832,
      "step": 18330
    },
    {
      "epoch": 9.609538784067086,
      "grad_norm": 0.041015625,
      "learning_rate": 0.018605924619520327,
      "loss": 0.6184,
      "num_input_tokens_seen": 11968000,
      "step": 18335
    },
    {
      "epoch": 9.61215932914046,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01859926485681554,
      "loss": 0.5418,
      "num_input_tokens_seen": 11970912,
      "step": 18340
    },
    {
      "epoch": 9.614779874213836,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0185926043411856,
      "loss": 0.3008,
      "num_input_tokens_seen": 11973472,
      "step": 18345
    },
    {
      "epoch": 9.617400419287211,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.01858594307402382,
      "loss": 0.4498,
      "num_input_tokens_seen": 11976864,
      "step": 18350
    },
    {
      "epoch": 9.620020964360586,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.01857928105672366,
      "loss": 0.5074,
      "num_input_tokens_seen": 11979872,
      "step": 18355
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.018572618290678734,
      "loss": 0.4095,
      "num_input_tokens_seen": 11983168,
      "step": 18360
    },
    {
      "epoch": 9.625262054507338,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.018565954777282824,
      "loss": 0.4411,
      "num_input_tokens_seen": 11986432,
      "step": 18365
    },
    {
      "epoch": 9.627882599580714,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.01855929051792986,
      "loss": 0.2554,
      "num_input_tokens_seen": 11988512,
      "step": 18370
    },
    {
      "epoch": 9.630503144654089,
      "grad_norm": 0.052734375,
      "learning_rate": 0.01855262551401392,
      "loss": 0.6317,
      "num_input_tokens_seen": 11991424,
      "step": 18375
    },
    {
      "epoch": 9.633123689727464,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.01854595976692926,
      "loss": 0.4029,
      "num_input_tokens_seen": 11994208,
      "step": 18380
    },
    {
      "epoch": 9.635744234800839,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.018539293278070267,
      "loss": 0.3581,
      "num_input_tokens_seen": 11997312,
      "step": 18385
    },
    {
      "epoch": 9.638364779874214,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.018532626048831498,
      "loss": 0.5465,
      "num_input_tokens_seen": 12000288,
      "step": 18390
    },
    {
      "epoch": 9.64098532494759,
      "grad_norm": 0.05859375,
      "learning_rate": 0.018525958080607665,
      "loss": 0.527,
      "num_input_tokens_seen": 12003328,
      "step": 18395
    },
    {
      "epoch": 9.643605870020965,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.018519289374793624,
      "loss": 0.4433,
      "num_input_tokens_seen": 12006176,
      "step": 18400
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0185126199327844,
      "loss": 0.5273,
      "num_input_tokens_seen": 12009856,
      "step": 18405
    },
    {
      "epoch": 9.648846960167715,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.018505949755975158,
      "loss": 0.3348,
      "num_input_tokens_seen": 12013376,
      "step": 18410
    },
    {
      "epoch": 9.65146750524109,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.018499278845761222,
      "loss": 0.5809,
      "num_input_tokens_seen": 12016192,
      "step": 18415
    },
    {
      "epoch": 9.654088050314465,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.018492607203538076,
      "loss": 0.528,
      "num_input_tokens_seen": 12019680,
      "step": 18420
    },
    {
      "epoch": 9.65670859538784,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.01848593483070135,
      "loss": 0.3787,
      "num_input_tokens_seen": 12023200,
      "step": 18425
    },
    {
      "epoch": 9.659329140461216,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.018479261728646827,
      "loss": 0.5141,
      "num_input_tokens_seen": 12026880,
      "step": 18430
    },
    {
      "epoch": 9.66194968553459,
      "grad_norm": 0.048828125,
      "learning_rate": 0.018472587898770442,
      "loss": 0.3973,
      "num_input_tokens_seen": 12029472,
      "step": 18435
    },
    {
      "epoch": 9.664570230607966,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.018465913342468287,
      "loss": 0.4517,
      "num_input_tokens_seen": 12032128,
      "step": 18440
    },
    {
      "epoch": 9.667190775681341,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.018459238061136602,
      "loss": 0.3538,
      "num_input_tokens_seen": 12034912,
      "step": 18445
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.018452562056171778,
      "loss": 0.3715,
      "num_input_tokens_seen": 12037888,
      "step": 18450
    },
    {
      "epoch": 9.672431865828091,
      "grad_norm": 0.068359375,
      "learning_rate": 0.018445885328970364,
      "loss": 0.369,
      "num_input_tokens_seen": 12040576,
      "step": 18455
    },
    {
      "epoch": 9.675052410901468,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01843920788092905,
      "loss": 0.5203,
      "num_input_tokens_seen": 12044096,
      "step": 18460
    },
    {
      "epoch": 9.677672955974844,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.01843252971344468,
      "loss": 0.3455,
      "num_input_tokens_seen": 12047424,
      "step": 18465
    },
    {
      "epoch": 9.680293501048219,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.018425850827914265,
      "loss": 0.3318,
      "num_input_tokens_seen": 12050560,
      "step": 18470
    },
    {
      "epoch": 9.682914046121594,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.018419171225734932,
      "loss": 0.3904,
      "num_input_tokens_seen": 12053952,
      "step": 18475
    },
    {
      "epoch": 9.685534591194969,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.01841249090830399,
      "loss": 0.3788,
      "num_input_tokens_seen": 12056416,
      "step": 18480
    },
    {
      "epoch": 9.688155136268344,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.018405809877018883,
      "loss": 0.4545,
      "num_input_tokens_seen": 12060448,
      "step": 18485
    },
    {
      "epoch": 9.69077568134172,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0183991281332772,
      "loss": 0.4832,
      "num_input_tokens_seen": 12064960,
      "step": 18490
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.018392445678476695,
      "loss": 0.3224,
      "num_input_tokens_seen": 12068192,
      "step": 18495
    },
    {
      "epoch": 9.69601677148847,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.018385762514015254,
      "loss": 0.3622,
      "num_input_tokens_seen": 12071200,
      "step": 18500
    },
    {
      "epoch": 9.698637316561845,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.018379078641290923,
      "loss": 0.4242,
      "num_input_tokens_seen": 12074112,
      "step": 18505
    },
    {
      "epoch": 9.70125786163522,
      "grad_norm": 0.138671875,
      "learning_rate": 0.018372394061701897,
      "loss": 0.4871,
      "num_input_tokens_seen": 12078048,
      "step": 18510
    },
    {
      "epoch": 9.703878406708595,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0183657087766465,
      "loss": 0.5057,
      "num_input_tokens_seen": 12080896,
      "step": 18515
    },
    {
      "epoch": 9.70649895178197,
      "grad_norm": 0.041015625,
      "learning_rate": 0.018359022787523226,
      "loss": 0.3426,
      "num_input_tokens_seen": 12083744,
      "step": 18520
    },
    {
      "epoch": 9.709119496855346,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01835233609573071,
      "loss": 0.4057,
      "num_input_tokens_seen": 12086496,
      "step": 18525
    },
    {
      "epoch": 9.71174004192872,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.018345648702667722,
      "loss": 0.368,
      "num_input_tokens_seen": 12089728,
      "step": 18530
    },
    {
      "epoch": 9.714360587002096,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.018338960609733203,
      "loss": 0.5389,
      "num_input_tokens_seen": 12093120,
      "step": 18535
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.018332271818326213,
      "loss": 0.508,
      "num_input_tokens_seen": 12095680,
      "step": 18540
    },
    {
      "epoch": 9.719601677148846,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.018325582329845973,
      "loss": 0.4343,
      "num_input_tokens_seen": 12099040,
      "step": 18545
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.01831889214569185,
      "loss": 0.3451,
      "num_input_tokens_seen": 12102112,
      "step": 18550
    },
    {
      "epoch": 9.724842767295598,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01831220126726336,
      "loss": 0.384,
      "num_input_tokens_seen": 12104544,
      "step": 18555
    },
    {
      "epoch": 9.727463312368974,
      "grad_norm": 0.0546875,
      "learning_rate": 0.018305509695960145,
      "loss": 0.3987,
      "num_input_tokens_seen": 12107712,
      "step": 18560
    },
    {
      "epoch": 9.730083857442349,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01829881743318202,
      "loss": 0.3679,
      "num_input_tokens_seen": 12110528,
      "step": 18565
    },
    {
      "epoch": 9.732704402515724,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01829212448032892,
      "loss": 0.3924,
      "num_input_tokens_seen": 12113600,
      "step": 18570
    },
    {
      "epoch": 9.735324947589099,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.018285430838800937,
      "loss": 0.4467,
      "num_input_tokens_seen": 12116896,
      "step": 18575
    },
    {
      "epoch": 9.737945492662474,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.018278736509998304,
      "loss": 0.3484,
      "num_input_tokens_seen": 12120832,
      "step": 18580
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 0.021484375,
      "learning_rate": 0.018272041495321394,
      "loss": 0.2995,
      "num_input_tokens_seen": 12124704,
      "step": 18585
    },
    {
      "epoch": 9.743186582809225,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.01826534579617074,
      "loss": 0.3623,
      "num_input_tokens_seen": 12129760,
      "step": 18590
    },
    {
      "epoch": 9.7458071278826,
      "grad_norm": 0.091796875,
      "learning_rate": 0.018258649413947,
      "loss": 0.4032,
      "num_input_tokens_seen": 12132096,
      "step": 18595
    },
    {
      "epoch": 9.748427672955975,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.018251952350050973,
      "loss": 0.5332,
      "num_input_tokens_seen": 12135008,
      "step": 18600
    },
    {
      "epoch": 9.75104821802935,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.018245254605883613,
      "loss": 0.4538,
      "num_input_tokens_seen": 12138144,
      "step": 18605
    },
    {
      "epoch": 9.753668763102725,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.018238556182846023,
      "loss": 0.5252,
      "num_input_tokens_seen": 12141568,
      "step": 18610
    },
    {
      "epoch": 9.7562893081761,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01823185708233942,
      "loss": 0.5218,
      "num_input_tokens_seen": 12145248,
      "step": 18615
    },
    {
      "epoch": 9.758909853249476,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.018225157305765183,
      "loss": 0.3881,
      "num_input_tokens_seen": 12149472,
      "step": 18620
    },
    {
      "epoch": 9.76153039832285,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.018218456854524837,
      "loss": 0.5155,
      "num_input_tokens_seen": 12152992,
      "step": 18625
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 0.0625,
      "learning_rate": 0.018211755730020037,
      "loss": 0.3919,
      "num_input_tokens_seen": 12155712,
      "step": 18630
    },
    {
      "epoch": 9.766771488469601,
      "grad_norm": 0.052734375,
      "learning_rate": 0.018205053933652578,
      "loss": 0.4989,
      "num_input_tokens_seen": 12158688,
      "step": 18635
    },
    {
      "epoch": 9.769392033542976,
      "grad_norm": 0.064453125,
      "learning_rate": 0.018198351466824404,
      "loss": 0.7349,
      "num_input_tokens_seen": 12161824,
      "step": 18640
    },
    {
      "epoch": 9.772012578616351,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.018191648330937588,
      "loss": 0.36,
      "num_input_tokens_seen": 12164992,
      "step": 18645
    },
    {
      "epoch": 9.774633123689728,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01818494452739435,
      "loss": 0.4009,
      "num_input_tokens_seen": 12167808,
      "step": 18650
    },
    {
      "epoch": 9.777253668763104,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.018178240057597064,
      "loss": 0.2816,
      "num_input_tokens_seen": 12170752,
      "step": 18655
    },
    {
      "epoch": 9.779874213836479,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.018171534922948207,
      "loss": 0.3388,
      "num_input_tokens_seen": 12173664,
      "step": 18660
    },
    {
      "epoch": 9.782494758909854,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01816482912485043,
      "loss": 0.4053,
      "num_input_tokens_seen": 12176960,
      "step": 18665
    },
    {
      "epoch": 9.785115303983229,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.018158122664706507,
      "loss": 0.5849,
      "num_input_tokens_seen": 12180096,
      "step": 18670
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 0.068359375,
      "learning_rate": 0.01815141554391935,
      "loss": 0.3687,
      "num_input_tokens_seen": 12183424,
      "step": 18675
    },
    {
      "epoch": 9.79035639412998,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.01814470776389202,
      "loss": 0.3517,
      "num_input_tokens_seen": 12186592,
      "step": 18680
    },
    {
      "epoch": 9.792976939203355,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.018137999326027696,
      "loss": 0.4976,
      "num_input_tokens_seen": 12190240,
      "step": 18685
    },
    {
      "epoch": 9.79559748427673,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.018131290231729713,
      "loss": 0.4348,
      "num_input_tokens_seen": 12193024,
      "step": 18690
    },
    {
      "epoch": 9.798218029350105,
      "grad_norm": 0.03515625,
      "learning_rate": 0.018124580482401544,
      "loss": 0.4495,
      "num_input_tokens_seen": 12196128,
      "step": 18695
    },
    {
      "epoch": 9.80083857442348,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01811787007944678,
      "loss": 0.4017,
      "num_input_tokens_seen": 12198880,
      "step": 18700
    },
    {
      "epoch": 9.803459119496855,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.018111159024269168,
      "loss": 0.3963,
      "num_input_tokens_seen": 12201760,
      "step": 18705
    },
    {
      "epoch": 9.80607966457023,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.018104447318272585,
      "loss": 0.637,
      "num_input_tokens_seen": 12206272,
      "step": 18710
    },
    {
      "epoch": 9.808700209643606,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01809773496286103,
      "loss": 0.4152,
      "num_input_tokens_seen": 12209472,
      "step": 18715
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.018091021959438673,
      "loss": 0.4528,
      "num_input_tokens_seen": 12212256,
      "step": 18720
    },
    {
      "epoch": 9.813941299790356,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.018084308309409783,
      "loss": 0.4689,
      "num_input_tokens_seen": 12215456,
      "step": 18725
    },
    {
      "epoch": 9.816561844863731,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.018077594014178775,
      "loss": 0.4245,
      "num_input_tokens_seen": 12219936,
      "step": 18730
    },
    {
      "epoch": 9.819182389937106,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.018070879075150217,
      "loss": 0.3983,
      "num_input_tokens_seen": 12223200,
      "step": 18735
    },
    {
      "epoch": 9.821802935010481,
      "grad_norm": 0.060546875,
      "learning_rate": 0.018064163493728795,
      "loss": 0.3988,
      "num_input_tokens_seen": 12226432,
      "step": 18740
    },
    {
      "epoch": 9.824423480083858,
      "grad_norm": 0.033203125,
      "learning_rate": 0.018057447271319322,
      "loss": 0.3938,
      "num_input_tokens_seen": 12229120,
      "step": 18745
    },
    {
      "epoch": 9.827044025157234,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.01805073040932676,
      "loss": 0.3638,
      "num_input_tokens_seen": 12233248,
      "step": 18750
    },
    {
      "epoch": 9.829664570230609,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.018044012909156203,
      "loss": 0.5093,
      "num_input_tokens_seen": 12237696,
      "step": 18755
    },
    {
      "epoch": 9.832285115303984,
      "grad_norm": 0.046875,
      "learning_rate": 0.018037294772212872,
      "loss": 0.2901,
      "num_input_tokens_seen": 12241504,
      "step": 18760
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.01803057599990213,
      "loss": 0.4258,
      "num_input_tokens_seen": 12244832,
      "step": 18765
    },
    {
      "epoch": 9.837526205450734,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.01802385659362946,
      "loss": 0.399,
      "num_input_tokens_seen": 12247776,
      "step": 18770
    },
    {
      "epoch": 9.84014675052411,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.01801713655480049,
      "loss": 0.4019,
      "num_input_tokens_seen": 12250560,
      "step": 18775
    },
    {
      "epoch": 9.842767295597485,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.018010415884820975,
      "loss": 0.4765,
      "num_input_tokens_seen": 12253504,
      "step": 18780
    },
    {
      "epoch": 9.84538784067086,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0180036945850968,
      "loss": 0.4108,
      "num_input_tokens_seen": 12256128,
      "step": 18785
    },
    {
      "epoch": 9.848008385744235,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017996972657033983,
      "loss": 0.5013,
      "num_input_tokens_seen": 12260032,
      "step": 18790
    },
    {
      "epoch": 9.85062893081761,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.017990250102038682,
      "loss": 0.3663,
      "num_input_tokens_seen": 12263136,
      "step": 18795
    },
    {
      "epoch": 9.853249475890985,
      "grad_norm": 0.12890625,
      "learning_rate": 0.01798352692151717,
      "loss": 0.4654,
      "num_input_tokens_seen": 12266496,
      "step": 18800
    },
    {
      "epoch": 9.85587002096436,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.017976803116875866,
      "loss": 0.3658,
      "num_input_tokens_seen": 12268768,
      "step": 18805
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017970078689521315,
      "loss": 0.4587,
      "num_input_tokens_seen": 12272160,
      "step": 18810
    },
    {
      "epoch": 9.86111111111111,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.017963353640860182,
      "loss": 0.5118,
      "num_input_tokens_seen": 12275776,
      "step": 18815
    },
    {
      "epoch": 9.863731656184486,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.017956627972299275,
      "loss": 0.399,
      "num_input_tokens_seen": 12278880,
      "step": 18820
    },
    {
      "epoch": 9.866352201257861,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01794990168524553,
      "loss": 0.4019,
      "num_input_tokens_seen": 12281536,
      "step": 18825
    },
    {
      "epoch": 9.868972746331236,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.017943174781106002,
      "loss": 0.5187,
      "num_input_tokens_seen": 12284416,
      "step": 18830
    },
    {
      "epoch": 9.871593291404611,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.017936447261287895,
      "loss": 0.3983,
      "num_input_tokens_seen": 12288224,
      "step": 18835
    },
    {
      "epoch": 9.874213836477988,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.017929719127198516,
      "loss": 0.3695,
      "num_input_tokens_seen": 12291520,
      "step": 18840
    },
    {
      "epoch": 9.876834381551364,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.017922990380245318,
      "loss": 0.5103,
      "num_input_tokens_seen": 12295264,
      "step": 18845
    },
    {
      "epoch": 9.879454926624739,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017916261021835883,
      "loss": 0.4362,
      "num_input_tokens_seen": 12297984,
      "step": 18850
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 0.076171875,
      "learning_rate": 0.017909531053377916,
      "loss": 0.4198,
      "num_input_tokens_seen": 12301184,
      "step": 18855
    },
    {
      "epoch": 9.884696016771489,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.017902800476279238,
      "loss": 0.4666,
      "num_input_tokens_seen": 12303968,
      "step": 18860
    },
    {
      "epoch": 9.887316561844864,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.017896069291947828,
      "loss": 0.3577,
      "num_input_tokens_seen": 12308256,
      "step": 18865
    },
    {
      "epoch": 9.88993710691824,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.017889337501791752,
      "loss": 0.5159,
      "num_input_tokens_seen": 12310816,
      "step": 18870
    },
    {
      "epoch": 9.892557651991615,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.017882605107219236,
      "loss": 0.4405,
      "num_input_tokens_seen": 12314752,
      "step": 18875
    },
    {
      "epoch": 9.89517819706499,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.017875872109638622,
      "loss": 0.3925,
      "num_input_tokens_seen": 12318272,
      "step": 18880
    },
    {
      "epoch": 9.897798742138365,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01786913851045837,
      "loss": 0.4769,
      "num_input_tokens_seen": 12322336,
      "step": 18885
    },
    {
      "epoch": 9.90041928721174,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.017862404311087075,
      "loss": 0.4095,
      "num_input_tokens_seen": 12325760,
      "step": 18890
    },
    {
      "epoch": 9.903039832285115,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.017855669512933454,
      "loss": 0.5497,
      "num_input_tokens_seen": 12330304,
      "step": 18895
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.01784893411740635,
      "loss": 0.4698,
      "num_input_tokens_seen": 12332960,
      "step": 18900
    },
    {
      "epoch": 9.908280922431866,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.017842198125914727,
      "loss": 0.4913,
      "num_input_tokens_seen": 12337536,
      "step": 18905
    },
    {
      "epoch": 9.91090146750524,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.01783546153986769,
      "loss": 0.4703,
      "num_input_tokens_seen": 12341120,
      "step": 18910
    },
    {
      "epoch": 9.913522012578616,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.017828724360674442,
      "loss": 0.3979,
      "num_input_tokens_seen": 12343904,
      "step": 18915
    },
    {
      "epoch": 9.916142557651991,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.017821986589744328,
      "loss": 0.5367,
      "num_input_tokens_seen": 12346912,
      "step": 18920
    },
    {
      "epoch": 9.918763102725366,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.01781524822848682,
      "loss": 0.3461,
      "num_input_tokens_seen": 12350272,
      "step": 18925
    },
    {
      "epoch": 9.921383647798741,
      "grad_norm": 0.2109375,
      "learning_rate": 0.017808509278311496,
      "loss": 0.503,
      "num_input_tokens_seen": 12352800,
      "step": 18930
    },
    {
      "epoch": 9.924004192872118,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01780176974062807,
      "loss": 0.3647,
      "num_input_tokens_seen": 12355616,
      "step": 18935
    },
    {
      "epoch": 9.926624737945493,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.017795029616846383,
      "loss": 0.4045,
      "num_input_tokens_seen": 12358784,
      "step": 18940
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.017788288908376387,
      "loss": 0.4777,
      "num_input_tokens_seen": 12361920,
      "step": 18945
    },
    {
      "epoch": 9.931865828092244,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.017781547616628163,
      "loss": 0.4316,
      "num_input_tokens_seen": 12365600,
      "step": 18950
    },
    {
      "epoch": 9.934486373165619,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.01777480574301191,
      "loss": 0.462,
      "num_input_tokens_seen": 12368000,
      "step": 18955
    },
    {
      "epoch": 9.937106918238994,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01776806328893795,
      "loss": 0.4373,
      "num_input_tokens_seen": 12370592,
      "step": 18960
    },
    {
      "epoch": 9.93972746331237,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.017761320255816733,
      "loss": 0.4649,
      "num_input_tokens_seen": 12373760,
      "step": 18965
    },
    {
      "epoch": 9.942348008385745,
      "grad_norm": 0.048828125,
      "learning_rate": 0.017754576645058823,
      "loss": 0.4445,
      "num_input_tokens_seen": 12377600,
      "step": 18970
    },
    {
      "epoch": 9.94496855345912,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0177478324580749,
      "loss": 0.3724,
      "num_input_tokens_seen": 12381152,
      "step": 18975
    },
    {
      "epoch": 9.947589098532495,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.01774108769627578,
      "loss": 0.4539,
      "num_input_tokens_seen": 12384096,
      "step": 18980
    },
    {
      "epoch": 9.95020964360587,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.01773434236107238,
      "loss": 0.4218,
      "num_input_tokens_seen": 12386848,
      "step": 18985
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.017727596453875755,
      "loss": 0.3901,
      "num_input_tokens_seen": 12389344,
      "step": 18990
    },
    {
      "epoch": 9.95545073375262,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.017720849976097075,
      "loss": 0.4478,
      "num_input_tokens_seen": 12392640,
      "step": 18995
    },
    {
      "epoch": 9.958071278825996,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.017714102929147615,
      "loss": 0.4476,
      "num_input_tokens_seen": 12397248,
      "step": 19000
    },
    {
      "epoch": 9.96069182389937,
      "grad_norm": 0.046875,
      "learning_rate": 0.017707355314438785,
      "loss": 0.4626,
      "num_input_tokens_seen": 12400352,
      "step": 19005
    },
    {
      "epoch": 9.963312368972746,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.01770060713338212,
      "loss": 0.3722,
      "num_input_tokens_seen": 12402944,
      "step": 19010
    },
    {
      "epoch": 9.965932914046121,
      "grad_norm": 0.0859375,
      "learning_rate": 0.017693858387389245,
      "loss": 0.415,
      "num_input_tokens_seen": 12405952,
      "step": 19015
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01768710907787193,
      "loss": 0.5732,
      "num_input_tokens_seen": 12409376,
      "step": 19020
    },
    {
      "epoch": 9.971174004192871,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.017680359206242054,
      "loss": 0.4684,
      "num_input_tokens_seen": 12413216,
      "step": 19025
    },
    {
      "epoch": 9.973794549266248,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.017673608773911605,
      "loss": 0.3657,
      "num_input_tokens_seen": 12416512,
      "step": 19030
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.017666857782292714,
      "loss": 0.4285,
      "num_input_tokens_seen": 12420192,
      "step": 19035
    },
    {
      "epoch": 9.979035639412999,
      "grad_norm": 0.208984375,
      "learning_rate": 0.017660106232797592,
      "loss": 0.49,
      "num_input_tokens_seen": 12422464,
      "step": 19040
    },
    {
      "epoch": 9.981656184486374,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.017653354126838595,
      "loss": 0.558,
      "num_input_tokens_seen": 12425248,
      "step": 19045
    },
    {
      "epoch": 9.984276729559749,
      "grad_norm": 0.0625,
      "learning_rate": 0.017646601465828186,
      "loss": 0.4034,
      "num_input_tokens_seen": 12428704,
      "step": 19050
    },
    {
      "epoch": 9.986897274633124,
      "grad_norm": 0.056640625,
      "learning_rate": 0.017639848251178945,
      "loss": 0.5155,
      "num_input_tokens_seen": 12432480,
      "step": 19055
    },
    {
      "epoch": 9.9895178197065,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.017633094484303568,
      "loss": 0.3751,
      "num_input_tokens_seen": 12434912,
      "step": 19060
    },
    {
      "epoch": 9.992138364779874,
      "grad_norm": 0.064453125,
      "learning_rate": 0.01762634016661486,
      "loss": 0.3658,
      "num_input_tokens_seen": 12438016,
      "step": 19065
    },
    {
      "epoch": 9.99475890985325,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.017619585299525755,
      "loss": 0.503,
      "num_input_tokens_seen": 12440672,
      "step": 19070
    },
    {
      "epoch": 9.997379454926625,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01761282988444929,
      "loss": 0.4292,
      "num_input_tokens_seen": 12443840,
      "step": 19075
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.01760607392279863,
      "loss": 0.4562,
      "num_input_tokens_seen": 12446408,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.4582311809062958,
      "eval_runtime": 13.7072,
      "eval_samples_per_second": 61.865,
      "eval_steps_per_second": 15.466,
      "num_input_tokens_seen": 12446408,
      "step": 19080
    },
    {
      "epoch": 10.002620545073375,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.017599317415987026,
      "loss": 0.3874,
      "num_input_tokens_seen": 12450472,
      "step": 19085
    },
    {
      "epoch": 10.00524109014675,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.01759256036542788,
      "loss": 0.3527,
      "num_input_tokens_seen": 12454056,
      "step": 19090
    },
    {
      "epoch": 10.007861635220126,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.017585802772534683,
      "loss": 0.4312,
      "num_input_tokens_seen": 12457576,
      "step": 19095
    },
    {
      "epoch": 10.0104821802935,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.01757904463872104,
      "loss": 0.4381,
      "num_input_tokens_seen": 12461608,
      "step": 19100
    },
    {
      "epoch": 10.013102725366876,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017572285965400686,
      "loss": 0.3374,
      "num_input_tokens_seen": 12464968,
      "step": 19105
    },
    {
      "epoch": 10.015723270440251,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.017565526753987454,
      "loss": 0.4872,
      "num_input_tokens_seen": 12468648,
      "step": 19110
    },
    {
      "epoch": 10.018343815513626,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.017558767005895298,
      "loss": 0.4407,
      "num_input_tokens_seen": 12471784,
      "step": 19115
    },
    {
      "epoch": 10.020964360587001,
      "grad_norm": 0.0546875,
      "learning_rate": 0.017552006722538267,
      "loss": 0.5552,
      "num_input_tokens_seen": 12474632,
      "step": 19120
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.017545245905330555,
      "loss": 0.5089,
      "num_input_tokens_seen": 12477832,
      "step": 19125
    },
    {
      "epoch": 10.026205450733753,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01753848455568643,
      "loss": 0.4391,
      "num_input_tokens_seen": 12482312,
      "step": 19130
    },
    {
      "epoch": 10.028825995807129,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0175317226750203,
      "loss": 0.3686,
      "num_input_tokens_seen": 12485768,
      "step": 19135
    },
    {
      "epoch": 10.031446540880504,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.01752496026474667,
      "loss": 0.3303,
      "num_input_tokens_seen": 12488776,
      "step": 19140
    },
    {
      "epoch": 10.034067085953879,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.017518197326280155,
      "loss": 0.4015,
      "num_input_tokens_seen": 12491464,
      "step": 19145
    },
    {
      "epoch": 10.036687631027254,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017511433861035493,
      "loss": 0.4339,
      "num_input_tokens_seen": 12494600,
      "step": 19150
    },
    {
      "epoch": 10.03930817610063,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.01750466987042752,
      "loss": 0.5996,
      "num_input_tokens_seen": 12498696,
      "step": 19155
    },
    {
      "epoch": 10.041928721174004,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.017497905355871177,
      "loss": 0.4204,
      "num_input_tokens_seen": 12501960,
      "step": 19160
    },
    {
      "epoch": 10.04454926624738,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.017491140318781538,
      "loss": 0.5631,
      "num_input_tokens_seen": 12505224,
      "step": 19165
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.017484374760573764,
      "loss": 0.463,
      "num_input_tokens_seen": 12508744,
      "step": 19170
    },
    {
      "epoch": 10.04979035639413,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.01747760868266313,
      "loss": 0.5046,
      "num_input_tokens_seen": 12511464,
      "step": 19175
    },
    {
      "epoch": 10.052410901467505,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.017470842086465033,
      "loss": 0.4356,
      "num_input_tokens_seen": 12513736,
      "step": 19180
    },
    {
      "epoch": 10.05503144654088,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.017464074973394954,
      "loss": 0.431,
      "num_input_tokens_seen": 12516584,
      "step": 19185
    },
    {
      "epoch": 10.057651991614255,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.017457307344868504,
      "loss": 0.3343,
      "num_input_tokens_seen": 12520680,
      "step": 19190
    },
    {
      "epoch": 10.06027253668763,
      "grad_norm": 0.0703125,
      "learning_rate": 0.017450539202301392,
      "loss": 0.4393,
      "num_input_tokens_seen": 12523624,
      "step": 19195
    },
    {
      "epoch": 10.062893081761006,
      "grad_norm": 0.0625,
      "learning_rate": 0.01744377054710943,
      "loss": 0.3636,
      "num_input_tokens_seen": 12526472,
      "step": 19200
    },
    {
      "epoch": 10.065513626834381,
      "grad_norm": 0.05078125,
      "learning_rate": 0.01743700138070856,
      "loss": 0.4398,
      "num_input_tokens_seen": 12529768,
      "step": 19205
    },
    {
      "epoch": 10.068134171907756,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.017430231704514798,
      "loss": 0.4258,
      "num_input_tokens_seen": 12532680,
      "step": 19210
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.017423461519944287,
      "loss": 0.3881,
      "num_input_tokens_seen": 12535944,
      "step": 19215
    },
    {
      "epoch": 10.073375262054507,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.017416690828413284,
      "loss": 0.5123,
      "num_input_tokens_seen": 12540008,
      "step": 19220
    },
    {
      "epoch": 10.075995807127883,
      "grad_norm": 0.080078125,
      "learning_rate": 0.01740991963133812,
      "loss": 0.2954,
      "num_input_tokens_seen": 12542952,
      "step": 19225
    },
    {
      "epoch": 10.078616352201259,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.017403147930135268,
      "loss": 0.3776,
      "num_input_tokens_seen": 12545672,
      "step": 19230
    },
    {
      "epoch": 10.081236897274634,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.017396375726221287,
      "loss": 0.3832,
      "num_input_tokens_seen": 12548808,
      "step": 19235
    },
    {
      "epoch": 10.083857442348009,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.01738960302101284,
      "loss": 0.4244,
      "num_input_tokens_seen": 12551944,
      "step": 19240
    },
    {
      "epoch": 10.086477987421384,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.017382829815926704,
      "loss": 0.4481,
      "num_input_tokens_seen": 12555048,
      "step": 19245
    },
    {
      "epoch": 10.08909853249476,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.017376056112379756,
      "loss": 0.3496,
      "num_input_tokens_seen": 12558312,
      "step": 19250
    },
    {
      "epoch": 10.091719077568134,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.017369281911788974,
      "loss": 0.4277,
      "num_input_tokens_seen": 12561992,
      "step": 19255
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01736250721557145,
      "loss": 0.4395,
      "num_input_tokens_seen": 12565544,
      "step": 19260
    },
    {
      "epoch": 10.096960167714885,
      "grad_norm": 0.064453125,
      "learning_rate": 0.017355732025144368,
      "loss": 0.4817,
      "num_input_tokens_seen": 12568232,
      "step": 19265
    },
    {
      "epoch": 10.09958071278826,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.01734895634192502,
      "loss": 0.3868,
      "num_input_tokens_seen": 12570568,
      "step": 19270
    },
    {
      "epoch": 10.102201257861635,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.017342180167330804,
      "loss": 0.3924,
      "num_input_tokens_seen": 12573448,
      "step": 19275
    },
    {
      "epoch": 10.10482180293501,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.01733540350277922,
      "loss": 0.3702,
      "num_input_tokens_seen": 12577256,
      "step": 19280
    },
    {
      "epoch": 10.107442348008385,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01732862634968787,
      "loss": 0.3941,
      "num_input_tokens_seen": 12580808,
      "step": 19285
    },
    {
      "epoch": 10.11006289308176,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.017321848709474445,
      "loss": 0.4263,
      "num_input_tokens_seen": 12585896,
      "step": 19290
    },
    {
      "epoch": 10.112683438155136,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.017315070583556767,
      "loss": 0.3719,
      "num_input_tokens_seen": 12589000,
      "step": 19295
    },
    {
      "epoch": 10.115303983228511,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01730829197335273,
      "loss": 0.4017,
      "num_input_tokens_seen": 12592104,
      "step": 19300
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 0.02734375,
      "learning_rate": 0.017301512880280347,
      "loss": 0.3405,
      "num_input_tokens_seen": 12595240,
      "step": 19305
    },
    {
      "epoch": 10.120545073375261,
      "grad_norm": 0.076171875,
      "learning_rate": 0.01729473330575773,
      "loss": 0.2958,
      "num_input_tokens_seen": 12597480,
      "step": 19310
    },
    {
      "epoch": 10.123165618448636,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.017287953251203088,
      "loss": 0.5025,
      "num_input_tokens_seen": 12600328,
      "step": 19315
    },
    {
      "epoch": 10.125786163522013,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.017281172718034723,
      "loss": 0.3851,
      "num_input_tokens_seen": 12602152,
      "step": 19320
    },
    {
      "epoch": 10.128406708595389,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.01727439170767105,
      "loss": 0.4663,
      "num_input_tokens_seen": 12605352,
      "step": 19325
    },
    {
      "epoch": 10.131027253668764,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.017267610221530578,
      "loss": 0.3337,
      "num_input_tokens_seen": 12607784,
      "step": 19330
    },
    {
      "epoch": 10.133647798742139,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.01726082826103192,
      "loss": 0.561,
      "num_input_tokens_seen": 12614056,
      "step": 19335
    },
    {
      "epoch": 10.136268343815514,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.017254045827593784,
      "loss": 0.4515,
      "num_input_tokens_seen": 12616552,
      "step": 19340
    },
    {
      "epoch": 10.13888888888889,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.01724726292263497,
      "loss": 0.4414,
      "num_input_tokens_seen": 12619976,
      "step": 19345
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.017240479547574402,
      "loss": 0.4033,
      "num_input_tokens_seen": 12623272,
      "step": 19350
    },
    {
      "epoch": 10.14412997903564,
      "grad_norm": 0.08203125,
      "learning_rate": 0.017233695703831067,
      "loss": 0.457,
      "num_input_tokens_seen": 12625960,
      "step": 19355
    },
    {
      "epoch": 10.146750524109015,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01722691139282407,
      "loss": 0.3899,
      "num_input_tokens_seen": 12630184,
      "step": 19360
    },
    {
      "epoch": 10.14937106918239,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.017220126615972626,
      "loss": 0.411,
      "num_input_tokens_seen": 12633224,
      "step": 19365
    },
    {
      "epoch": 10.151991614255765,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.017213341374696014,
      "loss": 0.4197,
      "num_input_tokens_seen": 12636584,
      "step": 19370
    },
    {
      "epoch": 10.15461215932914,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.01720655567041364,
      "loss": 0.4679,
      "num_input_tokens_seen": 12639944,
      "step": 19375
    },
    {
      "epoch": 10.157232704402515,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.017199769504544993,
      "loss": 0.4534,
      "num_input_tokens_seen": 12643624,
      "step": 19380
    },
    {
      "epoch": 10.15985324947589,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.017192982878509663,
      "loss": 0.2867,
      "num_input_tokens_seen": 12646792,
      "step": 19385
    },
    {
      "epoch": 10.162473794549266,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.01718619579372733,
      "loss": 0.4862,
      "num_input_tokens_seen": 12650696,
      "step": 19390
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 0.044921875,
      "learning_rate": 0.017179408251617786,
      "loss": 0.5251,
      "num_input_tokens_seen": 12653512,
      "step": 19395
    },
    {
      "epoch": 10.167714884696016,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01717262025360089,
      "loss": 0.3865,
      "num_input_tokens_seen": 12657160,
      "step": 19400
    },
    {
      "epoch": 10.170335429769391,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.017165831801096636,
      "loss": 0.4043,
      "num_input_tokens_seen": 12661096,
      "step": 19405
    },
    {
      "epoch": 10.172955974842766,
      "grad_norm": 0.03125,
      "learning_rate": 0.017159042895525067,
      "loss": 0.3629,
      "num_input_tokens_seen": 12664168,
      "step": 19410
    },
    {
      "epoch": 10.175576519916143,
      "grad_norm": 0.056640625,
      "learning_rate": 0.01715225353830636,
      "loss": 0.3812,
      "num_input_tokens_seen": 12667304,
      "step": 19415
    },
    {
      "epoch": 10.178197064989519,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.017145463730860767,
      "loss": 0.3548,
      "num_input_tokens_seen": 12670760,
      "step": 19420
    },
    {
      "epoch": 10.180817610062894,
      "grad_norm": 0.0625,
      "learning_rate": 0.017138673474608637,
      "loss": 0.4701,
      "num_input_tokens_seen": 12673384,
      "step": 19425
    },
    {
      "epoch": 10.183438155136269,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.017131882770970417,
      "loss": 0.4289,
      "num_input_tokens_seen": 12676168,
      "step": 19430
    },
    {
      "epoch": 10.186058700209644,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.017125091621366644,
      "loss": 0.3587,
      "num_input_tokens_seen": 12678984,
      "step": 19435
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.017118300027217946,
      "loss": 0.5617,
      "num_input_tokens_seen": 12682600,
      "step": 19440
    },
    {
      "epoch": 10.191299790356394,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.017111507989945048,
      "loss": 0.3316,
      "num_input_tokens_seen": 12685704,
      "step": 19445
    },
    {
      "epoch": 10.19392033542977,
      "grad_norm": 0.046875,
      "learning_rate": 0.01710471551096877,
      "loss": 0.3433,
      "num_input_tokens_seen": 12689256,
      "step": 19450
    },
    {
      "epoch": 10.196540880503145,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.017097922591710023,
      "loss": 0.3944,
      "num_input_tokens_seen": 12692968,
      "step": 19455
    },
    {
      "epoch": 10.19916142557652,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.017091129233589797,
      "loss": 0.3717,
      "num_input_tokens_seen": 12696712,
      "step": 19460
    },
    {
      "epoch": 10.201781970649895,
      "grad_norm": 0.091796875,
      "learning_rate": 0.017084335438029195,
      "loss": 0.4654,
      "num_input_tokens_seen": 12699848,
      "step": 19465
    },
    {
      "epoch": 10.20440251572327,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0170775412064494,
      "loss": 0.4023,
      "num_input_tokens_seen": 12703176,
      "step": 19470
    },
    {
      "epoch": 10.207023060796645,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017070746540271683,
      "loss": 0.5097,
      "num_input_tokens_seen": 12706280,
      "step": 19475
    },
    {
      "epoch": 10.20964360587002,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.017063951440917418,
      "loss": 0.5163,
      "num_input_tokens_seen": 12709224,
      "step": 19480
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017057155909808056,
      "loss": 0.3693,
      "num_input_tokens_seen": 12713064,
      "step": 19485
    },
    {
      "epoch": 10.214884696016771,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.017050359948365148,
      "loss": 0.441,
      "num_input_tokens_seen": 12716424,
      "step": 19490
    },
    {
      "epoch": 10.217505241090146,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.017043563558010334,
      "loss": 0.4177,
      "num_input_tokens_seen": 12719816,
      "step": 19495
    },
    {
      "epoch": 10.220125786163521,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.01703676674016533,
      "loss": 0.5023,
      "num_input_tokens_seen": 12723624,
      "step": 19500
    },
    {
      "epoch": 10.222746331236896,
      "grad_norm": 0.068359375,
      "learning_rate": 0.01702996949625197,
      "loss": 0.419,
      "num_input_tokens_seen": 12727240,
      "step": 19505
    },
    {
      "epoch": 10.225366876310272,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.017023171827692144,
      "loss": 0.3932,
      "num_input_tokens_seen": 12730792,
      "step": 19510
    },
    {
      "epoch": 10.227987421383649,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.017016373735907853,
      "loss": 0.4639,
      "num_input_tokens_seen": 12733928,
      "step": 19515
    },
    {
      "epoch": 10.230607966457024,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01700957522232118,
      "loss": 0.3555,
      "num_input_tokens_seen": 12737992,
      "step": 19520
    },
    {
      "epoch": 10.233228511530399,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0170027762883543,
      "loss": 0.4301,
      "num_input_tokens_seen": 12740456,
      "step": 19525
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.016995976935429466,
      "loss": 0.4689,
      "num_input_tokens_seen": 12743496,
      "step": 19530
    },
    {
      "epoch": 10.23846960167715,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.016989177164969036,
      "loss": 0.3689,
      "num_input_tokens_seen": 12746952,
      "step": 19535
    },
    {
      "epoch": 10.241090146750524,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.016982376978395428,
      "loss": 0.5196,
      "num_input_tokens_seen": 12749800,
      "step": 19540
    },
    {
      "epoch": 10.2437106918239,
      "grad_norm": 0.06640625,
      "learning_rate": 0.016975576377131176,
      "loss": 0.4168,
      "num_input_tokens_seen": 12752872,
      "step": 19545
    },
    {
      "epoch": 10.246331236897275,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.016968775362598887,
      "loss": 0.4721,
      "num_input_tokens_seen": 12755944,
      "step": 19550
    },
    {
      "epoch": 10.24895178197065,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.016961973936221248,
      "loss": 0.437,
      "num_input_tokens_seen": 12758344,
      "step": 19555
    },
    {
      "epoch": 10.251572327044025,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.01695517209942105,
      "loss": 0.3022,
      "num_input_tokens_seen": 12762312,
      "step": 19560
    },
    {
      "epoch": 10.2541928721174,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.01694836985362115,
      "loss": 0.4097,
      "num_input_tokens_seen": 12765416,
      "step": 19565
    },
    {
      "epoch": 10.256813417190775,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.016941567200244503,
      "loss": 0.3211,
      "num_input_tokens_seen": 12768552,
      "step": 19570
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.016934764140714154,
      "loss": 0.3697,
      "num_input_tokens_seen": 12770920,
      "step": 19575
    },
    {
      "epoch": 10.262054507337526,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01692796067645322,
      "loss": 0.4016,
      "num_input_tokens_seen": 12774184,
      "step": 19580
    },
    {
      "epoch": 10.264675052410901,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0169211568088849,
      "loss": 0.4526,
      "num_input_tokens_seen": 12777288,
      "step": 19585
    },
    {
      "epoch": 10.267295597484276,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.016914352539432506,
      "loss": 0.3947,
      "num_input_tokens_seen": 12780136,
      "step": 19590
    },
    {
      "epoch": 10.269916142557651,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.016907547869519392,
      "loss": 0.4591,
      "num_input_tokens_seen": 12783144,
      "step": 19595
    },
    {
      "epoch": 10.272536687631026,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.016900742800569028,
      "loss": 0.316,
      "num_input_tokens_seen": 12785992,
      "step": 19600
    },
    {
      "epoch": 10.275157232704402,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.016893937334004953,
      "loss": 0.4966,
      "num_input_tokens_seen": 12789000,
      "step": 19605
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0168871314712508,
      "loss": 0.4484,
      "num_input_tokens_seen": 12791976,
      "step": 19610
    },
    {
      "epoch": 10.280398322851154,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.016880325213730268,
      "loss": 0.4451,
      "num_input_tokens_seen": 12795976,
      "step": 19615
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 0.0390625,
      "learning_rate": 0.016873518562867153,
      "loss": 0.4669,
      "num_input_tokens_seen": 12798984,
      "step": 19620
    },
    {
      "epoch": 10.285639412997904,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.016866711520085332,
      "loss": 0.4239,
      "num_input_tokens_seen": 12801864,
      "step": 19625
    },
    {
      "epoch": 10.28825995807128,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.016859904086808757,
      "loss": 0.4465,
      "num_input_tokens_seen": 12804968,
      "step": 19630
    },
    {
      "epoch": 10.290880503144654,
      "grad_norm": 0.076171875,
      "learning_rate": 0.016853096264461465,
      "loss": 0.4139,
      "num_input_tokens_seen": 12807848,
      "step": 19635
    },
    {
      "epoch": 10.29350104821803,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.016846288054467574,
      "loss": 0.3665,
      "num_input_tokens_seen": 12810632,
      "step": 19640
    },
    {
      "epoch": 10.296121593291405,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.016839479458251285,
      "loss": 0.4182,
      "num_input_tokens_seen": 12814024,
      "step": 19645
    },
    {
      "epoch": 10.29874213836478,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.016832670477236883,
      "loss": 0.3719,
      "num_input_tokens_seen": 12817800,
      "step": 19650
    },
    {
      "epoch": 10.301362683438155,
      "grad_norm": 0.078125,
      "learning_rate": 0.016825861112848724,
      "loss": 0.4176,
      "num_input_tokens_seen": 12820456,
      "step": 19655
    },
    {
      "epoch": 10.30398322851153,
      "grad_norm": 0.068359375,
      "learning_rate": 0.016819051366511246,
      "loss": 0.4808,
      "num_input_tokens_seen": 12822792,
      "step": 19660
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.016812241239648983,
      "loss": 0.5737,
      "num_input_tokens_seen": 12826888,
      "step": 19665
    },
    {
      "epoch": 10.30922431865828,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.016805430733686524,
      "loss": 0.4266,
      "num_input_tokens_seen": 12829896,
      "step": 19670
    },
    {
      "epoch": 10.311844863731656,
      "grad_norm": 0.0546875,
      "learning_rate": 0.01679861985004855,
      "loss": 0.4957,
      "num_input_tokens_seen": 12833480,
      "step": 19675
    },
    {
      "epoch": 10.314465408805031,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.01679180859015983,
      "loss": 0.4284,
      "num_input_tokens_seen": 12836424,
      "step": 19680
    },
    {
      "epoch": 10.317085953878406,
      "grad_norm": 0.078125,
      "learning_rate": 0.01678499695544519,
      "loss": 0.5511,
      "num_input_tokens_seen": 12839464,
      "step": 19685
    },
    {
      "epoch": 10.319706498951781,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.016778184947329553,
      "loss": 0.3257,
      "num_input_tokens_seen": 12841896,
      "step": 19690
    },
    {
      "epoch": 10.322327044025156,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.016771372567237916,
      "loss": 0.4037,
      "num_input_tokens_seen": 12845768,
      "step": 19695
    },
    {
      "epoch": 10.324947589098532,
      "grad_norm": 0.0546875,
      "learning_rate": 0.01676455981659534,
      "loss": 0.3876,
      "num_input_tokens_seen": 12848680,
      "step": 19700
    },
    {
      "epoch": 10.327568134171909,
      "grad_norm": 0.091796875,
      "learning_rate": 0.016757746696826982,
      "loss": 0.4944,
      "num_input_tokens_seen": 12851464,
      "step": 19705
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 0.056640625,
      "learning_rate": 0.016750933209358068,
      "loss": 0.3032,
      "num_input_tokens_seen": 12854504,
      "step": 19710
    },
    {
      "epoch": 10.332809224318659,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.016744119355613905,
      "loss": 0.3246,
      "num_input_tokens_seen": 12857736,
      "step": 19715
    },
    {
      "epoch": 10.335429769392034,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.016737305137019872,
      "loss": 0.4128,
      "num_input_tokens_seen": 12860360,
      "step": 19720
    },
    {
      "epoch": 10.33805031446541,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.01673049055500142,
      "loss": 0.4037,
      "num_input_tokens_seen": 12863016,
      "step": 19725
    },
    {
      "epoch": 10.340670859538784,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.016723675610984094,
      "loss": 0.5075,
      "num_input_tokens_seen": 12865768,
      "step": 19730
    },
    {
      "epoch": 10.34329140461216,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.016716860306393487,
      "loss": 0.435,
      "num_input_tokens_seen": 12869256,
      "step": 19735
    },
    {
      "epoch": 10.345911949685535,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.016710044642655294,
      "loss": 0.4858,
      "num_input_tokens_seen": 12872456,
      "step": 19740
    },
    {
      "epoch": 10.34853249475891,
      "grad_norm": 0.076171875,
      "learning_rate": 0.01670322862119527,
      "loss": 0.3451,
      "num_input_tokens_seen": 12875720,
      "step": 19745
    },
    {
      "epoch": 10.351153039832285,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.016696412243439256,
      "loss": 0.3435,
      "num_input_tokens_seen": 12878088,
      "step": 19750
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.016689595510813143,
      "loss": 0.3789,
      "num_input_tokens_seen": 12880616,
      "step": 19755
    },
    {
      "epoch": 10.356394129979035,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01668277842474293,
      "loss": 0.5623,
      "num_input_tokens_seen": 12883144,
      "step": 19760
    },
    {
      "epoch": 10.35901467505241,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.016675960986654675,
      "loss": 0.3523,
      "num_input_tokens_seen": 12885832,
      "step": 19765
    },
    {
      "epoch": 10.361635220125786,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.016669143197974495,
      "loss": 0.4276,
      "num_input_tokens_seen": 12889000,
      "step": 19770
    },
    {
      "epoch": 10.364255765199161,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.016662325060128602,
      "loss": 0.4014,
      "num_input_tokens_seen": 12892616,
      "step": 19775
    },
    {
      "epoch": 10.366876310272536,
      "grad_norm": 0.060546875,
      "learning_rate": 0.01665550657454327,
      "loss": 0.4494,
      "num_input_tokens_seen": 12895112,
      "step": 19780
    },
    {
      "epoch": 10.369496855345911,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.016648687742644854,
      "loss": 0.4828,
      "num_input_tokens_seen": 12899176,
      "step": 19785
    },
    {
      "epoch": 10.372117400419286,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.01664186856585977,
      "loss": 0.4742,
      "num_input_tokens_seen": 12902184,
      "step": 19790
    },
    {
      "epoch": 10.374737945492662,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.016635049045614508,
      "loss": 0.3823,
      "num_input_tokens_seen": 12904712,
      "step": 19795
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 0.0546875,
      "learning_rate": 0.016628229183335647,
      "loss": 0.4774,
      "num_input_tokens_seen": 12907560,
      "step": 19800
    },
    {
      "epoch": 10.379979035639414,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.016621408980449812,
      "loss": 0.4604,
      "num_input_tokens_seen": 12910568,
      "step": 19805
    },
    {
      "epoch": 10.382599580712789,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.016614588438383718,
      "loss": 0.6223,
      "num_input_tokens_seen": 12913608,
      "step": 19810
    },
    {
      "epoch": 10.385220125786164,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.01660776755856414,
      "loss": 0.3589,
      "num_input_tokens_seen": 12917288,
      "step": 19815
    },
    {
      "epoch": 10.38784067085954,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.016600946342417936,
      "loss": 0.3974,
      "num_input_tokens_seen": 12920232,
      "step": 19820
    },
    {
      "epoch": 10.390461215932914,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.01659412479137202,
      "loss": 0.4507,
      "num_input_tokens_seen": 12923432,
      "step": 19825
    },
    {
      "epoch": 10.39308176100629,
      "grad_norm": 0.033203125,
      "learning_rate": 0.016587302906853382,
      "loss": 0.4168,
      "num_input_tokens_seen": 12927912,
      "step": 19830
    },
    {
      "epoch": 10.395702306079665,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.016580480690289084,
      "loss": 0.3918,
      "num_input_tokens_seen": 12931304,
      "step": 19835
    },
    {
      "epoch": 10.39832285115304,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.01657365814310626,
      "loss": 0.4255,
      "num_input_tokens_seen": 12934248,
      "step": 19840
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.016566835266732104,
      "loss": 0.4755,
      "num_input_tokens_seen": 12937608,
      "step": 19845
    },
    {
      "epoch": 10.40356394129979,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.01656001206259389,
      "loss": 0.3455,
      "num_input_tokens_seen": 12940360,
      "step": 19850
    },
    {
      "epoch": 10.406184486373165,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.016553188532118946,
      "loss": 0.3204,
      "num_input_tokens_seen": 12943048,
      "step": 19855
    },
    {
      "epoch": 10.40880503144654,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.016546364676734676,
      "loss": 0.2758,
      "num_input_tokens_seen": 12945704,
      "step": 19860
    },
    {
      "epoch": 10.411425576519916,
      "grad_norm": 0.0546875,
      "learning_rate": 0.016539540497868568,
      "loss": 0.3871,
      "num_input_tokens_seen": 12948616,
      "step": 19865
    },
    {
      "epoch": 10.414046121593291,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.016532715996948145,
      "loss": 0.2981,
      "num_input_tokens_seen": 12951880,
      "step": 19870
    },
    {
      "epoch": 10.416666666666666,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01652589117540102,
      "loss": 0.4796,
      "num_input_tokens_seen": 12954984,
      "step": 19875
    },
    {
      "epoch": 10.419287211740041,
      "grad_norm": 0.068359375,
      "learning_rate": 0.016519066034654875,
      "loss": 0.3724,
      "num_input_tokens_seen": 12957640,
      "step": 19880
    },
    {
      "epoch": 10.421907756813416,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.01651224057613744,
      "loss": 0.3182,
      "num_input_tokens_seen": 12960488,
      "step": 19885
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.016505414801276538,
      "loss": 0.5097,
      "num_input_tokens_seen": 12963560,
      "step": 19890
    },
    {
      "epoch": 10.427148846960169,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.016498588711500027,
      "loss": 0.3832,
      "num_input_tokens_seen": 12967208,
      "step": 19895
    },
    {
      "epoch": 10.429769392033544,
      "grad_norm": 0.044921875,
      "learning_rate": 0.016491762308235854,
      "loss": 0.3172,
      "num_input_tokens_seen": 12970248,
      "step": 19900
    },
    {
      "epoch": 10.432389937106919,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.01648493559291203,
      "loss": 0.4027,
      "num_input_tokens_seen": 12973480,
      "step": 19905
    },
    {
      "epoch": 10.435010482180294,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01647810856695662,
      "loss": 0.3834,
      "num_input_tokens_seen": 12976904,
      "step": 19910
    },
    {
      "epoch": 10.43763102725367,
      "grad_norm": 0.09765625,
      "learning_rate": 0.016471281231797755,
      "loss": 0.4801,
      "num_input_tokens_seen": 12979912,
      "step": 19915
    },
    {
      "epoch": 10.440251572327044,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.016464453588863646,
      "loss": 0.3914,
      "num_input_tokens_seen": 12982728,
      "step": 19920
    },
    {
      "epoch": 10.44287211740042,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.016457625639582554,
      "loss": 0.3208,
      "num_input_tokens_seen": 12986472,
      "step": 19925
    },
    {
      "epoch": 10.445492662473795,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.016450797385382805,
      "loss": 0.3763,
      "num_input_tokens_seen": 12990600,
      "step": 19930
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0164439688276928,
      "loss": 0.5161,
      "num_input_tokens_seen": 12994216,
      "step": 19935
    },
    {
      "epoch": 10.450733752620545,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01643713996794098,
      "loss": 0.4331,
      "num_input_tokens_seen": 12997256,
      "step": 19940
    },
    {
      "epoch": 10.45335429769392,
      "grad_norm": 0.046875,
      "learning_rate": 0.016430310807555883,
      "loss": 0.4008,
      "num_input_tokens_seen": 13001384,
      "step": 19945
    },
    {
      "epoch": 10.455974842767295,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.016423481347966083,
      "loss": 0.5704,
      "num_input_tokens_seen": 13004232,
      "step": 19950
    },
    {
      "epoch": 10.45859538784067,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.016416651590600222,
      "loss": 0.3918,
      "num_input_tokens_seen": 13007208,
      "step": 19955
    },
    {
      "epoch": 10.461215932914046,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.016409821536887012,
      "loss": 0.4356,
      "num_input_tokens_seen": 13010696,
      "step": 19960
    },
    {
      "epoch": 10.463836477987421,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.01640299118825522,
      "loss": 0.4601,
      "num_input_tokens_seen": 13013416,
      "step": 19965
    },
    {
      "epoch": 10.466457023060796,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.016396160546133676,
      "loss": 0.4028,
      "num_input_tokens_seen": 13016904,
      "step": 19970
    },
    {
      "epoch": 10.469077568134171,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.016389329611951275,
      "loss": 0.4274,
      "num_input_tokens_seen": 13019496,
      "step": 19975
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.016382498387136967,
      "loss": 0.4741,
      "num_input_tokens_seen": 13022184,
      "step": 19980
    },
    {
      "epoch": 10.474318658280922,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.016375666873119775,
      "loss": 0.4899,
      "num_input_tokens_seen": 13024968,
      "step": 19985
    },
    {
      "epoch": 10.476939203354299,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.016368835071328764,
      "loss": 0.3859,
      "num_input_tokens_seen": 13028328,
      "step": 19990
    },
    {
      "epoch": 10.479559748427674,
      "grad_norm": 0.076171875,
      "learning_rate": 0.016362002983193073,
      "loss": 0.412,
      "num_input_tokens_seen": 13032040,
      "step": 19995
    },
    {
      "epoch": 10.482180293501049,
      "grad_norm": 0.072265625,
      "learning_rate": 0.016355170610141894,
      "loss": 0.4012,
      "num_input_tokens_seen": 13035944,
      "step": 20000
    },
    {
      "epoch": 10.484800838574424,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.016348337953604485,
      "loss": 0.5056,
      "num_input_tokens_seen": 13042856,
      "step": 20005
    },
    {
      "epoch": 10.4874213836478,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.016341505015010162,
      "loss": 0.3825,
      "num_input_tokens_seen": 13046792,
      "step": 20010
    },
    {
      "epoch": 10.490041928721174,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.01633467179578829,
      "loss": 0.3841,
      "num_input_tokens_seen": 13050920,
      "step": 20015
    },
    {
      "epoch": 10.49266247379455,
      "grad_norm": 0.04296875,
      "learning_rate": 0.016327838297368308,
      "loss": 0.3971,
      "num_input_tokens_seen": 13058664,
      "step": 20020
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 0.03125,
      "learning_rate": 0.016321004521179704,
      "loss": 0.4625,
      "num_input_tokens_seen": 13062952,
      "step": 20025
    },
    {
      "epoch": 10.4979035639413,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01631417046865202,
      "loss": 0.2556,
      "num_input_tokens_seen": 13065768,
      "step": 20030
    },
    {
      "epoch": 10.500524109014675,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016307336141214875,
      "loss": 0.4724,
      "num_input_tokens_seen": 13069160,
      "step": 20035
    },
    {
      "epoch": 10.50314465408805,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.016300501540297916,
      "loss": 0.3617,
      "num_input_tokens_seen": 13072168,
      "step": 20040
    },
    {
      "epoch": 10.505765199161425,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.016293666667330876,
      "loss": 0.3452,
      "num_input_tokens_seen": 13074888,
      "step": 20045
    },
    {
      "epoch": 10.5083857442348,
      "grad_norm": 0.064453125,
      "learning_rate": 0.01628683152374353,
      "loss": 0.6467,
      "num_input_tokens_seen": 13078280,
      "step": 20050
    },
    {
      "epoch": 10.511006289308176,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.016279996110965703,
      "loss": 0.4747,
      "num_input_tokens_seen": 13081544,
      "step": 20055
    },
    {
      "epoch": 10.51362683438155,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0162731604304273,
      "loss": 0.4566,
      "num_input_tokens_seen": 13084360,
      "step": 20060
    },
    {
      "epoch": 10.516247379454926,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.016266324483558255,
      "loss": 0.5186,
      "num_input_tokens_seen": 13087464,
      "step": 20065
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.016259488271788577,
      "loss": 0.3656,
      "num_input_tokens_seen": 13091112,
      "step": 20070
    },
    {
      "epoch": 10.521488469601676,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.016252651796548324,
      "loss": 0.4759,
      "num_input_tokens_seen": 13094824,
      "step": 20075
    },
    {
      "epoch": 10.524109014675052,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0162458150592676,
      "loss": 0.3636,
      "num_input_tokens_seen": 13097768,
      "step": 20080
    },
    {
      "epoch": 10.526729559748428,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.016238978061376582,
      "loss": 0.4757,
      "num_input_tokens_seen": 13100712,
      "step": 20085
    },
    {
      "epoch": 10.529350104821804,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.016232140804305493,
      "loss": 0.328,
      "num_input_tokens_seen": 13103464,
      "step": 20090
    },
    {
      "epoch": 10.531970649895179,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.016225303289484596,
      "loss": 0.3721,
      "num_input_tokens_seen": 13106312,
      "step": 20095
    },
    {
      "epoch": 10.534591194968554,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.016218465518344233,
      "loss": 0.4076,
      "num_input_tokens_seen": 13108936,
      "step": 20100
    },
    {
      "epoch": 10.53721174004193,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01621162749231479,
      "loss": 0.3555,
      "num_input_tokens_seen": 13113544,
      "step": 20105
    },
    {
      "epoch": 10.539832285115304,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.01620478921282669,
      "loss": 0.5217,
      "num_input_tokens_seen": 13116648,
      "step": 20110
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016197950681310438,
      "loss": 0.2626,
      "num_input_tokens_seen": 13119816,
      "step": 20115
    },
    {
      "epoch": 10.545073375262055,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.016191111899196566,
      "loss": 0.6747,
      "num_input_tokens_seen": 13123656,
      "step": 20120
    },
    {
      "epoch": 10.54769392033543,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01618427286791568,
      "loss": 0.3396,
      "num_input_tokens_seen": 13126120,
      "step": 20125
    },
    {
      "epoch": 10.550314465408805,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.016177433588898414,
      "loss": 0.512,
      "num_input_tokens_seen": 13128968,
      "step": 20130
    },
    {
      "epoch": 10.55293501048218,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.016170594063575478,
      "loss": 0.3264,
      "num_input_tokens_seen": 13132168,
      "step": 20135
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 0.056640625,
      "learning_rate": 0.01616375429337762,
      "loss": 0.4701,
      "num_input_tokens_seen": 13135560,
      "step": 20140
    },
    {
      "epoch": 10.55817610062893,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01615691427973564,
      "loss": 0.4438,
      "num_input_tokens_seen": 13139400,
      "step": 20145
    },
    {
      "epoch": 10.560796645702306,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016150074024080384,
      "loss": 0.4594,
      "num_input_tokens_seen": 13144232,
      "step": 20150
    },
    {
      "epoch": 10.56341719077568,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.01614323352784277,
      "loss": 0.3225,
      "num_input_tokens_seen": 13147432,
      "step": 20155
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.016136392792453744,
      "loss": 0.3164,
      "num_input_tokens_seen": 13150632,
      "step": 20160
    },
    {
      "epoch": 10.568658280922431,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01612955181934431,
      "loss": 0.391,
      "num_input_tokens_seen": 13154152,
      "step": 20165
    },
    {
      "epoch": 10.571278825995806,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.016122710609945523,
      "loss": 0.3071,
      "num_input_tokens_seen": 13160552,
      "step": 20170
    },
    {
      "epoch": 10.573899371069182,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.01611586916568848,
      "loss": 0.4032,
      "num_input_tokens_seen": 13163208,
      "step": 20175
    },
    {
      "epoch": 10.576519916142558,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.016109027488004347,
      "loss": 0.467,
      "num_input_tokens_seen": 13166056,
      "step": 20180
    },
    {
      "epoch": 10.579140461215934,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016102185578324307,
      "loss": 0.3587,
      "num_input_tokens_seen": 13168712,
      "step": 20185
    },
    {
      "epoch": 10.581761006289309,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.016095343438079623,
      "loss": 0.4296,
      "num_input_tokens_seen": 13171400,
      "step": 20190
    },
    {
      "epoch": 10.584381551362684,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01608850106870159,
      "loss": 0.4185,
      "num_input_tokens_seen": 13174888,
      "step": 20195
    },
    {
      "epoch": 10.58700209643606,
      "grad_norm": 0.068359375,
      "learning_rate": 0.016081658471621546,
      "loss": 0.4023,
      "num_input_tokens_seen": 13178920,
      "step": 20200
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.016074815648270895,
      "loss": 0.4417,
      "num_input_tokens_seen": 13183176,
      "step": 20205
    },
    {
      "epoch": 10.59224318658281,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.016067972600081068,
      "loss": 0.35,
      "num_input_tokens_seen": 13186760,
      "step": 20210
    },
    {
      "epoch": 10.594863731656185,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.016061129328483557,
      "loss": 0.3587,
      "num_input_tokens_seen": 13189544,
      "step": 20215
    },
    {
      "epoch": 10.59748427672956,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.016054285834909903,
      "loss": 0.4355,
      "num_input_tokens_seen": 13193096,
      "step": 20220
    },
    {
      "epoch": 10.600104821802935,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.016047442120791677,
      "loss": 0.3318,
      "num_input_tokens_seen": 13195944,
      "step": 20225
    },
    {
      "epoch": 10.60272536687631,
      "grad_norm": 0.048828125,
      "learning_rate": 0.016040598187560512,
      "loss": 0.3595,
      "num_input_tokens_seen": 13198760,
      "step": 20230
    },
    {
      "epoch": 10.605345911949685,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.016033754036648074,
      "loss": 0.3945,
      "num_input_tokens_seen": 13201704,
      "step": 20235
    },
    {
      "epoch": 10.60796645702306,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.01602690966948608,
      "loss": 0.4502,
      "num_input_tokens_seen": 13204872,
      "step": 20240
    },
    {
      "epoch": 10.610587002096436,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01602006508750631,
      "loss": 0.3724,
      "num_input_tokens_seen": 13208200,
      "step": 20245
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 0.04296875,
      "learning_rate": 0.016013220292140553,
      "loss": 0.3637,
      "num_input_tokens_seen": 13211848,
      "step": 20250
    },
    {
      "epoch": 10.615828092243186,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01600637528482067,
      "loss": 0.3873,
      "num_input_tokens_seen": 13214760,
      "step": 20255
    },
    {
      "epoch": 10.618448637316561,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.015999530066978566,
      "loss": 0.3792,
      "num_input_tokens_seen": 13219016,
      "step": 20260
    },
    {
      "epoch": 10.621069182389936,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.01599268464004617,
      "loss": 0.3353,
      "num_input_tokens_seen": 13221800,
      "step": 20265
    },
    {
      "epoch": 10.623689727463312,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.015985839005455475,
      "loss": 0.3928,
      "num_input_tokens_seen": 13224936,
      "step": 20270
    },
    {
      "epoch": 10.626310272536688,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.015978993164638505,
      "loss": 0.4016,
      "num_input_tokens_seen": 13227560,
      "step": 20275
    },
    {
      "epoch": 10.628930817610064,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.015972147119027332,
      "loss": 0.375,
      "num_input_tokens_seen": 13230888,
      "step": 20280
    },
    {
      "epoch": 10.631551362683439,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.015965300870054076,
      "loss": 0.4742,
      "num_input_tokens_seen": 13234152,
      "step": 20285
    },
    {
      "epoch": 10.634171907756814,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.01595845441915089,
      "loss": 0.4288,
      "num_input_tokens_seen": 13237352,
      "step": 20290
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 0.064453125,
      "learning_rate": 0.01595160776774997,
      "loss": 0.4367,
      "num_input_tokens_seen": 13240264,
      "step": 20295
    },
    {
      "epoch": 10.639412997903564,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.01594476091728357,
      "loss": 0.3962,
      "num_input_tokens_seen": 13243528,
      "step": 20300
    },
    {
      "epoch": 10.64203354297694,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.01593791386918396,
      "loss": 0.4883,
      "num_input_tokens_seen": 13248712,
      "step": 20305
    },
    {
      "epoch": 10.644654088050315,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.015931066624883472,
      "loss": 0.4706,
      "num_input_tokens_seen": 13252712,
      "step": 20310
    },
    {
      "epoch": 10.64727463312369,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.01592421918581446,
      "loss": 0.421,
      "num_input_tokens_seen": 13256392,
      "step": 20315
    },
    {
      "epoch": 10.649895178197065,
      "grad_norm": 0.0546875,
      "learning_rate": 0.015917371553409346,
      "loss": 0.4719,
      "num_input_tokens_seen": 13259240,
      "step": 20320
    },
    {
      "epoch": 10.65251572327044,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.015910523729100567,
      "loss": 0.4347,
      "num_input_tokens_seen": 13263176,
      "step": 20325
    },
    {
      "epoch": 10.655136268343815,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.01590367571432061,
      "loss": 0.4828,
      "num_input_tokens_seen": 13266696,
      "step": 20330
    },
    {
      "epoch": 10.65775681341719,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.015896827510502004,
      "loss": 0.3527,
      "num_input_tokens_seen": 13269512,
      "step": 20335
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 0.046875,
      "learning_rate": 0.015889979119077313,
      "loss": 0.3067,
      "num_input_tokens_seen": 13272072,
      "step": 20340
    },
    {
      "epoch": 10.66299790356394,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01588313054147915,
      "loss": 0.4318,
      "num_input_tokens_seen": 13275848,
      "step": 20345
    },
    {
      "epoch": 10.665618448637316,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.015876281779140147,
      "loss": 0.4939,
      "num_input_tokens_seen": 13278440,
      "step": 20350
    },
    {
      "epoch": 10.668238993710691,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01586943283349299,
      "loss": 0.3761,
      "num_input_tokens_seen": 13281576,
      "step": 20355
    },
    {
      "epoch": 10.670859538784066,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.01586258370597041,
      "loss": 0.4155,
      "num_input_tokens_seen": 13284168,
      "step": 20360
    },
    {
      "epoch": 10.673480083857442,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.015855734398005156,
      "loss": 0.4457,
      "num_input_tokens_seen": 13286760,
      "step": 20365
    },
    {
      "epoch": 10.676100628930818,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.01584888491103003,
      "loss": 0.4648,
      "num_input_tokens_seen": 13289672,
      "step": 20370
    },
    {
      "epoch": 10.678721174004194,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.015842035246477863,
      "loss": 0.5318,
      "num_input_tokens_seen": 13292616,
      "step": 20375
    },
    {
      "epoch": 10.681341719077569,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.015835185405781535,
      "loss": 0.472,
      "num_input_tokens_seen": 13296424,
      "step": 20380
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.01582833539037394,
      "loss": 0.3507,
      "num_input_tokens_seen": 13299048,
      "step": 20385
    },
    {
      "epoch": 10.68658280922432,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01582148520168804,
      "loss": 0.405,
      "num_input_tokens_seen": 13302152,
      "step": 20390
    },
    {
      "epoch": 10.689203354297694,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0158146348411568,
      "loss": 0.3137,
      "num_input_tokens_seen": 13304840,
      "step": 20395
    },
    {
      "epoch": 10.69182389937107,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01580778431021325,
      "loss": 0.3421,
      "num_input_tokens_seen": 13307976,
      "step": 20400
    },
    {
      "epoch": 10.694444444444445,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01580093361029044,
      "loss": 0.4093,
      "num_input_tokens_seen": 13310856,
      "step": 20405
    },
    {
      "epoch": 10.69706498951782,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01579408274282145,
      "loss": 0.3956,
      "num_input_tokens_seen": 13313448,
      "step": 20410
    },
    {
      "epoch": 10.699685534591195,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.015787231709239413,
      "loss": 0.4856,
      "num_input_tokens_seen": 13316776,
      "step": 20415
    },
    {
      "epoch": 10.70230607966457,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.015780380510977483,
      "loss": 0.4923,
      "num_input_tokens_seen": 13319560,
      "step": 20420
    },
    {
      "epoch": 10.704926624737945,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.01577352914946885,
      "loss": 0.4261,
      "num_input_tokens_seen": 13322440,
      "step": 20425
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 0.046875,
      "learning_rate": 0.015766677626146752,
      "loss": 0.3845,
      "num_input_tokens_seen": 13325448,
      "step": 20430
    },
    {
      "epoch": 10.710167714884696,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.015759825942444433,
      "loss": 0.3401,
      "num_input_tokens_seen": 13328520,
      "step": 20435
    },
    {
      "epoch": 10.71278825995807,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.015752974099795203,
      "loss": 0.3254,
      "num_input_tokens_seen": 13331560,
      "step": 20440
    },
    {
      "epoch": 10.715408805031446,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.015746122099632377,
      "loss": 0.4213,
      "num_input_tokens_seen": 13334216,
      "step": 20445
    },
    {
      "epoch": 10.718029350104821,
      "grad_norm": 0.087890625,
      "learning_rate": 0.015739269943389315,
      "loss": 0.4687,
      "num_input_tokens_seen": 13337768,
      "step": 20450
    },
    {
      "epoch": 10.720649895178196,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.015732417632499423,
      "loss": 0.4258,
      "num_input_tokens_seen": 13341064,
      "step": 20455
    },
    {
      "epoch": 10.723270440251572,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.015725565168396117,
      "loss": 0.3323,
      "num_input_tokens_seen": 13343688,
      "step": 20460
    },
    {
      "epoch": 10.725890985324948,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01571871255251285,
      "loss": 0.3937,
      "num_input_tokens_seen": 13347048,
      "step": 20465
    },
    {
      "epoch": 10.728511530398324,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.015711859786283124,
      "loss": 0.4195,
      "num_input_tokens_seen": 13350952,
      "step": 20470
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 0.08203125,
      "learning_rate": 0.01570500687114045,
      "loss": 0.448,
      "num_input_tokens_seen": 13353672,
      "step": 20475
    },
    {
      "epoch": 10.733752620545074,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.01569815380851838,
      "loss": 0.5,
      "num_input_tokens_seen": 13357544,
      "step": 20480
    },
    {
      "epoch": 10.73637316561845,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.015691300599850494,
      "loss": 0.3725,
      "num_input_tokens_seen": 13361544,
      "step": 20485
    },
    {
      "epoch": 10.738993710691824,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.015684447246570418,
      "loss": 0.4974,
      "num_input_tokens_seen": 13365480,
      "step": 20490
    },
    {
      "epoch": 10.7416142557652,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.015677593750111782,
      "loss": 0.3429,
      "num_input_tokens_seen": 13368072,
      "step": 20495
    },
    {
      "epoch": 10.744234800838575,
      "grad_norm": 0.037109375,
      "learning_rate": 0.015670740111908263,
      "loss": 0.4182,
      "num_input_tokens_seen": 13370920,
      "step": 20500
    },
    {
      "epoch": 10.74685534591195,
      "grad_norm": 0.0703125,
      "learning_rate": 0.015663886333393565,
      "loss": 0.6431,
      "num_input_tokens_seen": 13373832,
      "step": 20505
    },
    {
      "epoch": 10.749475890985325,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.01565703241600142,
      "loss": 0.3735,
      "num_input_tokens_seen": 13376712,
      "step": 20510
    },
    {
      "epoch": 10.7520964360587,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.015650178361165588,
      "loss": 0.4407,
      "num_input_tokens_seen": 13379496,
      "step": 20515
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.01564332417031987,
      "loss": 0.3531,
      "num_input_tokens_seen": 13383688,
      "step": 20520
    },
    {
      "epoch": 10.75733752620545,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.015636469844898065,
      "loss": 0.3778,
      "num_input_tokens_seen": 13387112,
      "step": 20525
    },
    {
      "epoch": 10.759958071278826,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.015629615386334034,
      "loss": 0.4315,
      "num_input_tokens_seen": 13391048,
      "step": 20530
    },
    {
      "epoch": 10.7625786163522,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.015622760796061647,
      "loss": 0.4102,
      "num_input_tokens_seen": 13393448,
      "step": 20535
    },
    {
      "epoch": 10.765199161425576,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.015615906075514802,
      "loss": 0.5108,
      "num_input_tokens_seen": 13396040,
      "step": 20540
    },
    {
      "epoch": 10.767819706498951,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.015609051226127435,
      "loss": 0.4586,
      "num_input_tokens_seen": 13398824,
      "step": 20545
    },
    {
      "epoch": 10.770440251572326,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.015602196249333504,
      "loss": 0.4178,
      "num_input_tokens_seen": 13403368,
      "step": 20550
    },
    {
      "epoch": 10.773060796645701,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.01559534114656698,
      "loss": 0.4574,
      "num_input_tokens_seen": 13406376,
      "step": 20555
    },
    {
      "epoch": 10.775681341719078,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01558848591926189,
      "loss": 0.5198,
      "num_input_tokens_seen": 13411208,
      "step": 20560
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.015581630568852253,
      "loss": 0.3481,
      "num_input_tokens_seen": 13413704,
      "step": 20565
    },
    {
      "epoch": 10.780922431865829,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.015574775096772134,
      "loss": 0.4562,
      "num_input_tokens_seen": 13417128,
      "step": 20570
    },
    {
      "epoch": 10.783542976939204,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.01556791950445563,
      "loss": 0.4418,
      "num_input_tokens_seen": 13420136,
      "step": 20575
    },
    {
      "epoch": 10.786163522012579,
      "grad_norm": 0.08984375,
      "learning_rate": 0.015561063793336837,
      "loss": 0.4817,
      "num_input_tokens_seen": 13422472,
      "step": 20580
    },
    {
      "epoch": 10.788784067085954,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.015554207964849905,
      "loss": 0.3061,
      "num_input_tokens_seen": 13426280,
      "step": 20585
    },
    {
      "epoch": 10.79140461215933,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.015547352020428989,
      "loss": 0.3942,
      "num_input_tokens_seen": 13429416,
      "step": 20590
    },
    {
      "epoch": 10.794025157232705,
      "grad_norm": 0.05078125,
      "learning_rate": 0.015540495961508268,
      "loss": 0.4457,
      "num_input_tokens_seen": 13432776,
      "step": 20595
    },
    {
      "epoch": 10.79664570230608,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.015533639789521965,
      "loss": 0.4104,
      "num_input_tokens_seen": 13435976,
      "step": 20600
    },
    {
      "epoch": 10.799266247379455,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.015526783505904307,
      "loss": 0.411,
      "num_input_tokens_seen": 13439208,
      "step": 20605
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.015519927112089545,
      "loss": 0.3606,
      "num_input_tokens_seen": 13441928,
      "step": 20610
    },
    {
      "epoch": 10.804507337526205,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01551307060951197,
      "loss": 0.4,
      "num_input_tokens_seen": 13445000,
      "step": 20615
    },
    {
      "epoch": 10.80712788259958,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.01550621399960587,
      "loss": 0.479,
      "num_input_tokens_seen": 13449032,
      "step": 20620
    },
    {
      "epoch": 10.809748427672956,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.015499357283805578,
      "loss": 0.4299,
      "num_input_tokens_seen": 13452392,
      "step": 20625
    },
    {
      "epoch": 10.81236897274633,
      "grad_norm": 0.0390625,
      "learning_rate": 0.015492500463545441,
      "loss": 0.6622,
      "num_input_tokens_seen": 13455784,
      "step": 20630
    },
    {
      "epoch": 10.814989517819706,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.015485643540259822,
      "loss": 0.3849,
      "num_input_tokens_seen": 13459368,
      "step": 20635
    },
    {
      "epoch": 10.817610062893081,
      "grad_norm": 0.044921875,
      "learning_rate": 0.015478786515383116,
      "loss": 0.3875,
      "num_input_tokens_seen": 13462664,
      "step": 20640
    },
    {
      "epoch": 10.820230607966456,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.015471929390349737,
      "loss": 0.4655,
      "num_input_tokens_seen": 13464872,
      "step": 20645
    },
    {
      "epoch": 10.822851153039831,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.015465072166594108,
      "loss": 0.2681,
      "num_input_tokens_seen": 13467624,
      "step": 20650
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.015458214845550692,
      "loss": 0.4814,
      "num_input_tokens_seen": 13470856,
      "step": 20655
    },
    {
      "epoch": 10.828092243186584,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.015451357428653957,
      "loss": 0.362,
      "num_input_tokens_seen": 13474024,
      "step": 20660
    },
    {
      "epoch": 10.830712788259959,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.015444499917338397,
      "loss": 0.4544,
      "num_input_tokens_seen": 13478376,
      "step": 20665
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.015437642313038525,
      "loss": 0.3878,
      "num_input_tokens_seen": 13481480,
      "step": 20670
    },
    {
      "epoch": 10.835953878406709,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.015430784617188872,
      "loss": 0.3449,
      "num_input_tokens_seen": 13484456,
      "step": 20675
    },
    {
      "epoch": 10.838574423480084,
      "grad_norm": 0.076171875,
      "learning_rate": 0.015423926831223999,
      "loss": 0.4307,
      "num_input_tokens_seen": 13487560,
      "step": 20680
    },
    {
      "epoch": 10.84119496855346,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.015417068956578461,
      "loss": 0.4254,
      "num_input_tokens_seen": 13491272,
      "step": 20685
    },
    {
      "epoch": 10.843815513626835,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.015410210994686856,
      "loss": 0.3569,
      "num_input_tokens_seen": 13493992,
      "step": 20690
    },
    {
      "epoch": 10.84643605870021,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.015403352946983797,
      "loss": 0.455,
      "num_input_tokens_seen": 13496968,
      "step": 20695
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.015396494814903896,
      "loss": 0.5501,
      "num_input_tokens_seen": 13501032,
      "step": 20700
    },
    {
      "epoch": 10.85167714884696,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.015389636599881813,
      "loss": 0.5707,
      "num_input_tokens_seen": 13502984,
      "step": 20705
    },
    {
      "epoch": 10.854297693920335,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.015382778303352187,
      "loss": 0.4301,
      "num_input_tokens_seen": 13507816,
      "step": 20710
    },
    {
      "epoch": 10.85691823899371,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.015375919926749713,
      "loss": 0.4451,
      "num_input_tokens_seen": 13511368,
      "step": 20715
    },
    {
      "epoch": 10.859538784067086,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.015369061471509082,
      "loss": 0.4886,
      "num_input_tokens_seen": 13515720,
      "step": 20720
    },
    {
      "epoch": 10.86215932914046,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.015362202939064996,
      "loss": 0.4618,
      "num_input_tokens_seen": 13519784,
      "step": 20725
    },
    {
      "epoch": 10.864779874213836,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.015355344330852188,
      "loss": 0.3609,
      "num_input_tokens_seen": 13523304,
      "step": 20730
    },
    {
      "epoch": 10.867400419287211,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.015348485648305404,
      "loss": 0.6812,
      "num_input_tokens_seen": 13526120,
      "step": 20735
    },
    {
      "epoch": 10.870020964360586,
      "grad_norm": 0.072265625,
      "learning_rate": 0.015341626892859391,
      "loss": 0.381,
      "num_input_tokens_seen": 13529128,
      "step": 20740
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.015334768065948941,
      "loss": 0.4727,
      "num_input_tokens_seen": 13532584,
      "step": 20745
    },
    {
      "epoch": 10.875262054507338,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.015327909169008821,
      "loss": 0.4062,
      "num_input_tokens_seen": 13536104,
      "step": 20750
    },
    {
      "epoch": 10.877882599580714,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.015321050203473843,
      "loss": 0.3351,
      "num_input_tokens_seen": 13539528,
      "step": 20755
    },
    {
      "epoch": 10.880503144654089,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.015314191170778832,
      "loss": 0.479,
      "num_input_tokens_seen": 13542600,
      "step": 20760
    },
    {
      "epoch": 10.883123689727464,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01530733207235861,
      "loss": 0.3153,
      "num_input_tokens_seen": 13545704,
      "step": 20765
    },
    {
      "epoch": 10.885744234800839,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.015300472909648024,
      "loss": 0.4027,
      "num_input_tokens_seen": 13549096,
      "step": 20770
    },
    {
      "epoch": 10.888364779874214,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.015293613684081935,
      "loss": 0.4817,
      "num_input_tokens_seen": 13552232,
      "step": 20775
    },
    {
      "epoch": 10.89098532494759,
      "grad_norm": 0.09375,
      "learning_rate": 0.015286754397095207,
      "loss": 0.4891,
      "num_input_tokens_seen": 13554984,
      "step": 20780
    },
    {
      "epoch": 10.893605870020965,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.015279895050122736,
      "loss": 0.2523,
      "num_input_tokens_seen": 13561288,
      "step": 20785
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.015273035644599415,
      "loss": 0.4016,
      "num_input_tokens_seen": 13564072,
      "step": 20790
    },
    {
      "epoch": 10.898846960167715,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.015266176181960148,
      "loss": 0.4058,
      "num_input_tokens_seen": 13566856,
      "step": 20795
    },
    {
      "epoch": 10.90146750524109,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.015259316663639864,
      "loss": 0.4534,
      "num_input_tokens_seen": 13570696,
      "step": 20800
    },
    {
      "epoch": 10.904088050314465,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01525245709107349,
      "loss": 0.3186,
      "num_input_tokens_seen": 13573416,
      "step": 20805
    },
    {
      "epoch": 10.90670859538784,
      "grad_norm": 0.052734375,
      "learning_rate": 0.015245597465695973,
      "loss": 0.4791,
      "num_input_tokens_seen": 13576392,
      "step": 20810
    },
    {
      "epoch": 10.909329140461216,
      "grad_norm": 0.07421875,
      "learning_rate": 0.01523873778894227,
      "loss": 0.4884,
      "num_input_tokens_seen": 13579048,
      "step": 20815
    },
    {
      "epoch": 10.91194968553459,
      "grad_norm": 0.15625,
      "learning_rate": 0.015231878062247342,
      "loss": 0.5676,
      "num_input_tokens_seen": 13582920,
      "step": 20820
    },
    {
      "epoch": 10.914570230607966,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.015225018287046168,
      "loss": 0.327,
      "num_input_tokens_seen": 13585576,
      "step": 20825
    },
    {
      "epoch": 10.917190775681341,
      "grad_norm": 0.0859375,
      "learning_rate": 0.015218158464773741,
      "loss": 0.399,
      "num_input_tokens_seen": 13588360,
      "step": 20830
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015211298596865047,
      "loss": 0.4936,
      "num_input_tokens_seen": 13591272,
      "step": 20835
    },
    {
      "epoch": 10.922431865828091,
      "grad_norm": 0.048828125,
      "learning_rate": 0.015204438684755099,
      "loss": 0.3409,
      "num_input_tokens_seen": 13594184,
      "step": 20840
    },
    {
      "epoch": 10.925052410901468,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.015197578729878914,
      "loss": 0.4419,
      "num_input_tokens_seen": 13597352,
      "step": 20845
    },
    {
      "epoch": 10.927672955974844,
      "grad_norm": 0.0390625,
      "learning_rate": 0.015190718733671508,
      "loss": 0.4504,
      "num_input_tokens_seen": 13600008,
      "step": 20850
    },
    {
      "epoch": 10.930293501048219,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.015183858697567918,
      "loss": 0.3604,
      "num_input_tokens_seen": 13603656,
      "step": 20855
    },
    {
      "epoch": 10.932914046121594,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.01517699862300319,
      "loss": 0.4811,
      "num_input_tokens_seen": 13607176,
      "step": 20860
    },
    {
      "epoch": 10.935534591194969,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.015170138511412367,
      "loss": 0.5585,
      "num_input_tokens_seen": 13610600,
      "step": 20865
    },
    {
      "epoch": 10.938155136268344,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.015163278364230507,
      "loss": 0.3886,
      "num_input_tokens_seen": 13614472,
      "step": 20870
    },
    {
      "epoch": 10.94077568134172,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.01515641818289268,
      "loss": 0.3429,
      "num_input_tokens_seen": 13617384,
      "step": 20875
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.015149557968833953,
      "loss": 0.4525,
      "num_input_tokens_seen": 13620968,
      "step": 20880
    },
    {
      "epoch": 10.94601677148847,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.015142697723489402,
      "loss": 0.519,
      "num_input_tokens_seen": 13623752,
      "step": 20885
    },
    {
      "epoch": 10.948637316561845,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.015135837448294118,
      "loss": 0.4276,
      "num_input_tokens_seen": 13629000,
      "step": 20890
    },
    {
      "epoch": 10.95125786163522,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.015128977144683189,
      "loss": 0.389,
      "num_input_tokens_seen": 13632200,
      "step": 20895
    },
    {
      "epoch": 10.953878406708595,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.015122116814091712,
      "loss": 0.3886,
      "num_input_tokens_seen": 13634536,
      "step": 20900
    },
    {
      "epoch": 10.95649895178197,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.015115256457954793,
      "loss": 0.4265,
      "num_input_tokens_seen": 13638760,
      "step": 20905
    },
    {
      "epoch": 10.959119496855346,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.015108396077707533,
      "loss": 0.4152,
      "num_input_tokens_seen": 13641256,
      "step": 20910
    },
    {
      "epoch": 10.96174004192872,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.015101535674785053,
      "loss": 0.3977,
      "num_input_tokens_seen": 13644808,
      "step": 20915
    },
    {
      "epoch": 10.964360587002096,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015094675250622469,
      "loss": 0.4518,
      "num_input_tokens_seen": 13648168,
      "step": 20920
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0150878148066549,
      "loss": 0.4868,
      "num_input_tokens_seen": 13652520,
      "step": 20925
    },
    {
      "epoch": 10.969601677148846,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.015080954344317483,
      "loss": 0.4039,
      "num_input_tokens_seen": 13655656,
      "step": 20930
    },
    {
      "epoch": 10.972222222222221,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.015074093865045334,
      "loss": 0.4509,
      "num_input_tokens_seen": 13658440,
      "step": 20935
    },
    {
      "epoch": 10.974842767295598,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.015067233370273594,
      "loss": 0.3587,
      "num_input_tokens_seen": 13661160,
      "step": 20940
    },
    {
      "epoch": 10.977463312368974,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.015060372861437401,
      "loss": 0.4022,
      "num_input_tokens_seen": 13664936,
      "step": 20945
    },
    {
      "epoch": 10.980083857442349,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.015053512339971894,
      "loss": 0.4848,
      "num_input_tokens_seen": 13668072,
      "step": 20950
    },
    {
      "epoch": 10.982704402515724,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.015046651807312215,
      "loss": 0.3588,
      "num_input_tokens_seen": 13670856,
      "step": 20955
    },
    {
      "epoch": 10.985324947589099,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.015039791264893512,
      "loss": 0.3831,
      "num_input_tokens_seen": 13674792,
      "step": 20960
    },
    {
      "epoch": 10.987945492662474,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.015032930714150928,
      "loss": 0.4571,
      "num_input_tokens_seen": 13677544,
      "step": 20965
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.015026070156519617,
      "loss": 0.4406,
      "num_input_tokens_seen": 13681896,
      "step": 20970
    },
    {
      "epoch": 10.993186582809225,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.015019209593434725,
      "loss": 0.3275,
      "num_input_tokens_seen": 13685256,
      "step": 20975
    },
    {
      "epoch": 10.9958071278826,
      "grad_norm": 0.083984375,
      "learning_rate": 0.015012349026331409,
      "loss": 0.4744,
      "num_input_tokens_seen": 13688072,
      "step": 20980
    },
    {
      "epoch": 10.998427672955975,
      "grad_norm": 0.03515625,
      "learning_rate": 0.015005488456644813,
      "loss": 0.3546,
      "num_input_tokens_seen": 13690824,
      "step": 20985
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.46266764402389526,
      "eval_runtime": 13.6806,
      "eval_samples_per_second": 61.985,
      "eval_steps_per_second": 15.496,
      "num_input_tokens_seen": 13691904,
      "step": 20988
    },
    {
      "epoch": 11.00104821802935,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01499862788581009,
      "loss": 0.5156,
      "num_input_tokens_seen": 13692896,
      "step": 20990
    },
    {
      "epoch": 11.003668763102725,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.014991767315262408,
      "loss": 0.3278,
      "num_input_tokens_seen": 13696160,
      "step": 20995
    },
    {
      "epoch": 11.0062893081761,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.014984906746436906,
      "loss": 0.4902,
      "num_input_tokens_seen": 13699136,
      "step": 21000
    },
    {
      "epoch": 11.008909853249476,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.014978046180768737,
      "loss": 0.3748,
      "num_input_tokens_seen": 13702496,
      "step": 21005
    },
    {
      "epoch": 11.01153039832285,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01497118561969306,
      "loss": 0.4117,
      "num_input_tokens_seen": 13706048,
      "step": 21010
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.014964325064645019,
      "loss": 0.4568,
      "num_input_tokens_seen": 13708640,
      "step": 21015
    },
    {
      "epoch": 11.016771488469601,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.014957464517059772,
      "loss": 0.3298,
      "num_input_tokens_seen": 13712000,
      "step": 21020
    },
    {
      "epoch": 11.019392033542976,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.014950603978372466,
      "loss": 0.3897,
      "num_input_tokens_seen": 13715232,
      "step": 21025
    },
    {
      "epoch": 11.022012578616351,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.014943743450018235,
      "loss": 0.731,
      "num_input_tokens_seen": 13718368,
      "step": 21030
    },
    {
      "epoch": 11.024633123689727,
      "grad_norm": 0.08203125,
      "learning_rate": 0.014936882933432238,
      "loss": 0.4391,
      "num_input_tokens_seen": 13721824,
      "step": 21035
    },
    {
      "epoch": 11.027253668763104,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.014930022430049606,
      "loss": 0.4197,
      "num_input_tokens_seen": 13724480,
      "step": 21040
    },
    {
      "epoch": 11.029874213836479,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.01492316194130549,
      "loss": 0.3227,
      "num_input_tokens_seen": 13726656,
      "step": 21045
    },
    {
      "epoch": 11.032494758909854,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.01491630146863502,
      "loss": 0.3607,
      "num_input_tokens_seen": 13729184,
      "step": 21050
    },
    {
      "epoch": 11.035115303983229,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.014909441013473322,
      "loss": 0.4194,
      "num_input_tokens_seen": 13733600,
      "step": 21055
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.014902580577255534,
      "loss": 0.4858,
      "num_input_tokens_seen": 13736832,
      "step": 21060
    },
    {
      "epoch": 11.04035639412998,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.014895720161416773,
      "loss": 0.4014,
      "num_input_tokens_seen": 13739360,
      "step": 21065
    },
    {
      "epoch": 11.042976939203355,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.014888859767392168,
      "loss": 0.3189,
      "num_input_tokens_seen": 13743296,
      "step": 21070
    },
    {
      "epoch": 11.04559748427673,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.014881999396616833,
      "loss": 0.3138,
      "num_input_tokens_seen": 13747072,
      "step": 21075
    },
    {
      "epoch": 11.048218029350105,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01487513905052588,
      "loss": 0.3804,
      "num_input_tokens_seen": 13750208,
      "step": 21080
    },
    {
      "epoch": 11.05083857442348,
      "grad_norm": 0.0546875,
      "learning_rate": 0.014868278730554407,
      "loss": 0.3844,
      "num_input_tokens_seen": 13752928,
      "step": 21085
    },
    {
      "epoch": 11.053459119496855,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.014861418438137526,
      "loss": 0.3414,
      "num_input_tokens_seen": 13756736,
      "step": 21090
    },
    {
      "epoch": 11.05607966457023,
      "grad_norm": 0.0703125,
      "learning_rate": 0.014854558174710323,
      "loss": 0.4138,
      "num_input_tokens_seen": 13759520,
      "step": 21095
    },
    {
      "epoch": 11.058700209643606,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.014847697941707892,
      "loss": 0.3918,
      "num_input_tokens_seen": 13762592,
      "step": 21100
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.014840837740565313,
      "loss": 0.4802,
      "num_input_tokens_seen": 13766464,
      "step": 21105
    },
    {
      "epoch": 11.063941299790356,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.014833977572717668,
      "loss": 0.3619,
      "num_input_tokens_seen": 13772672,
      "step": 21110
    },
    {
      "epoch": 11.066561844863731,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.014827117439600019,
      "loss": 0.4263,
      "num_input_tokens_seen": 13776000,
      "step": 21115
    },
    {
      "epoch": 11.069182389937106,
      "grad_norm": 0.11328125,
      "learning_rate": 0.014820257342647424,
      "loss": 0.6515,
      "num_input_tokens_seen": 13778464,
      "step": 21120
    },
    {
      "epoch": 11.071802935010481,
      "grad_norm": 0.046875,
      "learning_rate": 0.014813397283294947,
      "loss": 0.4068,
      "num_input_tokens_seen": 13781536,
      "step": 21125
    },
    {
      "epoch": 11.074423480083857,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.014806537262977626,
      "loss": 0.3472,
      "num_input_tokens_seen": 13784800,
      "step": 21130
    },
    {
      "epoch": 11.077044025157234,
      "grad_norm": 0.021484375,
      "learning_rate": 0.014799677283130508,
      "loss": 0.4323,
      "num_input_tokens_seen": 13788384,
      "step": 21135
    },
    {
      "epoch": 11.079664570230609,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.014792817345188615,
      "loss": 0.5155,
      "num_input_tokens_seen": 13791744,
      "step": 21140
    },
    {
      "epoch": 11.082285115303984,
      "grad_norm": 0.0390625,
      "learning_rate": 0.014785957450586966,
      "loss": 0.3123,
      "num_input_tokens_seen": 13798304,
      "step": 21145
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.014779097600760576,
      "loss": 0.3916,
      "num_input_tokens_seen": 13801184,
      "step": 21150
    },
    {
      "epoch": 11.087526205450734,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01477223779714445,
      "loss": 0.387,
      "num_input_tokens_seen": 13805120,
      "step": 21155
    },
    {
      "epoch": 11.09014675052411,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.014765378041173578,
      "loss": 0.3655,
      "num_input_tokens_seen": 13808448,
      "step": 21160
    },
    {
      "epoch": 11.092767295597485,
      "grad_norm": 0.068359375,
      "learning_rate": 0.014758518334282947,
      "loss": 0.3872,
      "num_input_tokens_seen": 13812320,
      "step": 21165
    },
    {
      "epoch": 11.09538784067086,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.014751658677907523,
      "loss": 0.3452,
      "num_input_tokens_seen": 13815168,
      "step": 21170
    },
    {
      "epoch": 11.098008385744235,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.014744799073482267,
      "loss": 0.4463,
      "num_input_tokens_seen": 13817696,
      "step": 21175
    },
    {
      "epoch": 11.10062893081761,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.014737939522442137,
      "loss": 0.4651,
      "num_input_tokens_seen": 13820320,
      "step": 21180
    },
    {
      "epoch": 11.103249475890985,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.014731080026222065,
      "loss": 0.4254,
      "num_input_tokens_seen": 13823680,
      "step": 21185
    },
    {
      "epoch": 11.10587002096436,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.014724220586256989,
      "loss": 0.4036,
      "num_input_tokens_seen": 13826016,
      "step": 21190
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.014717361203981823,
      "loss": 0.4512,
      "num_input_tokens_seen": 13828416,
      "step": 21195
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.014710501880831459,
      "loss": 0.4522,
      "num_input_tokens_seen": 13830976,
      "step": 21200
    },
    {
      "epoch": 11.113731656184486,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.014703642618240803,
      "loss": 0.4524,
      "num_input_tokens_seen": 13833952,
      "step": 21205
    },
    {
      "epoch": 11.116352201257861,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.01469678341764473,
      "loss": 0.3006,
      "num_input_tokens_seen": 13836224,
      "step": 21210
    },
    {
      "epoch": 11.118972746331236,
      "grad_norm": 0.064453125,
      "learning_rate": 0.014689924280478108,
      "loss": 0.3753,
      "num_input_tokens_seen": 13840416,
      "step": 21215
    },
    {
      "epoch": 11.121593291404611,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.014683065208175791,
      "loss": 0.3347,
      "num_input_tokens_seen": 13843072,
      "step": 21220
    },
    {
      "epoch": 11.124213836477987,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.014676206202172613,
      "loss": 0.5116,
      "num_input_tokens_seen": 13845728,
      "step": 21225
    },
    {
      "epoch": 11.126834381551364,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0146693472639034,
      "loss": 0.339,
      "num_input_tokens_seen": 13849472,
      "step": 21230
    },
    {
      "epoch": 11.129454926624739,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.014662488394802973,
      "loss": 0.3573,
      "num_input_tokens_seen": 13852704,
      "step": 21235
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.014655629596306119,
      "loss": 0.4725,
      "num_input_tokens_seen": 13856000,
      "step": 21240
    },
    {
      "epoch": 11.134696016771489,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.014648770869847628,
      "loss": 0.4041,
      "num_input_tokens_seen": 13858944,
      "step": 21245
    },
    {
      "epoch": 11.137316561844864,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.014641912216862266,
      "loss": 0.3917,
      "num_input_tokens_seen": 13861504,
      "step": 21250
    },
    {
      "epoch": 11.13993710691824,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01463505363878478,
      "loss": 0.413,
      "num_input_tokens_seen": 13865088,
      "step": 21255
    },
    {
      "epoch": 11.142557651991615,
      "grad_norm": 0.0859375,
      "learning_rate": 0.01462819513704991,
      "loss": 0.4223,
      "num_input_tokens_seen": 13867488,
      "step": 21260
    },
    {
      "epoch": 11.14517819706499,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.014621336713092372,
      "loss": 0.4476,
      "num_input_tokens_seen": 13870912,
      "step": 21265
    },
    {
      "epoch": 11.147798742138365,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.014614478368346877,
      "loss": 0.3982,
      "num_input_tokens_seen": 13874080,
      "step": 21270
    },
    {
      "epoch": 11.15041928721174,
      "grad_norm": 0.0546875,
      "learning_rate": 0.014607620104248107,
      "loss": 0.509,
      "num_input_tokens_seen": 13877920,
      "step": 21275
    },
    {
      "epoch": 11.153039832285115,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.014600761922230742,
      "loss": 0.4546,
      "num_input_tokens_seen": 13880608,
      "step": 21280
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 0.111328125,
      "learning_rate": 0.014593903823729422,
      "loss": 0.5365,
      "num_input_tokens_seen": 13883936,
      "step": 21285
    },
    {
      "epoch": 11.158280922431866,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014587045810178787,
      "loss": 0.5249,
      "num_input_tokens_seen": 13887680,
      "step": 21290
    },
    {
      "epoch": 11.16090146750524,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.01458018788301346,
      "loss": 0.4451,
      "num_input_tokens_seen": 13890592,
      "step": 21295
    },
    {
      "epoch": 11.163522012578616,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.014573330043668037,
      "loss": 0.433,
      "num_input_tokens_seen": 13892704,
      "step": 21300
    },
    {
      "epoch": 11.166142557651991,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.014566472293577106,
      "loss": 0.2985,
      "num_input_tokens_seen": 13895616,
      "step": 21305
    },
    {
      "epoch": 11.168763102725366,
      "grad_norm": 0.056640625,
      "learning_rate": 0.01455961463417522,
      "loss": 0.4949,
      "num_input_tokens_seen": 13898688,
      "step": 21310
    },
    {
      "epoch": 11.171383647798741,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.014552757066896932,
      "loss": 0.3943,
      "num_input_tokens_seen": 13901760,
      "step": 21315
    },
    {
      "epoch": 11.174004192872117,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.014545899593176757,
      "loss": 0.5303,
      "num_input_tokens_seen": 13904416,
      "step": 21320
    },
    {
      "epoch": 11.176624737945493,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014539042214449212,
      "loss": 0.3765,
      "num_input_tokens_seen": 13908192,
      "step": 21325
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.014532184932148771,
      "loss": 0.5081,
      "num_input_tokens_seen": 13911392,
      "step": 21330
    },
    {
      "epoch": 11.181865828092244,
      "grad_norm": 0.048828125,
      "learning_rate": 0.014525327747709916,
      "loss": 0.299,
      "num_input_tokens_seen": 13914368,
      "step": 21335
    },
    {
      "epoch": 11.184486373165619,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.014518470662567074,
      "loss": 0.3617,
      "num_input_tokens_seen": 13916544,
      "step": 21340
    },
    {
      "epoch": 11.187106918238994,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.014511613678154674,
      "loss": 0.4659,
      "num_input_tokens_seen": 13919520,
      "step": 21345
    },
    {
      "epoch": 11.18972746331237,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.014504756795907124,
      "loss": 0.3406,
      "num_input_tokens_seen": 13922496,
      "step": 21350
    },
    {
      "epoch": 11.192348008385745,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0144979000172588,
      "loss": 0.3839,
      "num_input_tokens_seen": 13925440,
      "step": 21355
    },
    {
      "epoch": 11.19496855345912,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.014491043343644067,
      "loss": 0.4301,
      "num_input_tokens_seen": 13928544,
      "step": 21360
    },
    {
      "epoch": 11.197589098532495,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.014484186776497262,
      "loss": 0.3623,
      "num_input_tokens_seen": 13931200,
      "step": 21365
    },
    {
      "epoch": 11.20020964360587,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.014477330317252693,
      "loss": 0.44,
      "num_input_tokens_seen": 13933984,
      "step": 21370
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.014470473967344662,
      "loss": 0.3939,
      "num_input_tokens_seen": 13937184,
      "step": 21375
    },
    {
      "epoch": 11.20545073375262,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.014463617728207432,
      "loss": 0.4274,
      "num_input_tokens_seen": 13940864,
      "step": 21380
    },
    {
      "epoch": 11.208071278825996,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.014456761601275254,
      "loss": 0.4933,
      "num_input_tokens_seen": 13943904,
      "step": 21385
    },
    {
      "epoch": 11.21069182389937,
      "grad_norm": 0.04296875,
      "learning_rate": 0.014449905587982355,
      "loss": 0.3724,
      "num_input_tokens_seen": 13948192,
      "step": 21390
    },
    {
      "epoch": 11.213312368972746,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.014443049689762934,
      "loss": 0.264,
      "num_input_tokens_seen": 13950912,
      "step": 21395
    },
    {
      "epoch": 11.215932914046121,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.014436193908051157,
      "loss": 0.3767,
      "num_input_tokens_seen": 13953664,
      "step": 21400
    },
    {
      "epoch": 11.218553459119496,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014429338244281185,
      "loss": 0.3911,
      "num_input_tokens_seen": 13956576,
      "step": 21405
    },
    {
      "epoch": 11.221174004192871,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014422482699887142,
      "loss": 0.3812,
      "num_input_tokens_seen": 13959520,
      "step": 21410
    },
    {
      "epoch": 11.223794549266247,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.014415627276303128,
      "loss": 0.4454,
      "num_input_tokens_seen": 13963552,
      "step": 21415
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.014408771974963227,
      "loss": 0.4949,
      "num_input_tokens_seen": 13967072,
      "step": 21420
    },
    {
      "epoch": 11.229035639412999,
      "grad_norm": 0.041015625,
      "learning_rate": 0.014401916797301477,
      "loss": 0.333,
      "num_input_tokens_seen": 13970592,
      "step": 21425
    },
    {
      "epoch": 11.231656184486374,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.014395061744751912,
      "loss": 0.4275,
      "num_input_tokens_seen": 13973696,
      "step": 21430
    },
    {
      "epoch": 11.234276729559749,
      "grad_norm": 0.0625,
      "learning_rate": 0.014388206818748528,
      "loss": 0.4398,
      "num_input_tokens_seen": 13977248,
      "step": 21435
    },
    {
      "epoch": 11.236897274633124,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.014381352020725298,
      "loss": 0.4842,
      "num_input_tokens_seen": 13980704,
      "step": 21440
    },
    {
      "epoch": 11.2395178197065,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.014374497352116166,
      "loss": 0.4371,
      "num_input_tokens_seen": 13982976,
      "step": 21445
    },
    {
      "epoch": 11.242138364779874,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.014367642814355057,
      "loss": 0.5336,
      "num_input_tokens_seen": 13985888,
      "step": 21450
    },
    {
      "epoch": 11.24475890985325,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.014360788408875854,
      "loss": 0.3839,
      "num_input_tokens_seen": 13990048,
      "step": 21455
    },
    {
      "epoch": 11.247379454926625,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01435393413711242,
      "loss": 0.3622,
      "num_input_tokens_seen": 13993152,
      "step": 21460
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.014347080000498594,
      "loss": 0.4701,
      "num_input_tokens_seen": 13995776,
      "step": 21465
    },
    {
      "epoch": 11.252620545073375,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014340226000468187,
      "loss": 0.5304,
      "num_input_tokens_seen": 13999168,
      "step": 21470
    },
    {
      "epoch": 11.25524109014675,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.01433337213845497,
      "loss": 0.3724,
      "num_input_tokens_seen": 14002048,
      "step": 21475
    },
    {
      "epoch": 11.257861635220126,
      "grad_norm": 0.06640625,
      "learning_rate": 0.014326518415892702,
      "loss": 0.4075,
      "num_input_tokens_seen": 14004864,
      "step": 21480
    },
    {
      "epoch": 11.2604821802935,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.014319664834215096,
      "loss": 0.3119,
      "num_input_tokens_seen": 14008352,
      "step": 21485
    },
    {
      "epoch": 11.263102725366876,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.01431281139485584,
      "loss": 0.3711,
      "num_input_tokens_seen": 14011232,
      "step": 21490
    },
    {
      "epoch": 11.265723270440251,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.014305958099248605,
      "loss": 0.3998,
      "num_input_tokens_seen": 14014848,
      "step": 21495
    },
    {
      "epoch": 11.268343815513626,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.014299104948827013,
      "loss": 0.3796,
      "num_input_tokens_seen": 14017536,
      "step": 21500
    },
    {
      "epoch": 11.270964360587001,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.014292251945024677,
      "loss": 0.361,
      "num_input_tokens_seen": 14021376,
      "step": 21505
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014285399089275157,
      "loss": 0.4372,
      "num_input_tokens_seen": 14024224,
      "step": 21510
    },
    {
      "epoch": 11.276205450733753,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.01427854638301199,
      "loss": 0.3589,
      "num_input_tokens_seen": 14026848,
      "step": 21515
    },
    {
      "epoch": 11.278825995807129,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014271693827668694,
      "loss": 0.3378,
      "num_input_tokens_seen": 14030720,
      "step": 21520
    },
    {
      "epoch": 11.281446540880504,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.014264841424678737,
      "loss": 0.3593,
      "num_input_tokens_seen": 14033632,
      "step": 21525
    },
    {
      "epoch": 11.284067085953879,
      "grad_norm": 0.05078125,
      "learning_rate": 0.014257989175475569,
      "loss": 0.3923,
      "num_input_tokens_seen": 14036960,
      "step": 21530
    },
    {
      "epoch": 11.286687631027254,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.014251137081492603,
      "loss": 0.4316,
      "num_input_tokens_seen": 14039648,
      "step": 21535
    },
    {
      "epoch": 11.28930817610063,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.014244285144163212,
      "loss": 0.5359,
      "num_input_tokens_seen": 14043840,
      "step": 21540
    },
    {
      "epoch": 11.291928721174004,
      "grad_norm": 0.044921875,
      "learning_rate": 0.014237433364920744,
      "loss": 0.4432,
      "num_input_tokens_seen": 14047200,
      "step": 21545
    },
    {
      "epoch": 11.29454926624738,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.014230581745198522,
      "loss": 0.3405,
      "num_input_tokens_seen": 14050336,
      "step": 21550
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.014223730286429816,
      "loss": 0.4675,
      "num_input_tokens_seen": 14053312,
      "step": 21555
    },
    {
      "epoch": 11.29979035639413,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.014216878990047881,
      "loss": 0.3092,
      "num_input_tokens_seen": 14056256,
      "step": 21560
    },
    {
      "epoch": 11.302410901467505,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.014210027857485931,
      "loss": 0.4581,
      "num_input_tokens_seen": 14059168,
      "step": 21565
    },
    {
      "epoch": 11.30503144654088,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.014203176890177135,
      "loss": 0.3551,
      "num_input_tokens_seen": 14062720,
      "step": 21570
    },
    {
      "epoch": 11.307651991614255,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.014196326089554645,
      "loss": 0.3884,
      "num_input_tokens_seen": 14065952,
      "step": 21575
    },
    {
      "epoch": 11.31027253668763,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.014189475457051565,
      "loss": 0.4246,
      "num_input_tokens_seen": 14069664,
      "step": 21580
    },
    {
      "epoch": 11.312893081761006,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.014182624994100976,
      "loss": 0.4094,
      "num_input_tokens_seen": 14073024,
      "step": 21585
    },
    {
      "epoch": 11.315513626834381,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.014175774702135917,
      "loss": 0.38,
      "num_input_tokens_seen": 14076576,
      "step": 21590
    },
    {
      "epoch": 11.318134171907756,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.014168924582589379,
      "loss": 0.4701,
      "num_input_tokens_seen": 14080128,
      "step": 21595
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.014162074636894341,
      "loss": 0.438,
      "num_input_tokens_seen": 14084320,
      "step": 21600
    },
    {
      "epoch": 11.323375262054507,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.014155224866483728,
      "loss": 0.3213,
      "num_input_tokens_seen": 14087008,
      "step": 21605
    },
    {
      "epoch": 11.325995807127883,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.014148375272790436,
      "loss": 0.3645,
      "num_input_tokens_seen": 14090208,
      "step": 21610
    },
    {
      "epoch": 11.328616352201259,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.014141525857247319,
      "loss": 0.2781,
      "num_input_tokens_seen": 14093184,
      "step": 21615
    },
    {
      "epoch": 11.331236897274634,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.014134676621287206,
      "loss": 0.4728,
      "num_input_tokens_seen": 14100064,
      "step": 21620
    },
    {
      "epoch": 11.333857442348009,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.014127827566342862,
      "loss": 0.3932,
      "num_input_tokens_seen": 14103104,
      "step": 21625
    },
    {
      "epoch": 11.336477987421384,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.014120978693847049,
      "loss": 0.3607,
      "num_input_tokens_seen": 14106496,
      "step": 21630
    },
    {
      "epoch": 11.33909853249476,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014114130005232456,
      "loss": 0.5055,
      "num_input_tokens_seen": 14110368,
      "step": 21635
    },
    {
      "epoch": 11.341719077568134,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.014107281501931766,
      "loss": 0.5883,
      "num_input_tokens_seen": 14114944,
      "step": 21640
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.014100433185377598,
      "loss": 0.3584,
      "num_input_tokens_seen": 14118272,
      "step": 21645
    },
    {
      "epoch": 11.346960167714885,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01409358505700255,
      "loss": 0.3838,
      "num_input_tokens_seen": 14121472,
      "step": 21650
    },
    {
      "epoch": 11.34958071278826,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.014086737118239163,
      "loss": 0.327,
      "num_input_tokens_seen": 14123808,
      "step": 21655
    },
    {
      "epoch": 11.352201257861635,
      "grad_norm": 0.126953125,
      "learning_rate": 0.014079889370519951,
      "loss": 0.4069,
      "num_input_tokens_seen": 14126144,
      "step": 21660
    },
    {
      "epoch": 11.35482180293501,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.014073041815277387,
      "loss": 0.4172,
      "num_input_tokens_seen": 14129888,
      "step": 21665
    },
    {
      "epoch": 11.357442348008385,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.014066194453943897,
      "loss": 0.3138,
      "num_input_tokens_seen": 14132448,
      "step": 21670
    },
    {
      "epoch": 11.36006289308176,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01405934728795188,
      "loss": 0.3565,
      "num_input_tokens_seen": 14136736,
      "step": 21675
    },
    {
      "epoch": 11.362683438155136,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.014052500318733679,
      "loss": 0.4783,
      "num_input_tokens_seen": 14139840,
      "step": 21680
    },
    {
      "epoch": 11.365303983228511,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.014045653547721595,
      "loss": 0.3635,
      "num_input_tokens_seen": 14142528,
      "step": 21685
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 0.0625,
      "learning_rate": 0.014038806976347904,
      "loss": 0.3459,
      "num_input_tokens_seen": 14145984,
      "step": 21690
    },
    {
      "epoch": 11.370545073375261,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014031960606044827,
      "loss": 0.5057,
      "num_input_tokens_seen": 14149120,
      "step": 21695
    },
    {
      "epoch": 11.373165618448636,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.014025114438244547,
      "loss": 0.385,
      "num_input_tokens_seen": 14151936,
      "step": 21700
    },
    {
      "epoch": 11.375786163522013,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.01401826847437921,
      "loss": 0.4191,
      "num_input_tokens_seen": 14155040,
      "step": 21705
    },
    {
      "epoch": 11.378406708595389,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.014011422715880902,
      "loss": 0.4277,
      "num_input_tokens_seen": 14158336,
      "step": 21710
    },
    {
      "epoch": 11.381027253668764,
      "grad_norm": 0.111328125,
      "learning_rate": 0.014004577164181683,
      "loss": 0.378,
      "num_input_tokens_seen": 14161248,
      "step": 21715
    },
    {
      "epoch": 11.383647798742139,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.01399773182071357,
      "loss": 0.4238,
      "num_input_tokens_seen": 14164960,
      "step": 21720
    },
    {
      "epoch": 11.386268343815514,
      "grad_norm": 0.052734375,
      "learning_rate": 0.013990886686908519,
      "loss": 0.3404,
      "num_input_tokens_seen": 14167584,
      "step": 21725
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.013984041764198464,
      "loss": 0.3242,
      "num_input_tokens_seen": 14169888,
      "step": 21730
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.013977197054015286,
      "loss": 0.3866,
      "num_input_tokens_seen": 14173024,
      "step": 21735
    },
    {
      "epoch": 11.39412997903564,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.013970352557790808,
      "loss": 0.3307,
      "num_input_tokens_seen": 14176576,
      "step": 21740
    },
    {
      "epoch": 11.396750524109015,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.013963508276956832,
      "loss": 0.4456,
      "num_input_tokens_seen": 14180064,
      "step": 21745
    },
    {
      "epoch": 11.39937106918239,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.013956664212945096,
      "loss": 0.3998,
      "num_input_tokens_seen": 14183456,
      "step": 21750
    },
    {
      "epoch": 11.401991614255765,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.013949820367187307,
      "loss": 0.3485,
      "num_input_tokens_seen": 14186752,
      "step": 21755
    },
    {
      "epoch": 11.40461215932914,
      "grad_norm": 0.048828125,
      "learning_rate": 0.013942976741115113,
      "loss": 0.2768,
      "num_input_tokens_seen": 14189280,
      "step": 21760
    },
    {
      "epoch": 11.407232704402515,
      "grad_norm": 0.060546875,
      "learning_rate": 0.013936133336160133,
      "loss": 0.3796,
      "num_input_tokens_seen": 14193120,
      "step": 21765
    },
    {
      "epoch": 11.40985324947589,
      "grad_norm": 0.056640625,
      "learning_rate": 0.013929290153753918,
      "loss": 0.5016,
      "num_input_tokens_seen": 14196352,
      "step": 21770
    },
    {
      "epoch": 11.412473794549266,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.013922447195327984,
      "loss": 0.3969,
      "num_input_tokens_seen": 14199456,
      "step": 21775
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 0.078125,
      "learning_rate": 0.013915604462313802,
      "loss": 0.3941,
      "num_input_tokens_seen": 14201600,
      "step": 21780
    },
    {
      "epoch": 11.417714884696016,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0139087619561428,
      "loss": 0.4137,
      "num_input_tokens_seen": 14204672,
      "step": 21785
    },
    {
      "epoch": 11.420335429769391,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01390191967824635,
      "loss": 0.4679,
      "num_input_tokens_seen": 14208928,
      "step": 21790
    },
    {
      "epoch": 11.422955974842766,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01389507763005577,
      "loss": 0.2826,
      "num_input_tokens_seen": 14213888,
      "step": 21795
    },
    {
      "epoch": 11.425576519916142,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.013888235813002345,
      "loss": 0.3024,
      "num_input_tokens_seen": 14216704,
      "step": 21800
    },
    {
      "epoch": 11.428197064989519,
      "grad_norm": 0.08984375,
      "learning_rate": 0.013881394228517304,
      "loss": 0.4373,
      "num_input_tokens_seen": 14219616,
      "step": 21805
    },
    {
      "epoch": 11.430817610062894,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.01387455287803183,
      "loss": 0.3037,
      "num_input_tokens_seen": 14222368,
      "step": 21810
    },
    {
      "epoch": 11.433438155136269,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.013867711762977051,
      "loss": 0.4065,
      "num_input_tokens_seen": 14229888,
      "step": 21815
    },
    {
      "epoch": 11.436058700209644,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.013860870884784061,
      "loss": 0.3901,
      "num_input_tokens_seen": 14233440,
      "step": 21820
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.013854030244883884,
      "loss": 0.4003,
      "num_input_tokens_seen": 14235488,
      "step": 21825
    },
    {
      "epoch": 11.441299790356394,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0138471898447075,
      "loss": 0.4406,
      "num_input_tokens_seen": 14240128,
      "step": 21830
    },
    {
      "epoch": 11.44392033542977,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.013840349685685852,
      "loss": 0.3036,
      "num_input_tokens_seen": 14243264,
      "step": 21835
    },
    {
      "epoch": 11.446540880503145,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.013833509769249817,
      "loss": 0.509,
      "num_input_tokens_seen": 14247712,
      "step": 21840
    },
    {
      "epoch": 11.44916142557652,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01382667009683023,
      "loss": 0.409,
      "num_input_tokens_seen": 14251360,
      "step": 21845
    },
    {
      "epoch": 11.451781970649895,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.013819830669857883,
      "loss": 0.3762,
      "num_input_tokens_seen": 14254432,
      "step": 21850
    },
    {
      "epoch": 11.45440251572327,
      "grad_norm": 0.06640625,
      "learning_rate": 0.013812991489763485,
      "loss": 0.3485,
      "num_input_tokens_seen": 14257664,
      "step": 21855
    },
    {
      "epoch": 11.457023060796645,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.013806152557977727,
      "loss": 0.352,
      "num_input_tokens_seen": 14260960,
      "step": 21860
    },
    {
      "epoch": 11.45964360587002,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.013799313875931236,
      "loss": 0.5311,
      "num_input_tokens_seen": 14264960,
      "step": 21865
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.01379247544505458,
      "loss": 0.3941,
      "num_input_tokens_seen": 14267840,
      "step": 21870
    },
    {
      "epoch": 11.464884696016771,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.013785637266778292,
      "loss": 0.4474,
      "num_input_tokens_seen": 14270368,
      "step": 21875
    },
    {
      "epoch": 11.467505241090146,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.01377879934253283,
      "loss": 0.3201,
      "num_input_tokens_seen": 14273248,
      "step": 21880
    },
    {
      "epoch": 11.470125786163521,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01377196167374861,
      "loss": 0.4084,
      "num_input_tokens_seen": 14281472,
      "step": 21885
    },
    {
      "epoch": 11.472746331236896,
      "grad_norm": 0.080078125,
      "learning_rate": 0.013765124261856002,
      "loss": 0.4979,
      "num_input_tokens_seen": 14284864,
      "step": 21890
    },
    {
      "epoch": 11.475366876310272,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.013758287108285305,
      "loss": 0.4544,
      "num_input_tokens_seen": 14287968,
      "step": 21895
    },
    {
      "epoch": 11.477987421383649,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.013751450214466785,
      "loss": 0.4943,
      "num_input_tokens_seen": 14291392,
      "step": 21900
    },
    {
      "epoch": 11.480607966457024,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.013744613581830637,
      "loss": 0.3818,
      "num_input_tokens_seen": 14294432,
      "step": 21905
    },
    {
      "epoch": 11.483228511530399,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.013737777211807,
      "loss": 0.3006,
      "num_input_tokens_seen": 14299328,
      "step": 21910
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.013730941105825973,
      "loss": 0.3143,
      "num_input_tokens_seen": 14302272,
      "step": 21915
    },
    {
      "epoch": 11.48846960167715,
      "grad_norm": 0.03515625,
      "learning_rate": 0.013724105265317584,
      "loss": 0.4022,
      "num_input_tokens_seen": 14306432,
      "step": 21920
    },
    {
      "epoch": 11.491090146750524,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.013717269691711819,
      "loss": 0.5385,
      "num_input_tokens_seen": 14311680,
      "step": 21925
    },
    {
      "epoch": 11.4937106918239,
      "grad_norm": 0.06640625,
      "learning_rate": 0.013710434386438602,
      "loss": 0.4629,
      "num_input_tokens_seen": 14314208,
      "step": 21930
    },
    {
      "epoch": 11.496331236897275,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.013703599350927805,
      "loss": 0.5545,
      "num_input_tokens_seen": 14317696,
      "step": 21935
    },
    {
      "epoch": 11.49895178197065,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.013696764586609225,
      "loss": 0.4471,
      "num_input_tokens_seen": 14320864,
      "step": 21940
    },
    {
      "epoch": 11.501572327044025,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.01368993009491263,
      "loss": 0.4116,
      "num_input_tokens_seen": 14323520,
      "step": 21945
    },
    {
      "epoch": 11.5041928721174,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.013683095877267707,
      "loss": 0.5057,
      "num_input_tokens_seen": 14326400,
      "step": 21950
    },
    {
      "epoch": 11.506813417190775,
      "grad_norm": 0.115234375,
      "learning_rate": 0.013676261935104109,
      "loss": 0.2921,
      "num_input_tokens_seen": 14328992,
      "step": 21955
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.013669428269851413,
      "loss": 0.4162,
      "num_input_tokens_seen": 14331584,
      "step": 21960
    },
    {
      "epoch": 11.512054507337526,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.013662594882939139,
      "loss": 0.4154,
      "num_input_tokens_seen": 14335360,
      "step": 21965
    },
    {
      "epoch": 11.514675052410901,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013655761775796759,
      "loss": 0.5152,
      "num_input_tokens_seen": 14339360,
      "step": 21970
    },
    {
      "epoch": 11.517295597484276,
      "grad_norm": 0.072265625,
      "learning_rate": 0.013648928949853678,
      "loss": 0.4381,
      "num_input_tokens_seen": 14342112,
      "step": 21975
    },
    {
      "epoch": 11.519916142557651,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.013642096406539248,
      "loss": 0.3345,
      "num_input_tokens_seen": 14344960,
      "step": 21980
    },
    {
      "epoch": 11.522536687631026,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.013635264147282755,
      "loss": 0.4316,
      "num_input_tokens_seen": 14348384,
      "step": 21985
    },
    {
      "epoch": 11.525157232704402,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.01362843217351344,
      "loss": 0.5339,
      "num_input_tokens_seen": 14351616,
      "step": 21990
    },
    {
      "epoch": 11.527777777777779,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.013621600486660465,
      "loss": 0.4134,
      "num_input_tokens_seen": 14354848,
      "step": 21995
    },
    {
      "epoch": 11.530398322851154,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.013614769088152935,
      "loss": 0.4212,
      "num_input_tokens_seen": 14358496,
      "step": 22000
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 0.087890625,
      "learning_rate": 0.013607937979419912,
      "loss": 0.419,
      "num_input_tokens_seen": 14361728,
      "step": 22005
    },
    {
      "epoch": 11.535639412997904,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.013601107161890384,
      "loss": 0.3,
      "num_input_tokens_seen": 14365248,
      "step": 22010
    },
    {
      "epoch": 11.53825995807128,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.013594276636993275,
      "loss": 0.4095,
      "num_input_tokens_seen": 14368416,
      "step": 22015
    },
    {
      "epoch": 11.540880503144654,
      "grad_norm": 0.048828125,
      "learning_rate": 0.013587446406157463,
      "loss": 0.4984,
      "num_input_tokens_seen": 14371424,
      "step": 22020
    },
    {
      "epoch": 11.54350104821803,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.013580616470811744,
      "loss": 0.3809,
      "num_input_tokens_seen": 14374592,
      "step": 22025
    },
    {
      "epoch": 11.546121593291405,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.013573786832384863,
      "loss": 0.3815,
      "num_input_tokens_seen": 14377504,
      "step": 22030
    },
    {
      "epoch": 11.54874213836478,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01356695749230551,
      "loss": 0.4121,
      "num_input_tokens_seen": 14380960,
      "step": 22035
    },
    {
      "epoch": 11.551362683438155,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.013560128452002294,
      "loss": 0.3716,
      "num_input_tokens_seen": 14384608,
      "step": 22040
    },
    {
      "epoch": 11.55398322851153,
      "grad_norm": 0.060546875,
      "learning_rate": 0.013553299712903786,
      "loss": 0.3148,
      "num_input_tokens_seen": 14388160,
      "step": 22045
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.013546471276438475,
      "loss": 0.4858,
      "num_input_tokens_seen": 14391296,
      "step": 22050
    },
    {
      "epoch": 11.55922431865828,
      "grad_norm": 0.037109375,
      "learning_rate": 0.013539643144034783,
      "loss": 0.2781,
      "num_input_tokens_seen": 14394752,
      "step": 22055
    },
    {
      "epoch": 11.561844863731656,
      "grad_norm": 0.0625,
      "learning_rate": 0.01353281531712109,
      "loss": 0.3681,
      "num_input_tokens_seen": 14397376,
      "step": 22060
    },
    {
      "epoch": 11.564465408805031,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.013525987797125688,
      "loss": 0.4073,
      "num_input_tokens_seen": 14399872,
      "step": 22065
    },
    {
      "epoch": 11.567085953878406,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.013519160585476827,
      "loss": 0.4905,
      "num_input_tokens_seen": 14403552,
      "step": 22070
    },
    {
      "epoch": 11.569706498951781,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.013512333683602681,
      "loss": 0.3799,
      "num_input_tokens_seen": 14407552,
      "step": 22075
    },
    {
      "epoch": 11.572327044025156,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01350550709293135,
      "loss": 0.3945,
      "num_input_tokens_seen": 14410912,
      "step": 22080
    },
    {
      "epoch": 11.574947589098532,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.013498680814890883,
      "loss": 0.4345,
      "num_input_tokens_seen": 14413440,
      "step": 22085
    },
    {
      "epoch": 11.577568134171909,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.013491854850909268,
      "loss": 0.3337,
      "num_input_tokens_seen": 14417568,
      "step": 22090
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 0.046875,
      "learning_rate": 0.013485029202414407,
      "loss": 0.342,
      "num_input_tokens_seen": 14420736,
      "step": 22095
    },
    {
      "epoch": 11.582809224318659,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.013478203870834156,
      "loss": 0.4037,
      "num_input_tokens_seen": 14424224,
      "step": 22100
    },
    {
      "epoch": 11.585429769392034,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.013471378857596299,
      "loss": 0.3597,
      "num_input_tokens_seen": 14427264,
      "step": 22105
    },
    {
      "epoch": 11.58805031446541,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.013464554164128536,
      "loss": 0.4803,
      "num_input_tokens_seen": 14429504,
      "step": 22110
    },
    {
      "epoch": 11.590670859538784,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.01345772979185853,
      "loss": 0.4539,
      "num_input_tokens_seen": 14432672,
      "step": 22115
    },
    {
      "epoch": 11.59329140461216,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.013450905742213852,
      "loss": 0.3979,
      "num_input_tokens_seen": 14435072,
      "step": 22120
    },
    {
      "epoch": 11.595911949685535,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.013444082016622023,
      "loss": 0.3287,
      "num_input_tokens_seen": 14439424,
      "step": 22125
    },
    {
      "epoch": 11.59853249475891,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.013437258616510484,
      "loss": 0.3982,
      "num_input_tokens_seen": 14441888,
      "step": 22130
    },
    {
      "epoch": 11.601153039832285,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.013430435543306618,
      "loss": 0.433,
      "num_input_tokens_seen": 14444896,
      "step": 22135
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013423612798437727,
      "loss": 0.3839,
      "num_input_tokens_seen": 14447904,
      "step": 22140
    },
    {
      "epoch": 11.606394129979035,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.013416790383331052,
      "loss": 0.4449,
      "num_input_tokens_seen": 14450624,
      "step": 22145
    },
    {
      "epoch": 11.60901467505241,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01340996829941377,
      "loss": 0.4239,
      "num_input_tokens_seen": 14454240,
      "step": 22150
    },
    {
      "epoch": 11.611635220125786,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.013403146548112975,
      "loss": 0.3535,
      "num_input_tokens_seen": 14457088,
      "step": 22155
    },
    {
      "epoch": 11.614255765199161,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.01339632513085571,
      "loss": 0.4242,
      "num_input_tokens_seen": 14460608,
      "step": 22160
    },
    {
      "epoch": 11.616876310272536,
      "grad_norm": 0.037109375,
      "learning_rate": 0.013389504049068928,
      "loss": 0.4958,
      "num_input_tokens_seen": 14464064,
      "step": 22165
    },
    {
      "epoch": 11.619496855345911,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.013382683304179528,
      "loss": 0.4445,
      "num_input_tokens_seen": 14467584,
      "step": 22170
    },
    {
      "epoch": 11.622117400419286,
      "grad_norm": 0.052734375,
      "learning_rate": 0.013375862897614325,
      "loss": 0.3766,
      "num_input_tokens_seen": 14471328,
      "step": 22175
    },
    {
      "epoch": 11.624737945492662,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01336904283080008,
      "loss": 0.2824,
      "num_input_tokens_seen": 14475072,
      "step": 22180
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 0.07421875,
      "learning_rate": 0.013362223105163462,
      "loss": 0.3955,
      "num_input_tokens_seen": 14477984,
      "step": 22185
    },
    {
      "epoch": 11.629979035639414,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.013355403722131093,
      "loss": 0.2138,
      "num_input_tokens_seen": 14481984,
      "step": 22190
    },
    {
      "epoch": 11.632599580712789,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.013348584683129501,
      "loss": 0.4145,
      "num_input_tokens_seen": 14485760,
      "step": 22195
    },
    {
      "epoch": 11.635220125786164,
      "grad_norm": 0.046875,
      "learning_rate": 0.013341765989585148,
      "loss": 0.4606,
      "num_input_tokens_seen": 14489664,
      "step": 22200
    },
    {
      "epoch": 11.63784067085954,
      "grad_norm": 0.037109375,
      "learning_rate": 0.013334947642924436,
      "loss": 0.3013,
      "num_input_tokens_seen": 14493056,
      "step": 22205
    },
    {
      "epoch": 11.640461215932914,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.013328129644573677,
      "loss": 0.3123,
      "num_input_tokens_seen": 14495680,
      "step": 22210
    },
    {
      "epoch": 11.64308176100629,
      "grad_norm": 0.0546875,
      "learning_rate": 0.013321311995959128,
      "loss": 0.3658,
      "num_input_tokens_seen": 14498592,
      "step": 22215
    },
    {
      "epoch": 11.645702306079665,
      "grad_norm": 0.037109375,
      "learning_rate": 0.013314494698506959,
      "loss": 0.3952,
      "num_input_tokens_seen": 14501888,
      "step": 22220
    },
    {
      "epoch": 11.64832285115304,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.013307677753643265,
      "loss": 0.4436,
      "num_input_tokens_seen": 14505088,
      "step": 22225
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01330086116279408,
      "loss": 0.3788,
      "num_input_tokens_seen": 14507744,
      "step": 22230
    },
    {
      "epoch": 11.65356394129979,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.01329404492738535,
      "loss": 0.3886,
      "num_input_tokens_seen": 14510560,
      "step": 22235
    },
    {
      "epoch": 11.656184486373165,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.013287229048842962,
      "loss": 0.4236,
      "num_input_tokens_seen": 14514432,
      "step": 22240
    },
    {
      "epoch": 11.65880503144654,
      "grad_norm": 0.0625,
      "learning_rate": 0.013280413528592721,
      "loss": 0.4514,
      "num_input_tokens_seen": 14517664,
      "step": 22245
    },
    {
      "epoch": 11.661425576519916,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.01327359836806035,
      "loss": 0.4793,
      "num_input_tokens_seen": 14520704,
      "step": 22250
    },
    {
      "epoch": 11.664046121593291,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0132667835686715,
      "loss": 0.3334,
      "num_input_tokens_seen": 14525152,
      "step": 22255
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013259969131851757,
      "loss": 0.4356,
      "num_input_tokens_seen": 14529824,
      "step": 22260
    },
    {
      "epoch": 11.669287211740041,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.01325315505902662,
      "loss": 0.4132,
      "num_input_tokens_seen": 14532416,
      "step": 22265
    },
    {
      "epoch": 11.671907756813416,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.013246341351621515,
      "loss": 0.4847,
      "num_input_tokens_seen": 14535776,
      "step": 22270
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 0.052734375,
      "learning_rate": 0.013239528011061798,
      "loss": 0.4955,
      "num_input_tokens_seen": 14538624,
      "step": 22275
    },
    {
      "epoch": 11.677148846960169,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.013232715038772729,
      "loss": 0.341,
      "num_input_tokens_seen": 14542112,
      "step": 22280
    },
    {
      "epoch": 11.679769392033544,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.013225902436179514,
      "loss": 0.3636,
      "num_input_tokens_seen": 14545024,
      "step": 22285
    },
    {
      "epoch": 11.682389937106919,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.013219090204707268,
      "loss": 0.3861,
      "num_input_tokens_seen": 14549440,
      "step": 22290
    },
    {
      "epoch": 11.685010482180294,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013212278345781035,
      "loss": 0.3732,
      "num_input_tokens_seen": 14552224,
      "step": 22295
    },
    {
      "epoch": 11.68763102725367,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.013205466860825776,
      "loss": 0.3786,
      "num_input_tokens_seen": 14555040,
      "step": 22300
    },
    {
      "epoch": 11.690251572327044,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.01319865575126638,
      "loss": 0.3523,
      "num_input_tokens_seen": 14558016,
      "step": 22305
    },
    {
      "epoch": 11.69287211740042,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01319184501852765,
      "loss": 0.4043,
      "num_input_tokens_seen": 14560640,
      "step": 22310
    },
    {
      "epoch": 11.695492662473795,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.013185034664034307,
      "loss": 0.3139,
      "num_input_tokens_seen": 14563264,
      "step": 22315
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.013178224689211008,
      "loss": 0.4007,
      "num_input_tokens_seen": 14566432,
      "step": 22320
    },
    {
      "epoch": 11.700733752620545,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.013171415095482324,
      "loss": 0.3617,
      "num_input_tokens_seen": 14570176,
      "step": 22325
    },
    {
      "epoch": 11.70335429769392,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013164605884272743,
      "loss": 0.4909,
      "num_input_tokens_seen": 14573536,
      "step": 22330
    },
    {
      "epoch": 11.705974842767295,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.013157797057006664,
      "loss": 0.3548,
      "num_input_tokens_seen": 14576672,
      "step": 22335
    },
    {
      "epoch": 11.70859538784067,
      "grad_norm": 0.037109375,
      "learning_rate": 0.013150988615108428,
      "loss": 0.3668,
      "num_input_tokens_seen": 14579584,
      "step": 22340
    },
    {
      "epoch": 11.711215932914046,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.013144180560002276,
      "loss": 0.4407,
      "num_input_tokens_seen": 14582528,
      "step": 22345
    },
    {
      "epoch": 11.713836477987421,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01313737289311238,
      "loss": 0.4262,
      "num_input_tokens_seen": 14586464,
      "step": 22350
    },
    {
      "epoch": 11.716457023060796,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.013130565615862825,
      "loss": 0.3975,
      "num_input_tokens_seen": 14589440,
      "step": 22355
    },
    {
      "epoch": 11.719077568134171,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01312375872967762,
      "loss": 0.4427,
      "num_input_tokens_seen": 14593280,
      "step": 22360
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.01311695223598068,
      "loss": 0.409,
      "num_input_tokens_seen": 14596800,
      "step": 22365
    },
    {
      "epoch": 11.724318658280922,
      "grad_norm": 0.068359375,
      "learning_rate": 0.013110146136195848,
      "loss": 0.5228,
      "num_input_tokens_seen": 14599552,
      "step": 22370
    },
    {
      "epoch": 11.726939203354299,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.013103340431746885,
      "loss": 0.3764,
      "num_input_tokens_seen": 14601952,
      "step": 22375
    },
    {
      "epoch": 11.729559748427674,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.013096535124057462,
      "loss": 0.3822,
      "num_input_tokens_seen": 14606336,
      "step": 22380
    },
    {
      "epoch": 11.732180293501049,
      "grad_norm": 0.09765625,
      "learning_rate": 0.01308973021455118,
      "loss": 0.3159,
      "num_input_tokens_seen": 14608864,
      "step": 22385
    },
    {
      "epoch": 11.734800838574424,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.013082925704651544,
      "loss": 0.2975,
      "num_input_tokens_seen": 14611488,
      "step": 22390
    },
    {
      "epoch": 11.7374213836478,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.013076121595781975,
      "loss": 0.3799,
      "num_input_tokens_seen": 14615296,
      "step": 22395
    },
    {
      "epoch": 11.740041928721174,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01306931788936582,
      "loss": 0.3834,
      "num_input_tokens_seen": 14620096,
      "step": 22400
    },
    {
      "epoch": 11.74266247379455,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.013062514586826338,
      "loss": 0.6104,
      "num_input_tokens_seen": 14622912,
      "step": 22405
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.013055711689586698,
      "loss": 0.513,
      "num_input_tokens_seen": 14625568,
      "step": 22410
    },
    {
      "epoch": 11.7479035639413,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.013048909199069992,
      "loss": 0.4825,
      "num_input_tokens_seen": 14628576,
      "step": 22415
    },
    {
      "epoch": 11.750524109014675,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.013042107116699229,
      "loss": 0.3852,
      "num_input_tokens_seen": 14632096,
      "step": 22420
    },
    {
      "epoch": 11.75314465408805,
      "grad_norm": 0.08984375,
      "learning_rate": 0.013035305443897314,
      "loss": 0.271,
      "num_input_tokens_seen": 14636192,
      "step": 22425
    },
    {
      "epoch": 11.755765199161425,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.013028504182087085,
      "loss": 0.3678,
      "num_input_tokens_seen": 14638944,
      "step": 22430
    },
    {
      "epoch": 11.7583857442348,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.013021703332691289,
      "loss": 0.3632,
      "num_input_tokens_seen": 14641088,
      "step": 22435
    },
    {
      "epoch": 11.761006289308176,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.013014902897132589,
      "loss": 0.3147,
      "num_input_tokens_seen": 14644992,
      "step": 22440
    },
    {
      "epoch": 11.76362683438155,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.013008102876833558,
      "loss": 0.4085,
      "num_input_tokens_seen": 14648000,
      "step": 22445
    },
    {
      "epoch": 11.766247379454926,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.013001303273216674,
      "loss": 0.5484,
      "num_input_tokens_seen": 14651168,
      "step": 22450
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.012994504087704348,
      "loss": 0.4473,
      "num_input_tokens_seen": 14654528,
      "step": 22455
    },
    {
      "epoch": 11.771488469601676,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.012987705321718881,
      "loss": 0.4087,
      "num_input_tokens_seen": 14657088,
      "step": 22460
    },
    {
      "epoch": 11.774109014675052,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012980906976682508,
      "loss": 0.3411,
      "num_input_tokens_seen": 14660352,
      "step": 22465
    },
    {
      "epoch": 11.776729559748428,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.012974109054017358,
      "loss": 0.4879,
      "num_input_tokens_seen": 14664288,
      "step": 22470
    },
    {
      "epoch": 11.779350104821804,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.012967311555145486,
      "loss": 0.6432,
      "num_input_tokens_seen": 14668000,
      "step": 22475
    },
    {
      "epoch": 11.781970649895179,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.012960514481488841,
      "loss": 0.3645,
      "num_input_tokens_seen": 14671328,
      "step": 22480
    },
    {
      "epoch": 11.784591194968554,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.012953717834469301,
      "loss": 0.4193,
      "num_input_tokens_seen": 14674272,
      "step": 22485
    },
    {
      "epoch": 11.78721174004193,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.012946921615508642,
      "loss": 0.471,
      "num_input_tokens_seen": 14676896,
      "step": 22490
    },
    {
      "epoch": 11.789832285115304,
      "grad_norm": 0.1796875,
      "learning_rate": 0.012940125826028563,
      "loss": 0.4591,
      "num_input_tokens_seen": 14679840,
      "step": 22495
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.012933330467450656,
      "loss": 0.428,
      "num_input_tokens_seen": 14683456,
      "step": 22500
    },
    {
      "epoch": 11.795073375262055,
      "grad_norm": 0.052734375,
      "learning_rate": 0.012926535541196446,
      "loss": 0.5171,
      "num_input_tokens_seen": 14686112,
      "step": 22505
    },
    {
      "epoch": 11.79769392033543,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.012919741048687339,
      "loss": 0.2823,
      "num_input_tokens_seen": 14689024,
      "step": 22510
    },
    {
      "epoch": 11.800314465408805,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.012912946991344671,
      "loss": 0.4302,
      "num_input_tokens_seen": 14691712,
      "step": 22515
    },
    {
      "epoch": 11.80293501048218,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.012906153370589686,
      "loss": 0.5293,
      "num_input_tokens_seen": 14695104,
      "step": 22520
    },
    {
      "epoch": 11.805555555555555,
      "grad_norm": 0.087890625,
      "learning_rate": 0.012899360187843522,
      "loss": 0.4017,
      "num_input_tokens_seen": 14697952,
      "step": 22525
    },
    {
      "epoch": 11.80817610062893,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01289256744452725,
      "loss": 0.3657,
      "num_input_tokens_seen": 14700704,
      "step": 22530
    },
    {
      "epoch": 11.810796645702306,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.012885775142061821,
      "loss": 0.3777,
      "num_input_tokens_seen": 14703488,
      "step": 22535
    },
    {
      "epoch": 11.81341719077568,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.012878983281868109,
      "loss": 0.4748,
      "num_input_tokens_seen": 14706752,
      "step": 22540
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.012872191865366901,
      "loss": 0.3249,
      "num_input_tokens_seen": 14710176,
      "step": 22545
    },
    {
      "epoch": 11.818658280922431,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.012865400893978874,
      "loss": 0.5064,
      "num_input_tokens_seen": 14713376,
      "step": 22550
    },
    {
      "epoch": 11.821278825995806,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.01285861036912463,
      "loss": 0.4379,
      "num_input_tokens_seen": 14716864,
      "step": 22555
    },
    {
      "epoch": 11.823899371069182,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.012851820292224671,
      "loss": 0.2893,
      "num_input_tokens_seen": 14719776,
      "step": 22560
    },
    {
      "epoch": 11.826519916142558,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.012845030664699397,
      "loss": 0.4493,
      "num_input_tokens_seen": 14724512,
      "step": 22565
    },
    {
      "epoch": 11.829140461215934,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.012838241487969118,
      "loss": 0.398,
      "num_input_tokens_seen": 14727168,
      "step": 22570
    },
    {
      "epoch": 11.831761006289309,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.012831452763454062,
      "loss": 0.2995,
      "num_input_tokens_seen": 14730112,
      "step": 22575
    },
    {
      "epoch": 11.834381551362684,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.012824664492574344,
      "loss": 0.3823,
      "num_input_tokens_seen": 14733312,
      "step": 22580
    },
    {
      "epoch": 11.83700209643606,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01281787667675,
      "loss": 0.4273,
      "num_input_tokens_seen": 14737440,
      "step": 22585
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 0.076171875,
      "learning_rate": 0.012811089317400963,
      "loss": 0.4439,
      "num_input_tokens_seen": 14740672,
      "step": 22590
    },
    {
      "epoch": 11.84224318658281,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.012804302415947063,
      "loss": 0.4293,
      "num_input_tokens_seen": 14743552,
      "step": 22595
    },
    {
      "epoch": 11.844863731656185,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.012797515973808049,
      "loss": 0.4399,
      "num_input_tokens_seen": 14746816,
      "step": 22600
    },
    {
      "epoch": 11.84748427672956,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.012790729992403563,
      "loss": 0.398,
      "num_input_tokens_seen": 14749696,
      "step": 22605
    },
    {
      "epoch": 11.850104821802935,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.012783944473153163,
      "loss": 0.5903,
      "num_input_tokens_seen": 14753376,
      "step": 22610
    },
    {
      "epoch": 11.85272536687631,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0127771594174763,
      "loss": 0.527,
      "num_input_tokens_seen": 14756576,
      "step": 22615
    },
    {
      "epoch": 11.855345911949685,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01277037482679232,
      "loss": 0.4139,
      "num_input_tokens_seen": 14759552,
      "step": 22620
    },
    {
      "epoch": 11.85796645702306,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.012763590702520491,
      "loss": 0.4509,
      "num_input_tokens_seen": 14763424,
      "step": 22625
    },
    {
      "epoch": 11.860587002096436,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.012756807046079972,
      "loss": 0.4258,
      "num_input_tokens_seen": 14767392,
      "step": 22630
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.012750023858889828,
      "loss": 0.5077,
      "num_input_tokens_seen": 14770272,
      "step": 22635
    },
    {
      "epoch": 11.865828092243186,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.012743241142369025,
      "loss": 0.4299,
      "num_input_tokens_seen": 14773664,
      "step": 22640
    },
    {
      "epoch": 11.868448637316561,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.012736458897936431,
      "loss": 0.327,
      "num_input_tokens_seen": 14776640,
      "step": 22645
    },
    {
      "epoch": 11.871069182389936,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.012729677127010806,
      "loss": 0.4457,
      "num_input_tokens_seen": 14779584,
      "step": 22650
    },
    {
      "epoch": 11.873689727463312,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.01272289583101083,
      "loss": 0.5337,
      "num_input_tokens_seen": 14782784,
      "step": 22655
    },
    {
      "epoch": 11.876310272536688,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.012716115011355063,
      "loss": 0.5846,
      "num_input_tokens_seen": 14787040,
      "step": 22660
    },
    {
      "epoch": 11.878930817610064,
      "grad_norm": 0.06640625,
      "learning_rate": 0.012709334669461982,
      "loss": 0.4234,
      "num_input_tokens_seen": 14790560,
      "step": 22665
    },
    {
      "epoch": 11.881551362683439,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.012702554806749953,
      "loss": 0.4751,
      "num_input_tokens_seen": 14794976,
      "step": 22670
    },
    {
      "epoch": 11.884171907756814,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.012695775424637254,
      "loss": 0.4139,
      "num_input_tokens_seen": 14798368,
      "step": 22675
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.012688996524542043,
      "loss": 0.4046,
      "num_input_tokens_seen": 14801536,
      "step": 22680
    },
    {
      "epoch": 11.889412997903564,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.012682218107882393,
      "loss": 0.2987,
      "num_input_tokens_seen": 14804544,
      "step": 22685
    },
    {
      "epoch": 11.89203354297694,
      "grad_norm": 0.07421875,
      "learning_rate": 0.012675440176076275,
      "loss": 0.3129,
      "num_input_tokens_seen": 14807328,
      "step": 22690
    },
    {
      "epoch": 11.894654088050315,
      "grad_norm": 0.08203125,
      "learning_rate": 0.012668662730541548,
      "loss": 0.3732,
      "num_input_tokens_seen": 14810464,
      "step": 22695
    },
    {
      "epoch": 11.89727463312369,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.012661885772695987,
      "loss": 0.4573,
      "num_input_tokens_seen": 14814080,
      "step": 22700
    },
    {
      "epoch": 11.899895178197065,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.012655109303957244,
      "loss": 0.5309,
      "num_input_tokens_seen": 14818368,
      "step": 22705
    },
    {
      "epoch": 11.90251572327044,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.012648333325742882,
      "loss": 0.3769,
      "num_input_tokens_seen": 14821344,
      "step": 22710
    },
    {
      "epoch": 11.905136268343815,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.012641557839470356,
      "loss": 0.3248,
      "num_input_tokens_seen": 14824352,
      "step": 22715
    },
    {
      "epoch": 11.90775681341719,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.01263478284655703,
      "loss": 0.3905,
      "num_input_tokens_seen": 14828704,
      "step": 22720
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.012628008348420143,
      "loss": 0.4528,
      "num_input_tokens_seen": 14831456,
      "step": 22725
    },
    {
      "epoch": 11.91299790356394,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.012621234346476856,
      "loss": 0.3739,
      "num_input_tokens_seen": 14833856,
      "step": 22730
    },
    {
      "epoch": 11.915618448637316,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0126144608421442,
      "loss": 0.2853,
      "num_input_tokens_seen": 14837120,
      "step": 22735
    },
    {
      "epoch": 11.918238993710691,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.01260768783683912,
      "loss": 0.4095,
      "num_input_tokens_seen": 14841056,
      "step": 22740
    },
    {
      "epoch": 11.920859538784066,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.012600915331978452,
      "loss": 0.4079,
      "num_input_tokens_seen": 14844064,
      "step": 22745
    },
    {
      "epoch": 11.923480083857442,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.012594143328978924,
      "loss": 0.4279,
      "num_input_tokens_seen": 14846656,
      "step": 22750
    },
    {
      "epoch": 11.926100628930818,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.012587371829257168,
      "loss": 0.4464,
      "num_input_tokens_seen": 14849248,
      "step": 22755
    },
    {
      "epoch": 11.928721174004194,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.012580600834229703,
      "loss": 0.4224,
      "num_input_tokens_seen": 14852384,
      "step": 22760
    },
    {
      "epoch": 11.931341719077569,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.012573830345312936,
      "loss": 0.3431,
      "num_input_tokens_seen": 14854720,
      "step": 22765
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 0.052734375,
      "learning_rate": 0.012567060363923185,
      "loss": 0.3982,
      "num_input_tokens_seen": 14857696,
      "step": 22770
    },
    {
      "epoch": 11.93658280922432,
      "grad_norm": 0.06640625,
      "learning_rate": 0.012560290891476644,
      "loss": 0.4978,
      "num_input_tokens_seen": 14861216,
      "step": 22775
    },
    {
      "epoch": 11.939203354297694,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01255352192938942,
      "loss": 0.4136,
      "num_input_tokens_seen": 14864064,
      "step": 22780
    },
    {
      "epoch": 11.94182389937107,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.012546753479077494,
      "loss": 0.4298,
      "num_input_tokens_seen": 14866784,
      "step": 22785
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.012539985541956757,
      "loss": 0.3525,
      "num_input_tokens_seen": 14869664,
      "step": 22790
    },
    {
      "epoch": 11.94706498951782,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.012533218119442975,
      "loss": 0.4323,
      "num_input_tokens_seen": 14872640,
      "step": 22795
    },
    {
      "epoch": 11.949685534591195,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.012526451212951822,
      "loss": 0.4044,
      "num_input_tokens_seen": 14876512,
      "step": 22800
    },
    {
      "epoch": 11.95230607966457,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.012519684823898853,
      "loss": 0.4881,
      "num_input_tokens_seen": 14879776,
      "step": 22805
    },
    {
      "epoch": 11.954926624737945,
      "grad_norm": 0.05078125,
      "learning_rate": 0.012512918953699525,
      "loss": 0.3104,
      "num_input_tokens_seen": 14883552,
      "step": 22810
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.012506153603769181,
      "loss": 0.4045,
      "num_input_tokens_seen": 14885984,
      "step": 22815
    },
    {
      "epoch": 11.960167714884696,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.012499388775523045,
      "loss": 0.4379,
      "num_input_tokens_seen": 14888992,
      "step": 22820
    },
    {
      "epoch": 11.96278825995807,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.012492624470376253,
      "loss": 0.3787,
      "num_input_tokens_seen": 14891744,
      "step": 22825
    },
    {
      "epoch": 11.965408805031446,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01248586068974381,
      "loss": 0.5829,
      "num_input_tokens_seen": 14894880,
      "step": 22830
    },
    {
      "epoch": 11.968029350104821,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.012479097435040633,
      "loss": 0.4063,
      "num_input_tokens_seen": 14897760,
      "step": 22835
    },
    {
      "epoch": 11.970649895178196,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01247233470768151,
      "loss": 0.4035,
      "num_input_tokens_seen": 14900736,
      "step": 22840
    },
    {
      "epoch": 11.973270440251572,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.012465572509081134,
      "loss": 0.3676,
      "num_input_tokens_seen": 14904544,
      "step": 22845
    },
    {
      "epoch": 11.975890985324948,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.012458810840654073,
      "loss": 0.482,
      "num_input_tokens_seen": 14907648,
      "step": 22850
    },
    {
      "epoch": 11.978511530398324,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01245204970381479,
      "loss": 0.3165,
      "num_input_tokens_seen": 14910816,
      "step": 22855
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.012445289099977642,
      "loss": 0.3183,
      "num_input_tokens_seen": 14913632,
      "step": 22860
    },
    {
      "epoch": 11.983752620545074,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.012438529030556865,
      "loss": 0.4132,
      "num_input_tokens_seen": 14916768,
      "step": 22865
    },
    {
      "epoch": 11.98637316561845,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.012431769496966593,
      "loss": 0.3377,
      "num_input_tokens_seen": 14920896,
      "step": 22870
    },
    {
      "epoch": 11.988993710691824,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.012425010500620849,
      "loss": 0.3862,
      "num_input_tokens_seen": 14924192,
      "step": 22875
    },
    {
      "epoch": 11.9916142557652,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012418252042933528,
      "loss": 0.422,
      "num_input_tokens_seen": 14926784,
      "step": 22880
    },
    {
      "epoch": 11.994234800838575,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.012411494125318423,
      "loss": 0.4697,
      "num_input_tokens_seen": 14930432,
      "step": 22885
    },
    {
      "epoch": 11.99685534591195,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.012404736749189217,
      "loss": 0.4989,
      "num_input_tokens_seen": 14933664,
      "step": 22890
    },
    {
      "epoch": 11.999475890985325,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.012397979915959473,
      "loss": 0.3935,
      "num_input_tokens_seen": 14937088,
      "step": 22895
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.46974053978919983,
      "eval_runtime": 13.6621,
      "eval_samples_per_second": 62.069,
      "eval_steps_per_second": 15.517,
      "num_input_tokens_seen": 14937216,
      "step": 22896
    },
    {
      "epoch": 12.0020964360587,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.01239122362704265,
      "loss": 0.3837,
      "num_input_tokens_seen": 14939712,
      "step": 22900
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 0.0703125,
      "learning_rate": 0.01238446788385208,
      "loss": 0.3898,
      "num_input_tokens_seen": 14942272,
      "step": 22905
    },
    {
      "epoch": 12.00733752620545,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.012377712687800987,
      "loss": 0.464,
      "num_input_tokens_seen": 14945632,
      "step": 22910
    },
    {
      "epoch": 12.009958071278826,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.012370958040302483,
      "loss": 0.5701,
      "num_input_tokens_seen": 14949376,
      "step": 22915
    },
    {
      "epoch": 12.0125786163522,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.012364203942769563,
      "loss": 0.4279,
      "num_input_tokens_seen": 14951840,
      "step": 22920
    },
    {
      "epoch": 12.015199161425576,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.012357450396615108,
      "loss": 0.3081,
      "num_input_tokens_seen": 14956768,
      "step": 22925
    },
    {
      "epoch": 12.017819706498951,
      "grad_norm": 0.130859375,
      "learning_rate": 0.012350697403251888,
      "loss": 0.3219,
      "num_input_tokens_seen": 14960384,
      "step": 22930
    },
    {
      "epoch": 12.020440251572326,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.012343944964092537,
      "loss": 0.3196,
      "num_input_tokens_seen": 14963392,
      "step": 22935
    },
    {
      "epoch": 12.023060796645701,
      "grad_norm": 0.060546875,
      "learning_rate": 0.012337193080549605,
      "loss": 0.373,
      "num_input_tokens_seen": 14967392,
      "step": 22940
    },
    {
      "epoch": 12.025681341719078,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.012330441754035495,
      "loss": 0.3982,
      "num_input_tokens_seen": 14972704,
      "step": 22945
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.012323690985962516,
      "loss": 0.3605,
      "num_input_tokens_seen": 14975744,
      "step": 22950
    },
    {
      "epoch": 12.030922431865829,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.012316940777742851,
      "loss": 0.396,
      "num_input_tokens_seen": 14979104,
      "step": 22955
    },
    {
      "epoch": 12.033542976939204,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.01231019113078857,
      "loss": 0.4792,
      "num_input_tokens_seen": 14981600,
      "step": 22960
    },
    {
      "epoch": 12.036163522012579,
      "grad_norm": 0.064453125,
      "learning_rate": 0.012303442046511612,
      "loss": 0.368,
      "num_input_tokens_seen": 14985152,
      "step": 22965
    },
    {
      "epoch": 12.038784067085954,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.012296693526323815,
      "loss": 0.3051,
      "num_input_tokens_seen": 14989568,
      "step": 22970
    },
    {
      "epoch": 12.04140461215933,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.01228994557163689,
      "loss": 0.4023,
      "num_input_tokens_seen": 14992352,
      "step": 22975
    },
    {
      "epoch": 12.044025157232705,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.012283198183862436,
      "loss": 0.3017,
      "num_input_tokens_seen": 14995168,
      "step": 22980
    },
    {
      "epoch": 12.04664570230608,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01227645136441193,
      "loss": 0.5038,
      "num_input_tokens_seen": 14998240,
      "step": 22985
    },
    {
      "epoch": 12.049266247379455,
      "grad_norm": 0.083984375,
      "learning_rate": 0.012269705114696721,
      "loss": 0.4093,
      "num_input_tokens_seen": 15001248,
      "step": 22990
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.012262959436128054,
      "loss": 0.4626,
      "num_input_tokens_seen": 15004352,
      "step": 22995
    },
    {
      "epoch": 12.054507337526205,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.012256214330117046,
      "loss": 0.3537,
      "num_input_tokens_seen": 15007968,
      "step": 23000
    },
    {
      "epoch": 12.05712788259958,
      "grad_norm": 0.0625,
      "learning_rate": 0.0122494697980747,
      "loss": 0.3384,
      "num_input_tokens_seen": 15011648,
      "step": 23005
    },
    {
      "epoch": 12.059748427672956,
      "grad_norm": 0.095703125,
      "learning_rate": 0.01224272584141189,
      "loss": 0.3925,
      "num_input_tokens_seen": 15014176,
      "step": 23010
    },
    {
      "epoch": 12.06236897274633,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.012235982461539384,
      "loss": 0.4319,
      "num_input_tokens_seen": 15017184,
      "step": 23015
    },
    {
      "epoch": 12.064989517819706,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.012229239659867809,
      "loss": 0.3185,
      "num_input_tokens_seen": 15020416,
      "step": 23020
    },
    {
      "epoch": 12.067610062893081,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.012222497437807685,
      "loss": 0.3541,
      "num_input_tokens_seen": 15024800,
      "step": 23025
    },
    {
      "epoch": 12.070230607966456,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.012215755796769407,
      "loss": 0.2864,
      "num_input_tokens_seen": 15027456,
      "step": 23030
    },
    {
      "epoch": 12.072851153039831,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.012209014738163257,
      "loss": 0.5332,
      "num_input_tokens_seen": 15030208,
      "step": 23035
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 0.0703125,
      "learning_rate": 0.01220227426339938,
      "loss": 0.3602,
      "num_input_tokens_seen": 15033120,
      "step": 23040
    },
    {
      "epoch": 12.078092243186584,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.012195534373887814,
      "loss": 0.4364,
      "num_input_tokens_seen": 15036960,
      "step": 23045
    },
    {
      "epoch": 12.080712788259959,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.012188795071038459,
      "loss": 0.3836,
      "num_input_tokens_seen": 15040992,
      "step": 23050
    },
    {
      "epoch": 12.083333333333334,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0121820563562611,
      "loss": 0.3795,
      "num_input_tokens_seen": 15044096,
      "step": 23055
    },
    {
      "epoch": 12.085953878406709,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.012175318230965407,
      "loss": 0.4102,
      "num_input_tokens_seen": 15047680,
      "step": 23060
    },
    {
      "epoch": 12.088574423480084,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01216858069656091,
      "loss": 0.3305,
      "num_input_tokens_seen": 15051328,
      "step": 23065
    },
    {
      "epoch": 12.09119496855346,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.012161843754457032,
      "loss": 0.3301,
      "num_input_tokens_seen": 15054560,
      "step": 23070
    },
    {
      "epoch": 12.093815513626835,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.012155107406063063,
      "loss": 0.4109,
      "num_input_tokens_seen": 15057888,
      "step": 23075
    },
    {
      "epoch": 12.09643605870021,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.012148371652788165,
      "loss": 0.4368,
      "num_input_tokens_seen": 15061120,
      "step": 23080
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 0.0546875,
      "learning_rate": 0.012141636496041386,
      "loss": 0.3949,
      "num_input_tokens_seen": 15064224,
      "step": 23085
    },
    {
      "epoch": 12.10167714884696,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.012134901937231638,
      "loss": 0.3466,
      "num_input_tokens_seen": 15067360,
      "step": 23090
    },
    {
      "epoch": 12.104297693920335,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.012128167977767718,
      "loss": 0.3223,
      "num_input_tokens_seen": 15070688,
      "step": 23095
    },
    {
      "epoch": 12.10691823899371,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.012121434619058305,
      "loss": 0.3703,
      "num_input_tokens_seen": 15074304,
      "step": 23100
    },
    {
      "epoch": 12.109538784067086,
      "grad_norm": 0.06640625,
      "learning_rate": 0.012114701862511916,
      "loss": 0.4163,
      "num_input_tokens_seen": 15076768,
      "step": 23105
    },
    {
      "epoch": 12.11215932914046,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.012107969709536984,
      "loss": 0.3433,
      "num_input_tokens_seen": 15079232,
      "step": 23110
    },
    {
      "epoch": 12.114779874213836,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.012101238161541798,
      "loss": 0.3065,
      "num_input_tokens_seen": 15082688,
      "step": 23115
    },
    {
      "epoch": 12.117400419287211,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.012094507219934512,
      "loss": 0.3397,
      "num_input_tokens_seen": 15086400,
      "step": 23120
    },
    {
      "epoch": 12.120020964360586,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.012087776886123171,
      "loss": 0.3634,
      "num_input_tokens_seen": 15089184,
      "step": 23125
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.012081047161515686,
      "loss": 0.3024,
      "num_input_tokens_seen": 15092384,
      "step": 23130
    },
    {
      "epoch": 12.125262054507338,
      "grad_norm": 0.080078125,
      "learning_rate": 0.012074318047519825,
      "loss": 0.3773,
      "num_input_tokens_seen": 15095776,
      "step": 23135
    },
    {
      "epoch": 12.127882599580714,
      "grad_norm": 0.09765625,
      "learning_rate": 0.012067589545543256,
      "loss": 0.4634,
      "num_input_tokens_seen": 15098624,
      "step": 23140
    },
    {
      "epoch": 12.130503144654089,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.012060861656993497,
      "loss": 0.4201,
      "num_input_tokens_seen": 15100992,
      "step": 23145
    },
    {
      "epoch": 12.133123689727464,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01205413438327795,
      "loss": 0.3104,
      "num_input_tokens_seen": 15104352,
      "step": 23150
    },
    {
      "epoch": 12.135744234800839,
      "grad_norm": 0.08203125,
      "learning_rate": 0.012047407725803878,
      "loss": 0.3446,
      "num_input_tokens_seen": 15107200,
      "step": 23155
    },
    {
      "epoch": 12.138364779874214,
      "grad_norm": 0.08203125,
      "learning_rate": 0.012040681685978431,
      "loss": 0.4924,
      "num_input_tokens_seen": 15111008,
      "step": 23160
    },
    {
      "epoch": 12.14098532494759,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.012033956265208612,
      "loss": 0.3828,
      "num_input_tokens_seen": 15113440,
      "step": 23165
    },
    {
      "epoch": 12.143605870020965,
      "grad_norm": 0.072265625,
      "learning_rate": 0.012027231464901297,
      "loss": 0.3848,
      "num_input_tokens_seen": 15116256,
      "step": 23170
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 0.05078125,
      "learning_rate": 0.012020507286463249,
      "loss": 0.4186,
      "num_input_tokens_seen": 15119296,
      "step": 23175
    },
    {
      "epoch": 12.148846960167715,
      "grad_norm": 0.052734375,
      "learning_rate": 0.012013783731301086,
      "loss": 0.3045,
      "num_input_tokens_seen": 15123584,
      "step": 23180
    },
    {
      "epoch": 12.15146750524109,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0120070608008213,
      "loss": 0.3784,
      "num_input_tokens_seen": 15127008,
      "step": 23185
    },
    {
      "epoch": 12.154088050314465,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.012000338496430243,
      "loss": 0.3757,
      "num_input_tokens_seen": 15129728,
      "step": 23190
    },
    {
      "epoch": 12.15670859538784,
      "grad_norm": 0.044921875,
      "learning_rate": 0.011993616819534158,
      "loss": 0.4196,
      "num_input_tokens_seen": 15133152,
      "step": 23195
    },
    {
      "epoch": 12.159329140461216,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.011986895771539127,
      "loss": 0.4224,
      "num_input_tokens_seen": 15136960,
      "step": 23200
    },
    {
      "epoch": 12.16194968553459,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.01198017535385113,
      "loss": 0.4093,
      "num_input_tokens_seen": 15140128,
      "step": 23205
    },
    {
      "epoch": 12.164570230607966,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.011973455567875995,
      "loss": 0.4082,
      "num_input_tokens_seen": 15144128,
      "step": 23210
    },
    {
      "epoch": 12.167190775681341,
      "grad_norm": 0.08203125,
      "learning_rate": 0.011966736415019434,
      "loss": 0.4087,
      "num_input_tokens_seen": 15147968,
      "step": 23215
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.011960017896687004,
      "loss": 0.3263,
      "num_input_tokens_seen": 15150976,
      "step": 23220
    },
    {
      "epoch": 12.172431865828091,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.011953300014284145,
      "loss": 0.5317,
      "num_input_tokens_seen": 15154368,
      "step": 23225
    },
    {
      "epoch": 12.175052410901468,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.011946582769216169,
      "loss": 0.4357,
      "num_input_tokens_seen": 15157984,
      "step": 23230
    },
    {
      "epoch": 12.177672955974844,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.011939866162888238,
      "loss": 0.4615,
      "num_input_tokens_seen": 15162208,
      "step": 23235
    },
    {
      "epoch": 12.180293501048219,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.011933150196705397,
      "loss": 0.3909,
      "num_input_tokens_seen": 15165088,
      "step": 23240
    },
    {
      "epoch": 12.182914046121594,
      "grad_norm": 0.076171875,
      "learning_rate": 0.011926434872072551,
      "loss": 0.3186,
      "num_input_tokens_seen": 15167200,
      "step": 23245
    },
    {
      "epoch": 12.185534591194969,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.011919720190394459,
      "loss": 0.4251,
      "num_input_tokens_seen": 15170304,
      "step": 23250
    },
    {
      "epoch": 12.188155136268344,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.011913006153075759,
      "loss": 0.3889,
      "num_input_tokens_seen": 15173024,
      "step": 23255
    },
    {
      "epoch": 12.19077568134172,
      "grad_norm": 0.138671875,
      "learning_rate": 0.011906292761520958,
      "loss": 0.4906,
      "num_input_tokens_seen": 15176960,
      "step": 23260
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.011899580017134414,
      "loss": 0.4046,
      "num_input_tokens_seen": 15184576,
      "step": 23265
    },
    {
      "epoch": 12.19601677148847,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.011892867921320362,
      "loss": 0.4398,
      "num_input_tokens_seen": 15187840,
      "step": 23270
    },
    {
      "epoch": 12.198637316561845,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.011886156475482895,
      "loss": 0.3597,
      "num_input_tokens_seen": 15191072,
      "step": 23275
    },
    {
      "epoch": 12.20125786163522,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.01187944568102596,
      "loss": 0.3657,
      "num_input_tokens_seen": 15194592,
      "step": 23280
    },
    {
      "epoch": 12.203878406708595,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.011872735539353392,
      "loss": 0.3904,
      "num_input_tokens_seen": 15198240,
      "step": 23285
    },
    {
      "epoch": 12.20649895178197,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.01186602605186887,
      "loss": 0.4349,
      "num_input_tokens_seen": 15201056,
      "step": 23290
    },
    {
      "epoch": 12.209119496855346,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.011859317219975947,
      "loss": 0.3985,
      "num_input_tokens_seen": 15204320,
      "step": 23295
    },
    {
      "epoch": 12.21174004192872,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.011852609045078033,
      "loss": 0.3839,
      "num_input_tokens_seen": 15207264,
      "step": 23300
    },
    {
      "epoch": 12.214360587002096,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.011845901528578394,
      "loss": 0.3065,
      "num_input_tokens_seen": 15210112,
      "step": 23305
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011839194671880172,
      "loss": 0.3113,
      "num_input_tokens_seen": 15213504,
      "step": 23310
    },
    {
      "epoch": 12.219601677148846,
      "grad_norm": 0.056640625,
      "learning_rate": 0.011832488476386363,
      "loss": 0.4024,
      "num_input_tokens_seen": 15215808,
      "step": 23315
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 0.08984375,
      "learning_rate": 0.01182578294349983,
      "loss": 0.42,
      "num_input_tokens_seen": 15218304,
      "step": 23320
    },
    {
      "epoch": 12.224842767295598,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.01181907807462329,
      "loss": 0.5364,
      "num_input_tokens_seen": 15221024,
      "step": 23325
    },
    {
      "epoch": 12.227463312368974,
      "grad_norm": 0.0859375,
      "learning_rate": 0.011812373871159335,
      "loss": 0.4099,
      "num_input_tokens_seen": 15223872,
      "step": 23330
    },
    {
      "epoch": 12.230083857442349,
      "grad_norm": 0.076171875,
      "learning_rate": 0.01180567033451039,
      "loss": 0.3285,
      "num_input_tokens_seen": 15227168,
      "step": 23335
    },
    {
      "epoch": 12.232704402515724,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.011798967466078772,
      "loss": 0.4068,
      "num_input_tokens_seen": 15230816,
      "step": 23340
    },
    {
      "epoch": 12.235324947589099,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.011792265267266639,
      "loss": 0.3853,
      "num_input_tokens_seen": 15233568,
      "step": 23345
    },
    {
      "epoch": 12.237945492662474,
      "grad_norm": 0.095703125,
      "learning_rate": 0.011785563739476018,
      "loss": 0.2938,
      "num_input_tokens_seen": 15236064,
      "step": 23350
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 0.07421875,
      "learning_rate": 0.011778862884108794,
      "loss": 0.3385,
      "num_input_tokens_seen": 15239168,
      "step": 23355
    },
    {
      "epoch": 12.243186582809225,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0117721627025667,
      "loss": 0.3587,
      "num_input_tokens_seen": 15241984,
      "step": 23360
    },
    {
      "epoch": 12.2458071278826,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.011765463196251347,
      "loss": 0.3831,
      "num_input_tokens_seen": 15244800,
      "step": 23365
    },
    {
      "epoch": 12.248427672955975,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.011758764366564188,
      "loss": 0.3255,
      "num_input_tokens_seen": 15248416,
      "step": 23370
    },
    {
      "epoch": 12.25104821802935,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.01175206621490655,
      "loss": 0.3448,
      "num_input_tokens_seen": 15251008,
      "step": 23375
    },
    {
      "epoch": 12.253668763102725,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.011745368742679601,
      "loss": 0.4218,
      "num_input_tokens_seen": 15256992,
      "step": 23380
    },
    {
      "epoch": 12.2562893081761,
      "grad_norm": 0.107421875,
      "learning_rate": 0.011738671951284387,
      "loss": 0.3233,
      "num_input_tokens_seen": 15259712,
      "step": 23385
    },
    {
      "epoch": 12.258909853249476,
      "grad_norm": 0.056640625,
      "learning_rate": 0.01173197584212179,
      "loss": 0.5111,
      "num_input_tokens_seen": 15262880,
      "step": 23390
    },
    {
      "epoch": 12.26153039832285,
      "grad_norm": 0.07421875,
      "learning_rate": 0.01172528041659256,
      "loss": 0.4047,
      "num_input_tokens_seen": 15266048,
      "step": 23395
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.01171858567609731,
      "loss": 0.431,
      "num_input_tokens_seen": 15268928,
      "step": 23400
    },
    {
      "epoch": 12.266771488469601,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.011711891622036499,
      "loss": 0.4935,
      "num_input_tokens_seen": 15271872,
      "step": 23405
    },
    {
      "epoch": 12.269392033542976,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.011705198255810444,
      "loss": 0.4251,
      "num_input_tokens_seen": 15276704,
      "step": 23410
    },
    {
      "epoch": 12.272012578616351,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.011698505578819334,
      "loss": 0.3744,
      "num_input_tokens_seen": 15279648,
      "step": 23415
    },
    {
      "epoch": 12.274633123689728,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.011691813592463188,
      "loss": 0.3857,
      "num_input_tokens_seen": 15282176,
      "step": 23420
    },
    {
      "epoch": 12.277253668763104,
      "grad_norm": 0.064453125,
      "learning_rate": 0.011685122298141894,
      "loss": 0.3364,
      "num_input_tokens_seen": 15284640,
      "step": 23425
    },
    {
      "epoch": 12.279874213836479,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0116784316972552,
      "loss": 0.3565,
      "num_input_tokens_seen": 15287712,
      "step": 23430
    },
    {
      "epoch": 12.282494758909854,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.011671741791202696,
      "loss": 0.3558,
      "num_input_tokens_seen": 15291296,
      "step": 23435
    },
    {
      "epoch": 12.285115303983229,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.011665052581383843,
      "loss": 0.3703,
      "num_input_tokens_seen": 15294464,
      "step": 23440
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.011658364069197945,
      "loss": 0.2323,
      "num_input_tokens_seen": 15299392,
      "step": 23445
    },
    {
      "epoch": 12.29035639412998,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.011651676256044154,
      "loss": 0.4131,
      "num_input_tokens_seen": 15302496,
      "step": 23450
    },
    {
      "epoch": 12.292976939203355,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.011644989143321495,
      "loss": 0.4965,
      "num_input_tokens_seen": 15306464,
      "step": 23455
    },
    {
      "epoch": 12.29559748427673,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.011638302732428827,
      "loss": 0.3384,
      "num_input_tokens_seen": 15309792,
      "step": 23460
    },
    {
      "epoch": 12.298218029350105,
      "grad_norm": 0.0625,
      "learning_rate": 0.01163161702476488,
      "loss": 0.4072,
      "num_input_tokens_seen": 15312576,
      "step": 23465
    },
    {
      "epoch": 12.30083857442348,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.011624932021728223,
      "loss": 0.4,
      "num_input_tokens_seen": 15315296,
      "step": 23470
    },
    {
      "epoch": 12.303459119496855,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.01161824772471728,
      "loss": 0.4082,
      "num_input_tokens_seen": 15318112,
      "step": 23475
    },
    {
      "epoch": 12.30607966457023,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.01161156413513033,
      "loss": 0.4237,
      "num_input_tokens_seen": 15320896,
      "step": 23480
    },
    {
      "epoch": 12.308700209643606,
      "grad_norm": 0.0859375,
      "learning_rate": 0.011604881254365505,
      "loss": 0.2765,
      "num_input_tokens_seen": 15323904,
      "step": 23485
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.011598199083820789,
      "loss": 0.4699,
      "num_input_tokens_seen": 15326944,
      "step": 23490
    },
    {
      "epoch": 12.313941299790356,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.011591517624894016,
      "loss": 0.4047,
      "num_input_tokens_seen": 15331776,
      "step": 23495
    },
    {
      "epoch": 12.316561844863731,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.011584836878982875,
      "loss": 0.3874,
      "num_input_tokens_seen": 15334624,
      "step": 23500
    },
    {
      "epoch": 12.319182389937106,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.01157815684748489,
      "loss": 0.4297,
      "num_input_tokens_seen": 15339072,
      "step": 23505
    },
    {
      "epoch": 12.321802935010481,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.011571477531797455,
      "loss": 0.3723,
      "num_input_tokens_seen": 15342240,
      "step": 23510
    },
    {
      "epoch": 12.324423480083858,
      "grad_norm": 0.046875,
      "learning_rate": 0.011564798933317806,
      "loss": 0.357,
      "num_input_tokens_seen": 15346528,
      "step": 23515
    },
    {
      "epoch": 12.327044025157234,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01155812105344303,
      "loss": 0.3859,
      "num_input_tokens_seen": 15350144,
      "step": 23520
    },
    {
      "epoch": 12.329664570230609,
      "grad_norm": 0.0546875,
      "learning_rate": 0.011551443893570061,
      "loss": 0.3043,
      "num_input_tokens_seen": 15352832,
      "step": 23525
    },
    {
      "epoch": 12.332285115303984,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.011544767455095691,
      "loss": 0.4277,
      "num_input_tokens_seen": 15355392,
      "step": 23530
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.011538091739416547,
      "loss": 0.3034,
      "num_input_tokens_seen": 15357504,
      "step": 23535
    },
    {
      "epoch": 12.337526205450734,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.011531416747929115,
      "loss": 0.3067,
      "num_input_tokens_seen": 15360672,
      "step": 23540
    },
    {
      "epoch": 12.34014675052411,
      "grad_norm": 0.111328125,
      "learning_rate": 0.011524742482029728,
      "loss": 0.579,
      "num_input_tokens_seen": 15364128,
      "step": 23545
    },
    {
      "epoch": 12.342767295597485,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.011518068943114563,
      "loss": 0.4474,
      "num_input_tokens_seen": 15367072,
      "step": 23550
    },
    {
      "epoch": 12.34538784067086,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.011511396132579657,
      "loss": 0.3136,
      "num_input_tokens_seen": 15370848,
      "step": 23555
    },
    {
      "epoch": 12.348008385744235,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01150472405182088,
      "loss": 0.3815,
      "num_input_tokens_seen": 15375680,
      "step": 23560
    },
    {
      "epoch": 12.35062893081761,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01149805270223395,
      "loss": 0.481,
      "num_input_tokens_seen": 15378784,
      "step": 23565
    },
    {
      "epoch": 12.353249475890985,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.011491382085214441,
      "loss": 0.3743,
      "num_input_tokens_seen": 15381856,
      "step": 23570
    },
    {
      "epoch": 12.35587002096436,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.011484712202157776,
      "loss": 0.4587,
      "num_input_tokens_seen": 15385376,
      "step": 23575
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.01147804305445921,
      "loss": 0.3458,
      "num_input_tokens_seen": 15389568,
      "step": 23580
    },
    {
      "epoch": 12.36111111111111,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.011471374643513863,
      "loss": 0.3059,
      "num_input_tokens_seen": 15392160,
      "step": 23585
    },
    {
      "epoch": 12.363731656184486,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.01146470697071668,
      "loss": 0.4685,
      "num_input_tokens_seen": 15395520,
      "step": 23590
    },
    {
      "epoch": 12.366352201257861,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.011458040037462465,
      "loss": 0.4883,
      "num_input_tokens_seen": 15398592,
      "step": 23595
    },
    {
      "epoch": 12.368972746331236,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.011451373845145867,
      "loss": 0.4064,
      "num_input_tokens_seen": 15401824,
      "step": 23600
    },
    {
      "epoch": 12.371593291404611,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.011444708395161372,
      "loss": 0.4122,
      "num_input_tokens_seen": 15405536,
      "step": 23605
    },
    {
      "epoch": 12.374213836477987,
      "grad_norm": 0.06640625,
      "learning_rate": 0.011438043688903323,
      "loss": 0.4384,
      "num_input_tokens_seen": 15408800,
      "step": 23610
    },
    {
      "epoch": 12.376834381551364,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.011431379727765903,
      "loss": 0.4947,
      "num_input_tokens_seen": 15411392,
      "step": 23615
    },
    {
      "epoch": 12.379454926624739,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.011424716513143121,
      "loss": 0.4955,
      "num_input_tokens_seen": 15414144,
      "step": 23620
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.011418054046428861,
      "loss": 0.4769,
      "num_input_tokens_seen": 15417984,
      "step": 23625
    },
    {
      "epoch": 12.384696016771489,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.011411392329016825,
      "loss": 0.3294,
      "num_input_tokens_seen": 15421280,
      "step": 23630
    },
    {
      "epoch": 12.387316561844864,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.011404731362300576,
      "loss": 0.4713,
      "num_input_tokens_seen": 15423872,
      "step": 23635
    },
    {
      "epoch": 12.38993710691824,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.011398071147673515,
      "loss": 0.3675,
      "num_input_tokens_seen": 15426976,
      "step": 23640
    },
    {
      "epoch": 12.392557651991615,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.011391411686528867,
      "loss": 0.3518,
      "num_input_tokens_seen": 15430240,
      "step": 23645
    },
    {
      "epoch": 12.39517819706499,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.011384752980259727,
      "loss": 0.2909,
      "num_input_tokens_seen": 15433664,
      "step": 23650
    },
    {
      "epoch": 12.397798742138365,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.011378095030259022,
      "loss": 0.3952,
      "num_input_tokens_seen": 15437088,
      "step": 23655
    },
    {
      "epoch": 12.40041928721174,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.011371437837919513,
      "loss": 0.4118,
      "num_input_tokens_seen": 15440096,
      "step": 23660
    },
    {
      "epoch": 12.403039832285115,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.011364781404633812,
      "loss": 0.3941,
      "num_input_tokens_seen": 15444352,
      "step": 23665
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 0.0703125,
      "learning_rate": 0.011358125731794374,
      "loss": 0.4071,
      "num_input_tokens_seen": 15448736,
      "step": 23670
    },
    {
      "epoch": 12.408280922431866,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.011351470820793477,
      "loss": 0.3488,
      "num_input_tokens_seen": 15451712,
      "step": 23675
    },
    {
      "epoch": 12.41090146750524,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.011344816673023263,
      "loss": 0.327,
      "num_input_tokens_seen": 15454272,
      "step": 23680
    },
    {
      "epoch": 12.413522012578616,
      "grad_norm": 0.04296875,
      "learning_rate": 0.011338163289875698,
      "loss": 0.378,
      "num_input_tokens_seen": 15457856,
      "step": 23685
    },
    {
      "epoch": 12.416142557651991,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.011331510672742599,
      "loss": 0.3308,
      "num_input_tokens_seen": 15461408,
      "step": 23690
    },
    {
      "epoch": 12.418763102725366,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.011324858823015612,
      "loss": 0.4316,
      "num_input_tokens_seen": 15464992,
      "step": 23695
    },
    {
      "epoch": 12.421383647798741,
      "grad_norm": 0.0546875,
      "learning_rate": 0.011318207742086235,
      "loss": 0.4341,
      "num_input_tokens_seen": 15468096,
      "step": 23700
    },
    {
      "epoch": 12.424004192872117,
      "grad_norm": 0.076171875,
      "learning_rate": 0.011311557431345794,
      "loss": 0.3304,
      "num_input_tokens_seen": 15470752,
      "step": 23705
    },
    {
      "epoch": 12.426624737945493,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.011304907892185455,
      "loss": 0.4216,
      "num_input_tokens_seen": 15473952,
      "step": 23710
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.01129825912599623,
      "loss": 0.3094,
      "num_input_tokens_seen": 15477760,
      "step": 23715
    },
    {
      "epoch": 12.431865828092244,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.011291611134168962,
      "loss": 0.3413,
      "num_input_tokens_seen": 15480384,
      "step": 23720
    },
    {
      "epoch": 12.434486373165619,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.011284963918094346,
      "loss": 0.3882,
      "num_input_tokens_seen": 15482976,
      "step": 23725
    },
    {
      "epoch": 12.437106918238994,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.011278317479162886,
      "loss": 0.3062,
      "num_input_tokens_seen": 15486016,
      "step": 23730
    },
    {
      "epoch": 12.43972746331237,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.011271671818764952,
      "loss": 0.3374,
      "num_input_tokens_seen": 15489472,
      "step": 23735
    },
    {
      "epoch": 12.442348008385745,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.01126502693829074,
      "loss": 0.4777,
      "num_input_tokens_seen": 15493120,
      "step": 23740
    },
    {
      "epoch": 12.44496855345912,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.011258382839130282,
      "loss": 0.3882,
      "num_input_tokens_seen": 15495968,
      "step": 23745
    },
    {
      "epoch": 12.447589098532495,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.011251739522673443,
      "loss": 0.3944,
      "num_input_tokens_seen": 15498784,
      "step": 23750
    },
    {
      "epoch": 12.45020964360587,
      "grad_norm": 0.052734375,
      "learning_rate": 0.01124509699030994,
      "loss": 0.492,
      "num_input_tokens_seen": 15502784,
      "step": 23755
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.011238455243429308,
      "loss": 0.3672,
      "num_input_tokens_seen": 15505952,
      "step": 23760
    },
    {
      "epoch": 12.45545073375262,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.01123181428342092,
      "loss": 0.3739,
      "num_input_tokens_seen": 15509344,
      "step": 23765
    },
    {
      "epoch": 12.458071278825996,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.011225174111673997,
      "loss": 0.4762,
      "num_input_tokens_seen": 15512544,
      "step": 23770
    },
    {
      "epoch": 12.46069182389937,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.011218534729577583,
      "loss": 0.4907,
      "num_input_tokens_seen": 15519104,
      "step": 23775
    },
    {
      "epoch": 12.463312368972746,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.011211896138520564,
      "loss": 0.4202,
      "num_input_tokens_seen": 15522528,
      "step": 23780
    },
    {
      "epoch": 12.465932914046121,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.011205258339891661,
      "loss": 0.4349,
      "num_input_tokens_seen": 15525440,
      "step": 23785
    },
    {
      "epoch": 12.468553459119496,
      "grad_norm": 0.048828125,
      "learning_rate": 0.011198621335079411,
      "loss": 0.3996,
      "num_input_tokens_seen": 15528448,
      "step": 23790
    },
    {
      "epoch": 12.471174004192871,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.011191985125472214,
      "loss": 0.339,
      "num_input_tokens_seen": 15531392,
      "step": 23795
    },
    {
      "epoch": 12.473794549266247,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.011185349712458282,
      "loss": 0.291,
      "num_input_tokens_seen": 15534784,
      "step": 23800
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.011178715097425669,
      "loss": 0.3578,
      "num_input_tokens_seen": 15538144,
      "step": 23805
    },
    {
      "epoch": 12.479035639412999,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.011172081281762265,
      "loss": 0.3281,
      "num_input_tokens_seen": 15541600,
      "step": 23810
    },
    {
      "epoch": 12.481656184486374,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.01116544826685579,
      "loss": 0.492,
      "num_input_tokens_seen": 15545728,
      "step": 23815
    },
    {
      "epoch": 12.484276729559749,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.011158816054093785,
      "loss": 0.3247,
      "num_input_tokens_seen": 15548000,
      "step": 23820
    },
    {
      "epoch": 12.486897274633124,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.011152184644863642,
      "loss": 0.3433,
      "num_input_tokens_seen": 15550496,
      "step": 23825
    },
    {
      "epoch": 12.4895178197065,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01114555404055257,
      "loss": 0.4066,
      "num_input_tokens_seen": 15554304,
      "step": 23830
    },
    {
      "epoch": 12.492138364779874,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01113892424254762,
      "loss": 0.4123,
      "num_input_tokens_seen": 15557824,
      "step": 23835
    },
    {
      "epoch": 12.49475890985325,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.011132295252235674,
      "loss": 0.3846,
      "num_input_tokens_seen": 15560704,
      "step": 23840
    },
    {
      "epoch": 12.497379454926625,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01112566707100343,
      "loss": 0.4256,
      "num_input_tokens_seen": 15564832,
      "step": 23845
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.011119039700237436,
      "loss": 0.3307,
      "num_input_tokens_seen": 15567904,
      "step": 23850
    },
    {
      "epoch": 12.502620545073375,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.011112413141324058,
      "loss": 0.4755,
      "num_input_tokens_seen": 15571168,
      "step": 23855
    },
    {
      "epoch": 12.50524109014675,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.011105787395649502,
      "loss": 0.3268,
      "num_input_tokens_seen": 15574144,
      "step": 23860
    },
    {
      "epoch": 12.507861635220126,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.011099162464599792,
      "loss": 0.4337,
      "num_input_tokens_seen": 15576160,
      "step": 23865
    },
    {
      "epoch": 12.5104821802935,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.011092538349560798,
      "loss": 0.3749,
      "num_input_tokens_seen": 15579424,
      "step": 23870
    },
    {
      "epoch": 12.513102725366876,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.011085915051918199,
      "loss": 0.2509,
      "num_input_tokens_seen": 15584032,
      "step": 23875
    },
    {
      "epoch": 12.515723270440251,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.011079292573057512,
      "loss": 0.4546,
      "num_input_tokens_seen": 15588544,
      "step": 23880
    },
    {
      "epoch": 12.518343815513626,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.011072670914364091,
      "loss": 0.3405,
      "num_input_tokens_seen": 15594336,
      "step": 23885
    },
    {
      "epoch": 12.520964360587001,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.01106605007722311,
      "loss": 0.4347,
      "num_input_tokens_seen": 15597760,
      "step": 23890
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 0.07421875,
      "learning_rate": 0.011059430063019574,
      "loss": 0.3717,
      "num_input_tokens_seen": 15601824,
      "step": 23895
    },
    {
      "epoch": 12.526205450733752,
      "grad_norm": 0.052734375,
      "learning_rate": 0.011052810873138317,
      "loss": 0.4959,
      "num_input_tokens_seen": 15605824,
      "step": 23900
    },
    {
      "epoch": 12.528825995807129,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.011046192508963989,
      "loss": 0.362,
      "num_input_tokens_seen": 15609312,
      "step": 23905
    },
    {
      "epoch": 12.531446540880504,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.011039574971881076,
      "loss": 0.5533,
      "num_input_tokens_seen": 15613216,
      "step": 23910
    },
    {
      "epoch": 12.534067085953879,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.011032958263273902,
      "loss": 0.2987,
      "num_input_tokens_seen": 15616352,
      "step": 23915
    },
    {
      "epoch": 12.536687631027254,
      "grad_norm": 0.04296875,
      "learning_rate": 0.011026342384526594,
      "loss": 0.4083,
      "num_input_tokens_seen": 15619072,
      "step": 23920
    },
    {
      "epoch": 12.53930817610063,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.011019727337023135,
      "loss": 0.366,
      "num_input_tokens_seen": 15622144,
      "step": 23925
    },
    {
      "epoch": 12.541928721174004,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.011013113122147304,
      "loss": 0.3799,
      "num_input_tokens_seen": 15625152,
      "step": 23930
    },
    {
      "epoch": 12.54454926624738,
      "grad_norm": 0.08203125,
      "learning_rate": 0.011006499741282719,
      "loss": 0.443,
      "num_input_tokens_seen": 15628128,
      "step": 23935
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.010999887195812833,
      "loss": 0.4456,
      "num_input_tokens_seen": 15631328,
      "step": 23940
    },
    {
      "epoch": 12.54979035639413,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.010993275487120903,
      "loss": 0.4628,
      "num_input_tokens_seen": 15634304,
      "step": 23945
    },
    {
      "epoch": 12.552410901467505,
      "grad_norm": 0.04296875,
      "learning_rate": 0.010986664616590036,
      "loss": 0.2904,
      "num_input_tokens_seen": 15637984,
      "step": 23950
    },
    {
      "epoch": 12.55503144654088,
      "grad_norm": 0.06640625,
      "learning_rate": 0.010980054585603147,
      "loss": 0.3364,
      "num_input_tokens_seen": 15640672,
      "step": 23955
    },
    {
      "epoch": 12.557651991614255,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.01097344539554297,
      "loss": 0.401,
      "num_input_tokens_seen": 15644096,
      "step": 23960
    },
    {
      "epoch": 12.56027253668763,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.01096683704779208,
      "loss": 0.4325,
      "num_input_tokens_seen": 15647040,
      "step": 23965
    },
    {
      "epoch": 12.562893081761006,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.01096022954373287,
      "loss": 0.3691,
      "num_input_tokens_seen": 15649824,
      "step": 23970
    },
    {
      "epoch": 12.565513626834381,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.010953622884747546,
      "loss": 0.3081,
      "num_input_tokens_seen": 15652992,
      "step": 23975
    },
    {
      "epoch": 12.568134171907756,
      "grad_norm": 0.087890625,
      "learning_rate": 0.010947017072218157,
      "loss": 0.388,
      "num_input_tokens_seen": 15655680,
      "step": 23980
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.01094041210752656,
      "loss": 0.3025,
      "num_input_tokens_seen": 15658496,
      "step": 23985
    },
    {
      "epoch": 12.573375262054507,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.010933807992054428,
      "loss": 0.536,
      "num_input_tokens_seen": 15661248,
      "step": 23990
    },
    {
      "epoch": 12.575995807127882,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.01092720472718328,
      "loss": 0.4083,
      "num_input_tokens_seen": 15663840,
      "step": 23995
    },
    {
      "epoch": 12.578616352201259,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.010920602314294432,
      "loss": 0.4531,
      "num_input_tokens_seen": 15667744,
      "step": 24000
    },
    {
      "epoch": 12.581236897274634,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.010914000754769048,
      "loss": 0.4302,
      "num_input_tokens_seen": 15670592,
      "step": 24005
    },
    {
      "epoch": 12.583857442348009,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.01090740004998809,
      "loss": 0.4193,
      "num_input_tokens_seen": 15674688,
      "step": 24010
    },
    {
      "epoch": 12.586477987421384,
      "grad_norm": 0.0859375,
      "learning_rate": 0.010900800201332344,
      "loss": 0.3971,
      "num_input_tokens_seen": 15677440,
      "step": 24015
    },
    {
      "epoch": 12.58909853249476,
      "grad_norm": 0.072265625,
      "learning_rate": 0.010894201210182433,
      "loss": 0.4301,
      "num_input_tokens_seen": 15680416,
      "step": 24020
    },
    {
      "epoch": 12.591719077568134,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.010887603077918785,
      "loss": 0.4079,
      "num_input_tokens_seen": 15684800,
      "step": 24025
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01088100580592166,
      "loss": 0.4383,
      "num_input_tokens_seen": 15687936,
      "step": 24030
    },
    {
      "epoch": 12.596960167714885,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.010874409395571124,
      "loss": 0.2863,
      "num_input_tokens_seen": 15691104,
      "step": 24035
    },
    {
      "epoch": 12.59958071278826,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.010867813848247082,
      "loss": 0.452,
      "num_input_tokens_seen": 15694432,
      "step": 24040
    },
    {
      "epoch": 12.602201257861635,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.010861219165329235,
      "loss": 0.4483,
      "num_input_tokens_seen": 15697504,
      "step": 24045
    },
    {
      "epoch": 12.60482180293501,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.010854625348197122,
      "loss": 0.4384,
      "num_input_tokens_seen": 15701824,
      "step": 24050
    },
    {
      "epoch": 12.607442348008385,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010848032398230092,
      "loss": 0.3762,
      "num_input_tokens_seen": 15704736,
      "step": 24055
    },
    {
      "epoch": 12.61006289308176,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.01084144031680732,
      "loss": 0.325,
      "num_input_tokens_seen": 15707744,
      "step": 24060
    },
    {
      "epoch": 12.612683438155136,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.010834849105307784,
      "loss": 0.4623,
      "num_input_tokens_seen": 15710208,
      "step": 24065
    },
    {
      "epoch": 12.615303983228511,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.010828258765110305,
      "loss": 0.4108,
      "num_input_tokens_seen": 15713792,
      "step": 24070
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 0.068359375,
      "learning_rate": 0.010821669297593497,
      "loss": 0.4006,
      "num_input_tokens_seen": 15717792,
      "step": 24075
    },
    {
      "epoch": 12.620545073375261,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.010815080704135798,
      "loss": 0.3921,
      "num_input_tokens_seen": 15720896,
      "step": 24080
    },
    {
      "epoch": 12.623165618448636,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.010808492986115475,
      "loss": 0.2725,
      "num_input_tokens_seen": 15723904,
      "step": 24085
    },
    {
      "epoch": 12.625786163522012,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.010801906144910597,
      "loss": 0.4139,
      "num_input_tokens_seen": 15727168,
      "step": 24090
    },
    {
      "epoch": 12.628406708595389,
      "grad_norm": 0.0625,
      "learning_rate": 0.010795320181899066,
      "loss": 0.5227,
      "num_input_tokens_seen": 15730688,
      "step": 24095
    },
    {
      "epoch": 12.631027253668764,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.010788735098458579,
      "loss": 0.4784,
      "num_input_tokens_seen": 15733664,
      "step": 24100
    },
    {
      "epoch": 12.633647798742139,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.010782150895966663,
      "loss": 0.4496,
      "num_input_tokens_seen": 15736352,
      "step": 24105
    },
    {
      "epoch": 12.636268343815514,
      "grad_norm": 0.07421875,
      "learning_rate": 0.01077556757580066,
      "loss": 0.3211,
      "num_input_tokens_seen": 15739648,
      "step": 24110
    },
    {
      "epoch": 12.63888888888889,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.010768985139337725,
      "loss": 0.411,
      "num_input_tokens_seen": 15742464,
      "step": 24115
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.01076240358795483,
      "loss": 0.4379,
      "num_input_tokens_seen": 15745984,
      "step": 24120
    },
    {
      "epoch": 12.64412997903564,
      "grad_norm": 0.06640625,
      "learning_rate": 0.010755822923028761,
      "loss": 0.3299,
      "num_input_tokens_seen": 15748608,
      "step": 24125
    },
    {
      "epoch": 12.646750524109015,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.010749243145936116,
      "loss": 0.4228,
      "num_input_tokens_seen": 15751520,
      "step": 24130
    },
    {
      "epoch": 12.64937106918239,
      "grad_norm": 0.037109375,
      "learning_rate": 0.010742664258053309,
      "loss": 0.3794,
      "num_input_tokens_seen": 15754912,
      "step": 24135
    },
    {
      "epoch": 12.651991614255765,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.010736086260756568,
      "loss": 0.5372,
      "num_input_tokens_seen": 15758688,
      "step": 24140
    },
    {
      "epoch": 12.65461215932914,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.010729509155421934,
      "loss": 0.2757,
      "num_input_tokens_seen": 15761248,
      "step": 24145
    },
    {
      "epoch": 12.657232704402515,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.010722932943425268,
      "loss": 0.5333,
      "num_input_tokens_seen": 15765088,
      "step": 24150
    },
    {
      "epoch": 12.65985324947589,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.010716357626142238,
      "loss": 0.4213,
      "num_input_tokens_seen": 15768896,
      "step": 24155
    },
    {
      "epoch": 12.662473794549266,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.010709783204948317,
      "loss": 0.3349,
      "num_input_tokens_seen": 15771584,
      "step": 24160
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 0.07421875,
      "learning_rate": 0.010703209681218805,
      "loss": 0.4922,
      "num_input_tokens_seen": 15774080,
      "step": 24165
    },
    {
      "epoch": 12.667714884696016,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010696637056328806,
      "loss": 0.5381,
      "num_input_tokens_seen": 15777280,
      "step": 24170
    },
    {
      "epoch": 12.670335429769391,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.010690065331653242,
      "loss": 0.5084,
      "num_input_tokens_seen": 15780544,
      "step": 24175
    },
    {
      "epoch": 12.672955974842766,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.010683494508566838,
      "loss": 0.4752,
      "num_input_tokens_seen": 15783680,
      "step": 24180
    },
    {
      "epoch": 12.675576519916142,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.010676924588444143,
      "loss": 0.3443,
      "num_input_tokens_seen": 15786752,
      "step": 24185
    },
    {
      "epoch": 12.678197064989519,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.010670355572659499,
      "loss": 0.3362,
      "num_input_tokens_seen": 15790368,
      "step": 24190
    },
    {
      "epoch": 12.680817610062894,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.010663787462587074,
      "loss": 0.4244,
      "num_input_tokens_seen": 15793120,
      "step": 24195
    },
    {
      "epoch": 12.683438155136269,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.01065722025960084,
      "loss": 0.3232,
      "num_input_tokens_seen": 15797600,
      "step": 24200
    },
    {
      "epoch": 12.686058700209644,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.010650653965074586,
      "loss": 0.4194,
      "num_input_tokens_seen": 15802304,
      "step": 24205
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.010644088580381905,
      "loss": 0.3756,
      "num_input_tokens_seen": 15804672,
      "step": 24210
    },
    {
      "epoch": 12.691299790356394,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.010637524106896192,
      "loss": 0.3605,
      "num_input_tokens_seen": 15807168,
      "step": 24215
    },
    {
      "epoch": 12.69392033542977,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.01063096054599067,
      "loss": 0.2757,
      "num_input_tokens_seen": 15809824,
      "step": 24220
    },
    {
      "epoch": 12.696540880503145,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.010624397899038353,
      "loss": 0.4976,
      "num_input_tokens_seen": 15813408,
      "step": 24225
    },
    {
      "epoch": 12.69916142557652,
      "grad_norm": 0.091796875,
      "learning_rate": 0.01061783616741208,
      "loss": 0.4582,
      "num_input_tokens_seen": 15816096,
      "step": 24230
    },
    {
      "epoch": 12.701781970649895,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.010611275352484484,
      "loss": 0.2998,
      "num_input_tokens_seen": 15819328,
      "step": 24235
    },
    {
      "epoch": 12.70440251572327,
      "grad_norm": 0.177734375,
      "learning_rate": 0.010604715455628019,
      "loss": 0.3567,
      "num_input_tokens_seen": 15822176,
      "step": 24240
    },
    {
      "epoch": 12.707023060796645,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.010598156478214937,
      "loss": 0.5989,
      "num_input_tokens_seen": 15824768,
      "step": 24245
    },
    {
      "epoch": 12.70964360587002,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.010591598421617296,
      "loss": 0.4924,
      "num_input_tokens_seen": 15828608,
      "step": 24250
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.010585041287206973,
      "loss": 0.3993,
      "num_input_tokens_seen": 15830880,
      "step": 24255
    },
    {
      "epoch": 12.714884696016771,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.010578485076355643,
      "loss": 0.4237,
      "num_input_tokens_seen": 15834112,
      "step": 24260
    },
    {
      "epoch": 12.717505241090146,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.010571929790434791,
      "loss": 0.2809,
      "num_input_tokens_seen": 15836832,
      "step": 24265
    },
    {
      "epoch": 12.720125786163521,
      "grad_norm": 0.07421875,
      "learning_rate": 0.010565375430815717,
      "loss": 0.3364,
      "num_input_tokens_seen": 15839840,
      "step": 24270
    },
    {
      "epoch": 12.722746331236896,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.010558821998869504,
      "loss": 0.3879,
      "num_input_tokens_seen": 15842848,
      "step": 24275
    },
    {
      "epoch": 12.725366876310272,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.010552269495967059,
      "loss": 0.4442,
      "num_input_tokens_seen": 15846496,
      "step": 24280
    },
    {
      "epoch": 12.727987421383649,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.010545717923479097,
      "loss": 0.4307,
      "num_input_tokens_seen": 15852832,
      "step": 24285
    },
    {
      "epoch": 12.730607966457024,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.010539167282776122,
      "loss": 0.3584,
      "num_input_tokens_seen": 15856416,
      "step": 24290
    },
    {
      "epoch": 12.733228511530399,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.010532617575228468,
      "loss": 0.4351,
      "num_input_tokens_seen": 15859104,
      "step": 24295
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.010526068802206249,
      "loss": 0.3106,
      "num_input_tokens_seen": 15861792,
      "step": 24300
    },
    {
      "epoch": 12.73846960167715,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.010519520965079389,
      "loss": 0.4545,
      "num_input_tokens_seen": 15864320,
      "step": 24305
    },
    {
      "epoch": 12.741090146750524,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.010512974065217633,
      "loss": 0.2607,
      "num_input_tokens_seen": 15867168,
      "step": 24310
    },
    {
      "epoch": 12.7437106918239,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.010506428103990509,
      "loss": 0.3834,
      "num_input_tokens_seen": 15870048,
      "step": 24315
    },
    {
      "epoch": 12.746331236897275,
      "grad_norm": 0.078125,
      "learning_rate": 0.01049988308276736,
      "loss": 0.3775,
      "num_input_tokens_seen": 15873024,
      "step": 24320
    },
    {
      "epoch": 12.74895178197065,
      "grad_norm": 0.0625,
      "learning_rate": 0.010493339002917337,
      "loss": 0.3833,
      "num_input_tokens_seen": 15875648,
      "step": 24325
    },
    {
      "epoch": 12.751572327044025,
      "grad_norm": 0.0625,
      "learning_rate": 0.010486795865809372,
      "loss": 0.3071,
      "num_input_tokens_seen": 15878240,
      "step": 24330
    },
    {
      "epoch": 12.7541928721174,
      "grad_norm": 0.06640625,
      "learning_rate": 0.010480253672812226,
      "loss": 0.4557,
      "num_input_tokens_seen": 15881152,
      "step": 24335
    },
    {
      "epoch": 12.756813417190775,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.010473712425294443,
      "loss": 0.403,
      "num_input_tokens_seen": 15884384,
      "step": 24340
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.010467172124624382,
      "loss": 0.5083,
      "num_input_tokens_seen": 15887232,
      "step": 24345
    },
    {
      "epoch": 12.762054507337526,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.010460632772170205,
      "loss": 0.3387,
      "num_input_tokens_seen": 15889344,
      "step": 24350
    },
    {
      "epoch": 12.764675052410901,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.010454094369299867,
      "loss": 0.5488,
      "num_input_tokens_seen": 15892704,
      "step": 24355
    },
    {
      "epoch": 12.767295597484276,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.010447556917381116,
      "loss": 0.6313,
      "num_input_tokens_seen": 15895840,
      "step": 24360
    },
    {
      "epoch": 12.769916142557651,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010441020417781525,
      "loss": 0.3468,
      "num_input_tokens_seen": 15898720,
      "step": 24365
    },
    {
      "epoch": 12.772536687631026,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.010434484871868448,
      "loss": 0.2765,
      "num_input_tokens_seen": 15901408,
      "step": 24370
    },
    {
      "epoch": 12.775157232704402,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.010427950281009055,
      "loss": 0.4063,
      "num_input_tokens_seen": 15904160,
      "step": 24375
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.010421416646570302,
      "loss": 0.4208,
      "num_input_tokens_seen": 15907488,
      "step": 24380
    },
    {
      "epoch": 12.780398322851154,
      "grad_norm": 0.044921875,
      "learning_rate": 0.010414883969918947,
      "loss": 0.3572,
      "num_input_tokens_seen": 15910784,
      "step": 24385
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.010408352252421559,
      "loss": 0.4368,
      "num_input_tokens_seen": 15914016,
      "step": 24390
    },
    {
      "epoch": 12.785639412997904,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.010401821495444491,
      "loss": 0.4122,
      "num_input_tokens_seen": 15916672,
      "step": 24395
    },
    {
      "epoch": 12.78825995807128,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.010395291700353911,
      "loss": 0.392,
      "num_input_tokens_seen": 15920288,
      "step": 24400
    },
    {
      "epoch": 12.790880503144654,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.010388762868515772,
      "loss": 0.5373,
      "num_input_tokens_seen": 15924736,
      "step": 24405
    },
    {
      "epoch": 12.79350104821803,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.010382235001295839,
      "loss": 0.3439,
      "num_input_tokens_seen": 15928544,
      "step": 24410
    },
    {
      "epoch": 12.796121593291405,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.010375708100059658,
      "loss": 0.2882,
      "num_input_tokens_seen": 15932864,
      "step": 24415
    },
    {
      "epoch": 12.79874213836478,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.010369182166172582,
      "loss": 0.36,
      "num_input_tokens_seen": 15936512,
      "step": 24420
    },
    {
      "epoch": 12.801362683438155,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.010362657200999765,
      "loss": 0.4261,
      "num_input_tokens_seen": 15939392,
      "step": 24425
    },
    {
      "epoch": 12.80398322851153,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.01035613320590616,
      "loss": 0.3744,
      "num_input_tokens_seen": 15942272,
      "step": 24430
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.010349610182256507,
      "loss": 0.4743,
      "num_input_tokens_seen": 15945280,
      "step": 24435
    },
    {
      "epoch": 12.80922431865828,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.010343088131415355,
      "loss": 0.4026,
      "num_input_tokens_seen": 15948704,
      "step": 24440
    },
    {
      "epoch": 12.811844863731656,
      "grad_norm": 0.05078125,
      "learning_rate": 0.010336567054747031,
      "loss": 0.359,
      "num_input_tokens_seen": 15952320,
      "step": 24445
    },
    {
      "epoch": 12.814465408805031,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.010330046953615677,
      "loss": 0.3399,
      "num_input_tokens_seen": 15956576,
      "step": 24450
    },
    {
      "epoch": 12.817085953878406,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.010323527829385221,
      "loss": 0.3718,
      "num_input_tokens_seen": 15959232,
      "step": 24455
    },
    {
      "epoch": 12.819706498951781,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.010317009683419393,
      "loss": 0.3737,
      "num_input_tokens_seen": 15962720,
      "step": 24460
    },
    {
      "epoch": 12.822327044025156,
      "grad_norm": 0.046875,
      "learning_rate": 0.010310492517081715,
      "loss": 0.3603,
      "num_input_tokens_seen": 15966240,
      "step": 24465
    },
    {
      "epoch": 12.824947589098532,
      "grad_norm": 0.06640625,
      "learning_rate": 0.010303976331735501,
      "loss": 0.5069,
      "num_input_tokens_seen": 15968800,
      "step": 24470
    },
    {
      "epoch": 12.827568134171909,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.010297461128743858,
      "loss": 0.4375,
      "num_input_tokens_seen": 15971680,
      "step": 24475
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.010290946909469703,
      "loss": 0.4647,
      "num_input_tokens_seen": 15975552,
      "step": 24480
    },
    {
      "epoch": 12.832809224318659,
      "grad_norm": 0.08203125,
      "learning_rate": 0.010284433675275727,
      "loss": 0.4114,
      "num_input_tokens_seen": 15978656,
      "step": 24485
    },
    {
      "epoch": 12.835429769392034,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.010277921427524431,
      "loss": 0.4155,
      "num_input_tokens_seen": 15981856,
      "step": 24490
    },
    {
      "epoch": 12.83805031446541,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.010271410167578102,
      "loss": 0.4952,
      "num_input_tokens_seen": 15984736,
      "step": 24495
    },
    {
      "epoch": 12.840670859538784,
      "grad_norm": 0.068359375,
      "learning_rate": 0.010264899896798815,
      "loss": 0.3449,
      "num_input_tokens_seen": 15987424,
      "step": 24500
    },
    {
      "epoch": 12.84329140461216,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.010258390616548449,
      "loss": 0.637,
      "num_input_tokens_seen": 15991296,
      "step": 24505
    },
    {
      "epoch": 12.845911949685535,
      "grad_norm": 0.064453125,
      "learning_rate": 0.010251882328188671,
      "loss": 0.5707,
      "num_input_tokens_seen": 15993760,
      "step": 24510
    },
    {
      "epoch": 12.84853249475891,
      "grad_norm": 0.064453125,
      "learning_rate": 0.01024537503308094,
      "loss": 0.2892,
      "num_input_tokens_seen": 15996288,
      "step": 24515
    },
    {
      "epoch": 12.851153039832285,
      "grad_norm": 0.052734375,
      "learning_rate": 0.010238868732586512,
      "loss": 0.4451,
      "num_input_tokens_seen": 15999328,
      "step": 24520
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.010232363428066429,
      "loss": 0.3778,
      "num_input_tokens_seen": 16002336,
      "step": 24525
    },
    {
      "epoch": 12.856394129979035,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.010225859120881519,
      "loss": 0.2696,
      "num_input_tokens_seen": 16005440,
      "step": 24530
    },
    {
      "epoch": 12.85901467505241,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.010219355812392417,
      "loss": 0.5093,
      "num_input_tokens_seen": 16009376,
      "step": 24535
    },
    {
      "epoch": 12.861635220125786,
      "grad_norm": 0.083984375,
      "learning_rate": 0.010212853503959535,
      "loss": 0.3837,
      "num_input_tokens_seen": 16012768,
      "step": 24540
    },
    {
      "epoch": 12.864255765199161,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.01020635219694309,
      "loss": 0.3664,
      "num_input_tokens_seen": 16015136,
      "step": 24545
    },
    {
      "epoch": 12.866876310272536,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.010199851892703074,
      "loss": 0.4279,
      "num_input_tokens_seen": 16018688,
      "step": 24550
    },
    {
      "epoch": 12.869496855345911,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.010193352592599283,
      "loss": 0.6124,
      "num_input_tokens_seen": 16021504,
      "step": 24555
    },
    {
      "epoch": 12.872117400419286,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.010186854297991292,
      "loss": 0.3742,
      "num_input_tokens_seen": 16024736,
      "step": 24560
    },
    {
      "epoch": 12.874737945492662,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.010180357010238466,
      "loss": 0.4105,
      "num_input_tokens_seen": 16027296,
      "step": 24565
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01017386073069997,
      "loss": 0.3289,
      "num_input_tokens_seen": 16029920,
      "step": 24570
    },
    {
      "epoch": 12.879979035639414,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.010167365460734748,
      "loss": 0.3277,
      "num_input_tokens_seen": 16032672,
      "step": 24575
    },
    {
      "epoch": 12.882599580712789,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.010160871201701544,
      "loss": 0.2803,
      "num_input_tokens_seen": 16035040,
      "step": 24580
    },
    {
      "epoch": 12.885220125786164,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.010154377954958872,
      "loss": 0.3512,
      "num_input_tokens_seen": 16040064,
      "step": 24585
    },
    {
      "epoch": 12.88784067085954,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010147885721865053,
      "loss": 0.3466,
      "num_input_tokens_seen": 16042784,
      "step": 24590
    },
    {
      "epoch": 12.890461215932914,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01014139450377818,
      "loss": 0.3389,
      "num_input_tokens_seen": 16045952,
      "step": 24595
    },
    {
      "epoch": 12.89308176100629,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.01013490430205615,
      "loss": 0.4068,
      "num_input_tokens_seen": 16049248,
      "step": 24600
    },
    {
      "epoch": 12.895702306079665,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.010128415118056634,
      "loss": 0.4567,
      "num_input_tokens_seen": 16052576,
      "step": 24605
    },
    {
      "epoch": 12.89832285115304,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.010121926953137104,
      "loss": 0.6095,
      "num_input_tokens_seen": 16055936,
      "step": 24610
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0101154398086548,
      "loss": 0.4811,
      "num_input_tokens_seen": 16058176,
      "step": 24615
    },
    {
      "epoch": 12.90356394129979,
      "grad_norm": 0.06640625,
      "learning_rate": 0.010108953685966758,
      "loss": 0.4447,
      "num_input_tokens_seen": 16061568,
      "step": 24620
    },
    {
      "epoch": 12.906184486373165,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.010102468586429807,
      "loss": 0.4656,
      "num_input_tokens_seen": 16065536,
      "step": 24625
    },
    {
      "epoch": 12.90880503144654,
      "grad_norm": 0.05078125,
      "learning_rate": 0.010095984511400554,
      "loss": 0.4275,
      "num_input_tokens_seen": 16068192,
      "step": 24630
    },
    {
      "epoch": 12.911425576519916,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.010089501462235393,
      "loss": 0.4323,
      "num_input_tokens_seen": 16071008,
      "step": 24635
    },
    {
      "epoch": 12.914046121593291,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010083019440290511,
      "loss": 0.3989,
      "num_input_tokens_seen": 16074688,
      "step": 24640
    },
    {
      "epoch": 12.916666666666666,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.010076538446921858,
      "loss": 0.3571,
      "num_input_tokens_seen": 16076928,
      "step": 24645
    },
    {
      "epoch": 12.919287211740041,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.010070058483485194,
      "loss": 0.487,
      "num_input_tokens_seen": 16079904,
      "step": 24650
    },
    {
      "epoch": 12.921907756813416,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01006357955133605,
      "loss": 0.3077,
      "num_input_tokens_seen": 16083168,
      "step": 24655
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.010057101651829745,
      "loss": 0.5184,
      "num_input_tokens_seen": 16086592,
      "step": 24660
    },
    {
      "epoch": 12.927148846960169,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.010050624786321391,
      "loss": 0.4367,
      "num_input_tokens_seen": 16089088,
      "step": 24665
    },
    {
      "epoch": 12.929769392033544,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.010044148956165861,
      "loss": 0.7171,
      "num_input_tokens_seen": 16091936,
      "step": 24670
    },
    {
      "epoch": 12.932389937106919,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.010037674162717827,
      "loss": 0.39,
      "num_input_tokens_seen": 16094336,
      "step": 24675
    },
    {
      "epoch": 12.935010482180294,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.01003120040733175,
      "loss": 0.3998,
      "num_input_tokens_seen": 16098304,
      "step": 24680
    },
    {
      "epoch": 12.93763102725367,
      "grad_norm": 0.048828125,
      "learning_rate": 0.010024727691361856,
      "loss": 0.3989,
      "num_input_tokens_seen": 16101632,
      "step": 24685
    },
    {
      "epoch": 12.940251572327044,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.010018256016162172,
      "loss": 0.3451,
      "num_input_tokens_seen": 16105024,
      "step": 24690
    },
    {
      "epoch": 12.94287211740042,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.010011785383086499,
      "loss": 0.3312,
      "num_input_tokens_seen": 16108096,
      "step": 24695
    },
    {
      "epoch": 12.945492662473795,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.010005315793488406,
      "loss": 0.4173,
      "num_input_tokens_seen": 16111648,
      "step": 24700
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.009998847248721271,
      "loss": 0.4593,
      "num_input_tokens_seen": 16114752,
      "step": 24705
    },
    {
      "epoch": 12.950733752620545,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.009992379750138234,
      "loss": 0.4857,
      "num_input_tokens_seen": 16119616,
      "step": 24710
    },
    {
      "epoch": 12.95335429769392,
      "grad_norm": 0.052734375,
      "learning_rate": 0.009985913299092224,
      "loss": 0.45,
      "num_input_tokens_seen": 16123232,
      "step": 24715
    },
    {
      "epoch": 12.955974842767295,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.009979447896935946,
      "loss": 0.4262,
      "num_input_tokens_seen": 16127552,
      "step": 24720
    },
    {
      "epoch": 12.95859538784067,
      "grad_norm": 0.078125,
      "learning_rate": 0.009972983545021897,
      "loss": 0.3832,
      "num_input_tokens_seen": 16130368,
      "step": 24725
    },
    {
      "epoch": 12.961215932914046,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.009966520244702336,
      "loss": 0.3639,
      "num_input_tokens_seen": 16134208,
      "step": 24730
    },
    {
      "epoch": 12.963836477987421,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00996005799732931,
      "loss": 0.3575,
      "num_input_tokens_seen": 16137120,
      "step": 24735
    },
    {
      "epoch": 12.966457023060796,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.009953596804254655,
      "loss": 0.4472,
      "num_input_tokens_seen": 16140576,
      "step": 24740
    },
    {
      "epoch": 12.969077568134171,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.009947136666829977,
      "loss": 0.4448,
      "num_input_tokens_seen": 16143808,
      "step": 24745
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.009940677586406666,
      "loss": 0.4808,
      "num_input_tokens_seen": 16147136,
      "step": 24750
    },
    {
      "epoch": 12.974318658280922,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00993421956433588,
      "loss": 0.3844,
      "num_input_tokens_seen": 16150272,
      "step": 24755
    },
    {
      "epoch": 12.976939203354299,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.009927762601968568,
      "loss": 0.4548,
      "num_input_tokens_seen": 16153920,
      "step": 24760
    },
    {
      "epoch": 12.979559748427674,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00992130670065545,
      "loss": 0.3384,
      "num_input_tokens_seen": 16157280,
      "step": 24765
    },
    {
      "epoch": 12.982180293501049,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.009914851861747032,
      "loss": 0.4202,
      "num_input_tokens_seen": 16161056,
      "step": 24770
    },
    {
      "epoch": 12.984800838574424,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00990839808659359,
      "loss": 0.3966,
      "num_input_tokens_seen": 16163776,
      "step": 24775
    },
    {
      "epoch": 12.9874213836478,
      "grad_norm": 0.0625,
      "learning_rate": 0.009901945376545184,
      "loss": 0.3582,
      "num_input_tokens_seen": 16166176,
      "step": 24780
    },
    {
      "epoch": 12.990041928721174,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.00989549373295164,
      "loss": 0.3311,
      "num_input_tokens_seen": 16168704,
      "step": 24785
    },
    {
      "epoch": 12.99266247379455,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.00988904315716257,
      "loss": 0.4078,
      "num_input_tokens_seen": 16171584,
      "step": 24790
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.009882593650527365,
      "loss": 0.4859,
      "num_input_tokens_seen": 16174656,
      "step": 24795
    },
    {
      "epoch": 12.9979035639413,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.009876145214395183,
      "loss": 0.5209,
      "num_input_tokens_seen": 16177344,
      "step": 24800
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.4719694256782532,
      "eval_runtime": 13.6425,
      "eval_samples_per_second": 62.159,
      "eval_steps_per_second": 15.54,
      "num_input_tokens_seen": 16179624,
      "step": 24804
    },
    {
      "epoch": 13.000524109014675,
      "grad_norm": 0.078125,
      "learning_rate": 0.00986969785011497,
      "loss": 0.4761,
      "num_input_tokens_seen": 16180040,
      "step": 24805
    },
    {
      "epoch": 13.00314465408805,
      "grad_norm": 0.111328125,
      "learning_rate": 0.009863251559035438,
      "loss": 0.4057,
      "num_input_tokens_seen": 16183176,
      "step": 24810
    },
    {
      "epoch": 13.005765199161425,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.009856806342505074,
      "loss": 0.3571,
      "num_input_tokens_seen": 16185928,
      "step": 24815
    },
    {
      "epoch": 13.0083857442348,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.009850362201872142,
      "loss": 0.456,
      "num_input_tokens_seen": 16189160,
      "step": 24820
    },
    {
      "epoch": 13.011006289308176,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.009843919138484694,
      "loss": 0.2573,
      "num_input_tokens_seen": 16192392,
      "step": 24825
    },
    {
      "epoch": 13.01362683438155,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.009837477153690535,
      "loss": 0.4064,
      "num_input_tokens_seen": 16195752,
      "step": 24830
    },
    {
      "epoch": 13.016247379454926,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.00983103624883726,
      "loss": 0.3345,
      "num_input_tokens_seen": 16199176,
      "step": 24835
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.009824596425272234,
      "loss": 0.4588,
      "num_input_tokens_seen": 16201768,
      "step": 24840
    },
    {
      "epoch": 13.021488469601676,
      "grad_norm": 0.0703125,
      "learning_rate": 0.009818157684342587,
      "loss": 0.3481,
      "num_input_tokens_seen": 16204680,
      "step": 24845
    },
    {
      "epoch": 13.024109014675052,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.009811720027395237,
      "loss": 0.3821,
      "num_input_tokens_seen": 16208264,
      "step": 24850
    },
    {
      "epoch": 13.026729559748428,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.009805283455776862,
      "loss": 0.2894,
      "num_input_tokens_seen": 16210536,
      "step": 24855
    },
    {
      "epoch": 13.029350104821804,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.009798847970833928,
      "loss": 0.5319,
      "num_input_tokens_seen": 16214376,
      "step": 24860
    },
    {
      "epoch": 13.031970649895179,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.009792413573912664,
      "loss": 0.4928,
      "num_input_tokens_seen": 16217416,
      "step": 24865
    },
    {
      "epoch": 13.034591194968554,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.009785980266359063,
      "loss": 0.4142,
      "num_input_tokens_seen": 16220584,
      "step": 24870
    },
    {
      "epoch": 13.03721174004193,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.009779548049518909,
      "loss": 0.3191,
      "num_input_tokens_seen": 16224040,
      "step": 24875
    },
    {
      "epoch": 13.039832285115304,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.009773116924737741,
      "loss": 0.4006,
      "num_input_tokens_seen": 16227016,
      "step": 24880
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.009766686893360884,
      "loss": 0.3999,
      "num_input_tokens_seen": 16229768,
      "step": 24885
    },
    {
      "epoch": 13.045073375262055,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.009760257956733421,
      "loss": 0.2949,
      "num_input_tokens_seen": 16233224,
      "step": 24890
    },
    {
      "epoch": 13.04769392033543,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.009753830116200221,
      "loss": 0.308,
      "num_input_tokens_seen": 16236264,
      "step": 24895
    },
    {
      "epoch": 13.050314465408805,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.009747403373105904,
      "loss": 0.3521,
      "num_input_tokens_seen": 16238984,
      "step": 24900
    },
    {
      "epoch": 13.05293501048218,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.00974097772879488,
      "loss": 0.3327,
      "num_input_tokens_seen": 16241608,
      "step": 24905
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.009734553184611314,
      "loss": 0.3833,
      "num_input_tokens_seen": 16244328,
      "step": 24910
    },
    {
      "epoch": 13.05817610062893,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.009728129741899152,
      "loss": 0.4315,
      "num_input_tokens_seen": 16247912,
      "step": 24915
    },
    {
      "epoch": 13.060796645702306,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.009721707402002101,
      "loss": 0.4221,
      "num_input_tokens_seen": 16250920,
      "step": 24920
    },
    {
      "epoch": 13.06341719077568,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00971528616626365,
      "loss": 0.4261,
      "num_input_tokens_seen": 16254728,
      "step": 24925
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.009708866036027037,
      "loss": 0.3147,
      "num_input_tokens_seen": 16257640,
      "step": 24930
    },
    {
      "epoch": 13.068658280922431,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.009702447012635285,
      "loss": 0.4527,
      "num_input_tokens_seen": 16260808,
      "step": 24935
    },
    {
      "epoch": 13.071278825995806,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.009696029097431181,
      "loss": 0.3252,
      "num_input_tokens_seen": 16264264,
      "step": 24940
    },
    {
      "epoch": 13.073899371069182,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.009689612291757278,
      "loss": 0.3512,
      "num_input_tokens_seen": 16267528,
      "step": 24945
    },
    {
      "epoch": 13.076519916142558,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.009683196596955905,
      "loss": 0.4974,
      "num_input_tokens_seen": 16270088,
      "step": 24950
    },
    {
      "epoch": 13.079140461215934,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.00967678201436914,
      "loss": 0.4386,
      "num_input_tokens_seen": 16273320,
      "step": 24955
    },
    {
      "epoch": 13.081761006289309,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.009670368545338847,
      "loss": 0.3648,
      "num_input_tokens_seen": 16276232,
      "step": 24960
    },
    {
      "epoch": 13.084381551362684,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.009663956191206653,
      "loss": 0.4506,
      "num_input_tokens_seen": 16279272,
      "step": 24965
    },
    {
      "epoch": 13.08700209643606,
      "grad_norm": 0.052734375,
      "learning_rate": 0.009657544953313944,
      "loss": 0.3519,
      "num_input_tokens_seen": 16282024,
      "step": 24970
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.009651134833001883,
      "loss": 0.3834,
      "num_input_tokens_seen": 16284808,
      "step": 24975
    },
    {
      "epoch": 13.09224318658281,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.009644725831611394,
      "loss": 0.5725,
      "num_input_tokens_seen": 16288200,
      "step": 24980
    },
    {
      "epoch": 13.094863731656185,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.009638317950483167,
      "loss": 0.349,
      "num_input_tokens_seen": 16291176,
      "step": 24985
    },
    {
      "epoch": 13.09748427672956,
      "grad_norm": 0.078125,
      "learning_rate": 0.009631911190957647,
      "loss": 0.4177,
      "num_input_tokens_seen": 16293896,
      "step": 24990
    },
    {
      "epoch": 13.100104821802935,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.009625505554375068,
      "loss": 0.3315,
      "num_input_tokens_seen": 16297192,
      "step": 24995
    },
    {
      "epoch": 13.10272536687631,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.009619101042075409,
      "loss": 0.4284,
      "num_input_tokens_seen": 16300680,
      "step": 25000
    },
    {
      "epoch": 13.105345911949685,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.009612697655398428,
      "loss": 0.2006,
      "num_input_tokens_seen": 16303080,
      "step": 25005
    },
    {
      "epoch": 13.10796645702306,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.009606295395683638,
      "loss": 0.3655,
      "num_input_tokens_seen": 16305224,
      "step": 25010
    },
    {
      "epoch": 13.110587002096436,
      "grad_norm": 0.068359375,
      "learning_rate": 0.00959989426427031,
      "loss": 0.4905,
      "num_input_tokens_seen": 16308168,
      "step": 25015
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 0.041015625,
      "learning_rate": 0.009593494262497497,
      "loss": 0.4118,
      "num_input_tokens_seen": 16312072,
      "step": 25020
    },
    {
      "epoch": 13.115828092243186,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.009587095391704003,
      "loss": 0.3956,
      "num_input_tokens_seen": 16315112,
      "step": 25025
    },
    {
      "epoch": 13.118448637316561,
      "grad_norm": 0.07421875,
      "learning_rate": 0.009580697653228398,
      "loss": 0.4879,
      "num_input_tokens_seen": 16317704,
      "step": 25030
    },
    {
      "epoch": 13.121069182389936,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.009574301048409024,
      "loss": 0.4053,
      "num_input_tokens_seen": 16321288,
      "step": 25035
    },
    {
      "epoch": 13.123689727463312,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.009567905578583962,
      "loss": 0.2771,
      "num_input_tokens_seen": 16324360,
      "step": 25040
    },
    {
      "epoch": 13.126310272536688,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.009561511245091088,
      "loss": 0.3224,
      "num_input_tokens_seen": 16327784,
      "step": 25045
    },
    {
      "epoch": 13.128930817610064,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.00955511804926801,
      "loss": 0.4788,
      "num_input_tokens_seen": 16331848,
      "step": 25050
    },
    {
      "epoch": 13.131551362683439,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.009548725992452116,
      "loss": 0.4104,
      "num_input_tokens_seen": 16335592,
      "step": 25055
    },
    {
      "epoch": 13.134171907756814,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.009542335075980558,
      "loss": 0.5043,
      "num_input_tokens_seen": 16340200,
      "step": 25060
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.00953594530119024,
      "loss": 0.4491,
      "num_input_tokens_seen": 16343208,
      "step": 25065
    },
    {
      "epoch": 13.139412997903564,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.009529556669417823,
      "loss": 0.3078,
      "num_input_tokens_seen": 16346088,
      "step": 25070
    },
    {
      "epoch": 13.14203354297694,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.009523169181999743,
      "loss": 0.387,
      "num_input_tokens_seen": 16349768,
      "step": 25075
    },
    {
      "epoch": 13.144654088050315,
      "grad_norm": 0.103515625,
      "learning_rate": 0.009516782840272185,
      "loss": 0.3691,
      "num_input_tokens_seen": 16353960,
      "step": 25080
    },
    {
      "epoch": 13.14727463312369,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.009510397645571102,
      "loss": 0.4941,
      "num_input_tokens_seen": 16356904,
      "step": 25085
    },
    {
      "epoch": 13.149895178197065,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0095040135992322,
      "loss": 0.2972,
      "num_input_tokens_seen": 16360040,
      "step": 25090
    },
    {
      "epoch": 13.15251572327044,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00949763070259096,
      "loss": 0.3811,
      "num_input_tokens_seen": 16363496,
      "step": 25095
    },
    {
      "epoch": 13.155136268343815,
      "grad_norm": 0.103515625,
      "learning_rate": 0.009491248956982596,
      "loss": 0.4146,
      "num_input_tokens_seen": 16366376,
      "step": 25100
    },
    {
      "epoch": 13.15775681341719,
      "grad_norm": 0.046875,
      "learning_rate": 0.009484868363742103,
      "loss": 0.4737,
      "num_input_tokens_seen": 16370184,
      "step": 25105
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.00947848892420423,
      "loss": 0.3612,
      "num_input_tokens_seen": 16374152,
      "step": 25110
    },
    {
      "epoch": 13.16299790356394,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.009472110639703478,
      "loss": 0.3728,
      "num_input_tokens_seen": 16376904,
      "step": 25115
    },
    {
      "epoch": 13.165618448637316,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.009465733511574124,
      "loss": 0.3498,
      "num_input_tokens_seen": 16379784,
      "step": 25120
    },
    {
      "epoch": 13.168238993710691,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.009459357541150173,
      "loss": 0.3882,
      "num_input_tokens_seen": 16382888,
      "step": 25125
    },
    {
      "epoch": 13.170859538784066,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.009452982729765413,
      "loss": 0.3361,
      "num_input_tokens_seen": 16386472,
      "step": 25130
    },
    {
      "epoch": 13.173480083857442,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.009446609078753384,
      "loss": 0.5275,
      "num_input_tokens_seen": 16389608,
      "step": 25135
    },
    {
      "epoch": 13.176100628930818,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00944023658944738,
      "loss": 0.3211,
      "num_input_tokens_seen": 16393032,
      "step": 25140
    },
    {
      "epoch": 13.178721174004194,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.009433865263180448,
      "loss": 0.3176,
      "num_input_tokens_seen": 16396712,
      "step": 25145
    },
    {
      "epoch": 13.181341719077569,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00942749510128541,
      "loss": 0.4161,
      "num_input_tokens_seen": 16399272,
      "step": 25150
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.009421126105094818,
      "loss": 0.3541,
      "num_input_tokens_seen": 16402088,
      "step": 25155
    },
    {
      "epoch": 13.18658280922432,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.009414758275940996,
      "loss": 0.5352,
      "num_input_tokens_seen": 16406600,
      "step": 25160
    },
    {
      "epoch": 13.189203354297694,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.009408391615156024,
      "loss": 0.3499,
      "num_input_tokens_seen": 16409224,
      "step": 25165
    },
    {
      "epoch": 13.19182389937107,
      "grad_norm": 0.08203125,
      "learning_rate": 0.009402026124071731,
      "loss": 0.6463,
      "num_input_tokens_seen": 16412136,
      "step": 25170
    },
    {
      "epoch": 13.194444444444445,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.009395661804019712,
      "loss": 0.365,
      "num_input_tokens_seen": 16416392,
      "step": 25175
    },
    {
      "epoch": 13.19706498951782,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.009389298656331308,
      "loss": 0.4211,
      "num_input_tokens_seen": 16419688,
      "step": 25180
    },
    {
      "epoch": 13.199685534591195,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00938293668233761,
      "loss": 0.3975,
      "num_input_tokens_seen": 16422856,
      "step": 25185
    },
    {
      "epoch": 13.20230607966457,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.009376575883369476,
      "loss": 0.3789,
      "num_input_tokens_seen": 16425608,
      "step": 25190
    },
    {
      "epoch": 13.204926624737945,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.009370216260757508,
      "loss": 0.3685,
      "num_input_tokens_seen": 16429288,
      "step": 25195
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.009363857815832079,
      "loss": 0.4134,
      "num_input_tokens_seen": 16432744,
      "step": 25200
    },
    {
      "epoch": 13.210167714884696,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.009357500549923285,
      "loss": 0.2597,
      "num_input_tokens_seen": 16435944,
      "step": 25205
    },
    {
      "epoch": 13.21278825995807,
      "grad_norm": 0.076171875,
      "learning_rate": 0.009351144464361012,
      "loss": 0.2662,
      "num_input_tokens_seen": 16438760,
      "step": 25210
    },
    {
      "epoch": 13.215408805031446,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.009344789560474864,
      "loss": 0.3498,
      "num_input_tokens_seen": 16442280,
      "step": 25215
    },
    {
      "epoch": 13.218029350104821,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.009338435839594227,
      "loss": 0.3347,
      "num_input_tokens_seen": 16445384,
      "step": 25220
    },
    {
      "epoch": 13.220649895178196,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.009332083303048216,
      "loss": 0.243,
      "num_input_tokens_seen": 16448040,
      "step": 25225
    },
    {
      "epoch": 13.223270440251572,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.009325731952165722,
      "loss": 0.3572,
      "num_input_tokens_seen": 16451432,
      "step": 25230
    },
    {
      "epoch": 13.225890985324948,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.009319381788275368,
      "loss": 0.324,
      "num_input_tokens_seen": 16455624,
      "step": 25235
    },
    {
      "epoch": 13.228511530398324,
      "grad_norm": 0.076171875,
      "learning_rate": 0.009313032812705532,
      "loss": 0.2953,
      "num_input_tokens_seen": 16459112,
      "step": 25240
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.009306685026784353,
      "loss": 0.3699,
      "num_input_tokens_seen": 16462312,
      "step": 25245
    },
    {
      "epoch": 13.233752620545074,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.00930033843183971,
      "loss": 0.5271,
      "num_input_tokens_seen": 16465128,
      "step": 25250
    },
    {
      "epoch": 13.23637316561845,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.009293993029199246,
      "loss": 0.3565,
      "num_input_tokens_seen": 16469320,
      "step": 25255
    },
    {
      "epoch": 13.238993710691824,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00928764882019034,
      "loss": 0.5172,
      "num_input_tokens_seen": 16472296,
      "step": 25260
    },
    {
      "epoch": 13.2416142557652,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.009281305806140136,
      "loss": 0.4248,
      "num_input_tokens_seen": 16475240,
      "step": 25265
    },
    {
      "epoch": 13.244234800838575,
      "grad_norm": 0.072265625,
      "learning_rate": 0.009274963988375512,
      "loss": 0.3979,
      "num_input_tokens_seen": 16477736,
      "step": 25270
    },
    {
      "epoch": 13.24685534591195,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.009268623368223105,
      "loss": 0.4334,
      "num_input_tokens_seen": 16480776,
      "step": 25275
    },
    {
      "epoch": 13.249475890985325,
      "grad_norm": 0.11328125,
      "learning_rate": 0.009262283947009304,
      "loss": 0.3528,
      "num_input_tokens_seen": 16484584,
      "step": 25280
    },
    {
      "epoch": 13.2520964360587,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00925594572606024,
      "loss": 0.3743,
      "num_input_tokens_seen": 16487400,
      "step": 25285
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.009249608706701797,
      "loss": 0.3333,
      "num_input_tokens_seen": 16490024,
      "step": 25290
    },
    {
      "epoch": 13.25733752620545,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.009243272890259616,
      "loss": 0.4143,
      "num_input_tokens_seen": 16493064,
      "step": 25295
    },
    {
      "epoch": 13.259958071278826,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.009236938278059067,
      "loss": 0.3025,
      "num_input_tokens_seen": 16496136,
      "step": 25300
    },
    {
      "epoch": 13.2625786163522,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.009230604871425279,
      "loss": 0.342,
      "num_input_tokens_seen": 16500328,
      "step": 25305
    },
    {
      "epoch": 13.265199161425576,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.009224272671683131,
      "loss": 0.4131,
      "num_input_tokens_seen": 16503688,
      "step": 25310
    },
    {
      "epoch": 13.267819706498951,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.009217941680157249,
      "loss": 0.405,
      "num_input_tokens_seen": 16507112,
      "step": 25315
    },
    {
      "epoch": 13.270440251572326,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.009211611898172005,
      "loss": 0.3995,
      "num_input_tokens_seen": 16510920,
      "step": 25320
    },
    {
      "epoch": 13.273060796645701,
      "grad_norm": 0.068359375,
      "learning_rate": 0.009205283327051514,
      "loss": 0.4103,
      "num_input_tokens_seen": 16513736,
      "step": 25325
    },
    {
      "epoch": 13.275681341719078,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.009198955968119637,
      "loss": 0.4063,
      "num_input_tokens_seen": 16516744,
      "step": 25330
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.009192629822699994,
      "loss": 0.4327,
      "num_input_tokens_seen": 16519976,
      "step": 25335
    },
    {
      "epoch": 13.280922431865829,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00918630489211593,
      "loss": 0.3651,
      "num_input_tokens_seen": 16523112,
      "step": 25340
    },
    {
      "epoch": 13.283542976939204,
      "grad_norm": 0.083984375,
      "learning_rate": 0.009179981177690566,
      "loss": 0.4628,
      "num_input_tokens_seen": 16526376,
      "step": 25345
    },
    {
      "epoch": 13.286163522012579,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.009173658680746741,
      "loss": 0.4012,
      "num_input_tokens_seen": 16533160,
      "step": 25350
    },
    {
      "epoch": 13.288784067085954,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.009167337402607043,
      "loss": 0.3183,
      "num_input_tokens_seen": 16536680,
      "step": 25355
    },
    {
      "epoch": 13.29140461215933,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.009161017344593823,
      "loss": 0.3414,
      "num_input_tokens_seen": 16539656,
      "step": 25360
    },
    {
      "epoch": 13.294025157232705,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.009154698508029155,
      "loss": 0.3573,
      "num_input_tokens_seen": 16542376,
      "step": 25365
    },
    {
      "epoch": 13.29664570230608,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.009148380894234875,
      "loss": 0.3618,
      "num_input_tokens_seen": 16545352,
      "step": 25370
    },
    {
      "epoch": 13.299266247379455,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.009142064504532553,
      "loss": 0.4135,
      "num_input_tokens_seen": 16548424,
      "step": 25375
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00913574934024351,
      "loss": 0.3838,
      "num_input_tokens_seen": 16551592,
      "step": 25380
    },
    {
      "epoch": 13.304507337526205,
      "grad_norm": 0.048828125,
      "learning_rate": 0.009129435402688797,
      "loss": 0.2949,
      "num_input_tokens_seen": 16554568,
      "step": 25385
    },
    {
      "epoch": 13.30712788259958,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.009123122693189224,
      "loss": 0.359,
      "num_input_tokens_seen": 16557416,
      "step": 25390
    },
    {
      "epoch": 13.309748427672956,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.009116811213065335,
      "loss": 0.3533,
      "num_input_tokens_seen": 16560424,
      "step": 25395
    },
    {
      "epoch": 13.31236897274633,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.009110500963637424,
      "loss": 0.3178,
      "num_input_tokens_seen": 16563880,
      "step": 25400
    },
    {
      "epoch": 13.314989517819706,
      "grad_norm": 0.109375,
      "learning_rate": 0.009104191946225524,
      "loss": 0.3208,
      "num_input_tokens_seen": 16567432,
      "step": 25405
    },
    {
      "epoch": 13.317610062893081,
      "grad_norm": 0.080078125,
      "learning_rate": 0.009097884162149399,
      "loss": 0.4637,
      "num_input_tokens_seen": 16570888,
      "step": 25410
    },
    {
      "epoch": 13.320230607966456,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.009091577612728574,
      "loss": 0.3319,
      "num_input_tokens_seen": 16573384,
      "step": 25415
    },
    {
      "epoch": 13.322851153039831,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.009085272299282304,
      "loss": 0.4192,
      "num_input_tokens_seen": 16576584,
      "step": 25420
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.009078968223129594,
      "loss": 0.4159,
      "num_input_tokens_seen": 16580584,
      "step": 25425
    },
    {
      "epoch": 13.328092243186584,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.009072665385589177,
      "loss": 0.4625,
      "num_input_tokens_seen": 16583848,
      "step": 25430
    },
    {
      "epoch": 13.330712788259959,
      "grad_norm": 0.1171875,
      "learning_rate": 0.009066363787979547,
      "loss": 0.2883,
      "num_input_tokens_seen": 16587272,
      "step": 25435
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.009060063431618913,
      "loss": 0.37,
      "num_input_tokens_seen": 16590088,
      "step": 25440
    },
    {
      "epoch": 13.335953878406709,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.009053764317825239,
      "loss": 0.366,
      "num_input_tokens_seen": 16593320,
      "step": 25445
    },
    {
      "epoch": 13.338574423480084,
      "grad_norm": 0.0625,
      "learning_rate": 0.009047466447916232,
      "loss": 0.3985,
      "num_input_tokens_seen": 16596328,
      "step": 25450
    },
    {
      "epoch": 13.34119496855346,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.009041169823209337,
      "loss": 0.5377,
      "num_input_tokens_seen": 16599080,
      "step": 25455
    },
    {
      "epoch": 13.343815513626835,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.009034874445021733,
      "loss": 0.3294,
      "num_input_tokens_seen": 16602056,
      "step": 25460
    },
    {
      "epoch": 13.34643605870021,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.009028580314670343,
      "loss": 0.4248,
      "num_input_tokens_seen": 16606216,
      "step": 25465
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.009022287433471827,
      "loss": 0.3551,
      "num_input_tokens_seen": 16609064,
      "step": 25470
    },
    {
      "epoch": 13.35167714884696,
      "grad_norm": 0.095703125,
      "learning_rate": 0.009015995802742574,
      "loss": 0.4548,
      "num_input_tokens_seen": 16611560,
      "step": 25475
    },
    {
      "epoch": 13.354297693920335,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.009009705423798737,
      "loss": 0.2975,
      "num_input_tokens_seen": 16614408,
      "step": 25480
    },
    {
      "epoch": 13.35691823899371,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.00900341629795618,
      "loss": 0.3621,
      "num_input_tokens_seen": 16617256,
      "step": 25485
    },
    {
      "epoch": 13.359538784067086,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.008997128426530528,
      "loss": 0.4186,
      "num_input_tokens_seen": 16620232,
      "step": 25490
    },
    {
      "epoch": 13.36215932914046,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.008990841810837119,
      "loss": 0.4151,
      "num_input_tokens_seen": 16623912,
      "step": 25495
    },
    {
      "epoch": 13.364779874213836,
      "grad_norm": 0.126953125,
      "learning_rate": 0.008984556452191048,
      "loss": 0.4783,
      "num_input_tokens_seen": 16626440,
      "step": 25500
    },
    {
      "epoch": 13.367400419287211,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00897827235190714,
      "loss": 0.5339,
      "num_input_tokens_seen": 16630440,
      "step": 25505
    },
    {
      "epoch": 13.370020964360586,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.008971989511299955,
      "loss": 0.3685,
      "num_input_tokens_seen": 16633352,
      "step": 25510
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.008965707931683792,
      "loss": 0.4002,
      "num_input_tokens_seen": 16636392,
      "step": 25515
    },
    {
      "epoch": 13.375262054507338,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.008959427614372696,
      "loss": 0.3859,
      "num_input_tokens_seen": 16639688,
      "step": 25520
    },
    {
      "epoch": 13.377882599580714,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.008953148560680418,
      "loss": 0.3452,
      "num_input_tokens_seen": 16643176,
      "step": 25525
    },
    {
      "epoch": 13.380503144654089,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.008946870771920476,
      "loss": 0.3801,
      "num_input_tokens_seen": 16646952,
      "step": 25530
    },
    {
      "epoch": 13.383123689727464,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.008940594249406115,
      "loss": 0.3851,
      "num_input_tokens_seen": 16650792,
      "step": 25535
    },
    {
      "epoch": 13.385744234800839,
      "grad_norm": 0.0703125,
      "learning_rate": 0.008934318994450305,
      "loss": 0.3313,
      "num_input_tokens_seen": 16654568,
      "step": 25540
    },
    {
      "epoch": 13.388364779874214,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.008928045008365762,
      "loss": 0.4693,
      "num_input_tokens_seen": 16657608,
      "step": 25545
    },
    {
      "epoch": 13.39098532494759,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.008921772292464934,
      "loss": 0.3901,
      "num_input_tokens_seen": 16660360,
      "step": 25550
    },
    {
      "epoch": 13.393605870020965,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.008915500848059992,
      "loss": 0.294,
      "num_input_tokens_seen": 16663048,
      "step": 25555
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.00890923067646286,
      "loss": 0.3471,
      "num_input_tokens_seen": 16667080,
      "step": 25560
    },
    {
      "epoch": 13.398846960167715,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.008902961778985181,
      "loss": 0.3524,
      "num_input_tokens_seen": 16669608,
      "step": 25565
    },
    {
      "epoch": 13.40146750524109,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.008896694156938343,
      "loss": 0.3755,
      "num_input_tokens_seen": 16676968,
      "step": 25570
    },
    {
      "epoch": 13.404088050314465,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.008890427811633453,
      "loss": 0.3839,
      "num_input_tokens_seen": 16680552,
      "step": 25575
    },
    {
      "epoch": 13.40670859538784,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.008884162744381368,
      "loss": 0.341,
      "num_input_tokens_seen": 16683880,
      "step": 25580
    },
    {
      "epoch": 13.409329140461216,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.008877898956492662,
      "loss": 0.4464,
      "num_input_tokens_seen": 16686312,
      "step": 25585
    },
    {
      "epoch": 13.41194968553459,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.008871636449277645,
      "loss": 0.3819,
      "num_input_tokens_seen": 16688776,
      "step": 25590
    },
    {
      "epoch": 13.414570230607966,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.00886537522404637,
      "loss": 0.2783,
      "num_input_tokens_seen": 16692104,
      "step": 25595
    },
    {
      "epoch": 13.417190775681341,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.008859115282108612,
      "loss": 0.3626,
      "num_input_tokens_seen": 16694600,
      "step": 25600
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00885285662477388,
      "loss": 0.3863,
      "num_input_tokens_seen": 16697384,
      "step": 25605
    },
    {
      "epoch": 13.422431865828091,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.008846599253351406,
      "loss": 0.405,
      "num_input_tokens_seen": 16700776,
      "step": 25610
    },
    {
      "epoch": 13.425052410901468,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.008840343169150172,
      "loss": 0.3742,
      "num_input_tokens_seen": 16703272,
      "step": 25615
    },
    {
      "epoch": 13.427672955974844,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.008834088373478868,
      "loss": 0.3016,
      "num_input_tokens_seen": 16706440,
      "step": 25620
    },
    {
      "epoch": 13.430293501048219,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.008827834867645936,
      "loss": 0.3756,
      "num_input_tokens_seen": 16709256,
      "step": 25625
    },
    {
      "epoch": 13.432914046121594,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.00882158265295953,
      "loss": 0.3615,
      "num_input_tokens_seen": 16712712,
      "step": 25630
    },
    {
      "epoch": 13.435534591194969,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.008815331730727552,
      "loss": 0.3727,
      "num_input_tokens_seen": 16715464,
      "step": 25635
    },
    {
      "epoch": 13.438155136268344,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.008809082102257617,
      "loss": 0.5799,
      "num_input_tokens_seen": 16719304,
      "step": 25640
    },
    {
      "epoch": 13.44077568134172,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.008802833768857071,
      "loss": 0.4298,
      "num_input_tokens_seen": 16722760,
      "step": 25645
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 0.060546875,
      "learning_rate": 0.008796586731833003,
      "loss": 0.4002,
      "num_input_tokens_seen": 16726152,
      "step": 25650
    },
    {
      "epoch": 13.44601677148847,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.008790340992492218,
      "loss": 0.393,
      "num_input_tokens_seen": 16728808,
      "step": 25655
    },
    {
      "epoch": 13.448637316561845,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.008784096552141253,
      "loss": 0.3656,
      "num_input_tokens_seen": 16731240,
      "step": 25660
    },
    {
      "epoch": 13.45125786163522,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008777853412086378,
      "loss": 0.5064,
      "num_input_tokens_seen": 16735112,
      "step": 25665
    },
    {
      "epoch": 13.453878406708595,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.008771611573633578,
      "loss": 0.3972,
      "num_input_tokens_seen": 16738568,
      "step": 25670
    },
    {
      "epoch": 13.45649895178197,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.00876537103808858,
      "loss": 0.3713,
      "num_input_tokens_seen": 16742056,
      "step": 25675
    },
    {
      "epoch": 13.459119496855346,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.008759131806756837,
      "loss": 0.3626,
      "num_input_tokens_seen": 16744616,
      "step": 25680
    },
    {
      "epoch": 13.46174004192872,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.008752893880943514,
      "loss": 0.427,
      "num_input_tokens_seen": 16747464,
      "step": 25685
    },
    {
      "epoch": 13.464360587002096,
      "grad_norm": 0.033203125,
      "learning_rate": 0.008746657261953528,
      "loss": 0.3897,
      "num_input_tokens_seen": 16752104,
      "step": 25690
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.008740421951091497,
      "loss": 0.3349,
      "num_input_tokens_seen": 16755720,
      "step": 25695
    },
    {
      "epoch": 13.469601677148846,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.008734187949661784,
      "loss": 0.4628,
      "num_input_tokens_seen": 16758600,
      "step": 25700
    },
    {
      "epoch": 13.472222222222221,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.008727955258968462,
      "loss": 0.365,
      "num_input_tokens_seen": 16761352,
      "step": 25705
    },
    {
      "epoch": 13.474842767295598,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.008721723880315346,
      "loss": 0.4625,
      "num_input_tokens_seen": 16765512,
      "step": 25710
    },
    {
      "epoch": 13.477463312368974,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.008715493815005965,
      "loss": 0.4847,
      "num_input_tokens_seen": 16769064,
      "step": 25715
    },
    {
      "epoch": 13.480083857442349,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.008709265064343589,
      "loss": 0.4881,
      "num_input_tokens_seen": 16772232,
      "step": 25720
    },
    {
      "epoch": 13.482704402515724,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.008703037629631191,
      "loss": 0.3189,
      "num_input_tokens_seen": 16776616,
      "step": 25725
    },
    {
      "epoch": 13.485324947589099,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00869681151217148,
      "loss": 0.3728,
      "num_input_tokens_seen": 16779496,
      "step": 25730
    },
    {
      "epoch": 13.487945492662474,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.008690586713266885,
      "loss": 0.3137,
      "num_input_tokens_seen": 16783368,
      "step": 25735
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.00868436323421957,
      "loss": 0.3762,
      "num_input_tokens_seen": 16785928,
      "step": 25740
    },
    {
      "epoch": 13.493186582809225,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.00867814107633142,
      "loss": 0.3749,
      "num_input_tokens_seen": 16789320,
      "step": 25745
    },
    {
      "epoch": 13.4958071278826,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.008671920240904032,
      "loss": 0.4392,
      "num_input_tokens_seen": 16793928,
      "step": 25750
    },
    {
      "epoch": 13.498427672955975,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00866570072923873,
      "loss": 0.4318,
      "num_input_tokens_seen": 16796872,
      "step": 25755
    },
    {
      "epoch": 13.50104821802935,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00865948254263657,
      "loss": 0.3362,
      "num_input_tokens_seen": 16800552,
      "step": 25760
    },
    {
      "epoch": 13.503668763102725,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.008653265682398326,
      "loss": 0.3735,
      "num_input_tokens_seen": 16803880,
      "step": 25765
    },
    {
      "epoch": 13.5062893081761,
      "grad_norm": 0.076171875,
      "learning_rate": 0.008647050149824493,
      "loss": 0.3789,
      "num_input_tokens_seen": 16806600,
      "step": 25770
    },
    {
      "epoch": 13.508909853249476,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.008640835946215297,
      "loss": 0.3533,
      "num_input_tokens_seen": 16809576,
      "step": 25775
    },
    {
      "epoch": 13.51153039832285,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.008634623072870672,
      "loss": 0.4603,
      "num_input_tokens_seen": 16812776,
      "step": 25780
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.008628411531090273,
      "loss": 0.3474,
      "num_input_tokens_seen": 16815496,
      "step": 25785
    },
    {
      "epoch": 13.516771488469601,
      "grad_norm": 0.095703125,
      "learning_rate": 0.00862220132217349,
      "loss": 0.3877,
      "num_input_tokens_seen": 16818984,
      "step": 25790
    },
    {
      "epoch": 13.519392033542976,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.008615992447419433,
      "loss": 0.2938,
      "num_input_tokens_seen": 16825096,
      "step": 25795
    },
    {
      "epoch": 13.522012578616351,
      "grad_norm": 0.125,
      "learning_rate": 0.008609784908126925,
      "loss": 0.4492,
      "num_input_tokens_seen": 16827880,
      "step": 25800
    },
    {
      "epoch": 13.524633123689728,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.008603578705594512,
      "loss": 0.4457,
      "num_input_tokens_seen": 16830504,
      "step": 25805
    },
    {
      "epoch": 13.527253668763104,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.008597373841120455,
      "loss": 0.3413,
      "num_input_tokens_seen": 16833128,
      "step": 25810
    },
    {
      "epoch": 13.529874213836479,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.008591170316002743,
      "loss": 0.4091,
      "num_input_tokens_seen": 16835592,
      "step": 25815
    },
    {
      "epoch": 13.532494758909854,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.008584968131539087,
      "loss": 0.478,
      "num_input_tokens_seen": 16839752,
      "step": 25820
    },
    {
      "epoch": 13.535115303983229,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.008578767289026916,
      "loss": 0.3642,
      "num_input_tokens_seen": 16842568,
      "step": 25825
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.008572567789763367,
      "loss": 0.3921,
      "num_input_tokens_seen": 16845320,
      "step": 25830
    },
    {
      "epoch": 13.54035639412998,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.008566369635045312,
      "loss": 0.3902,
      "num_input_tokens_seen": 16848552,
      "step": 25835
    },
    {
      "epoch": 13.542976939203355,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.008560172826169325,
      "loss": 0.53,
      "num_input_tokens_seen": 16853608,
      "step": 25840
    },
    {
      "epoch": 13.54559748427673,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.008553977364431713,
      "loss": 0.3514,
      "num_input_tokens_seen": 16856648,
      "step": 25845
    },
    {
      "epoch": 13.548218029350105,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0085477832511285,
      "loss": 0.4223,
      "num_input_tokens_seen": 16861864,
      "step": 25850
    },
    {
      "epoch": 13.55083857442348,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.008541590487555421,
      "loss": 0.3441,
      "num_input_tokens_seen": 16865256,
      "step": 25855
    },
    {
      "epoch": 13.553459119496855,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.008535399075007933,
      "loss": 0.3767,
      "num_input_tokens_seen": 16868936,
      "step": 25860
    },
    {
      "epoch": 13.55607966457023,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0085292090147812,
      "loss": 0.3957,
      "num_input_tokens_seen": 16872200,
      "step": 25865
    },
    {
      "epoch": 13.558700209643606,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.008523020308170119,
      "loss": 0.39,
      "num_input_tokens_seen": 16878824,
      "step": 25870
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.008516832956469296,
      "loss": 0.3229,
      "num_input_tokens_seen": 16883080,
      "step": 25875
    },
    {
      "epoch": 13.563941299790356,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00851064696097306,
      "loss": 0.4977,
      "num_input_tokens_seen": 16887336,
      "step": 25880
    },
    {
      "epoch": 13.566561844863731,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.008504462322975442,
      "loss": 0.4164,
      "num_input_tokens_seen": 16890760,
      "step": 25885
    },
    {
      "epoch": 13.569182389937106,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.008498279043770204,
      "loss": 0.3344,
      "num_input_tokens_seen": 16893928,
      "step": 25890
    },
    {
      "epoch": 13.571802935010481,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.00849209712465081,
      "loss": 0.286,
      "num_input_tokens_seen": 16897032,
      "step": 25895
    },
    {
      "epoch": 13.574423480083858,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.00848591656691045,
      "loss": 0.3386,
      "num_input_tokens_seen": 16900648,
      "step": 25900
    },
    {
      "epoch": 13.577044025157234,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.008479737371842034,
      "loss": 0.4687,
      "num_input_tokens_seen": 16903880,
      "step": 25905
    },
    {
      "epoch": 13.579664570230609,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.008473559540738166,
      "loss": 0.4601,
      "num_input_tokens_seen": 16907816,
      "step": 25910
    },
    {
      "epoch": 13.582285115303984,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.008467383074891187,
      "loss": 0.3203,
      "num_input_tokens_seen": 16910664,
      "step": 25915
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.008461207975593145,
      "loss": 0.3925,
      "num_input_tokens_seen": 16914408,
      "step": 25920
    },
    {
      "epoch": 13.587526205450734,
      "grad_norm": 0.07421875,
      "learning_rate": 0.008455034244135792,
      "loss": 0.4956,
      "num_input_tokens_seen": 16918120,
      "step": 25925
    },
    {
      "epoch": 13.59014675052411,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.008448861881810602,
      "loss": 0.4575,
      "num_input_tokens_seen": 16920680,
      "step": 25930
    },
    {
      "epoch": 13.592767295597485,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.008442690889908776,
      "loss": 0.3351,
      "num_input_tokens_seen": 16923592,
      "step": 25935
    },
    {
      "epoch": 13.59538784067086,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.008436521269721201,
      "loss": 0.3291,
      "num_input_tokens_seen": 16926920,
      "step": 25940
    },
    {
      "epoch": 13.598008385744235,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.008430353022538493,
      "loss": 0.3903,
      "num_input_tokens_seen": 16930056,
      "step": 25945
    },
    {
      "epoch": 13.60062893081761,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.008424186149650991,
      "loss": 0.3968,
      "num_input_tokens_seen": 16932424,
      "step": 25950
    },
    {
      "epoch": 13.603249475890985,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00841802065234872,
      "loss": 0.3322,
      "num_input_tokens_seen": 16936200,
      "step": 25955
    },
    {
      "epoch": 13.60587002096436,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.008411856531921441,
      "loss": 0.3528,
      "num_input_tokens_seen": 16939176,
      "step": 25960
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00840569378965861,
      "loss": 0.3732,
      "num_input_tokens_seen": 16943048,
      "step": 25965
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.008399532426849404,
      "loss": 0.3597,
      "num_input_tokens_seen": 16946056,
      "step": 25970
    },
    {
      "epoch": 13.613731656184486,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.008393372444782719,
      "loss": 0.4231,
      "num_input_tokens_seen": 16948712,
      "step": 25975
    },
    {
      "epoch": 13.616352201257861,
      "grad_norm": 0.080078125,
      "learning_rate": 0.008387213844747141,
      "loss": 0.4178,
      "num_input_tokens_seen": 16952168,
      "step": 25980
    },
    {
      "epoch": 13.618972746331236,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00838105662803099,
      "loss": 0.3827,
      "num_input_tokens_seen": 16955496,
      "step": 25985
    },
    {
      "epoch": 13.621593291404611,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.008374900795922274,
      "loss": 0.3925,
      "num_input_tokens_seen": 16958472,
      "step": 25990
    },
    {
      "epoch": 13.624213836477988,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.008368746349708727,
      "loss": 0.3202,
      "num_input_tokens_seen": 16961832,
      "step": 25995
    },
    {
      "epoch": 13.626834381551364,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.00836259329067779,
      "loss": 0.3172,
      "num_input_tokens_seen": 16964552,
      "step": 26000
    },
    {
      "epoch": 13.629454926624739,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00835644162011662,
      "loss": 0.3341,
      "num_input_tokens_seen": 16967688,
      "step": 26005
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.008350291339312064,
      "loss": 0.3888,
      "num_input_tokens_seen": 16971176,
      "step": 26010
    },
    {
      "epoch": 13.634696016771489,
      "grad_norm": 0.078125,
      "learning_rate": 0.008344142449550701,
      "loss": 0.3555,
      "num_input_tokens_seen": 16974568,
      "step": 26015
    },
    {
      "epoch": 13.637316561844864,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0083379949521188,
      "loss": 0.4429,
      "num_input_tokens_seen": 16978728,
      "step": 26020
    },
    {
      "epoch": 13.63993710691824,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.008331848848302348,
      "loss": 0.2809,
      "num_input_tokens_seen": 16981928,
      "step": 26025
    },
    {
      "epoch": 13.642557651991615,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.008325704139387043,
      "loss": 0.3832,
      "num_input_tokens_seen": 16984104,
      "step": 26030
    },
    {
      "epoch": 13.64517819706499,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.008319560826658292,
      "loss": 0.4306,
      "num_input_tokens_seen": 16986952,
      "step": 26035
    },
    {
      "epoch": 13.647798742138365,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.008313418911401204,
      "loss": 0.3041,
      "num_input_tokens_seen": 16993160,
      "step": 26040
    },
    {
      "epoch": 13.65041928721174,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00830727839490059,
      "loss": 0.4108,
      "num_input_tokens_seen": 16995560,
      "step": 26045
    },
    {
      "epoch": 13.653039832285115,
      "grad_norm": 0.080078125,
      "learning_rate": 0.00830113927844098,
      "loss": 0.4443,
      "num_input_tokens_seen": 16999016,
      "step": 26050
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00829500156330661,
      "loss": 0.2927,
      "num_input_tokens_seen": 17001960,
      "step": 26055
    },
    {
      "epoch": 13.658280922431866,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.008288865250781421,
      "loss": 0.4355,
      "num_input_tokens_seen": 17004936,
      "step": 26060
    },
    {
      "epoch": 13.66090146750524,
      "grad_norm": 0.09375,
      "learning_rate": 0.008282730342149058,
      "loss": 0.3301,
      "num_input_tokens_seen": 17008584,
      "step": 26065
    },
    {
      "epoch": 13.663522012578616,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.00827659683869287,
      "loss": 0.3889,
      "num_input_tokens_seen": 17011048,
      "step": 26070
    },
    {
      "epoch": 13.666142557651991,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.008270464741695913,
      "loss": 0.3004,
      "num_input_tokens_seen": 17013384,
      "step": 26075
    },
    {
      "epoch": 13.668763102725366,
      "grad_norm": 0.08203125,
      "learning_rate": 0.008264334052440961,
      "loss": 0.3803,
      "num_input_tokens_seen": 17016584,
      "step": 26080
    },
    {
      "epoch": 13.671383647798741,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.008258204772210481,
      "loss": 0.4111,
      "num_input_tokens_seen": 17019368,
      "step": 26085
    },
    {
      "epoch": 13.674004192872118,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.008252076902286652,
      "loss": 0.3239,
      "num_input_tokens_seen": 17022184,
      "step": 26090
    },
    {
      "epoch": 13.676624737945493,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.008245950443951353,
      "loss": 0.3795,
      "num_input_tokens_seen": 17025288,
      "step": 26095
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.008239825398486158,
      "loss": 0.3371,
      "num_input_tokens_seen": 17028424,
      "step": 26100
    },
    {
      "epoch": 13.681865828092244,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.008233701767172369,
      "loss": 0.2954,
      "num_input_tokens_seen": 17031720,
      "step": 26105
    },
    {
      "epoch": 13.684486373165619,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.008227579551290973,
      "loss": 0.436,
      "num_input_tokens_seen": 17034888,
      "step": 26110
    },
    {
      "epoch": 13.687106918238994,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.00822145875212268,
      "loss": 0.3988,
      "num_input_tokens_seen": 17041576,
      "step": 26115
    },
    {
      "epoch": 13.68972746331237,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00821533937094788,
      "loss": 0.3864,
      "num_input_tokens_seen": 17044424,
      "step": 26120
    },
    {
      "epoch": 13.692348008385745,
      "grad_norm": 0.04296875,
      "learning_rate": 0.008209221409046674,
      "loss": 0.3159,
      "num_input_tokens_seen": 17047976,
      "step": 26125
    },
    {
      "epoch": 13.69496855345912,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.008203104867698877,
      "loss": 0.5439,
      "num_input_tokens_seen": 17051208,
      "step": 26130
    },
    {
      "epoch": 13.697589098532495,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.008196989748184001,
      "loss": 0.3399,
      "num_input_tokens_seen": 17053640,
      "step": 26135
    },
    {
      "epoch": 13.70020964360587,
      "grad_norm": 0.119140625,
      "learning_rate": 0.008190876051781251,
      "loss": 0.3332,
      "num_input_tokens_seen": 17056968,
      "step": 26140
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.008184763779769564,
      "loss": 0.3925,
      "num_input_tokens_seen": 17059688,
      "step": 26145
    },
    {
      "epoch": 13.70545073375262,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.00817865293342753,
      "loss": 0.4481,
      "num_input_tokens_seen": 17064712,
      "step": 26150
    },
    {
      "epoch": 13.708071278825996,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00817254351403348,
      "loss": 0.3348,
      "num_input_tokens_seen": 17067528,
      "step": 26155
    },
    {
      "epoch": 13.71069182389937,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.008166435522865435,
      "loss": 0.3444,
      "num_input_tokens_seen": 17070952,
      "step": 26160
    },
    {
      "epoch": 13.713312368972746,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.008160328961201118,
      "loss": 0.4832,
      "num_input_tokens_seen": 17073992,
      "step": 26165
    },
    {
      "epoch": 13.715932914046121,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.008154223830317955,
      "loss": 0.3862,
      "num_input_tokens_seen": 17078376,
      "step": 26170
    },
    {
      "epoch": 13.718553459119496,
      "grad_norm": 0.08984375,
      "learning_rate": 0.008148120131493068,
      "loss": 0.4548,
      "num_input_tokens_seen": 17082952,
      "step": 26175
    },
    {
      "epoch": 13.721174004192871,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.008142017866003272,
      "loss": 0.402,
      "num_input_tokens_seen": 17085640,
      "step": 26180
    },
    {
      "epoch": 13.723794549266248,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0081359170351251,
      "loss": 0.2886,
      "num_input_tokens_seen": 17089096,
      "step": 26185
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 0.166015625,
      "learning_rate": 0.008129817640134774,
      "loss": 0.5362,
      "num_input_tokens_seen": 17091752,
      "step": 26190
    },
    {
      "epoch": 13.729035639412999,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008123719682308224,
      "loss": 0.4115,
      "num_input_tokens_seen": 17094824,
      "step": 26195
    },
    {
      "epoch": 13.731656184486374,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.008117623162921061,
      "loss": 0.4929,
      "num_input_tokens_seen": 17097992,
      "step": 26200
    },
    {
      "epoch": 13.734276729559749,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.008111528083248623,
      "loss": 0.4752,
      "num_input_tokens_seen": 17100968,
      "step": 26205
    },
    {
      "epoch": 13.736897274633124,
      "grad_norm": 0.080078125,
      "learning_rate": 0.008105434444565912,
      "loss": 0.3917,
      "num_input_tokens_seen": 17103496,
      "step": 26210
    },
    {
      "epoch": 13.7395178197065,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.008099342248147661,
      "loss": 0.3655,
      "num_input_tokens_seen": 17106920,
      "step": 26215
    },
    {
      "epoch": 13.742138364779874,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.008093251495268284,
      "loss": 0.34,
      "num_input_tokens_seen": 17110472,
      "step": 26220
    },
    {
      "epoch": 13.74475890985325,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.008087162187201903,
      "loss": 0.4601,
      "num_input_tokens_seen": 17113640,
      "step": 26225
    },
    {
      "epoch": 13.747379454926625,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.008081074325222325,
      "loss": 0.5031,
      "num_input_tokens_seen": 17116808,
      "step": 26230
    },
    {
      "epoch": 13.75,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00807498791060306,
      "loss": 0.3454,
      "num_input_tokens_seen": 17119144,
      "step": 26235
    },
    {
      "epoch": 13.752620545073375,
      "grad_norm": 0.076171875,
      "learning_rate": 0.00806890294461732,
      "loss": 0.3249,
      "num_input_tokens_seen": 17122056,
      "step": 26240
    },
    {
      "epoch": 13.75524109014675,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.008062819428538008,
      "loss": 0.3251,
      "num_input_tokens_seen": 17125352,
      "step": 26245
    },
    {
      "epoch": 13.757861635220126,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.008056737363637734,
      "loss": 0.3575,
      "num_input_tokens_seen": 17128456,
      "step": 26250
    },
    {
      "epoch": 13.7604821802935,
      "grad_norm": 0.06640625,
      "learning_rate": 0.008050656751188785,
      "loss": 0.2732,
      "num_input_tokens_seen": 17131496,
      "step": 26255
    },
    {
      "epoch": 13.763102725366876,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.008044577592463168,
      "loss": 0.2697,
      "num_input_tokens_seen": 17135528,
      "step": 26260
    },
    {
      "epoch": 13.765723270440251,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.008038499888732559,
      "loss": 0.4108,
      "num_input_tokens_seen": 17138728,
      "step": 26265
    },
    {
      "epoch": 13.768343815513626,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.008032423641268353,
      "loss": 0.3971,
      "num_input_tokens_seen": 17142600,
      "step": 26270
    },
    {
      "epoch": 13.770964360587001,
      "grad_norm": 0.150390625,
      "learning_rate": 0.008026348851341639,
      "loss": 0.4483,
      "num_input_tokens_seen": 17145448,
      "step": 26275
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.008020275520223177,
      "loss": 0.2975,
      "num_input_tokens_seen": 17148712,
      "step": 26280
    },
    {
      "epoch": 13.776205450733752,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.008014203649183448,
      "loss": 0.3126,
      "num_input_tokens_seen": 17151144,
      "step": 26285
    },
    {
      "epoch": 13.778825995807129,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.008008133239492623,
      "loss": 0.3092,
      "num_input_tokens_seen": 17153864,
      "step": 26290
    },
    {
      "epoch": 13.781446540880504,
      "grad_norm": 0.09765625,
      "learning_rate": 0.008002064292420551,
      "loss": 0.4184,
      "num_input_tokens_seen": 17156968,
      "step": 26295
    },
    {
      "epoch": 13.784067085953879,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.007995996809236794,
      "loss": 0.4439,
      "num_input_tokens_seen": 17160520,
      "step": 26300
    },
    {
      "epoch": 13.786687631027254,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.007989930791210605,
      "loss": 0.3216,
      "num_input_tokens_seen": 17163368,
      "step": 26305
    },
    {
      "epoch": 13.78930817610063,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.007983866239610912,
      "loss": 0.535,
      "num_input_tokens_seen": 17166472,
      "step": 26310
    },
    {
      "epoch": 13.791928721174004,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.007977803155706356,
      "loss": 0.4767,
      "num_input_tokens_seen": 17170088,
      "step": 26315
    },
    {
      "epoch": 13.79454926624738,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.007971741540765274,
      "loss": 0.3746,
      "num_input_tokens_seen": 17172840,
      "step": 26320
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.007965681396055673,
      "loss": 0.3467,
      "num_input_tokens_seen": 17175752,
      "step": 26325
    },
    {
      "epoch": 13.79979035639413,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.007959622722845278,
      "loss": 0.445,
      "num_input_tokens_seen": 17178504,
      "step": 26330
    },
    {
      "epoch": 13.802410901467505,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.007953565522401482,
      "loss": 0.412,
      "num_input_tokens_seen": 17180648,
      "step": 26335
    },
    {
      "epoch": 13.80503144654088,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00794750979599139,
      "loss": 0.4859,
      "num_input_tokens_seen": 17184648,
      "step": 26340
    },
    {
      "epoch": 13.807651991614255,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.007941455544881794,
      "loss": 0.499,
      "num_input_tokens_seen": 17188264,
      "step": 26345
    },
    {
      "epoch": 13.81027253668763,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.007935402770339165,
      "loss": 0.2905,
      "num_input_tokens_seen": 17191464,
      "step": 26350
    },
    {
      "epoch": 13.812893081761006,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.007929351473629687,
      "loss": 0.3804,
      "num_input_tokens_seen": 17194792,
      "step": 26355
    },
    {
      "epoch": 13.815513626834381,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.007923301656019205,
      "loss": 0.3669,
      "num_input_tokens_seen": 17197864,
      "step": 26360
    },
    {
      "epoch": 13.818134171907756,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.007917253318773283,
      "loss": 0.4215,
      "num_input_tokens_seen": 17200840,
      "step": 26365
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.007911206463157162,
      "loss": 0.3446,
      "num_input_tokens_seen": 17204296,
      "step": 26370
    },
    {
      "epoch": 13.823375262054507,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.007905161090435783,
      "loss": 0.3748,
      "num_input_tokens_seen": 17206920,
      "step": 26375
    },
    {
      "epoch": 13.825995807127882,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.007899117201873754,
      "loss": 0.4125,
      "num_input_tokens_seen": 17209672,
      "step": 26380
    },
    {
      "epoch": 13.828616352201259,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.007893074798735402,
      "loss": 0.5434,
      "num_input_tokens_seen": 17212296,
      "step": 26385
    },
    {
      "epoch": 13.831236897274634,
      "grad_norm": 0.07421875,
      "learning_rate": 0.007887033882284716,
      "loss": 0.5258,
      "num_input_tokens_seen": 17215880,
      "step": 26390
    },
    {
      "epoch": 13.833857442348009,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007880994453785393,
      "loss": 0.4155,
      "num_input_tokens_seen": 17218760,
      "step": 26395
    },
    {
      "epoch": 13.836477987421384,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.007874956514500812,
      "loss": 0.528,
      "num_input_tokens_seen": 17222408,
      "step": 26400
    },
    {
      "epoch": 13.83909853249476,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.007868920065694049,
      "loss": 0.4429,
      "num_input_tokens_seen": 17225800,
      "step": 26405
    },
    {
      "epoch": 13.841719077568134,
      "grad_norm": 0.0703125,
      "learning_rate": 0.007862885108627848,
      "loss": 0.303,
      "num_input_tokens_seen": 17229256,
      "step": 26410
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.007856851644564655,
      "loss": 0.4826,
      "num_input_tokens_seen": 17232104,
      "step": 26415
    },
    {
      "epoch": 13.846960167714885,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.007850819674766606,
      "loss": 0.6392,
      "num_input_tokens_seen": 17235528,
      "step": 26420
    },
    {
      "epoch": 13.84958071278826,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.007844789200495516,
      "loss": 0.357,
      "num_input_tokens_seen": 17238088,
      "step": 26425
    },
    {
      "epoch": 13.852201257861635,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007838760223012902,
      "loss": 0.4079,
      "num_input_tokens_seen": 17241288,
      "step": 26430
    },
    {
      "epoch": 13.85482180293501,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.007832732743579948,
      "loss": 0.349,
      "num_input_tokens_seen": 17244840,
      "step": 26435
    },
    {
      "epoch": 13.857442348008385,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.007826706763457533,
      "loss": 0.4159,
      "num_input_tokens_seen": 17247688,
      "step": 26440
    },
    {
      "epoch": 13.86006289308176,
      "grad_norm": 0.107421875,
      "learning_rate": 0.007820682283906222,
      "loss": 0.4727,
      "num_input_tokens_seen": 17252360,
      "step": 26445
    },
    {
      "epoch": 13.862683438155136,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.007814659306186274,
      "loss": 0.4484,
      "num_input_tokens_seen": 17255368,
      "step": 26450
    },
    {
      "epoch": 13.865303983228511,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.007808637831557623,
      "loss": 0.4367,
      "num_input_tokens_seen": 17257832,
      "step": 26455
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0078026178612798985,
      "loss": 0.2847,
      "num_input_tokens_seen": 17261480,
      "step": 26460
    },
    {
      "epoch": 13.870545073375261,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.007796599396612407,
      "loss": 0.464,
      "num_input_tokens_seen": 17264360,
      "step": 26465
    },
    {
      "epoch": 13.873165618448636,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.007790582438814136,
      "loss": 0.3451,
      "num_input_tokens_seen": 17268104,
      "step": 26470
    },
    {
      "epoch": 13.875786163522012,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.00778456698914377,
      "loss": 0.3352,
      "num_input_tokens_seen": 17271048,
      "step": 26475
    },
    {
      "epoch": 13.878406708595389,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.00777855304885967,
      "loss": 0.3506,
      "num_input_tokens_seen": 17274056,
      "step": 26480
    },
    {
      "epoch": 13.881027253668764,
      "grad_norm": 0.103515625,
      "learning_rate": 0.007772540619219892,
      "loss": 0.3921,
      "num_input_tokens_seen": 17277192,
      "step": 26485
    },
    {
      "epoch": 13.883647798742139,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.007766529701482161,
      "loss": 0.3875,
      "num_input_tokens_seen": 17281128,
      "step": 26490
    },
    {
      "epoch": 13.886268343815514,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0077605202969038885,
      "loss": 0.3267,
      "num_input_tokens_seen": 17283848,
      "step": 26495
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.007754512406742175,
      "loss": 0.3326,
      "num_input_tokens_seen": 17287208,
      "step": 26500
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.007748506032253805,
      "loss": 0.336,
      "num_input_tokens_seen": 17290696,
      "step": 26505
    },
    {
      "epoch": 13.89412997903564,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.007742501174695253,
      "loss": 0.3601,
      "num_input_tokens_seen": 17294600,
      "step": 26510
    },
    {
      "epoch": 13.896750524109015,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.007736497835322655,
      "loss": 0.4367,
      "num_input_tokens_seen": 17297768,
      "step": 26515
    },
    {
      "epoch": 13.89937106918239,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.007730496015391838,
      "loss": 0.3807,
      "num_input_tokens_seen": 17300808,
      "step": 26520
    },
    {
      "epoch": 13.901991614255765,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.007724495716158319,
      "loss": 0.4056,
      "num_input_tokens_seen": 17303432,
      "step": 26525
    },
    {
      "epoch": 13.90461215932914,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.007718496938877294,
      "loss": 0.3712,
      "num_input_tokens_seen": 17306152,
      "step": 26530
    },
    {
      "epoch": 13.907232704402515,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.007712499684803636,
      "loss": 0.371,
      "num_input_tokens_seen": 17308840,
      "step": 26535
    },
    {
      "epoch": 13.90985324947589,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.007706503955191909,
      "loss": 0.4548,
      "num_input_tokens_seen": 17311944,
      "step": 26540
    },
    {
      "epoch": 13.912473794549266,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.007700509751296348,
      "loss": 0.3984,
      "num_input_tokens_seen": 17314888,
      "step": 26545
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.007694517074370864,
      "loss": 0.4288,
      "num_input_tokens_seen": 17318248,
      "step": 26550
    },
    {
      "epoch": 13.917714884696016,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00768852592566906,
      "loss": 0.4016,
      "num_input_tokens_seen": 17321320,
      "step": 26555
    },
    {
      "epoch": 13.920335429769391,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.007682536306444223,
      "loss": 0.3599,
      "num_input_tokens_seen": 17324680,
      "step": 26560
    },
    {
      "epoch": 13.922955974842766,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.007676548217949312,
      "loss": 0.3707,
      "num_input_tokens_seen": 17328072,
      "step": 26565
    },
    {
      "epoch": 13.925576519916142,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0076705616614369575,
      "loss": 0.4084,
      "num_input_tokens_seen": 17331272,
      "step": 26570
    },
    {
      "epoch": 13.928197064989519,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.007664576638159492,
      "loss": 0.3531,
      "num_input_tokens_seen": 17334920,
      "step": 26575
    },
    {
      "epoch": 13.930817610062894,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.007658593149368901,
      "loss": 0.4049,
      "num_input_tokens_seen": 17338664,
      "step": 26580
    },
    {
      "epoch": 13.933438155136269,
      "grad_norm": 0.083984375,
      "learning_rate": 0.007652611196316866,
      "loss": 0.4712,
      "num_input_tokens_seen": 17341000,
      "step": 26585
    },
    {
      "epoch": 13.936058700209644,
      "grad_norm": 0.05078125,
      "learning_rate": 0.007646630780254753,
      "loss": 0.3365,
      "num_input_tokens_seen": 17344456,
      "step": 26590
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.007640651902433583,
      "loss": 0.4727,
      "num_input_tokens_seen": 17347976,
      "step": 26595
    },
    {
      "epoch": 13.941299790356394,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0076346745641040735,
      "loss": 0.455,
      "num_input_tokens_seen": 17351336,
      "step": 26600
    },
    {
      "epoch": 13.94392033542977,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007628698766516625,
      "loss": 0.3359,
      "num_input_tokens_seen": 17356744,
      "step": 26605
    },
    {
      "epoch": 13.946540880503145,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0076227245109212905,
      "loss": 0.3887,
      "num_input_tokens_seen": 17359240,
      "step": 26610
    },
    {
      "epoch": 13.94916142557652,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0076167517985678244,
      "loss": 0.4108,
      "num_input_tokens_seen": 17363592,
      "step": 26615
    },
    {
      "epoch": 13.951781970649895,
      "grad_norm": 0.103515625,
      "learning_rate": 0.007610780630705656,
      "loss": 0.5326,
      "num_input_tokens_seen": 17365992,
      "step": 26620
    },
    {
      "epoch": 13.95440251572327,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.007604811008583873,
      "loss": 0.3997,
      "num_input_tokens_seen": 17368520,
      "step": 26625
    },
    {
      "epoch": 13.957023060796645,
      "grad_norm": 0.064453125,
      "learning_rate": 0.007598842933451265,
      "loss": 0.3198,
      "num_input_tokens_seen": 17371880,
      "step": 26630
    },
    {
      "epoch": 13.95964360587002,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00759287640655627,
      "loss": 0.332,
      "num_input_tokens_seen": 17374600,
      "step": 26635
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 0.08203125,
      "learning_rate": 0.007586911429147029,
      "loss": 0.3239,
      "num_input_tokens_seen": 17378600,
      "step": 26640
    },
    {
      "epoch": 13.964884696016771,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.007580948002471348,
      "loss": 0.354,
      "num_input_tokens_seen": 17381416,
      "step": 26645
    },
    {
      "epoch": 13.967505241090146,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.007574986127776698,
      "loss": 0.3255,
      "num_input_tokens_seen": 17384264,
      "step": 26650
    },
    {
      "epoch": 13.970125786163521,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.007569025806310241,
      "loss": 0.4715,
      "num_input_tokens_seen": 17387912,
      "step": 26655
    },
    {
      "epoch": 13.972746331236896,
      "grad_norm": 0.15234375,
      "learning_rate": 0.007563067039318815,
      "loss": 0.3068,
      "num_input_tokens_seen": 17390600,
      "step": 26660
    },
    {
      "epoch": 13.975366876310272,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.007557109828048912,
      "loss": 0.3645,
      "num_input_tokens_seen": 17393704,
      "step": 26665
    },
    {
      "epoch": 13.977987421383649,
      "grad_norm": 0.06640625,
      "learning_rate": 0.007551154173746726,
      "loss": 0.4428,
      "num_input_tokens_seen": 17396808,
      "step": 26670
    },
    {
      "epoch": 13.980607966457024,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.007545200077658102,
      "loss": 0.4124,
      "num_input_tokens_seen": 17400040,
      "step": 26675
    },
    {
      "epoch": 13.983228511530399,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.007539247541028574,
      "loss": 0.2991,
      "num_input_tokens_seen": 17403624,
      "step": 26680
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.007533296565103341,
      "loss": 0.5719,
      "num_input_tokens_seen": 17406760,
      "step": 26685
    },
    {
      "epoch": 13.98846960167715,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.007527347151127288,
      "loss": 0.3924,
      "num_input_tokens_seen": 17409832,
      "step": 26690
    },
    {
      "epoch": 13.991090146750524,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007521399300344953,
      "loss": 0.3723,
      "num_input_tokens_seen": 17413672,
      "step": 26695
    },
    {
      "epoch": 13.9937106918239,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00751545301400057,
      "loss": 0.4993,
      "num_input_tokens_seen": 17416936,
      "step": 26700
    },
    {
      "epoch": 13.996331236897275,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.007509508293338022,
      "loss": 0.4188,
      "num_input_tokens_seen": 17420520,
      "step": 26705
    },
    {
      "epoch": 13.99895178197065,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.007503565139600882,
      "loss": 0.3421,
      "num_input_tokens_seen": 17424840,
      "step": 26710
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.47550317645072937,
      "eval_runtime": 13.6573,
      "eval_samples_per_second": 62.092,
      "eval_steps_per_second": 15.523,
      "num_input_tokens_seen": 17425368,
      "step": 26712
    },
    {
      "epoch": 14.001572327044025,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.007497623554032398,
      "loss": 0.3675,
      "num_input_tokens_seen": 17427448,
      "step": 26715
    },
    {
      "epoch": 14.0041928721174,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00749168353787547,
      "loss": 0.3748,
      "num_input_tokens_seen": 17430008,
      "step": 26720
    },
    {
      "epoch": 14.006813417190775,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0074857450923726905,
      "loss": 0.4481,
      "num_input_tokens_seen": 17433080,
      "step": 26725
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.007479808218766305,
      "loss": 0.4005,
      "num_input_tokens_seen": 17436088,
      "step": 26730
    },
    {
      "epoch": 14.012054507337526,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.007473872918298246,
      "loss": 0.2889,
      "num_input_tokens_seen": 17438680,
      "step": 26735
    },
    {
      "epoch": 14.014675052410901,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00746793919221011,
      "loss": 0.3309,
      "num_input_tokens_seen": 17442424,
      "step": 26740
    },
    {
      "epoch": 14.017295597484276,
      "grad_norm": 0.109375,
      "learning_rate": 0.007462007041743169,
      "loss": 0.3672,
      "num_input_tokens_seen": 17445784,
      "step": 26745
    },
    {
      "epoch": 14.019916142557651,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0074560764681383595,
      "loss": 0.2669,
      "num_input_tokens_seen": 17448280,
      "step": 26750
    },
    {
      "epoch": 14.022536687631026,
      "grad_norm": 0.119140625,
      "learning_rate": 0.00745014747263628,
      "loss": 0.4133,
      "num_input_tokens_seen": 17452248,
      "step": 26755
    },
    {
      "epoch": 14.025157232704403,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.007444220056477215,
      "loss": 0.2792,
      "num_input_tokens_seen": 17455672,
      "step": 26760
    },
    {
      "epoch": 14.027777777777779,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.007438294220901114,
      "loss": 0.2713,
      "num_input_tokens_seen": 17458328,
      "step": 26765
    },
    {
      "epoch": 14.030398322851154,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.007432369967147595,
      "loss": 0.406,
      "num_input_tokens_seen": 17460856,
      "step": 26770
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.007426447296455946,
      "loss": 0.3424,
      "num_input_tokens_seen": 17464440,
      "step": 26775
    },
    {
      "epoch": 14.035639412997904,
      "grad_norm": 0.076171875,
      "learning_rate": 0.00742052621006512,
      "loss": 0.3969,
      "num_input_tokens_seen": 17467512,
      "step": 26780
    },
    {
      "epoch": 14.03825995807128,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.007414606709213734,
      "loss": 0.3769,
      "num_input_tokens_seen": 17471224,
      "step": 26785
    },
    {
      "epoch": 14.040880503144654,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.007408688795140086,
      "loss": 0.2918,
      "num_input_tokens_seen": 17474616,
      "step": 26790
    },
    {
      "epoch": 14.04350104821803,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.007402772469082135,
      "loss": 0.314,
      "num_input_tokens_seen": 17477880,
      "step": 26795
    },
    {
      "epoch": 14.046121593291405,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007396857732277517,
      "loss": 0.5086,
      "num_input_tokens_seen": 17481016,
      "step": 26800
    },
    {
      "epoch": 14.04874213836478,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.007390944585963519,
      "loss": 0.3079,
      "num_input_tokens_seen": 17484568,
      "step": 26805
    },
    {
      "epoch": 14.051362683438155,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.007385033031377099,
      "loss": 0.3273,
      "num_input_tokens_seen": 17491032,
      "step": 26810
    },
    {
      "epoch": 14.05398322851153,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.007379123069754893,
      "loss": 0.2572,
      "num_input_tokens_seen": 17494136,
      "step": 26815
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 0.08203125,
      "learning_rate": 0.007373214702333198,
      "loss": 0.4362,
      "num_input_tokens_seen": 17496920,
      "step": 26820
    },
    {
      "epoch": 14.05922431865828,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.007367307930347981,
      "loss": 0.2595,
      "num_input_tokens_seen": 17500024,
      "step": 26825
    },
    {
      "epoch": 14.061844863731656,
      "grad_norm": 0.078125,
      "learning_rate": 0.007361402755034867,
      "loss": 0.3503,
      "num_input_tokens_seen": 17503416,
      "step": 26830
    },
    {
      "epoch": 14.064465408805031,
      "grad_norm": 0.10546875,
      "learning_rate": 0.007355499177629145,
      "loss": 0.285,
      "num_input_tokens_seen": 17506648,
      "step": 26835
    },
    {
      "epoch": 14.067085953878406,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.007349597199365783,
      "loss": 0.3292,
      "num_input_tokens_seen": 17509496,
      "step": 26840
    },
    {
      "epoch": 14.069706498951781,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.007343696821479405,
      "loss": 0.3973,
      "num_input_tokens_seen": 17513464,
      "step": 26845
    },
    {
      "epoch": 14.072327044025156,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.007337798045204304,
      "loss": 0.2479,
      "num_input_tokens_seen": 17516568,
      "step": 26850
    },
    {
      "epoch": 14.074947589098532,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.007331900871774443,
      "loss": 0.4357,
      "num_input_tokens_seen": 17520440,
      "step": 26855
    },
    {
      "epoch": 14.077568134171909,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.00732600530242344,
      "loss": 0.375,
      "num_input_tokens_seen": 17523224,
      "step": 26860
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00732011133838457,
      "loss": 0.3647,
      "num_input_tokens_seen": 17530008,
      "step": 26865
    },
    {
      "epoch": 14.082809224318659,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.00731421898089079,
      "loss": 0.3637,
      "num_input_tokens_seen": 17533048,
      "step": 26870
    },
    {
      "epoch": 14.085429769392034,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.007308328231174718,
      "loss": 0.2923,
      "num_input_tokens_seen": 17535608,
      "step": 26875
    },
    {
      "epoch": 14.08805031446541,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0073024390904686326,
      "loss": 0.323,
      "num_input_tokens_seen": 17539096,
      "step": 26880
    },
    {
      "epoch": 14.090670859538784,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.007296551560004472,
      "loss": 0.4681,
      "num_input_tokens_seen": 17542936,
      "step": 26885
    },
    {
      "epoch": 14.09329140461216,
      "grad_norm": 0.076171875,
      "learning_rate": 0.007290665641013833,
      "loss": 0.4404,
      "num_input_tokens_seen": 17545688,
      "step": 26890
    },
    {
      "epoch": 14.095911949685535,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.007284781334727989,
      "loss": 0.3585,
      "num_input_tokens_seen": 17548536,
      "step": 26895
    },
    {
      "epoch": 14.09853249475891,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00727889864237787,
      "loss": 0.3334,
      "num_input_tokens_seen": 17552056,
      "step": 26900
    },
    {
      "epoch": 14.101153039832285,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.007273017565194074,
      "loss": 0.3178,
      "num_input_tokens_seen": 17555512,
      "step": 26905
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0072671381044068445,
      "loss": 0.3275,
      "num_input_tokens_seen": 17559320,
      "step": 26910
    },
    {
      "epoch": 14.106394129979035,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0072612602612461075,
      "loss": 0.2765,
      "num_input_tokens_seen": 17562616,
      "step": 26915
    },
    {
      "epoch": 14.10901467505241,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.007255384036941429,
      "loss": 0.4084,
      "num_input_tokens_seen": 17566200,
      "step": 26920
    },
    {
      "epoch": 14.111635220125786,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.007249509432722056,
      "loss": 0.3329,
      "num_input_tokens_seen": 17569496,
      "step": 26925
    },
    {
      "epoch": 14.114255765199161,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.007243636449816886,
      "loss": 0.3789,
      "num_input_tokens_seen": 17571832,
      "step": 26930
    },
    {
      "epoch": 14.116876310272536,
      "grad_norm": 0.09375,
      "learning_rate": 0.007237765089454489,
      "loss": 0.3641,
      "num_input_tokens_seen": 17574744,
      "step": 26935
    },
    {
      "epoch": 14.119496855345911,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0072318953528630716,
      "loss": 0.3871,
      "num_input_tokens_seen": 17578232,
      "step": 26940
    },
    {
      "epoch": 14.122117400419286,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.007226027241270531,
      "loss": 0.4351,
      "num_input_tokens_seen": 17581688,
      "step": 26945
    },
    {
      "epoch": 14.124737945492662,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0072201607559043935,
      "loss": 0.3706,
      "num_input_tokens_seen": 17585336,
      "step": 26950
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.00721429589799187,
      "loss": 0.3262,
      "num_input_tokens_seen": 17588120,
      "step": 26955
    },
    {
      "epoch": 14.129979035639414,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.007208432668759825,
      "loss": 0.378,
      "num_input_tokens_seen": 17591928,
      "step": 26960
    },
    {
      "epoch": 14.132599580712789,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.007202571069434772,
      "loss": 0.2809,
      "num_input_tokens_seen": 17596120,
      "step": 26965
    },
    {
      "epoch": 14.135220125786164,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00719671110124289,
      "loss": 0.4464,
      "num_input_tokens_seen": 17599416,
      "step": 26970
    },
    {
      "epoch": 14.13784067085954,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.007190852765410029,
      "loss": 0.4417,
      "num_input_tokens_seen": 17603192,
      "step": 26975
    },
    {
      "epoch": 14.140461215932914,
      "grad_norm": 0.091796875,
      "learning_rate": 0.007184996063161671,
      "loss": 0.3618,
      "num_input_tokens_seen": 17606520,
      "step": 26980
    },
    {
      "epoch": 14.14308176100629,
      "grad_norm": 0.095703125,
      "learning_rate": 0.007179140995722984,
      "loss": 0.3484,
      "num_input_tokens_seen": 17609112,
      "step": 26985
    },
    {
      "epoch": 14.145702306079665,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.007173287564318771,
      "loss": 0.3961,
      "num_input_tokens_seen": 17612440,
      "step": 26990
    },
    {
      "epoch": 14.14832285115304,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007167435770173507,
      "loss": 0.3353,
      "num_input_tokens_seen": 17615384,
      "step": 26995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.007161585614511327,
      "loss": 0.2875,
      "num_input_tokens_seen": 17618456,
      "step": 27000
    },
    {
      "epoch": 14.15356394129979,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.007155737098556004,
      "loss": 0.6207,
      "num_input_tokens_seen": 17621720,
      "step": 27005
    },
    {
      "epoch": 14.156184486373165,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00714989022353099,
      "loss": 0.3948,
      "num_input_tokens_seen": 17625112,
      "step": 27010
    },
    {
      "epoch": 14.15880503144654,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.007144044990659388,
      "loss": 0.4852,
      "num_input_tokens_seen": 17628824,
      "step": 27015
    },
    {
      "epoch": 14.161425576519916,
      "grad_norm": 0.146484375,
      "learning_rate": 0.007138201401163943,
      "loss": 0.3107,
      "num_input_tokens_seen": 17632568,
      "step": 27020
    },
    {
      "epoch": 14.164046121593291,
      "grad_norm": 0.046875,
      "learning_rate": 0.007132359456267072,
      "loss": 0.3264,
      "num_input_tokens_seen": 17635928,
      "step": 27025
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.007126519157190849,
      "loss": 0.4282,
      "num_input_tokens_seen": 17639832,
      "step": 27030
    },
    {
      "epoch": 14.169287211740041,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.007120680505156988,
      "loss": 0.3266,
      "num_input_tokens_seen": 17643160,
      "step": 27035
    },
    {
      "epoch": 14.171907756813416,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.007114843501386878,
      "loss": 0.4666,
      "num_input_tokens_seen": 17646040,
      "step": 27040
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0071090081471015445,
      "loss": 0.3347,
      "num_input_tokens_seen": 17648600,
      "step": 27045
    },
    {
      "epoch": 14.177148846960169,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0071031744435216815,
      "loss": 0.3786,
      "num_input_tokens_seen": 17652024,
      "step": 27050
    },
    {
      "epoch": 14.179769392033544,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0070973423918676325,
      "loss": 0.324,
      "num_input_tokens_seen": 17655160,
      "step": 27055
    },
    {
      "epoch": 14.182389937106919,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.007091511993359403,
      "loss": 0.3765,
      "num_input_tokens_seen": 17658712,
      "step": 27060
    },
    {
      "epoch": 14.185010482180294,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.007085683249216639,
      "loss": 0.3101,
      "num_input_tokens_seen": 17661432,
      "step": 27065
    },
    {
      "epoch": 14.18763102725367,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.007079856160658643,
      "loss": 0.4418,
      "num_input_tokens_seen": 17664408,
      "step": 27070
    },
    {
      "epoch": 14.190251572327044,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.00707403072890438,
      "loss": 0.3174,
      "num_input_tokens_seen": 17667608,
      "step": 27075
    },
    {
      "epoch": 14.19287211740042,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.007068206955172465,
      "loss": 0.3922,
      "num_input_tokens_seen": 17670872,
      "step": 27080
    },
    {
      "epoch": 14.195492662473795,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0070623848406811685,
      "loss": 0.3845,
      "num_input_tokens_seen": 17673368,
      "step": 27085
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.007056564386648401,
      "loss": 0.2882,
      "num_input_tokens_seen": 17675800,
      "step": 27090
    },
    {
      "epoch": 14.200733752620545,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0070507455942917475,
      "loss": 0.3277,
      "num_input_tokens_seen": 17679288,
      "step": 27095
    },
    {
      "epoch": 14.20335429769392,
      "grad_norm": 0.142578125,
      "learning_rate": 0.007044928464828419,
      "loss": 0.422,
      "num_input_tokens_seen": 17682392,
      "step": 27100
    },
    {
      "epoch": 14.205974842767295,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.007039112999475299,
      "loss": 0.3696,
      "num_input_tokens_seen": 17685176,
      "step": 27105
    },
    {
      "epoch": 14.20859538784067,
      "grad_norm": 0.078125,
      "learning_rate": 0.007033299199448918,
      "loss": 0.3036,
      "num_input_tokens_seen": 17687768,
      "step": 27110
    },
    {
      "epoch": 14.211215932914046,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.00702748706596546,
      "loss": 0.2832,
      "num_input_tokens_seen": 17690712,
      "step": 27115
    },
    {
      "epoch": 14.213836477987421,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007021676600240754,
      "loss": 0.5951,
      "num_input_tokens_seen": 17693944,
      "step": 27120
    },
    {
      "epoch": 14.216457023060796,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.007015867803490274,
      "loss": 0.3431,
      "num_input_tokens_seen": 17697528,
      "step": 27125
    },
    {
      "epoch": 14.219077568134171,
      "grad_norm": 0.091796875,
      "learning_rate": 0.007010060676929163,
      "loss": 0.4222,
      "num_input_tokens_seen": 17700472,
      "step": 27130
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 0.078125,
      "learning_rate": 0.007004255221772204,
      "loss": 0.3344,
      "num_input_tokens_seen": 17704024,
      "step": 27135
    },
    {
      "epoch": 14.224318658280922,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.006998451439233839,
      "loss": 0.3084,
      "num_input_tokens_seen": 17706840,
      "step": 27140
    },
    {
      "epoch": 14.226939203354299,
      "grad_norm": 0.07421875,
      "learning_rate": 0.006992649330528146,
      "loss": 0.4781,
      "num_input_tokens_seen": 17709592,
      "step": 27145
    },
    {
      "epoch": 14.229559748427674,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.006986848896868852,
      "loss": 0.4006,
      "num_input_tokens_seen": 17713112,
      "step": 27150
    },
    {
      "epoch": 14.232180293501049,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0069810501394693535,
      "loss": 0.4583,
      "num_input_tokens_seen": 17716056,
      "step": 27155
    },
    {
      "epoch": 14.234800838574424,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.006975253059542679,
      "loss": 0.3706,
      "num_input_tokens_seen": 17719704,
      "step": 27160
    },
    {
      "epoch": 14.2374213836478,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.006969457658301514,
      "loss": 0.3469,
      "num_input_tokens_seen": 17722456,
      "step": 27165
    },
    {
      "epoch": 14.240041928721174,
      "grad_norm": 0.05078125,
      "learning_rate": 0.006963663936958196,
      "loss": 0.3657,
      "num_input_tokens_seen": 17725400,
      "step": 27170
    },
    {
      "epoch": 14.24266247379455,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.006957871896724697,
      "loss": 0.4805,
      "num_input_tokens_seen": 17728440,
      "step": 27175
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 0.220703125,
      "learning_rate": 0.0069520815388126435,
      "loss": 0.2941,
      "num_input_tokens_seen": 17731192,
      "step": 27180
    },
    {
      "epoch": 14.2479035639413,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.006946292864433317,
      "loss": 0.3081,
      "num_input_tokens_seen": 17735096,
      "step": 27185
    },
    {
      "epoch": 14.250524109014675,
      "grad_norm": 0.083984375,
      "learning_rate": 0.00694050587479764,
      "loss": 0.4235,
      "num_input_tokens_seen": 17739000,
      "step": 27190
    },
    {
      "epoch": 14.25314465408805,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0069347205711161935,
      "loss": 0.3088,
      "num_input_tokens_seen": 17741784,
      "step": 27195
    },
    {
      "epoch": 14.255765199161425,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.00692893695459919,
      "loss": 0.4639,
      "num_input_tokens_seen": 17744440,
      "step": 27200
    },
    {
      "epoch": 14.2583857442348,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.006923155026456488,
      "loss": 0.5218,
      "num_input_tokens_seen": 17747352,
      "step": 27205
    },
    {
      "epoch": 14.261006289308176,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.006917374787897612,
      "loss": 0.3527,
      "num_input_tokens_seen": 17750104,
      "step": 27210
    },
    {
      "epoch": 14.26362683438155,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.006911596240131718,
      "loss": 0.3136,
      "num_input_tokens_seen": 17752728,
      "step": 27215
    },
    {
      "epoch": 14.266247379454926,
      "grad_norm": 0.064453125,
      "learning_rate": 0.006905819384367618,
      "loss": 0.304,
      "num_input_tokens_seen": 17756248,
      "step": 27220
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.006900044221813755,
      "loss": 0.338,
      "num_input_tokens_seen": 17759704,
      "step": 27225
    },
    {
      "epoch": 14.271488469601676,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.006894270753678236,
      "loss": 0.342,
      "num_input_tokens_seen": 17763064,
      "step": 27230
    },
    {
      "epoch": 14.274109014675052,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.006888498981168796,
      "loss": 0.4452,
      "num_input_tokens_seen": 17767384,
      "step": 27235
    },
    {
      "epoch": 14.276729559748428,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.006882728905492827,
      "loss": 0.4353,
      "num_input_tokens_seen": 17770328,
      "step": 27240
    },
    {
      "epoch": 14.279350104821804,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.006876960527857366,
      "loss": 0.3345,
      "num_input_tokens_seen": 17774072,
      "step": 27245
    },
    {
      "epoch": 14.281970649895179,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.006871193849469094,
      "loss": 0.2199,
      "num_input_tokens_seen": 17776344,
      "step": 27250
    },
    {
      "epoch": 14.284591194968554,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.006865428871534331,
      "loss": 0.4193,
      "num_input_tokens_seen": 17780536,
      "step": 27255
    },
    {
      "epoch": 14.28721174004193,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.006859665595259037,
      "loss": 0.4238,
      "num_input_tokens_seen": 17784024,
      "step": 27260
    },
    {
      "epoch": 14.289832285115304,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.006853904021848833,
      "loss": 0.5258,
      "num_input_tokens_seen": 17788248,
      "step": 27265
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.006848144152508971,
      "loss": 0.3572,
      "num_input_tokens_seen": 17790872,
      "step": 27270
    },
    {
      "epoch": 14.295073375262055,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.006842385988444356,
      "loss": 0.3583,
      "num_input_tokens_seen": 17793912,
      "step": 27275
    },
    {
      "epoch": 14.29769392033543,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.00683662953085952,
      "loss": 0.3519,
      "num_input_tokens_seen": 17797336,
      "step": 27280
    },
    {
      "epoch": 14.300314465408805,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0068308747809586595,
      "loss": 0.4086,
      "num_input_tokens_seen": 17800728,
      "step": 27285
    },
    {
      "epoch": 14.30293501048218,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.00682512173994559,
      "loss": 0.5303,
      "num_input_tokens_seen": 17803448,
      "step": 27290
    },
    {
      "epoch": 14.305555555555555,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.00681937040902379,
      "loss": 0.302,
      "num_input_tokens_seen": 17807896,
      "step": 27295
    },
    {
      "epoch": 14.30817610062893,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.006813620789396376,
      "loss": 0.3628,
      "num_input_tokens_seen": 17811032,
      "step": 27300
    },
    {
      "epoch": 14.310796645702306,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.006807872882266092,
      "loss": 0.3917,
      "num_input_tokens_seen": 17815640,
      "step": 27305
    },
    {
      "epoch": 14.31341719077568,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.006802126688835339,
      "loss": 0.3512,
      "num_input_tokens_seen": 17818744,
      "step": 27310
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 0.072265625,
      "learning_rate": 0.006796382210306163,
      "loss": 0.3122,
      "num_input_tokens_seen": 17821432,
      "step": 27315
    },
    {
      "epoch": 14.318658280922431,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.006790639447880232,
      "loss": 0.3975,
      "num_input_tokens_seen": 17824312,
      "step": 27320
    },
    {
      "epoch": 14.321278825995806,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.00678489840275887,
      "loss": 0.2713,
      "num_input_tokens_seen": 17827288,
      "step": 27325
    },
    {
      "epoch": 14.323899371069182,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.006779159076143046,
      "loss": 0.3252,
      "num_input_tokens_seen": 17830392,
      "step": 27330
    },
    {
      "epoch": 14.326519916142558,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.006773421469233348,
      "loss": 0.35,
      "num_input_tokens_seen": 17833432,
      "step": 27335
    },
    {
      "epoch": 14.329140461215934,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.006767685583230026,
      "loss": 0.386,
      "num_input_tokens_seen": 17837112,
      "step": 27340
    },
    {
      "epoch": 14.331761006289309,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.006761951419332967,
      "loss": 0.3124,
      "num_input_tokens_seen": 17840088,
      "step": 27345
    },
    {
      "epoch": 14.334381551362684,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.006756218978741679,
      "loss": 0.3267,
      "num_input_tokens_seen": 17843000,
      "step": 27350
    },
    {
      "epoch": 14.33700209643606,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.006750488262655338,
      "loss": 0.3777,
      "num_input_tokens_seen": 17846872,
      "step": 27355
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 0.0703125,
      "learning_rate": 0.006744759272272729,
      "loss": 0.3902,
      "num_input_tokens_seen": 17850232,
      "step": 27360
    },
    {
      "epoch": 14.34224318658281,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.006739032008792302,
      "loss": 0.5985,
      "num_input_tokens_seen": 17854360,
      "step": 27365
    },
    {
      "epoch": 14.344863731656185,
      "grad_norm": 0.06640625,
      "learning_rate": 0.006733306473412137,
      "loss": 0.3786,
      "num_input_tokens_seen": 17858040,
      "step": 27370
    },
    {
      "epoch": 14.34748427672956,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0067275826673299415,
      "loss": 0.4519,
      "num_input_tokens_seen": 17861400,
      "step": 27375
    },
    {
      "epoch": 14.350104821802935,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.00672186059174308,
      "loss": 0.5922,
      "num_input_tokens_seen": 17865176,
      "step": 27380
    },
    {
      "epoch": 14.35272536687631,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.006716140247848536,
      "loss": 0.4565,
      "num_input_tokens_seen": 17867544,
      "step": 27385
    },
    {
      "epoch": 14.355345911949685,
      "grad_norm": 0.076171875,
      "learning_rate": 0.006710421636842943,
      "loss": 0.3164,
      "num_input_tokens_seen": 17870584,
      "step": 27390
    },
    {
      "epoch": 14.35796645702306,
      "grad_norm": 0.068359375,
      "learning_rate": 0.006704704759922573,
      "loss": 0.3381,
      "num_input_tokens_seen": 17873560,
      "step": 27395
    },
    {
      "epoch": 14.360587002096436,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.006698989618283335,
      "loss": 0.4578,
      "num_input_tokens_seen": 17876824,
      "step": 27400
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0066932762131207595,
      "loss": 0.3887,
      "num_input_tokens_seen": 17879928,
      "step": 27405
    },
    {
      "epoch": 14.365828092243186,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0066875645456300386,
      "loss": 0.4294,
      "num_input_tokens_seen": 17883672,
      "step": 27410
    },
    {
      "epoch": 14.368448637316561,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.006681854617005975,
      "loss": 0.329,
      "num_input_tokens_seen": 17887032,
      "step": 27415
    },
    {
      "epoch": 14.371069182389936,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.006676146428443027,
      "loss": 0.3491,
      "num_input_tokens_seen": 17889816,
      "step": 27420
    },
    {
      "epoch": 14.373689727463312,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.006670439981135283,
      "loss": 0.3831,
      "num_input_tokens_seen": 17893016,
      "step": 27425
    },
    {
      "epoch": 14.376310272536688,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.006664735276276471,
      "loss": 0.3327,
      "num_input_tokens_seen": 17896056,
      "step": 27430
    },
    {
      "epoch": 14.378930817610064,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.006659032315059947,
      "loss": 0.4601,
      "num_input_tokens_seen": 17899224,
      "step": 27435
    },
    {
      "epoch": 14.381551362683439,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.006653331098678697,
      "loss": 0.4662,
      "num_input_tokens_seen": 17902840,
      "step": 27440
    },
    {
      "epoch": 14.384171907756814,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.006647631628325358,
      "loss": 0.4324,
      "num_input_tokens_seen": 17905272,
      "step": 27445
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.006641933905192194,
      "loss": 0.4069,
      "num_input_tokens_seen": 17907992,
      "step": 27450
    },
    {
      "epoch": 14.389412997903564,
      "grad_norm": 0.064453125,
      "learning_rate": 0.006636237930471109,
      "loss": 0.316,
      "num_input_tokens_seen": 17910904,
      "step": 27455
    },
    {
      "epoch": 14.39203354297694,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.006630543705353629,
      "loss": 0.3148,
      "num_input_tokens_seen": 17914904,
      "step": 27460
    },
    {
      "epoch": 14.394654088050315,
      "grad_norm": 0.078125,
      "learning_rate": 0.006624851231030918,
      "loss": 0.3649,
      "num_input_tokens_seen": 17918104,
      "step": 27465
    },
    {
      "epoch": 14.39727463312369,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.006619160508693778,
      "loss": 0.3828,
      "num_input_tokens_seen": 17921720,
      "step": 27470
    },
    {
      "epoch": 14.399895178197065,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.006613471539532648,
      "loss": 0.3766,
      "num_input_tokens_seen": 17924984,
      "step": 27475
    },
    {
      "epoch": 14.40251572327044,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.006607784324737594,
      "loss": 0.373,
      "num_input_tokens_seen": 17927672,
      "step": 27480
    },
    {
      "epoch": 14.405136268343815,
      "grad_norm": 0.041015625,
      "learning_rate": 0.006602098865498319,
      "loss": 0.478,
      "num_input_tokens_seen": 17931128,
      "step": 27485
    },
    {
      "epoch": 14.40775681341719,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.006596415163004154,
      "loss": 0.385,
      "num_input_tokens_seen": 17934936,
      "step": 27490
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0065907332184440585,
      "loss": 0.2662,
      "num_input_tokens_seen": 17937432,
      "step": 27495
    },
    {
      "epoch": 14.41299790356394,
      "grad_norm": 0.1328125,
      "learning_rate": 0.006585053033006635,
      "loss": 0.3635,
      "num_input_tokens_seen": 17939800,
      "step": 27500
    },
    {
      "epoch": 14.415618448637316,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.006579374607880116,
      "loss": 0.3776,
      "num_input_tokens_seen": 17942808,
      "step": 27505
    },
    {
      "epoch": 14.418238993710691,
      "grad_norm": 0.087890625,
      "learning_rate": 0.006573697944252366,
      "loss": 0.3875,
      "num_input_tokens_seen": 17945592,
      "step": 27510
    },
    {
      "epoch": 14.420859538784066,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.006568023043310873,
      "loss": 0.444,
      "num_input_tokens_seen": 17948536,
      "step": 27515
    },
    {
      "epoch": 14.423480083857442,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0065623499062427565,
      "loss": 0.378,
      "num_input_tokens_seen": 17952408,
      "step": 27520
    },
    {
      "epoch": 14.426100628930818,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0065566785342347795,
      "loss": 0.3862,
      "num_input_tokens_seen": 17955896,
      "step": 27525
    },
    {
      "epoch": 14.428721174004194,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.006551008928473325,
      "loss": 0.4407,
      "num_input_tokens_seen": 17963128,
      "step": 27530
    },
    {
      "epoch": 14.431341719077569,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.006545341090144419,
      "loss": 0.3949,
      "num_input_tokens_seen": 17965688,
      "step": 27535
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0065396750204337,
      "loss": 0.2855,
      "num_input_tokens_seen": 17968888,
      "step": 27540
    },
    {
      "epoch": 14.43658280922432,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0065340107205264395,
      "loss": 0.3501,
      "num_input_tokens_seen": 17972120,
      "step": 27545
    },
    {
      "epoch": 14.439203354297694,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.006528348191607552,
      "loss": 0.3272,
      "num_input_tokens_seen": 17975608,
      "step": 27550
    },
    {
      "epoch": 14.44182389937107,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.006522687434861574,
      "loss": 0.4762,
      "num_input_tokens_seen": 17979128,
      "step": 27555
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.006517028451472669,
      "loss": 0.4166,
      "num_input_tokens_seen": 17982456,
      "step": 27560
    },
    {
      "epoch": 14.44706498951782,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00651137124262464,
      "loss": 0.4615,
      "num_input_tokens_seen": 17985720,
      "step": 27565
    },
    {
      "epoch": 14.449685534591195,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0065057158095009025,
      "loss": 0.3656,
      "num_input_tokens_seen": 17989272,
      "step": 27570
    },
    {
      "epoch": 14.45230607966457,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.006500062153284504,
      "loss": 0.4448,
      "num_input_tokens_seen": 17992408,
      "step": 27575
    },
    {
      "epoch": 14.454926624737945,
      "grad_norm": 0.0625,
      "learning_rate": 0.006494410275158132,
      "loss": 0.3504,
      "num_input_tokens_seen": 17996568,
      "step": 27580
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0064887601763040925,
      "loss": 0.3405,
      "num_input_tokens_seen": 17999832,
      "step": 27585
    },
    {
      "epoch": 14.460167714884696,
      "grad_norm": 0.052734375,
      "learning_rate": 0.006483111857904329,
      "loss": 0.4724,
      "num_input_tokens_seen": 18003704,
      "step": 27590
    },
    {
      "epoch": 14.46278825995807,
      "grad_norm": 0.078125,
      "learning_rate": 0.006477465321140392,
      "loss": 0.3153,
      "num_input_tokens_seen": 18006328,
      "step": 27595
    },
    {
      "epoch": 14.465408805031446,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.006471820567193487,
      "loss": 0.4183,
      "num_input_tokens_seen": 18010008,
      "step": 27600
    },
    {
      "epoch": 14.468029350104821,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.006466177597244418,
      "loss": 0.4319,
      "num_input_tokens_seen": 18013880,
      "step": 27605
    },
    {
      "epoch": 14.470649895178196,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.006460536412473637,
      "loss": 0.3772,
      "num_input_tokens_seen": 18016824,
      "step": 27610
    },
    {
      "epoch": 14.473270440251572,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.006454897014061221,
      "loss": 0.3734,
      "num_input_tokens_seen": 18020472,
      "step": 27615
    },
    {
      "epoch": 14.475890985324948,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.006449259403186856,
      "loss": 0.2746,
      "num_input_tokens_seen": 18023128,
      "step": 27620
    },
    {
      "epoch": 14.478511530398324,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.006443623581029877,
      "loss": 0.4259,
      "num_input_tokens_seen": 18026872,
      "step": 27625
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 0.064453125,
      "learning_rate": 0.006437989548769224,
      "loss": 0.4684,
      "num_input_tokens_seen": 18030200,
      "step": 27630
    },
    {
      "epoch": 14.483752620545074,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.006432357307583477,
      "loss": 0.2777,
      "num_input_tokens_seen": 18033528,
      "step": 27635
    },
    {
      "epoch": 14.48637316561845,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.006426726858650836,
      "loss": 0.3806,
      "num_input_tokens_seen": 18036088,
      "step": 27640
    },
    {
      "epoch": 14.488993710691824,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.006421098203149131,
      "loss": 0.3752,
      "num_input_tokens_seen": 18040408,
      "step": 27645
    },
    {
      "epoch": 14.4916142557652,
      "grad_norm": 0.083984375,
      "learning_rate": 0.006415471342255805,
      "loss": 0.31,
      "num_input_tokens_seen": 18043096,
      "step": 27650
    },
    {
      "epoch": 14.494234800838575,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.006409846277147942,
      "loss": 0.3943,
      "num_input_tokens_seen": 18046232,
      "step": 27655
    },
    {
      "epoch": 14.49685534591195,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.00640422300900223,
      "loss": 0.4311,
      "num_input_tokens_seen": 18048792,
      "step": 27660
    },
    {
      "epoch": 14.499475890985325,
      "grad_norm": 0.064453125,
      "learning_rate": 0.006398601538995,
      "loss": 0.3914,
      "num_input_tokens_seen": 18052696,
      "step": 27665
    },
    {
      "epoch": 14.5020964360587,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.006392981868302202,
      "loss": 0.3079,
      "num_input_tokens_seen": 18055256,
      "step": 27670
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 0.060546875,
      "learning_rate": 0.006387363998099399,
      "loss": 0.3231,
      "num_input_tokens_seen": 18057784,
      "step": 27675
    },
    {
      "epoch": 14.50733752620545,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.006381747929561789,
      "loss": 0.497,
      "num_input_tokens_seen": 18060920,
      "step": 27680
    },
    {
      "epoch": 14.509958071278826,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.006376133663864195,
      "loss": 0.3702,
      "num_input_tokens_seen": 18064184,
      "step": 27685
    },
    {
      "epoch": 14.5125786163522,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.006370521202181047,
      "loss": 0.441,
      "num_input_tokens_seen": 18067672,
      "step": 27690
    },
    {
      "epoch": 14.515199161425576,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0063649105456864185,
      "loss": 0.3809,
      "num_input_tokens_seen": 18071032,
      "step": 27695
    },
    {
      "epoch": 14.517819706498951,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0063593016955539846,
      "loss": 0.584,
      "num_input_tokens_seen": 18074200,
      "step": 27700
    },
    {
      "epoch": 14.520440251572326,
      "grad_norm": 0.087890625,
      "learning_rate": 0.006353694652957057,
      "loss": 0.3422,
      "num_input_tokens_seen": 18076824,
      "step": 27705
    },
    {
      "epoch": 14.523060796645701,
      "grad_norm": 0.13671875,
      "learning_rate": 0.006348089419068567,
      "loss": 0.4035,
      "num_input_tokens_seen": 18079512,
      "step": 27710
    },
    {
      "epoch": 14.525681341719078,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0063424859950610675,
      "loss": 0.3244,
      "num_input_tokens_seen": 18082424,
      "step": 27715
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.006336884382106724,
      "loss": 0.3393,
      "num_input_tokens_seen": 18085400,
      "step": 27720
    },
    {
      "epoch": 14.530922431865829,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.006331284581377339,
      "loss": 0.5253,
      "num_input_tokens_seen": 18088664,
      "step": 27725
    },
    {
      "epoch": 14.533542976939204,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.006325686594044315,
      "loss": 0.3811,
      "num_input_tokens_seen": 18091512,
      "step": 27730
    },
    {
      "epoch": 14.536163522012579,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.006320090421278695,
      "loss": 0.5259,
      "num_input_tokens_seen": 18093752,
      "step": 27735
    },
    {
      "epoch": 14.538784067085954,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0063144960642511384,
      "loss": 0.4378,
      "num_input_tokens_seen": 18096440,
      "step": 27740
    },
    {
      "epoch": 14.54140461215933,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00630890352413191,
      "loss": 0.3037,
      "num_input_tokens_seen": 18099384,
      "step": 27745
    },
    {
      "epoch": 14.544025157232705,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006303312802090916,
      "loss": 0.3583,
      "num_input_tokens_seen": 18102744,
      "step": 27750
    },
    {
      "epoch": 14.54664570230608,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.006297723899297663,
      "loss": 0.4313,
      "num_input_tokens_seen": 18105720,
      "step": 27755
    },
    {
      "epoch": 14.549266247379455,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.006292136816921289,
      "loss": 0.2715,
      "num_input_tokens_seen": 18108536,
      "step": 27760
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.006286551556130547,
      "loss": 0.3964,
      "num_input_tokens_seen": 18110904,
      "step": 27765
    },
    {
      "epoch": 14.554507337526205,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.006280968118093817,
      "loss": 0.3369,
      "num_input_tokens_seen": 18113944,
      "step": 27770
    },
    {
      "epoch": 14.55712788259958,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.006275386503979085,
      "loss": 0.4133,
      "num_input_tokens_seen": 18117304,
      "step": 27775
    },
    {
      "epoch": 14.559748427672956,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.006269806714953956,
      "loss": 0.4172,
      "num_input_tokens_seen": 18119704,
      "step": 27780
    },
    {
      "epoch": 14.56236897274633,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.006264228752185663,
      "loss": 0.4798,
      "num_input_tokens_seen": 18124312,
      "step": 27785
    },
    {
      "epoch": 14.564989517819706,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.006258652616841051,
      "loss": 0.3431,
      "num_input_tokens_seen": 18128408,
      "step": 27790
    },
    {
      "epoch": 14.567610062893081,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.006253078310086586,
      "loss": 0.3333,
      "num_input_tokens_seen": 18131416,
      "step": 27795
    },
    {
      "epoch": 14.570230607966456,
      "grad_norm": 0.0625,
      "learning_rate": 0.006247505833088352,
      "loss": 0.4653,
      "num_input_tokens_seen": 18134872,
      "step": 27800
    },
    {
      "epoch": 14.572851153039831,
      "grad_norm": 0.04296875,
      "learning_rate": 0.006241935187012045,
      "loss": 0.2573,
      "num_input_tokens_seen": 18138776,
      "step": 27805
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0062363663730229755,
      "loss": 0.4096,
      "num_input_tokens_seen": 18141752,
      "step": 27810
    },
    {
      "epoch": 14.578092243186584,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.006230799392286077,
      "loss": 0.2742,
      "num_input_tokens_seen": 18144152,
      "step": 27815
    },
    {
      "epoch": 14.580712788259959,
      "grad_norm": 0.1015625,
      "learning_rate": 0.006225234245965904,
      "loss": 0.3806,
      "num_input_tokens_seen": 18147192,
      "step": 27820
    },
    {
      "epoch": 14.583333333333334,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0062196709352266245,
      "loss": 0.3352,
      "num_input_tokens_seen": 18150456,
      "step": 27825
    },
    {
      "epoch": 14.585953878406709,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0062141094612320116,
      "loss": 0.3667,
      "num_input_tokens_seen": 18153464,
      "step": 27830
    },
    {
      "epoch": 14.588574423480084,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.006208549825145461,
      "loss": 0.4358,
      "num_input_tokens_seen": 18157336,
      "step": 27835
    },
    {
      "epoch": 14.59119496855346,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.006202992028129989,
      "loss": 0.3063,
      "num_input_tokens_seen": 18161016,
      "step": 27840
    },
    {
      "epoch": 14.593815513626835,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.006197436071348223,
      "loss": 0.4146,
      "num_input_tokens_seen": 18163480,
      "step": 27845
    },
    {
      "epoch": 14.59643605870021,
      "grad_norm": 0.09375,
      "learning_rate": 0.006191881955962411,
      "loss": 0.3591,
      "num_input_tokens_seen": 18166584,
      "step": 27850
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 0.072265625,
      "learning_rate": 0.006186329683134405,
      "loss": 0.3101,
      "num_input_tokens_seen": 18169144,
      "step": 27855
    },
    {
      "epoch": 14.60167714884696,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.006180779254025672,
      "loss": 0.423,
      "num_input_tokens_seen": 18172184,
      "step": 27860
    },
    {
      "epoch": 14.604297693920335,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.006175230669797306,
      "loss": 0.3424,
      "num_input_tokens_seen": 18174840,
      "step": 27865
    },
    {
      "epoch": 14.60691823899371,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.006169683931610004,
      "loss": 0.2995,
      "num_input_tokens_seen": 18178584,
      "step": 27870
    },
    {
      "epoch": 14.609538784067086,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.006164139040624083,
      "loss": 0.3768,
      "num_input_tokens_seen": 18184792,
      "step": 27875
    },
    {
      "epoch": 14.61215932914046,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.006158595997999474,
      "loss": 0.3713,
      "num_input_tokens_seen": 18188248,
      "step": 27880
    },
    {
      "epoch": 14.614779874213836,
      "grad_norm": 0.078125,
      "learning_rate": 0.006153054804895715,
      "loss": 0.3094,
      "num_input_tokens_seen": 18191160,
      "step": 27885
    },
    {
      "epoch": 14.617400419287211,
      "grad_norm": 0.06640625,
      "learning_rate": 0.006147515462471953,
      "loss": 0.313,
      "num_input_tokens_seen": 18195000,
      "step": 27890
    },
    {
      "epoch": 14.620020964360586,
      "grad_norm": 0.103515625,
      "learning_rate": 0.006141977971886962,
      "loss": 0.3075,
      "num_input_tokens_seen": 18198232,
      "step": 27895
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.006136442334299122,
      "loss": 0.4708,
      "num_input_tokens_seen": 18200856,
      "step": 27900
    },
    {
      "epoch": 14.625262054507338,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0061309085508664315,
      "loss": 0.3285,
      "num_input_tokens_seen": 18203256,
      "step": 27905
    },
    {
      "epoch": 14.627882599580714,
      "grad_norm": 0.205078125,
      "learning_rate": 0.006125376622746487,
      "loss": 0.4712,
      "num_input_tokens_seen": 18206008,
      "step": 27910
    },
    {
      "epoch": 14.630503144654089,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.006119846551096502,
      "loss": 0.4259,
      "num_input_tokens_seen": 18208984,
      "step": 27915
    },
    {
      "epoch": 14.633123689727464,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0061143183370733056,
      "loss": 0.2775,
      "num_input_tokens_seen": 18213848,
      "step": 27920
    },
    {
      "epoch": 14.635744234800839,
      "grad_norm": 0.0703125,
      "learning_rate": 0.006108791981833341,
      "loss": 0.3874,
      "num_input_tokens_seen": 18217496,
      "step": 27925
    },
    {
      "epoch": 14.638364779874214,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.006103267486532664,
      "loss": 0.3639,
      "num_input_tokens_seen": 18220536,
      "step": 27930
    },
    {
      "epoch": 14.64098532494759,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.006097744852326923,
      "loss": 0.488,
      "num_input_tokens_seen": 18224888,
      "step": 27935
    },
    {
      "epoch": 14.643605870020965,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006092224080371403,
      "loss": 0.2806,
      "num_input_tokens_seen": 18229368,
      "step": 27940
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.006086705171820974,
      "loss": 0.3751,
      "num_input_tokens_seen": 18232760,
      "step": 27945
    },
    {
      "epoch": 14.648846960167715,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.006081188127830135,
      "loss": 0.3909,
      "num_input_tokens_seen": 18235576,
      "step": 27950
    },
    {
      "epoch": 14.65146750524109,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.006075672949552989,
      "loss": 0.3899,
      "num_input_tokens_seen": 18239224,
      "step": 27955
    },
    {
      "epoch": 14.654088050314465,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.006070159638143252,
      "loss": 0.3639,
      "num_input_tokens_seen": 18241912,
      "step": 27960
    },
    {
      "epoch": 14.65670859538784,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.006064648194754238,
      "loss": 0.3078,
      "num_input_tokens_seen": 18244632,
      "step": 27965
    },
    {
      "epoch": 14.659329140461216,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.006059138620538886,
      "loss": 0.3338,
      "num_input_tokens_seen": 18247768,
      "step": 27970
    },
    {
      "epoch": 14.66194968553459,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.006053630916649725,
      "loss": 0.3325,
      "num_input_tokens_seen": 18250456,
      "step": 27975
    },
    {
      "epoch": 14.664570230607966,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.006048125084238913,
      "loss": 0.373,
      "num_input_tokens_seen": 18253688,
      "step": 27980
    },
    {
      "epoch": 14.667190775681341,
      "grad_norm": 0.05078125,
      "learning_rate": 0.006042621124458211,
      "loss": 0.315,
      "num_input_tokens_seen": 18256952,
      "step": 27985
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.006037119038458971,
      "loss": 0.2942,
      "num_input_tokens_seen": 18263064,
      "step": 27990
    },
    {
      "epoch": 14.672431865828091,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.006031618827392175,
      "loss": 0.3297,
      "num_input_tokens_seen": 18265848,
      "step": 27995
    },
    {
      "epoch": 14.675052410901468,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.006026120492408408,
      "loss": 0.5397,
      "num_input_tokens_seen": 18269208,
      "step": 28000
    },
    {
      "epoch": 14.677672955974844,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.006020624034657851,
      "loss": 0.3491,
      "num_input_tokens_seen": 18274136,
      "step": 28005
    },
    {
      "epoch": 14.680293501048219,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.006015129455290309,
      "loss": 0.3149,
      "num_input_tokens_seen": 18277016,
      "step": 28010
    },
    {
      "epoch": 14.682914046121594,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.006009636755455175,
      "loss": 0.3085,
      "num_input_tokens_seen": 18279640,
      "step": 28015
    },
    {
      "epoch": 14.685534591194969,
      "grad_norm": 0.109375,
      "learning_rate": 0.0060041459363014615,
      "loss": 0.3217,
      "num_input_tokens_seen": 18281816,
      "step": 28020
    },
    {
      "epoch": 14.688155136268344,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.005998656998977795,
      "loss": 0.491,
      "num_input_tokens_seen": 18284568,
      "step": 28025
    },
    {
      "epoch": 14.69077568134172,
      "grad_norm": 0.0625,
      "learning_rate": 0.0059931699446323875,
      "loss": 0.3614,
      "num_input_tokens_seen": 18288792,
      "step": 28030
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.005987684774413072,
      "loss": 0.3853,
      "num_input_tokens_seen": 18292344,
      "step": 28035
    },
    {
      "epoch": 14.69601677148847,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.005982201489467288,
      "loss": 0.3662,
      "num_input_tokens_seen": 18296184,
      "step": 28040
    },
    {
      "epoch": 14.698637316561845,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.005976720090942066,
      "loss": 0.4321,
      "num_input_tokens_seen": 18299960,
      "step": 28045
    },
    {
      "epoch": 14.70125786163522,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.005971240579984059,
      "loss": 0.3896,
      "num_input_tokens_seen": 18304024,
      "step": 28050
    },
    {
      "epoch": 14.703878406708595,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.005965762957739522,
      "loss": 0.3876,
      "num_input_tokens_seen": 18306808,
      "step": 28055
    },
    {
      "epoch": 14.70649895178197,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.005960287225354302,
      "loss": 0.3108,
      "num_input_tokens_seen": 18308984,
      "step": 28060
    },
    {
      "epoch": 14.709119496855346,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.005954813383973868,
      "loss": 0.3502,
      "num_input_tokens_seen": 18311416,
      "step": 28065
    },
    {
      "epoch": 14.71174004192872,
      "grad_norm": 0.142578125,
      "learning_rate": 0.005949341434743277,
      "loss": 0.4921,
      "num_input_tokens_seen": 18314552,
      "step": 28070
    },
    {
      "epoch": 14.714360587002096,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.005943871378807202,
      "loss": 0.313,
      "num_input_tokens_seen": 18317752,
      "step": 28075
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005938403217309917,
      "loss": 0.4306,
      "num_input_tokens_seen": 18320888,
      "step": 28080
    },
    {
      "epoch": 14.719601677148846,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0059329369513953035,
      "loss": 0.5381,
      "num_input_tokens_seen": 18324600,
      "step": 28085
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.005927472582206837,
      "loss": 0.3486,
      "num_input_tokens_seen": 18328184,
      "step": 28090
    },
    {
      "epoch": 14.724842767295598,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.005922010110887596,
      "loss": 0.3299,
      "num_input_tokens_seen": 18330680,
      "step": 28095
    },
    {
      "epoch": 14.727463312368974,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.005916549538580274,
      "loss": 0.4027,
      "num_input_tokens_seen": 18334552,
      "step": 28100
    },
    {
      "epoch": 14.730083857442349,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.005911090866427157,
      "loss": 0.41,
      "num_input_tokens_seen": 18338168,
      "step": 28105
    },
    {
      "epoch": 14.732704402515724,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.005905634095570147,
      "loss": 0.3437,
      "num_input_tokens_seen": 18341528,
      "step": 28110
    },
    {
      "epoch": 14.735324947589099,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.005900179227150724,
      "loss": 0.3578,
      "num_input_tokens_seen": 18344920,
      "step": 28115
    },
    {
      "epoch": 14.737945492662474,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0058947262623099935,
      "loss": 0.3004,
      "num_input_tokens_seen": 18348312,
      "step": 28120
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.005889275202188646,
      "loss": 0.3324,
      "num_input_tokens_seen": 18351160,
      "step": 28125
    },
    {
      "epoch": 14.743186582809225,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.005883826047926987,
      "loss": 0.4395,
      "num_input_tokens_seen": 18354104,
      "step": 28130
    },
    {
      "epoch": 14.7458071278826,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0058783788006649155,
      "loss": 0.3813,
      "num_input_tokens_seen": 18357784,
      "step": 28135
    },
    {
      "epoch": 14.748427672955975,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.005872933461541939,
      "loss": 0.4301,
      "num_input_tokens_seen": 18360984,
      "step": 28140
    },
    {
      "epoch": 14.75104821802935,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0058674900316971566,
      "loss": 0.3767,
      "num_input_tokens_seen": 18365080,
      "step": 28145
    },
    {
      "epoch": 14.753668763102725,
      "grad_norm": 0.08203125,
      "learning_rate": 0.005862048512269263,
      "loss": 0.4215,
      "num_input_tokens_seen": 18367800,
      "step": 28150
    },
    {
      "epoch": 14.7562893081761,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00585660890439657,
      "loss": 0.5184,
      "num_input_tokens_seen": 18370392,
      "step": 28155
    },
    {
      "epoch": 14.758909853249476,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.005851171209216982,
      "loss": 0.384,
      "num_input_tokens_seen": 18373560,
      "step": 28160
    },
    {
      "epoch": 14.76153039832285,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.005845735427868001,
      "loss": 0.3094,
      "num_input_tokens_seen": 18378296,
      "step": 28165
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 0.08984375,
      "learning_rate": 0.005840301561486742,
      "loss": 0.3825,
      "num_input_tokens_seen": 18381016,
      "step": 28170
    },
    {
      "epoch": 14.766771488469601,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.005834869611209885,
      "loss": 0.4136,
      "num_input_tokens_seen": 18384248,
      "step": 28175
    },
    {
      "epoch": 14.769392033542976,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.005829439578173745,
      "loss": 0.4038,
      "num_input_tokens_seen": 18386648,
      "step": 28180
    },
    {
      "epoch": 14.772012578616351,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0058240114635142205,
      "loss": 0.3368,
      "num_input_tokens_seen": 18390232,
      "step": 28185
    },
    {
      "epoch": 14.774633123689728,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.005818585268366814,
      "loss": 0.4158,
      "num_input_tokens_seen": 18393624,
      "step": 28190
    },
    {
      "epoch": 14.777253668763104,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.005813160993866629,
      "loss": 0.3317,
      "num_input_tokens_seen": 18397176,
      "step": 28195
    },
    {
      "epoch": 14.779874213836479,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.005807738641148352,
      "loss": 0.4015,
      "num_input_tokens_seen": 18400472,
      "step": 28200
    },
    {
      "epoch": 14.782494758909854,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.005802318211346278,
      "loss": 0.3365,
      "num_input_tokens_seen": 18403224,
      "step": 28205
    },
    {
      "epoch": 14.785115303983229,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0057968997055943,
      "loss": 0.2882,
      "num_input_tokens_seen": 18406200,
      "step": 28210
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 0.09765625,
      "learning_rate": 0.005791483125025908,
      "loss": 0.3922,
      "num_input_tokens_seen": 18409432,
      "step": 28215
    },
    {
      "epoch": 14.79035639412998,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0057860684707741975,
      "loss": 0.3171,
      "num_input_tokens_seen": 18411736,
      "step": 28220
    },
    {
      "epoch": 14.792976939203355,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.005780655743971843,
      "loss": 0.409,
      "num_input_tokens_seen": 18415448,
      "step": 28225
    },
    {
      "epoch": 14.79559748427673,
      "grad_norm": 0.068359375,
      "learning_rate": 0.005775244945751124,
      "loss": 0.4548,
      "num_input_tokens_seen": 18419032,
      "step": 28230
    },
    {
      "epoch": 14.798218029350105,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.005769836077243918,
      "loss": 0.4174,
      "num_input_tokens_seen": 18422136,
      "step": 28235
    },
    {
      "epoch": 14.80083857442348,
      "grad_norm": 0.078125,
      "learning_rate": 0.005764429139581703,
      "loss": 0.5214,
      "num_input_tokens_seen": 18424696,
      "step": 28240
    },
    {
      "epoch": 14.803459119496855,
      "grad_norm": 0.072265625,
      "learning_rate": 0.005759024133895547,
      "loss": 0.4055,
      "num_input_tokens_seen": 18427704,
      "step": 28245
    },
    {
      "epoch": 14.80607966457023,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.005753621061316121,
      "loss": 0.3131,
      "num_input_tokens_seen": 18430328,
      "step": 28250
    },
    {
      "epoch": 14.808700209643606,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.00574821992297368,
      "loss": 0.4298,
      "num_input_tokens_seen": 18432920,
      "step": 28255
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.005742820719998076,
      "loss": 0.4956,
      "num_input_tokens_seen": 18435896,
      "step": 28260
    },
    {
      "epoch": 14.813941299790356,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.005737423453518765,
      "loss": 0.3832,
      "num_input_tokens_seen": 18438712,
      "step": 28265
    },
    {
      "epoch": 14.816561844863731,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.005732028124664797,
      "loss": 0.3856,
      "num_input_tokens_seen": 18441592,
      "step": 28270
    },
    {
      "epoch": 14.819182389937106,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.005726634734564814,
      "loss": 0.3521,
      "num_input_tokens_seen": 18444120,
      "step": 28275
    },
    {
      "epoch": 14.821802935010481,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0057212432843470485,
      "loss": 0.3091,
      "num_input_tokens_seen": 18447640,
      "step": 28280
    },
    {
      "epoch": 14.824423480083858,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.005715853775139327,
      "loss": 0.433,
      "num_input_tokens_seen": 18452312,
      "step": 28285
    },
    {
      "epoch": 14.827044025157234,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.005710466208069073,
      "loss": 0.3461,
      "num_input_tokens_seen": 18455704,
      "step": 28290
    },
    {
      "epoch": 14.829664570230609,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.00570508058426331,
      "loss": 0.3502,
      "num_input_tokens_seen": 18459384,
      "step": 28295
    },
    {
      "epoch": 14.832285115303984,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.005699696904848652,
      "loss": 0.5,
      "num_input_tokens_seen": 18463512,
      "step": 28300
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 0.13671875,
      "learning_rate": 0.005694315170951292,
      "loss": 0.4607,
      "num_input_tokens_seen": 18466520,
      "step": 28305
    },
    {
      "epoch": 14.837526205450734,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005688935383697039,
      "loss": 0.324,
      "num_input_tokens_seen": 18469240,
      "step": 28310
    },
    {
      "epoch": 14.84014675052411,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.005683557544211273,
      "loss": 0.4,
      "num_input_tokens_seen": 18473272,
      "step": 28315
    },
    {
      "epoch": 14.842767295597485,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.005678181653618981,
      "loss": 0.4076,
      "num_input_tokens_seen": 18478392,
      "step": 28320
    },
    {
      "epoch": 14.84538784067086,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005672807713044742,
      "loss": 0.3145,
      "num_input_tokens_seen": 18481048,
      "step": 28325
    },
    {
      "epoch": 14.848008385744235,
      "grad_norm": 0.068359375,
      "learning_rate": 0.005667435723612716,
      "loss": 0.4364,
      "num_input_tokens_seen": 18484376,
      "step": 28330
    },
    {
      "epoch": 14.85062893081761,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0056620656864466625,
      "loss": 0.2793,
      "num_input_tokens_seen": 18487928,
      "step": 28335
    },
    {
      "epoch": 14.853249475890985,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.005656697602669941,
      "loss": 0.5052,
      "num_input_tokens_seen": 18491192,
      "step": 28340
    },
    {
      "epoch": 14.85587002096436,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.005651331473405482,
      "loss": 0.3336,
      "num_input_tokens_seen": 18494744,
      "step": 28345
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 0.060546875,
      "learning_rate": 0.005645967299775824,
      "loss": 0.4231,
      "num_input_tokens_seen": 18498584,
      "step": 28350
    },
    {
      "epoch": 14.86111111111111,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.005640605082903096,
      "loss": 0.4537,
      "num_input_tokens_seen": 18501400,
      "step": 28355
    },
    {
      "epoch": 14.863731656184486,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.005635244823909001,
      "loss": 0.3589,
      "num_input_tokens_seen": 18503864,
      "step": 28360
    },
    {
      "epoch": 14.866352201257861,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0056298865239148505,
      "loss": 0.3463,
      "num_input_tokens_seen": 18506712,
      "step": 28365
    },
    {
      "epoch": 14.868972746331236,
      "grad_norm": 0.087890625,
      "learning_rate": 0.005624530184041544,
      "loss": 0.4051,
      "num_input_tokens_seen": 18510360,
      "step": 28370
    },
    {
      "epoch": 14.871593291404611,
      "grad_norm": 0.0625,
      "learning_rate": 0.005619175805409556,
      "loss": 0.3811,
      "num_input_tokens_seen": 18513432,
      "step": 28375
    },
    {
      "epoch": 14.874213836477988,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.005613823389138973,
      "loss": 0.3833,
      "num_input_tokens_seen": 18516632,
      "step": 28380
    },
    {
      "epoch": 14.876834381551364,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.005608472936349449,
      "loss": 0.4559,
      "num_input_tokens_seen": 18521624,
      "step": 28385
    },
    {
      "epoch": 14.879454926624739,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.005603124448160244,
      "loss": 0.4439,
      "num_input_tokens_seen": 18524760,
      "step": 28390
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 0.15234375,
      "learning_rate": 0.005597777925690203,
      "loss": 0.3472,
      "num_input_tokens_seen": 18527416,
      "step": 28395
    },
    {
      "epoch": 14.884696016771489,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.005592433370057748,
      "loss": 0.4192,
      "num_input_tokens_seen": 18531032,
      "step": 28400
    },
    {
      "epoch": 14.887316561844864,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.005587090782380911,
      "loss": 0.3906,
      "num_input_tokens_seen": 18534744,
      "step": 28405
    },
    {
      "epoch": 14.88993710691824,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.005581750163777288,
      "loss": 0.4022,
      "num_input_tokens_seen": 18538392,
      "step": 28410
    },
    {
      "epoch": 14.892557651991615,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.005576411515364082,
      "loss": 0.3277,
      "num_input_tokens_seen": 18541400,
      "step": 28415
    },
    {
      "epoch": 14.89517819706499,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.005571074838258078,
      "loss": 0.362,
      "num_input_tokens_seen": 18544696,
      "step": 28420
    },
    {
      "epoch": 14.897798742138365,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00556574013357565,
      "loss": 0.2892,
      "num_input_tokens_seen": 18547864,
      "step": 28425
    },
    {
      "epoch": 14.90041928721174,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.005560407402432751,
      "loss": 0.3713,
      "num_input_tokens_seen": 18550584,
      "step": 28430
    },
    {
      "epoch": 14.903039832285115,
      "grad_norm": 0.173828125,
      "learning_rate": 0.005555076645944936,
      "loss": 0.4703,
      "num_input_tokens_seen": 18553624,
      "step": 28435
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.005549747865227328,
      "loss": 0.461,
      "num_input_tokens_seen": 18556472,
      "step": 28440
    },
    {
      "epoch": 14.908280922431866,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.005544421061394651,
      "loss": 0.2751,
      "num_input_tokens_seen": 18558840,
      "step": 28445
    },
    {
      "epoch": 14.91090146750524,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.005539096235561214,
      "loss": 0.374,
      "num_input_tokens_seen": 18562008,
      "step": 28450
    },
    {
      "epoch": 14.913522012578616,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.005533773388840916,
      "loss": 0.4516,
      "num_input_tokens_seen": 18564952,
      "step": 28455
    },
    {
      "epoch": 14.916142557651991,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.005528452522347226,
      "loss": 0.3833,
      "num_input_tokens_seen": 18567800,
      "step": 28460
    },
    {
      "epoch": 14.918763102725366,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.005523133637193206,
      "loss": 0.4375,
      "num_input_tokens_seen": 18571096,
      "step": 28465
    },
    {
      "epoch": 14.921383647798741,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.005517816734491513,
      "loss": 0.3078,
      "num_input_tokens_seen": 18574136,
      "step": 28470
    },
    {
      "epoch": 14.924004192872118,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.005512501815354379,
      "loss": 0.3711,
      "num_input_tokens_seen": 18576664,
      "step": 28475
    },
    {
      "epoch": 14.926624737945493,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.005507188880893632,
      "loss": 0.3622,
      "num_input_tokens_seen": 18580824,
      "step": 28480
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.005501877932220672,
      "loss": 0.3197,
      "num_input_tokens_seen": 18583864,
      "step": 28485
    },
    {
      "epoch": 14.931865828092244,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.005496568970446483,
      "loss": 0.3434,
      "num_input_tokens_seen": 18586680,
      "step": 28490
    },
    {
      "epoch": 14.934486373165619,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0054912619966816425,
      "loss": 0.3074,
      "num_input_tokens_seen": 18590776,
      "step": 28495
    },
    {
      "epoch": 14.937106918238994,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.005485957012036313,
      "loss": 0.2875,
      "num_input_tokens_seen": 18594616,
      "step": 28500
    },
    {
      "epoch": 14.93972746331237,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.005480654017620235,
      "loss": 0.4547,
      "num_input_tokens_seen": 18597528,
      "step": 28505
    },
    {
      "epoch": 14.942348008385745,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.005475353014542737,
      "loss": 0.196,
      "num_input_tokens_seen": 18599992,
      "step": 28510
    },
    {
      "epoch": 14.94496855345912,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.005470054003912727,
      "loss": 0.3332,
      "num_input_tokens_seen": 18602872,
      "step": 28515
    },
    {
      "epoch": 14.947589098532495,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.00546475698683869,
      "loss": 0.3559,
      "num_input_tokens_seen": 18606360,
      "step": 28520
    },
    {
      "epoch": 14.95020964360587,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.005459461964428708,
      "loss": 0.3239,
      "num_input_tokens_seen": 18608824,
      "step": 28525
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0054541689377904405,
      "loss": 0.3557,
      "num_input_tokens_seen": 18611864,
      "step": 28530
    },
    {
      "epoch": 14.95545073375262,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.005448877908031129,
      "loss": 0.4459,
      "num_input_tokens_seen": 18615480,
      "step": 28535
    },
    {
      "epoch": 14.958071278825996,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.005443588876257597,
      "loss": 0.4051,
      "num_input_tokens_seen": 18618392,
      "step": 28540
    },
    {
      "epoch": 14.96069182389937,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.005438301843576239,
      "loss": 0.4723,
      "num_input_tokens_seen": 18622200,
      "step": 28545
    },
    {
      "epoch": 14.963312368972746,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00543301681109305,
      "loss": 0.3765,
      "num_input_tokens_seen": 18625080,
      "step": 28550
    },
    {
      "epoch": 14.965932914046121,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0054277337799135995,
      "loss": 0.4047,
      "num_input_tokens_seen": 18627992,
      "step": 28555
    },
    {
      "epoch": 14.968553459119496,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.005422452751143035,
      "loss": 0.3602,
      "num_input_tokens_seen": 18631192,
      "step": 28560
    },
    {
      "epoch": 14.971174004192871,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0054171737258860994,
      "loss": 0.3181,
      "num_input_tokens_seen": 18634904,
      "step": 28565
    },
    {
      "epoch": 14.973794549266248,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.005411896705247084,
      "loss": 0.3824,
      "num_input_tokens_seen": 18637784,
      "step": 28570
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 0.158203125,
      "learning_rate": 0.00540662169032989,
      "loss": 0.3466,
      "num_input_tokens_seen": 18640120,
      "step": 28575
    },
    {
      "epoch": 14.979035639412999,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.005401348682237992,
      "loss": 0.5228,
      "num_input_tokens_seen": 18643672,
      "step": 28580
    },
    {
      "epoch": 14.981656184486374,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.005396077682074441,
      "loss": 0.4694,
      "num_input_tokens_seen": 18646872,
      "step": 28585
    },
    {
      "epoch": 14.984276729559749,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.005390808690941879,
      "loss": 0.2593,
      "num_input_tokens_seen": 18650744,
      "step": 28590
    },
    {
      "epoch": 14.986897274633124,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.005385541709942509,
      "loss": 0.3732,
      "num_input_tokens_seen": 18654296,
      "step": 28595
    },
    {
      "epoch": 14.9895178197065,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0053802767401781235,
      "loss": 0.3491,
      "num_input_tokens_seen": 18657560,
      "step": 28600
    },
    {
      "epoch": 14.992138364779874,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.005375013782750094,
      "loss": 0.3693,
      "num_input_tokens_seen": 18660728,
      "step": 28605
    },
    {
      "epoch": 14.99475890985325,
      "grad_norm": 0.076171875,
      "learning_rate": 0.005369752838759374,
      "loss": 0.3287,
      "num_input_tokens_seen": 18663352,
      "step": 28610
    },
    {
      "epoch": 14.997379454926625,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.005364493909306499,
      "loss": 0.3649,
      "num_input_tokens_seen": 18665816,
      "step": 28615
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.005359236995491564,
      "loss": 0.2642,
      "num_input_tokens_seen": 18668536,
      "step": 28620
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.48977452516555786,
      "eval_runtime": 13.6609,
      "eval_samples_per_second": 62.075,
      "eval_steps_per_second": 15.519,
      "num_input_tokens_seen": 18668536,
      "step": 28620
    },
    {
      "epoch": 15.002620545073375,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.005353982098414266,
      "loss": 0.5735,
      "num_input_tokens_seen": 18672024,
      "step": 28625
    },
    {
      "epoch": 15.00524109014675,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.005348729219173862,
      "loss": 0.435,
      "num_input_tokens_seen": 18674712,
      "step": 28630
    },
    {
      "epoch": 15.007861635220126,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.005343478358869197,
      "loss": 0.3349,
      "num_input_tokens_seen": 18682488,
      "step": 28635
    },
    {
      "epoch": 15.0104821802935,
      "grad_norm": 0.0546875,
      "learning_rate": 0.005338229518598693,
      "loss": 0.3566,
      "num_input_tokens_seen": 18685816,
      "step": 28640
    },
    {
      "epoch": 15.013102725366876,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.005332982699460351,
      "loss": 0.3605,
      "num_input_tokens_seen": 18688760,
      "step": 28645
    },
    {
      "epoch": 15.015723270440251,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.005327737902551739,
      "loss": 0.2798,
      "num_input_tokens_seen": 18691160,
      "step": 28650
    },
    {
      "epoch": 15.018343815513626,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.005322495128970009,
      "loss": 0.4079,
      "num_input_tokens_seen": 18694616,
      "step": 28655
    },
    {
      "epoch": 15.020964360587001,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00531725437981189,
      "loss": 0.3113,
      "num_input_tokens_seen": 18697656,
      "step": 28660
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.005312015656173686,
      "loss": 0.2906,
      "num_input_tokens_seen": 18700536,
      "step": 28665
    },
    {
      "epoch": 15.026205450733753,
      "grad_norm": 0.154296875,
      "learning_rate": 0.005306778959151289,
      "loss": 0.4583,
      "num_input_tokens_seen": 18703960,
      "step": 28670
    },
    {
      "epoch": 15.028825995807129,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.005301544289840141,
      "loss": 0.2944,
      "num_input_tokens_seen": 18706744,
      "step": 28675
    },
    {
      "epoch": 15.031446540880504,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.005296311649335286,
      "loss": 0.385,
      "num_input_tokens_seen": 18710424,
      "step": 28680
    },
    {
      "epoch": 15.034067085953879,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005291081038731324,
      "loss": 0.4817,
      "num_input_tokens_seen": 18715640,
      "step": 28685
    },
    {
      "epoch": 15.036687631027254,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.005285852459122444,
      "loss": 0.453,
      "num_input_tokens_seen": 18718680,
      "step": 28690
    },
    {
      "epoch": 15.03930817610063,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005280625911602409,
      "loss": 0.3835,
      "num_input_tokens_seen": 18721688,
      "step": 28695
    },
    {
      "epoch": 15.041928721174004,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.005275401397264545,
      "loss": 0.3461,
      "num_input_tokens_seen": 18724920,
      "step": 28700
    },
    {
      "epoch": 15.04454926624738,
      "grad_norm": 0.169921875,
      "learning_rate": 0.005270178917201763,
      "loss": 0.479,
      "num_input_tokens_seen": 18728376,
      "step": 28705
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 0.0703125,
      "learning_rate": 0.005264958472506554,
      "loss": 0.2705,
      "num_input_tokens_seen": 18731704,
      "step": 28710
    },
    {
      "epoch": 15.04979035639413,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.005259740064270964,
      "loss": 0.4221,
      "num_input_tokens_seen": 18734968,
      "step": 28715
    },
    {
      "epoch": 15.052410901467505,
      "grad_norm": 0.1171875,
      "learning_rate": 0.00525452369358663,
      "loss": 0.41,
      "num_input_tokens_seen": 18738680,
      "step": 28720
    },
    {
      "epoch": 15.05503144654088,
      "grad_norm": 0.05859375,
      "learning_rate": 0.005249309361544761,
      "loss": 0.3182,
      "num_input_tokens_seen": 18742136,
      "step": 28725
    },
    {
      "epoch": 15.057651991614255,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.005244097069236128,
      "loss": 0.2793,
      "num_input_tokens_seen": 18745016,
      "step": 28730
    },
    {
      "epoch": 15.06027253668763,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0052388868177510855,
      "loss": 0.3831,
      "num_input_tokens_seen": 18748920,
      "step": 28735
    },
    {
      "epoch": 15.062893081761006,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.005233678608179566,
      "loss": 0.3559,
      "num_input_tokens_seen": 18753560,
      "step": 28740
    },
    {
      "epoch": 15.065513626834381,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.005228472441611057,
      "loss": 0.4792,
      "num_input_tokens_seen": 18756952,
      "step": 28745
    },
    {
      "epoch": 15.068134171907756,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0052232683191346385,
      "loss": 0.4198,
      "num_input_tokens_seen": 18760472,
      "step": 28750
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.005218066241838943,
      "loss": 0.3675,
      "num_input_tokens_seen": 18763928,
      "step": 28755
    },
    {
      "epoch": 15.073375262054507,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0052128662108121915,
      "loss": 0.3585,
      "num_input_tokens_seen": 18767512,
      "step": 28760
    },
    {
      "epoch": 15.075995807127883,
      "grad_norm": 0.064453125,
      "learning_rate": 0.005207668227142177,
      "loss": 0.4872,
      "num_input_tokens_seen": 18774040,
      "step": 28765
    },
    {
      "epoch": 15.078616352201259,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.005202472291916247,
      "loss": 0.3395,
      "num_input_tokens_seen": 18777336,
      "step": 28770
    },
    {
      "epoch": 15.081236897274634,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0051972784062213435,
      "loss": 0.2696,
      "num_input_tokens_seen": 18780504,
      "step": 28775
    },
    {
      "epoch": 15.083857442348009,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.005192086571143956,
      "loss": 0.5152,
      "num_input_tokens_seen": 18783800,
      "step": 28780
    },
    {
      "epoch": 15.086477987421384,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.005186896787770163,
      "loss": 0.3316,
      "num_input_tokens_seen": 18786968,
      "step": 28785
    },
    {
      "epoch": 15.08909853249476,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.00518170905718561,
      "loss": 0.2796,
      "num_input_tokens_seen": 18789720,
      "step": 28790
    },
    {
      "epoch": 15.091719077568134,
      "grad_norm": 0.1328125,
      "learning_rate": 0.005176523380475513,
      "loss": 0.2991,
      "num_input_tokens_seen": 18792952,
      "step": 28795
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 0.0703125,
      "learning_rate": 0.005171339758724648,
      "loss": 0.3108,
      "num_input_tokens_seen": 18795480,
      "step": 28800
    },
    {
      "epoch": 15.096960167714885,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.005166158193017378,
      "loss": 0.2476,
      "num_input_tokens_seen": 18798648,
      "step": 28805
    },
    {
      "epoch": 15.09958071278826,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0051609786844376195,
      "loss": 0.4107,
      "num_input_tokens_seen": 18802680,
      "step": 28810
    },
    {
      "epoch": 15.102201257861635,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.005155801234068871,
      "loss": 0.3694,
      "num_input_tokens_seen": 18805112,
      "step": 28815
    },
    {
      "epoch": 15.10482180293501,
      "grad_norm": 0.0859375,
      "learning_rate": 0.005150625842994196,
      "loss": 0.3423,
      "num_input_tokens_seen": 18809272,
      "step": 28820
    },
    {
      "epoch": 15.107442348008385,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.00514545251229623,
      "loss": 0.3567,
      "num_input_tokens_seen": 18812440,
      "step": 28825
    },
    {
      "epoch": 15.11006289308176,
      "grad_norm": 0.072265625,
      "learning_rate": 0.005140281243057174,
      "loss": 0.4625,
      "num_input_tokens_seen": 18815288,
      "step": 28830
    },
    {
      "epoch": 15.112683438155136,
      "grad_norm": 0.111328125,
      "learning_rate": 0.005135112036358791,
      "loss": 0.4678,
      "num_input_tokens_seen": 18818424,
      "step": 28835
    },
    {
      "epoch": 15.115303983228511,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0051299448932824255,
      "loss": 0.4039,
      "num_input_tokens_seen": 18821144,
      "step": 28840
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0051247798149089835,
      "loss": 0.2767,
      "num_input_tokens_seen": 18824632,
      "step": 28845
    },
    {
      "epoch": 15.120545073375261,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.005119616802318947,
      "loss": 0.4046,
      "num_input_tokens_seen": 18827224,
      "step": 28850
    },
    {
      "epoch": 15.123165618448636,
      "grad_norm": 0.068359375,
      "learning_rate": 0.005114455856592356,
      "loss": 0.3637,
      "num_input_tokens_seen": 18830072,
      "step": 28855
    },
    {
      "epoch": 15.125786163522013,
      "grad_norm": 0.13671875,
      "learning_rate": 0.005109296978808813,
      "loss": 0.4367,
      "num_input_tokens_seen": 18832856,
      "step": 28860
    },
    {
      "epoch": 15.128406708595389,
      "grad_norm": 0.220703125,
      "learning_rate": 0.005104140170047503,
      "loss": 0.5216,
      "num_input_tokens_seen": 18835544,
      "step": 28865
    },
    {
      "epoch": 15.131027253668764,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005098985431387172,
      "loss": 0.3911,
      "num_input_tokens_seen": 18838360,
      "step": 28870
    },
    {
      "epoch": 15.133647798742139,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0050938327639061315,
      "loss": 0.4992,
      "num_input_tokens_seen": 18842168,
      "step": 28875
    },
    {
      "epoch": 15.136268343815514,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.005088682168682266,
      "loss": 0.4037,
      "num_input_tokens_seen": 18845496,
      "step": 28880
    },
    {
      "epoch": 15.13888888888889,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.005083533646793016,
      "loss": 0.3067,
      "num_input_tokens_seen": 18850232,
      "step": 28885
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 0.10546875,
      "learning_rate": 0.005078387199315391,
      "loss": 0.3704,
      "num_input_tokens_seen": 18852536,
      "step": 28890
    },
    {
      "epoch": 15.14412997903564,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.005073242827325971,
      "loss": 0.422,
      "num_input_tokens_seen": 18855544,
      "step": 28895
    },
    {
      "epoch": 15.146750524109015,
      "grad_norm": 0.087890625,
      "learning_rate": 0.005068100531900902,
      "loss": 0.4494,
      "num_input_tokens_seen": 18858584,
      "step": 28900
    },
    {
      "epoch": 15.14937106918239,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0050629603141158965,
      "loss": 0.3563,
      "num_input_tokens_seen": 18861688,
      "step": 28905
    },
    {
      "epoch": 15.151991614255765,
      "grad_norm": 0.103515625,
      "learning_rate": 0.005057822175046225,
      "loss": 0.3704,
      "num_input_tokens_seen": 18864056,
      "step": 28910
    },
    {
      "epoch": 15.15461215932914,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.005052686115766724,
      "loss": 0.3571,
      "num_input_tokens_seen": 18867960,
      "step": 28915
    },
    {
      "epoch": 15.157232704402515,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.005047552137351803,
      "loss": 0.2594,
      "num_input_tokens_seen": 18870328,
      "step": 28920
    },
    {
      "epoch": 15.15985324947589,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00504242024087543,
      "loss": 0.3807,
      "num_input_tokens_seen": 18873976,
      "step": 28925
    },
    {
      "epoch": 15.162473794549266,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.005037290427411148,
      "loss": 0.2378,
      "num_input_tokens_seen": 18877112,
      "step": 28930
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.005032162698032047,
      "loss": 0.364,
      "num_input_tokens_seen": 18879480,
      "step": 28935
    },
    {
      "epoch": 15.167714884696016,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.005027037053810784,
      "loss": 0.2841,
      "num_input_tokens_seen": 18882072,
      "step": 28940
    },
    {
      "epoch": 15.170335429769391,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.005021913495819592,
      "loss": 0.4412,
      "num_input_tokens_seen": 18884760,
      "step": 28945
    },
    {
      "epoch": 15.172955974842766,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.005016792025130262,
      "loss": 0.3612,
      "num_input_tokens_seen": 18887736,
      "step": 28950
    },
    {
      "epoch": 15.175576519916143,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.005011672642814146,
      "loss": 0.3198,
      "num_input_tokens_seen": 18890584,
      "step": 28955
    },
    {
      "epoch": 15.178197064989519,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.005006555349942164,
      "loss": 0.3385,
      "num_input_tokens_seen": 18893208,
      "step": 28960
    },
    {
      "epoch": 15.180817610062894,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.005001440147584793,
      "loss": 0.3222,
      "num_input_tokens_seen": 18895864,
      "step": 28965
    },
    {
      "epoch": 15.183438155136269,
      "grad_norm": 0.080078125,
      "learning_rate": 0.004996327036812067,
      "loss": 0.3509,
      "num_input_tokens_seen": 18898712,
      "step": 28970
    },
    {
      "epoch": 15.186058700209644,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.004991216018693598,
      "loss": 0.4022,
      "num_input_tokens_seen": 18901624,
      "step": 28975
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 0.095703125,
      "learning_rate": 0.004986107094298551,
      "loss": 0.2752,
      "num_input_tokens_seen": 18905176,
      "step": 28980
    },
    {
      "epoch": 15.191299790356394,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.004981000264695661,
      "loss": 0.2396,
      "num_input_tokens_seen": 18907864,
      "step": 28985
    },
    {
      "epoch": 15.19392033542977,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004975895530953206,
      "loss": 0.285,
      "num_input_tokens_seen": 18911064,
      "step": 28990
    },
    {
      "epoch": 15.196540880503145,
      "grad_norm": 0.06640625,
      "learning_rate": 0.004970792894139053,
      "loss": 0.427,
      "num_input_tokens_seen": 18914136,
      "step": 28995
    },
    {
      "epoch": 15.19916142557652,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.004965692355320599,
      "loss": 0.3331,
      "num_input_tokens_seen": 18917048,
      "step": 29000
    },
    {
      "epoch": 15.201781970649895,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.00496059391556483,
      "loss": 0.4507,
      "num_input_tokens_seen": 18919864,
      "step": 29005
    },
    {
      "epoch": 15.20440251572327,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00495549757593828,
      "loss": 0.3582,
      "num_input_tokens_seen": 18922968,
      "step": 29010
    },
    {
      "epoch": 15.207023060796645,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.004950403337507041,
      "loss": 0.3342,
      "num_input_tokens_seen": 18926328,
      "step": 29015
    },
    {
      "epoch": 15.20964360587002,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.004945311201336776,
      "loss": 0.3266,
      "num_input_tokens_seen": 18929368,
      "step": 29020
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.004940221168492692,
      "loss": 0.331,
      "num_input_tokens_seen": 18932056,
      "step": 29025
    },
    {
      "epoch": 15.214884696016771,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.004935133240039572,
      "loss": 0.4505,
      "num_input_tokens_seen": 18935896,
      "step": 29030
    },
    {
      "epoch": 15.217505241090146,
      "grad_norm": 0.09375,
      "learning_rate": 0.004930047417041753,
      "loss": 0.3122,
      "num_input_tokens_seen": 18938968,
      "step": 29035
    },
    {
      "epoch": 15.220125786163521,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.004924963700563135,
      "loss": 0.4183,
      "num_input_tokens_seen": 18943000,
      "step": 29040
    },
    {
      "epoch": 15.222746331236896,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.004919882091667164,
      "loss": 0.3978,
      "num_input_tokens_seen": 18946104,
      "step": 29045
    },
    {
      "epoch": 15.225366876310272,
      "grad_norm": 0.068359375,
      "learning_rate": 0.004914802591416863,
      "loss": 0.3489,
      "num_input_tokens_seen": 18950328,
      "step": 29050
    },
    {
      "epoch": 15.227987421383649,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.004909725200874799,
      "loss": 0.3059,
      "num_input_tokens_seen": 18953144,
      "step": 29055
    },
    {
      "epoch": 15.230607966457024,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.004904649921103105,
      "loss": 0.4523,
      "num_input_tokens_seen": 18960248,
      "step": 29060
    },
    {
      "epoch": 15.233228511530399,
      "grad_norm": 0.109375,
      "learning_rate": 0.00489957675316348,
      "loss": 0.3123,
      "num_input_tokens_seen": 18962744,
      "step": 29065
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00489450569811716,
      "loss": 0.3009,
      "num_input_tokens_seen": 18965816,
      "step": 29070
    },
    {
      "epoch": 15.23846960167715,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.004889436757024959,
      "loss": 0.3623,
      "num_input_tokens_seen": 18969560,
      "step": 29075
    },
    {
      "epoch": 15.241090146750524,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.004884369930947247,
      "loss": 0.2093,
      "num_input_tokens_seen": 18972792,
      "step": 29080
    },
    {
      "epoch": 15.2437106918239,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.004879305220943933,
      "loss": 0.2643,
      "num_input_tokens_seen": 18975640,
      "step": 29085
    },
    {
      "epoch": 15.246331236897275,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.004874242628074511,
      "loss": 0.3029,
      "num_input_tokens_seen": 18978840,
      "step": 29090
    },
    {
      "epoch": 15.24895178197065,
      "grad_norm": 0.109375,
      "learning_rate": 0.004869182153398005,
      "loss": 0.2925,
      "num_input_tokens_seen": 18981464,
      "step": 29095
    },
    {
      "epoch": 15.251572327044025,
      "grad_norm": 0.06640625,
      "learning_rate": 0.004864123797973013,
      "loss": 0.3072,
      "num_input_tokens_seen": 18985368,
      "step": 29100
    },
    {
      "epoch": 15.2541928721174,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0048590675628576874,
      "loss": 0.3721,
      "num_input_tokens_seen": 18988120,
      "step": 29105
    },
    {
      "epoch": 15.256813417190775,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.004854013449109739,
      "loss": 0.4506,
      "num_input_tokens_seen": 18991640,
      "step": 29110
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.00484896145778642,
      "loss": 0.3708,
      "num_input_tokens_seen": 18994552,
      "step": 29115
    },
    {
      "epoch": 15.262054507337526,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00484391158994456,
      "loss": 0.4844,
      "num_input_tokens_seen": 18999224,
      "step": 29120
    },
    {
      "epoch": 15.264675052410901,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.004838863846640523,
      "loss": 0.6082,
      "num_input_tokens_seen": 19002648,
      "step": 29125
    },
    {
      "epoch": 15.267295597484276,
      "grad_norm": 0.09375,
      "learning_rate": 0.004833818228930245,
      "loss": 0.313,
      "num_input_tokens_seen": 19005976,
      "step": 29130
    },
    {
      "epoch": 15.269916142557651,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.004828774737869217,
      "loss": 0.3846,
      "num_input_tokens_seen": 19009144,
      "step": 29135
    },
    {
      "epoch": 15.272536687631026,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00482373337451247,
      "loss": 0.2947,
      "num_input_tokens_seen": 19011896,
      "step": 29140
    },
    {
      "epoch": 15.275157232704402,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.004818694139914608,
      "loss": 0.4995,
      "num_input_tokens_seen": 19015256,
      "step": 29145
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.004813657035129772,
      "loss": 0.4263,
      "num_input_tokens_seen": 19018360,
      "step": 29150
    },
    {
      "epoch": 15.280398322851154,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.004808622061211673,
      "loss": 0.2499,
      "num_input_tokens_seen": 19022136,
      "step": 29155
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.004803589219213569,
      "loss": 0.4147,
      "num_input_tokens_seen": 19024920,
      "step": 29160
    },
    {
      "epoch": 15.285639412997904,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.00479855851018828,
      "loss": 0.3098,
      "num_input_tokens_seen": 19029464,
      "step": 29165
    },
    {
      "epoch": 15.28825995807128,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.004793529935188165,
      "loss": 0.3823,
      "num_input_tokens_seen": 19033272,
      "step": 29170
    },
    {
      "epoch": 15.290880503144654,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.004788503495265141,
      "loss": 0.3993,
      "num_input_tokens_seen": 19035736,
      "step": 29175
    },
    {
      "epoch": 15.29350104821803,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.004783479191470688,
      "loss": 0.4025,
      "num_input_tokens_seen": 19038904,
      "step": 29180
    },
    {
      "epoch": 15.296121593291405,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.004778457024855832,
      "loss": 0.3232,
      "num_input_tokens_seen": 19042424,
      "step": 29185
    },
    {
      "epoch": 15.29874213836478,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.004773436996471151,
      "loss": 0.2569,
      "num_input_tokens_seen": 19045784,
      "step": 29190
    },
    {
      "epoch": 15.301362683438155,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.004768419107366786,
      "loss": 0.3452,
      "num_input_tokens_seen": 19049016,
      "step": 29195
    },
    {
      "epoch": 15.30398322851153,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.004763403358592415,
      "loss": 0.3089,
      "num_input_tokens_seen": 19052536,
      "step": 29200
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.004758389751197272,
      "loss": 0.2814,
      "num_input_tokens_seen": 19055480,
      "step": 29205
    },
    {
      "epoch": 15.30922431865828,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.00475337828623015,
      "loss": 0.4065,
      "num_input_tokens_seen": 19059416,
      "step": 29210
    },
    {
      "epoch": 15.311844863731656,
      "grad_norm": 0.1171875,
      "learning_rate": 0.004748368964739391,
      "loss": 0.2952,
      "num_input_tokens_seen": 19062360,
      "step": 29215
    },
    {
      "epoch": 15.314465408805031,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0047433617877728936,
      "loss": 0.3385,
      "num_input_tokens_seen": 19065400,
      "step": 29220
    },
    {
      "epoch": 15.317085953878406,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0047383567563780956,
      "loss": 0.3447,
      "num_input_tokens_seen": 19068376,
      "step": 29225
    },
    {
      "epoch": 15.319706498951781,
      "grad_norm": 0.12890625,
      "learning_rate": 0.004733353871601988,
      "loss": 0.2644,
      "num_input_tokens_seen": 19071032,
      "step": 29230
    },
    {
      "epoch": 15.322327044025156,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.004728353134491121,
      "loss": 0.3664,
      "num_input_tokens_seen": 19073944,
      "step": 29235
    },
    {
      "epoch": 15.324947589098532,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.004723354546091595,
      "loss": 0.3485,
      "num_input_tokens_seen": 19076760,
      "step": 29240
    },
    {
      "epoch": 15.327568134171909,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.004718358107449061,
      "loss": 0.4627,
      "num_input_tokens_seen": 19079480,
      "step": 29245
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00471336381960871,
      "loss": 0.4319,
      "num_input_tokens_seen": 19081944,
      "step": 29250
    },
    {
      "epoch": 15.332809224318659,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.004708371683615288,
      "loss": 0.3707,
      "num_input_tokens_seen": 19085272,
      "step": 29255
    },
    {
      "epoch": 15.335429769392034,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.004703381700513097,
      "loss": 0.3851,
      "num_input_tokens_seen": 19088152,
      "step": 29260
    },
    {
      "epoch": 15.33805031446541,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.004698393871345984,
      "loss": 0.4956,
      "num_input_tokens_seen": 19091576,
      "step": 29265
    },
    {
      "epoch": 15.340670859538784,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.004693408197157345,
      "loss": 0.4383,
      "num_input_tokens_seen": 19095672,
      "step": 29270
    },
    {
      "epoch": 15.34329140461216,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.004688424678990134,
      "loss": 0.4102,
      "num_input_tokens_seen": 19099416,
      "step": 29275
    },
    {
      "epoch": 15.345911949685535,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.004683443317886839,
      "loss": 0.3518,
      "num_input_tokens_seen": 19102296,
      "step": 29280
    },
    {
      "epoch": 15.34853249475891,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0046784641148894975,
      "loss": 0.3236,
      "num_input_tokens_seen": 19105432,
      "step": 29285
    },
    {
      "epoch": 15.351153039832285,
      "grad_norm": 0.154296875,
      "learning_rate": 0.004673487071039711,
      "loss": 0.2849,
      "num_input_tokens_seen": 19108632,
      "step": 29290
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.004668512187378618,
      "loss": 0.5167,
      "num_input_tokens_seen": 19111832,
      "step": 29295
    },
    {
      "epoch": 15.356394129979035,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.004663539464946913,
      "loss": 0.3376,
      "num_input_tokens_seen": 19115288,
      "step": 29300
    },
    {
      "epoch": 15.35901467505241,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.004658568904784826,
      "loss": 0.3202,
      "num_input_tokens_seen": 19118456,
      "step": 29305
    },
    {
      "epoch": 15.361635220125786,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.004653600507932137,
      "loss": 0.3433,
      "num_input_tokens_seen": 19121272,
      "step": 29310
    },
    {
      "epoch": 15.364255765199161,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.004648634275428185,
      "loss": 0.2746,
      "num_input_tokens_seen": 19124440,
      "step": 29315
    },
    {
      "epoch": 15.366876310272536,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0046436702083118495,
      "loss": 0.3786,
      "num_input_tokens_seen": 19127672,
      "step": 29320
    },
    {
      "epoch": 15.369496855345911,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0046387083076215585,
      "loss": 0.446,
      "num_input_tokens_seen": 19130776,
      "step": 29325
    },
    {
      "epoch": 15.372117400419286,
      "grad_norm": 0.09375,
      "learning_rate": 0.004633748574395276,
      "loss": 0.3477,
      "num_input_tokens_seen": 19133720,
      "step": 29330
    },
    {
      "epoch": 15.374737945492662,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0046287910096705345,
      "loss": 0.4009,
      "num_input_tokens_seen": 19137112,
      "step": 29335
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.004623835614484387,
      "loss": 0.3883,
      "num_input_tokens_seen": 19140408,
      "step": 29340
    },
    {
      "epoch": 15.379979035639414,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00461888238987345,
      "loss": 0.3758,
      "num_input_tokens_seen": 19143672,
      "step": 29345
    },
    {
      "epoch": 15.382599580712789,
      "grad_norm": 0.091796875,
      "learning_rate": 0.004613931336873885,
      "loss": 0.3141,
      "num_input_tokens_seen": 19147288,
      "step": 29350
    },
    {
      "epoch": 15.385220125786164,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.004608982456521397,
      "loss": 0.434,
      "num_input_tokens_seen": 19149688,
      "step": 29355
    },
    {
      "epoch": 15.38784067085954,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.004604035749851228,
      "loss": 0.3816,
      "num_input_tokens_seen": 19153752,
      "step": 29360
    },
    {
      "epoch": 15.390461215932914,
      "grad_norm": 0.0703125,
      "learning_rate": 0.004599091217898181,
      "loss": 0.4439,
      "num_input_tokens_seen": 19157208,
      "step": 29365
    },
    {
      "epoch": 15.39308176100629,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0045941488616965875,
      "loss": 0.4322,
      "num_input_tokens_seen": 19161656,
      "step": 29370
    },
    {
      "epoch": 15.395702306079665,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.004589208682280335,
      "loss": 0.3671,
      "num_input_tokens_seen": 19164824,
      "step": 29375
    },
    {
      "epoch": 15.39832285115304,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.004584270680682859,
      "loss": 0.3526,
      "num_input_tokens_seen": 19167480,
      "step": 29380
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 0.08203125,
      "learning_rate": 0.004579334857937122,
      "loss": 0.284,
      "num_input_tokens_seen": 19169752,
      "step": 29385
    },
    {
      "epoch": 15.40356394129979,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.004574401215075653,
      "loss": 0.4625,
      "num_input_tokens_seen": 19172696,
      "step": 29390
    },
    {
      "epoch": 15.406184486373165,
      "grad_norm": 0.064453125,
      "learning_rate": 0.004569469753130504,
      "loss": 0.3812,
      "num_input_tokens_seen": 19176600,
      "step": 29395
    },
    {
      "epoch": 15.40880503144654,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.004564540473133285,
      "loss": 0.2391,
      "num_input_tokens_seen": 19179000,
      "step": 29400
    },
    {
      "epoch": 15.411425576519916,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.004559613376115148,
      "loss": 0.4132,
      "num_input_tokens_seen": 19182264,
      "step": 29405
    },
    {
      "epoch": 15.414046121593291,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.004554688463106778,
      "loss": 0.3332,
      "num_input_tokens_seen": 19185240,
      "step": 29410
    },
    {
      "epoch": 15.416666666666666,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004549765735138417,
      "loss": 0.3401,
      "num_input_tokens_seen": 19188920,
      "step": 29415
    },
    {
      "epoch": 15.419287211740041,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.004544845193239845,
      "loss": 0.3827,
      "num_input_tokens_seen": 19191672,
      "step": 29420
    },
    {
      "epoch": 15.421907756813416,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.004539926838440376,
      "loss": 0.3277,
      "num_input_tokens_seen": 19194744,
      "step": 29425
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0045350106717688765,
      "loss": 0.4496,
      "num_input_tokens_seen": 19197848,
      "step": 29430
    },
    {
      "epoch": 15.427148846960169,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0045300966942537595,
      "loss": 0.2612,
      "num_input_tokens_seen": 19201432,
      "step": 29435
    },
    {
      "epoch": 15.429769392033544,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.004525184906922961,
      "loss": 0.3181,
      "num_input_tokens_seen": 19205304,
      "step": 29440
    },
    {
      "epoch": 15.432389937106919,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.004520275310803978,
      "loss": 0.3956,
      "num_input_tokens_seen": 19208888,
      "step": 29445
    },
    {
      "epoch": 15.435010482180294,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.004515367906923845,
      "loss": 0.3598,
      "num_input_tokens_seen": 19212216,
      "step": 29450
    },
    {
      "epoch": 15.43763102725367,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0045104626963091265,
      "loss": 0.4516,
      "num_input_tokens_seen": 19215736,
      "step": 29455
    },
    {
      "epoch": 15.440251572327044,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.004505559679985944,
      "loss": 0.3433,
      "num_input_tokens_seen": 19219224,
      "step": 29460
    },
    {
      "epoch": 15.44287211740042,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.004500658858979947,
      "loss": 0.4033,
      "num_input_tokens_seen": 19222424,
      "step": 29465
    },
    {
      "epoch": 15.445492662473795,
      "grad_norm": 0.095703125,
      "learning_rate": 0.004495760234316333,
      "loss": 0.4051,
      "num_input_tokens_seen": 19225784,
      "step": 29470
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.004490863807019839,
      "loss": 0.2546,
      "num_input_tokens_seen": 19228600,
      "step": 29475
    },
    {
      "epoch": 15.450733752620545,
      "grad_norm": 0.091796875,
      "learning_rate": 0.004485969578114749,
      "loss": 0.3503,
      "num_input_tokens_seen": 19231864,
      "step": 29480
    },
    {
      "epoch": 15.45335429769392,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.004481077548624871,
      "loss": 0.4419,
      "num_input_tokens_seen": 19235480,
      "step": 29485
    },
    {
      "epoch": 15.455974842767295,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.004476187719573559,
      "loss": 0.4493,
      "num_input_tokens_seen": 19238904,
      "step": 29490
    },
    {
      "epoch": 15.45859538784067,
      "grad_norm": 0.064453125,
      "learning_rate": 0.004471300091983715,
      "loss": 0.3051,
      "num_input_tokens_seen": 19242040,
      "step": 29495
    },
    {
      "epoch": 15.461215932914046,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.004466414666877773,
      "loss": 0.353,
      "num_input_tokens_seen": 19245336,
      "step": 29500
    },
    {
      "epoch": 15.463836477987421,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0044615314452777135,
      "loss": 0.3827,
      "num_input_tokens_seen": 19250264,
      "step": 29505
    },
    {
      "epoch": 15.466457023060796,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.004456650428205043,
      "loss": 0.6528,
      "num_input_tokens_seen": 19253592,
      "step": 29510
    },
    {
      "epoch": 15.469077568134171,
      "grad_norm": 0.130859375,
      "learning_rate": 0.004451771616680821,
      "loss": 0.3176,
      "num_input_tokens_seen": 19256728,
      "step": 29515
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 0.12890625,
      "learning_rate": 0.004446895011725632,
      "loss": 0.3503,
      "num_input_tokens_seen": 19259640,
      "step": 29520
    },
    {
      "epoch": 15.474318658280922,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0044420206143596114,
      "loss": 0.3181,
      "num_input_tokens_seen": 19263128,
      "step": 29525
    },
    {
      "epoch": 15.476939203354299,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.004437148425602426,
      "loss": 0.3441,
      "num_input_tokens_seen": 19267928,
      "step": 29530
    },
    {
      "epoch": 15.479559748427674,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0044322784464732885,
      "loss": 0.3623,
      "num_input_tokens_seen": 19270552,
      "step": 29535
    },
    {
      "epoch": 15.482180293501049,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.004427410677990934,
      "loss": 0.3376,
      "num_input_tokens_seen": 19274072,
      "step": 29540
    },
    {
      "epoch": 15.484800838574424,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.004422545121173645,
      "loss": 0.338,
      "num_input_tokens_seen": 19277048,
      "step": 29545
    },
    {
      "epoch": 15.4874213836478,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0044176817770392425,
      "loss": 0.3498,
      "num_input_tokens_seen": 19279576,
      "step": 29550
    },
    {
      "epoch": 15.490041928721174,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.004412820646605081,
      "loss": 0.3371,
      "num_input_tokens_seen": 19282200,
      "step": 29555
    },
    {
      "epoch": 15.49266247379455,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.004407961730888063,
      "loss": 0.4582,
      "num_input_tokens_seen": 19285336,
      "step": 29560
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.004403105030904608,
      "loss": 0.3874,
      "num_input_tokens_seen": 19288344,
      "step": 29565
    },
    {
      "epoch": 15.4979035639413,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.004398250547670681,
      "loss": 0.3293,
      "num_input_tokens_seen": 19291928,
      "step": 29570
    },
    {
      "epoch": 15.500524109014675,
      "grad_norm": 0.208984375,
      "learning_rate": 0.004393398282201788,
      "loss": 0.2847,
      "num_input_tokens_seen": 19295736,
      "step": 29575
    },
    {
      "epoch": 15.50314465408805,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.004388548235512968,
      "loss": 0.3138,
      "num_input_tokens_seen": 19298744,
      "step": 29580
    },
    {
      "epoch": 15.505765199161425,
      "grad_norm": 0.103515625,
      "learning_rate": 0.004383700408618795,
      "loss": 0.3383,
      "num_input_tokens_seen": 19301368,
      "step": 29585
    },
    {
      "epoch": 15.5083857442348,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004378854802533384,
      "loss": 0.3025,
      "num_input_tokens_seen": 19304952,
      "step": 29590
    },
    {
      "epoch": 15.511006289308176,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.004374011418270377,
      "loss": 0.4366,
      "num_input_tokens_seen": 19308088,
      "step": 29595
    },
    {
      "epoch": 15.51362683438155,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.004369170256842947,
      "loss": 0.3538,
      "num_input_tokens_seen": 19310840,
      "step": 29600
    },
    {
      "epoch": 15.516247379454926,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0043643313192638184,
      "loss": 0.3718,
      "num_input_tokens_seen": 19313720,
      "step": 29605
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.004359494606545238,
      "loss": 0.4541,
      "num_input_tokens_seen": 19318424,
      "step": 29610
    },
    {
      "epoch": 15.521488469601676,
      "grad_norm": 0.08203125,
      "learning_rate": 0.004354660119699,
      "loss": 0.2808,
      "num_input_tokens_seen": 19322488,
      "step": 29615
    },
    {
      "epoch": 15.524109014675052,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.004349827859736414,
      "loss": 0.5207,
      "num_input_tokens_seen": 19325592,
      "step": 29620
    },
    {
      "epoch": 15.526729559748428,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.004344997827668331,
      "loss": 0.3803,
      "num_input_tokens_seen": 19328408,
      "step": 29625
    },
    {
      "epoch": 15.529350104821804,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.004340170024505144,
      "loss": 0.4751,
      "num_input_tokens_seen": 19330776,
      "step": 29630
    },
    {
      "epoch": 15.531970649895179,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.004335344451256776,
      "loss": 0.3584,
      "num_input_tokens_seen": 19334360,
      "step": 29635
    },
    {
      "epoch": 15.534591194968554,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.004330521108932684,
      "loss": 0.3065,
      "num_input_tokens_seen": 19336664,
      "step": 29640
    },
    {
      "epoch": 15.53721174004193,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.004325699998541848,
      "loss": 0.2955,
      "num_input_tokens_seen": 19339608,
      "step": 29645
    },
    {
      "epoch": 15.539832285115304,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.004320881121092797,
      "loss": 0.3443,
      "num_input_tokens_seen": 19342712,
      "step": 29650
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.004316064477593576,
      "loss": 0.3298,
      "num_input_tokens_seen": 19346296,
      "step": 29655
    },
    {
      "epoch": 15.545073375262055,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.004311250069051779,
      "loss": 0.2634,
      "num_input_tokens_seen": 19349656,
      "step": 29660
    },
    {
      "epoch": 15.54769392033543,
      "grad_norm": 0.095703125,
      "learning_rate": 0.004306437896474523,
      "loss": 0.3354,
      "num_input_tokens_seen": 19352280,
      "step": 29665
    },
    {
      "epoch": 15.550314465408805,
      "grad_norm": 0.078125,
      "learning_rate": 0.004301627960868467,
      "loss": 0.3853,
      "num_input_tokens_seen": 19355128,
      "step": 29670
    },
    {
      "epoch": 15.55293501048218,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.004296820263239786,
      "loss": 0.4901,
      "num_input_tokens_seen": 19359096,
      "step": 29675
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.004292014804594194,
      "loss": 0.3655,
      "num_input_tokens_seen": 19361784,
      "step": 29680
    },
    {
      "epoch": 15.55817610062893,
      "grad_norm": 0.13671875,
      "learning_rate": 0.004287211585936942,
      "loss": 0.3083,
      "num_input_tokens_seen": 19364824,
      "step": 29685
    },
    {
      "epoch": 15.560796645702306,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0042824106082728075,
      "loss": 0.4401,
      "num_input_tokens_seen": 19368056,
      "step": 29690
    },
    {
      "epoch": 15.56341719077568,
      "grad_norm": 0.052734375,
      "learning_rate": 0.004277611872606105,
      "loss": 0.3442,
      "num_input_tokens_seen": 19371448,
      "step": 29695
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.004272815379940667,
      "loss": 0.465,
      "num_input_tokens_seen": 19375448,
      "step": 29700
    },
    {
      "epoch": 15.568658280922431,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.004268021131279875,
      "loss": 0.3593,
      "num_input_tokens_seen": 19379256,
      "step": 29705
    },
    {
      "epoch": 15.571278825995806,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0042632291276266205,
      "loss": 0.3864,
      "num_input_tokens_seen": 19383032,
      "step": 29710
    },
    {
      "epoch": 15.573899371069182,
      "grad_norm": 0.078125,
      "learning_rate": 0.00425843936998334,
      "loss": 0.3646,
      "num_input_tokens_seen": 19387640,
      "step": 29715
    },
    {
      "epoch": 15.576519916142558,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0042536518593520025,
      "loss": 0.377,
      "num_input_tokens_seen": 19389944,
      "step": 29720
    },
    {
      "epoch": 15.579140461215934,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0042488665967340905,
      "loss": 0.4099,
      "num_input_tokens_seen": 19393464,
      "step": 29725
    },
    {
      "epoch": 15.581761006289309,
      "grad_norm": 0.103515625,
      "learning_rate": 0.004244083583130632,
      "loss": 0.2566,
      "num_input_tokens_seen": 19395864,
      "step": 29730
    },
    {
      "epoch": 15.584381551362684,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.004239302819542181,
      "loss": 0.3505,
      "num_input_tokens_seen": 19398552,
      "step": 29735
    },
    {
      "epoch": 15.58700209643606,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.004234524306968812,
      "loss": 0.3303,
      "num_input_tokens_seen": 19401688,
      "step": 29740
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 0.107421875,
      "learning_rate": 0.004229748046410139,
      "loss": 0.3367,
      "num_input_tokens_seen": 19404472,
      "step": 29745
    },
    {
      "epoch": 15.59224318658281,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.004224974038865305,
      "loss": 0.3424,
      "num_input_tokens_seen": 19407544,
      "step": 29750
    },
    {
      "epoch": 15.594863731656185,
      "grad_norm": 0.07421875,
      "learning_rate": 0.004220202285332969,
      "loss": 0.378,
      "num_input_tokens_seen": 19410488,
      "step": 29755
    },
    {
      "epoch": 15.59748427672956,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.004215432786811334,
      "loss": 0.5009,
      "num_input_tokens_seen": 19414360,
      "step": 29760
    },
    {
      "epoch": 15.600104821802935,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.004210665544298127,
      "loss": 0.3458,
      "num_input_tokens_seen": 19418072,
      "step": 29765
    },
    {
      "epoch": 15.60272536687631,
      "grad_norm": 0.06640625,
      "learning_rate": 0.004205900558790593,
      "loss": 0.3771,
      "num_input_tokens_seen": 19421016,
      "step": 29770
    },
    {
      "epoch": 15.605345911949685,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00420113783128552,
      "loss": 0.4252,
      "num_input_tokens_seen": 19424568,
      "step": 29775
    },
    {
      "epoch": 15.60796645702306,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.004196377362779209,
      "loss": 0.3436,
      "num_input_tokens_seen": 19427352,
      "step": 29780
    },
    {
      "epoch": 15.610587002096436,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.004191619154267499,
      "loss": 0.3343,
      "num_input_tokens_seen": 19430328,
      "step": 29785
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0041868632067457575,
      "loss": 0.2985,
      "num_input_tokens_seen": 19433848,
      "step": 29790
    },
    {
      "epoch": 15.615828092243186,
      "grad_norm": 0.119140625,
      "learning_rate": 0.004182109521208866,
      "loss": 0.3962,
      "num_input_tokens_seen": 19436888,
      "step": 29795
    },
    {
      "epoch": 15.618448637316561,
      "grad_norm": 0.109375,
      "learning_rate": 0.004177358098651251,
      "loss": 0.4698,
      "num_input_tokens_seen": 19441176,
      "step": 29800
    },
    {
      "epoch": 15.621069182389936,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.004172608940066846,
      "loss": 0.3676,
      "num_input_tokens_seen": 19444664,
      "step": 29805
    },
    {
      "epoch": 15.623689727463312,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0041678620464491245,
      "loss": 0.3479,
      "num_input_tokens_seen": 19447736,
      "step": 29810
    },
    {
      "epoch": 15.626310272536688,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.004163117418791083,
      "loss": 0.3955,
      "num_input_tokens_seen": 19451384,
      "step": 29815
    },
    {
      "epoch": 15.628930817610064,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.004158375058085249,
      "loss": 0.3792,
      "num_input_tokens_seen": 19454552,
      "step": 29820
    },
    {
      "epoch": 15.631551362683439,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.004153634965323661,
      "loss": 0.3362,
      "num_input_tokens_seen": 19457336,
      "step": 29825
    },
    {
      "epoch": 15.634171907756814,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0041488971414979006,
      "loss": 0.3688,
      "num_input_tokens_seen": 19460344,
      "step": 29830
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 0.111328125,
      "learning_rate": 0.0041441615875990575,
      "loss": 0.3666,
      "num_input_tokens_seen": 19462744,
      "step": 29835
    },
    {
      "epoch": 15.639412997903564,
      "grad_norm": 0.078125,
      "learning_rate": 0.004139428304617761,
      "loss": 0.373,
      "num_input_tokens_seen": 19465880,
      "step": 29840
    },
    {
      "epoch": 15.64203354297694,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.004134697293544157,
      "loss": 0.4184,
      "num_input_tokens_seen": 19468632,
      "step": 29845
    },
    {
      "epoch": 15.644654088050315,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.004129968555367928,
      "loss": 0.3629,
      "num_input_tokens_seen": 19471736,
      "step": 29850
    },
    {
      "epoch": 15.64727463312369,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.004125242091078266,
      "loss": 0.3197,
      "num_input_tokens_seen": 19474424,
      "step": 29855
    },
    {
      "epoch": 15.649895178197065,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0041205179016638865,
      "loss": 0.3078,
      "num_input_tokens_seen": 19477400,
      "step": 29860
    },
    {
      "epoch": 15.65251572327044,
      "grad_norm": 0.09765625,
      "learning_rate": 0.004115795988113042,
      "loss": 0.3513,
      "num_input_tokens_seen": 19480024,
      "step": 29865
    },
    {
      "epoch": 15.655136268343815,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.004111076351413505,
      "loss": 0.3966,
      "num_input_tokens_seen": 19483256,
      "step": 29870
    },
    {
      "epoch": 15.65775681341719,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.004106358992552571,
      "loss": 0.3088,
      "num_input_tokens_seen": 19486392,
      "step": 29875
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.004101643912517054,
      "loss": 0.3453,
      "num_input_tokens_seen": 19489464,
      "step": 29880
    },
    {
      "epoch": 15.66299790356394,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0040969311122932926,
      "loss": 0.3431,
      "num_input_tokens_seen": 19492952,
      "step": 29885
    },
    {
      "epoch": 15.665618448637316,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0040922205928671535,
      "loss": 0.4347,
      "num_input_tokens_seen": 19495928,
      "step": 29890
    },
    {
      "epoch": 15.668238993710691,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0040875123552240235,
      "loss": 0.3301,
      "num_input_tokens_seen": 19499800,
      "step": 29895
    },
    {
      "epoch": 15.670859538784066,
      "grad_norm": 0.0859375,
      "learning_rate": 0.004082806400348813,
      "loss": 0.3126,
      "num_input_tokens_seen": 19504408,
      "step": 29900
    },
    {
      "epoch": 15.673480083857442,
      "grad_norm": 0.05859375,
      "learning_rate": 0.004078102729225957,
      "loss": 0.308,
      "num_input_tokens_seen": 19507576,
      "step": 29905
    },
    {
      "epoch": 15.676100628930818,
      "grad_norm": 0.13671875,
      "learning_rate": 0.004073401342839407,
      "loss": 0.5203,
      "num_input_tokens_seen": 19510712,
      "step": 29910
    },
    {
      "epoch": 15.678721174004194,
      "grad_norm": 0.0703125,
      "learning_rate": 0.004068702242172634,
      "loss": 0.2976,
      "num_input_tokens_seen": 19514168,
      "step": 29915
    },
    {
      "epoch": 15.681341719077569,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.004064005428208641,
      "loss": 0.3124,
      "num_input_tokens_seen": 19517304,
      "step": 29920
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 0.07421875,
      "learning_rate": 0.004059310901929948,
      "loss": 0.3432,
      "num_input_tokens_seen": 19520504,
      "step": 29925
    },
    {
      "epoch": 15.68658280922432,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.004054618664318603,
      "loss": 0.3356,
      "num_input_tokens_seen": 19527000,
      "step": 29930
    },
    {
      "epoch": 15.689203354297694,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.004049928716356159,
      "loss": 0.394,
      "num_input_tokens_seen": 19530008,
      "step": 29935
    },
    {
      "epoch": 15.69182389937107,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0040452410590237,
      "loss": 0.2642,
      "num_input_tokens_seen": 19533208,
      "step": 29940
    },
    {
      "epoch": 15.694444444444445,
      "grad_norm": 0.072265625,
      "learning_rate": 0.004040555693301831,
      "loss": 0.5758,
      "num_input_tokens_seen": 19536248,
      "step": 29945
    },
    {
      "epoch": 15.69706498951782,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.004035872620170681,
      "loss": 0.3385,
      "num_input_tokens_seen": 19539480,
      "step": 29950
    },
    {
      "epoch": 15.699685534591195,
      "grad_norm": 0.126953125,
      "learning_rate": 0.004031191840609897,
      "loss": 0.3999,
      "num_input_tokens_seen": 19542584,
      "step": 29955
    },
    {
      "epoch": 15.70230607966457,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0040265133555986415,
      "loss": 0.4584,
      "num_input_tokens_seen": 19546296,
      "step": 29960
    },
    {
      "epoch": 15.704926624737945,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0040218371661155956,
      "loss": 0.4023,
      "num_input_tokens_seen": 19549400,
      "step": 29965
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.004017163273138968,
      "loss": 0.3872,
      "num_input_tokens_seen": 19552984,
      "step": 29970
    },
    {
      "epoch": 15.710167714884696,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.004012491677646487,
      "loss": 0.4705,
      "num_input_tokens_seen": 19555448,
      "step": 29975
    },
    {
      "epoch": 15.71278825995807,
      "grad_norm": 0.078125,
      "learning_rate": 0.004007822380615393,
      "loss": 0.3156,
      "num_input_tokens_seen": 19558136,
      "step": 29980
    },
    {
      "epoch": 15.715408805031446,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.004003155383022455,
      "loss": 0.2854,
      "num_input_tokens_seen": 19560824,
      "step": 29985
    },
    {
      "epoch": 15.718029350104821,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.003998490685843954,
      "loss": 0.2734,
      "num_input_tokens_seen": 19564472,
      "step": 29990
    },
    {
      "epoch": 15.720649895178196,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.003993828290055685,
      "loss": 0.4918,
      "num_input_tokens_seen": 19567960,
      "step": 29995
    },
    {
      "epoch": 15.723270440251572,
      "grad_norm": 0.193359375,
      "learning_rate": 0.003989168196632973,
      "loss": 0.3356,
      "num_input_tokens_seen": 19570616,
      "step": 30000
    },
    {
      "epoch": 15.725890985324948,
      "grad_norm": 0.1328125,
      "learning_rate": 0.003984510406550655,
      "loss": 0.4804,
      "num_input_tokens_seen": 19573816,
      "step": 30005
    },
    {
      "epoch": 15.728511530398324,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.003979854920783094,
      "loss": 0.415,
      "num_input_tokens_seen": 19577272,
      "step": 30010
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.003975201740304154,
      "loss": 0.2972,
      "num_input_tokens_seen": 19580408,
      "step": 30015
    },
    {
      "epoch": 15.733752620545074,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.003970550866087238,
      "loss": 0.341,
      "num_input_tokens_seen": 19583640,
      "step": 30020
    },
    {
      "epoch": 15.73637316561845,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.003965902299105245,
      "loss": 0.3436,
      "num_input_tokens_seen": 19587128,
      "step": 30025
    },
    {
      "epoch": 15.738993710691824,
      "grad_norm": 0.076171875,
      "learning_rate": 0.003961256040330608,
      "loss": 0.3757,
      "num_input_tokens_seen": 19590584,
      "step": 30030
    },
    {
      "epoch": 15.7416142557652,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.003956612090735274,
      "loss": 0.3773,
      "num_input_tokens_seen": 19594264,
      "step": 30035
    },
    {
      "epoch": 15.744234800838575,
      "grad_norm": 0.154296875,
      "learning_rate": 0.003951970451290697,
      "loss": 0.3832,
      "num_input_tokens_seen": 19597176,
      "step": 30040
    },
    {
      "epoch": 15.74685534591195,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.003947331122967863,
      "loss": 0.4587,
      "num_input_tokens_seen": 19599896,
      "step": 30045
    },
    {
      "epoch": 15.749475890985325,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.003942694106737259,
      "loss": 0.3392,
      "num_input_tokens_seen": 19602904,
      "step": 30050
    },
    {
      "epoch": 15.7520964360587,
      "grad_norm": 0.05078125,
      "learning_rate": 0.003938059403568898,
      "loss": 0.3935,
      "num_input_tokens_seen": 19606456,
      "step": 30055
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.00393342701443231,
      "loss": 0.3066,
      "num_input_tokens_seen": 19609304,
      "step": 30060
    },
    {
      "epoch": 15.75733752620545,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.00392879694029654,
      "loss": 0.3539,
      "num_input_tokens_seen": 19613304,
      "step": 30065
    },
    {
      "epoch": 15.759958071278826,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.003924169182130139,
      "loss": 0.3934,
      "num_input_tokens_seen": 19616952,
      "step": 30070
    },
    {
      "epoch": 15.7625786163522,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00391954374090119,
      "loss": 0.3058,
      "num_input_tokens_seen": 19620056,
      "step": 30075
    },
    {
      "epoch": 15.765199161425576,
      "grad_norm": 0.08203125,
      "learning_rate": 0.003914920617577273,
      "loss": 0.3193,
      "num_input_tokens_seen": 19623096,
      "step": 30080
    },
    {
      "epoch": 15.767819706498951,
      "grad_norm": 0.095703125,
      "learning_rate": 0.003910299813125498,
      "loss": 0.5299,
      "num_input_tokens_seen": 19625624,
      "step": 30085
    },
    {
      "epoch": 15.770440251572326,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0039056813285124896,
      "loss": 0.3734,
      "num_input_tokens_seen": 19628312,
      "step": 30090
    },
    {
      "epoch": 15.773060796645701,
      "grad_norm": 0.1015625,
      "learning_rate": 0.003901065164704371,
      "loss": 0.3242,
      "num_input_tokens_seen": 19632056,
      "step": 30095
    },
    {
      "epoch": 15.775681341719078,
      "grad_norm": 0.08984375,
      "learning_rate": 0.003896451322666796,
      "loss": 0.3582,
      "num_input_tokens_seen": 19637560,
      "step": 30100
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 0.109375,
      "learning_rate": 0.0038918398033649334,
      "loss": 0.3691,
      "num_input_tokens_seen": 19640600,
      "step": 30105
    },
    {
      "epoch": 15.780922431865829,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00388723060776345,
      "loss": 0.3481,
      "num_input_tokens_seen": 19643608,
      "step": 30110
    },
    {
      "epoch": 15.783542976939204,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0038826237368265485,
      "loss": 0.4128,
      "num_input_tokens_seen": 19646040,
      "step": 30115
    },
    {
      "epoch": 15.786163522012579,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.003878019191517922,
      "loss": 0.3736,
      "num_input_tokens_seen": 19648600,
      "step": 30120
    },
    {
      "epoch": 15.788784067085954,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0038734169728007937,
      "loss": 0.2749,
      "num_input_tokens_seen": 19652280,
      "step": 30125
    },
    {
      "epoch": 15.79140461215933,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.003868817081637898,
      "loss": 0.3956,
      "num_input_tokens_seen": 19655736,
      "step": 30130
    },
    {
      "epoch": 15.794025157232705,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00386421951899148,
      "loss": 0.4061,
      "num_input_tokens_seen": 19658456,
      "step": 30135
    },
    {
      "epoch": 15.79664570230608,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0038596242858232907,
      "loss": 0.3522,
      "num_input_tokens_seen": 19661688,
      "step": 30140
    },
    {
      "epoch": 15.799266247379455,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0038550313830946095,
      "loss": 0.3777,
      "num_input_tokens_seen": 19665400,
      "step": 30145
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0038504408117662112,
      "loss": 0.3367,
      "num_input_tokens_seen": 19668472,
      "step": 30150
    },
    {
      "epoch": 15.804507337526205,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0038458525727983928,
      "loss": 0.2801,
      "num_input_tokens_seen": 19672088,
      "step": 30155
    },
    {
      "epoch": 15.80712788259958,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0038412666671509676,
      "loss": 0.3635,
      "num_input_tokens_seen": 19675864,
      "step": 30160
    },
    {
      "epoch": 15.809748427672956,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.003836683095783245,
      "loss": 0.3332,
      "num_input_tokens_seen": 19680216,
      "step": 30165
    },
    {
      "epoch": 15.81236897274633,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.003832101859654067,
      "loss": 0.5007,
      "num_input_tokens_seen": 19683480,
      "step": 30170
    },
    {
      "epoch": 15.814989517819706,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0038275229597217633,
      "loss": 0.4743,
      "num_input_tokens_seen": 19687160,
      "step": 30175
    },
    {
      "epoch": 15.817610062893081,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.003822946396944196,
      "loss": 0.4599,
      "num_input_tokens_seen": 19690424,
      "step": 30180
    },
    {
      "epoch": 15.820230607966456,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0038183721722787277,
      "loss": 0.3571,
      "num_input_tokens_seen": 19693656,
      "step": 30185
    },
    {
      "epoch": 15.822851153039831,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.003813800286682238,
      "loss": 0.4814,
      "num_input_tokens_seen": 19696920,
      "step": 30190
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00380923074111111,
      "loss": 0.2503,
      "num_input_tokens_seen": 19700088,
      "step": 30195
    },
    {
      "epoch": 15.828092243186584,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.003804663536521234,
      "loss": 0.3284,
      "num_input_tokens_seen": 19703320,
      "step": 30200
    },
    {
      "epoch": 15.830712788259959,
      "grad_norm": 0.150390625,
      "learning_rate": 0.003800098673868024,
      "loss": 0.3699,
      "num_input_tokens_seen": 19706168,
      "step": 30205
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.003795536154106397,
      "loss": 0.2886,
      "num_input_tokens_seen": 19708984,
      "step": 30210
    },
    {
      "epoch": 15.835953878406709,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0037909759781907786,
      "loss": 0.3568,
      "num_input_tokens_seen": 19711928,
      "step": 30215
    },
    {
      "epoch": 15.838574423480084,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0037864181470751096,
      "loss": 0.3817,
      "num_input_tokens_seen": 19715640,
      "step": 30220
    },
    {
      "epoch": 15.84119496855346,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0037818626617128334,
      "loss": 0.3031,
      "num_input_tokens_seen": 19718616,
      "step": 30225
    },
    {
      "epoch": 15.843815513626835,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.003777309523056902,
      "loss": 0.3537,
      "num_input_tokens_seen": 19721688,
      "step": 30230
    },
    {
      "epoch": 15.84643605870021,
      "grad_norm": 0.078125,
      "learning_rate": 0.0037727587320597828,
      "loss": 0.3493,
      "num_input_tokens_seen": 19725272,
      "step": 30235
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.003768210289673449,
      "loss": 0.2985,
      "num_input_tokens_seen": 19727960,
      "step": 30240
    },
    {
      "epoch": 15.85167714884696,
      "grad_norm": 0.091796875,
      "learning_rate": 0.003763664196849387,
      "loss": 0.4267,
      "num_input_tokens_seen": 19730904,
      "step": 30245
    },
    {
      "epoch": 15.854297693920335,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.003759120454538586,
      "loss": 0.4325,
      "num_input_tokens_seen": 19734552,
      "step": 30250
    },
    {
      "epoch": 15.85691823899371,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0037545790636915382,
      "loss": 0.3566,
      "num_input_tokens_seen": 19737400,
      "step": 30255
    },
    {
      "epoch": 15.859538784067086,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.003750040025258256,
      "loss": 0.3303,
      "num_input_tokens_seen": 19740792,
      "step": 30260
    },
    {
      "epoch": 15.86215932914046,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.003745503340188253,
      "loss": 0.3689,
      "num_input_tokens_seen": 19744248,
      "step": 30265
    },
    {
      "epoch": 15.864779874213836,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0037409690094305586,
      "loss": 0.2515,
      "num_input_tokens_seen": 19746744,
      "step": 30270
    },
    {
      "epoch": 15.867400419287211,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.003736437033933697,
      "loss": 0.3618,
      "num_input_tokens_seen": 19749688,
      "step": 30275
    },
    {
      "epoch": 15.870020964360586,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0037319074146457036,
      "loss": 0.3078,
      "num_input_tokens_seen": 19753208,
      "step": 30280
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0037273801525141235,
      "loss": 0.455,
      "num_input_tokens_seen": 19756568,
      "step": 30285
    },
    {
      "epoch": 15.875262054507338,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0037228552484860128,
      "loss": 0.3649,
      "num_input_tokens_seen": 19759576,
      "step": 30290
    },
    {
      "epoch": 15.877882599580714,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0037183327035079255,
      "loss": 0.2974,
      "num_input_tokens_seen": 19763608,
      "step": 30295
    },
    {
      "epoch": 15.880503144654089,
      "grad_norm": 0.0625,
      "learning_rate": 0.0037138125185259352,
      "loss": 0.5546,
      "num_input_tokens_seen": 19766808,
      "step": 30300
    },
    {
      "epoch": 15.883123689727464,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.003709294694485605,
      "loss": 0.27,
      "num_input_tokens_seen": 19769752,
      "step": 30305
    },
    {
      "epoch": 15.885744234800839,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0037047792323320083,
      "loss": 0.5091,
      "num_input_tokens_seen": 19772920,
      "step": 30310
    },
    {
      "epoch": 15.888364779874214,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.003700266133009735,
      "loss": 0.3357,
      "num_input_tokens_seen": 19775512,
      "step": 30315
    },
    {
      "epoch": 15.89098532494759,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0036957553974628723,
      "loss": 0.4531,
      "num_input_tokens_seen": 19778968,
      "step": 30320
    },
    {
      "epoch": 15.893605870020965,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0036912470266350188,
      "loss": 0.31,
      "num_input_tokens_seen": 19781656,
      "step": 30325
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.003686741021469271,
      "loss": 0.4246,
      "num_input_tokens_seen": 19785112,
      "step": 30330
    },
    {
      "epoch": 15.898846960167715,
      "grad_norm": 0.064453125,
      "learning_rate": 0.003682237382908228,
      "loss": 0.3796,
      "num_input_tokens_seen": 19788088,
      "step": 30335
    },
    {
      "epoch": 15.90146750524109,
      "grad_norm": 0.09375,
      "learning_rate": 0.003677736111894004,
      "loss": 0.2571,
      "num_input_tokens_seen": 19791352,
      "step": 30340
    },
    {
      "epoch": 15.904088050314465,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.003673237209368216,
      "loss": 0.3638,
      "num_input_tokens_seen": 19795288,
      "step": 30345
    },
    {
      "epoch": 15.90670859538784,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0036687406762719844,
      "loss": 0.372,
      "num_input_tokens_seen": 19798328,
      "step": 30350
    },
    {
      "epoch": 15.909329140461216,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0036642465135459257,
      "loss": 0.2622,
      "num_input_tokens_seen": 19801432,
      "step": 30355
    },
    {
      "epoch": 15.91194968553459,
      "grad_norm": 0.09375,
      "learning_rate": 0.003659754722130175,
      "loss": 0.4387,
      "num_input_tokens_seen": 19804280,
      "step": 30360
    },
    {
      "epoch": 15.914570230607966,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.003655265302964356,
      "loss": 0.286,
      "num_input_tokens_seen": 19807768,
      "step": 30365
    },
    {
      "epoch": 15.917190775681341,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0036507782569876073,
      "loss": 0.3395,
      "num_input_tokens_seen": 19810616,
      "step": 30370
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 0.078125,
      "learning_rate": 0.0036462935851385696,
      "loss": 0.3064,
      "num_input_tokens_seen": 19816472,
      "step": 30375
    },
    {
      "epoch": 15.922431865828091,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0036418112883553878,
      "loss": 0.2876,
      "num_input_tokens_seen": 19819928,
      "step": 30380
    },
    {
      "epoch": 15.925052410901468,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.003637331367575698,
      "loss": 0.3735,
      "num_input_tokens_seen": 19823704,
      "step": 30385
    },
    {
      "epoch": 15.927672955974844,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0036328538237366606,
      "loss": 0.3972,
      "num_input_tokens_seen": 19826456,
      "step": 30390
    },
    {
      "epoch": 15.930293501048219,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0036283786577749138,
      "loss": 0.3735,
      "num_input_tokens_seen": 19829848,
      "step": 30395
    },
    {
      "epoch": 15.932914046121594,
      "grad_norm": 0.08984375,
      "learning_rate": 0.003623905870626618,
      "loss": 0.3941,
      "num_input_tokens_seen": 19832856,
      "step": 30400
    },
    {
      "epoch": 15.935534591194969,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0036194354632274332,
      "loss": 0.3348,
      "num_input_tokens_seen": 19835704,
      "step": 30405
    },
    {
      "epoch": 15.938155136268344,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.003614967436512508,
      "loss": 0.3245,
      "num_input_tokens_seen": 19839352,
      "step": 30410
    },
    {
      "epoch": 15.94077568134172,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.003610501791416514,
      "loss": 0.3567,
      "num_input_tokens_seen": 19842840,
      "step": 30415
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0036060385288736005,
      "loss": 0.344,
      "num_input_tokens_seen": 19846808,
      "step": 30420
    },
    {
      "epoch": 15.94601677148847,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.003601577649817439,
      "loss": 0.3507,
      "num_input_tokens_seen": 19849816,
      "step": 30425
    },
    {
      "epoch": 15.948637316561845,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.003597119155181197,
      "loss": 0.2843,
      "num_input_tokens_seen": 19852568,
      "step": 30430
    },
    {
      "epoch": 15.95125786163522,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.003592663045897532,
      "loss": 0.254,
      "num_input_tokens_seen": 19857976,
      "step": 30435
    },
    {
      "epoch": 15.953878406708595,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0035882093228986173,
      "loss": 0.3879,
      "num_input_tokens_seen": 19861432,
      "step": 30440
    },
    {
      "epoch": 15.95649895178197,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.003583757987116123,
      "loss": 0.4389,
      "num_input_tokens_seen": 19865144,
      "step": 30445
    },
    {
      "epoch": 15.959119496855346,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.003579309039481211,
      "loss": 0.3121,
      "num_input_tokens_seen": 19867928,
      "step": 30450
    },
    {
      "epoch": 15.96174004192872,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.003574862480924552,
      "loss": 0.3159,
      "num_input_tokens_seen": 19870456,
      "step": 30455
    },
    {
      "epoch": 15.964360587002096,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.003570418312376323,
      "loss": 0.3306,
      "num_input_tokens_seen": 19872824,
      "step": 30460
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0035659765347661825,
      "loss": 0.2908,
      "num_input_tokens_seen": 19875256,
      "step": 30465
    },
    {
      "epoch": 15.969601677148846,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.003561537149023305,
      "loss": 0.407,
      "num_input_tokens_seen": 19878328,
      "step": 30470
    },
    {
      "epoch": 15.972222222222221,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0035571001560763637,
      "loss": 0.3915,
      "num_input_tokens_seen": 19881784,
      "step": 30475
    },
    {
      "epoch": 15.974842767295598,
      "grad_norm": 0.138671875,
      "learning_rate": 0.003552665556853518,
      "loss": 0.3758,
      "num_input_tokens_seen": 19885144,
      "step": 30480
    },
    {
      "epoch": 15.977463312368974,
      "grad_norm": 0.078125,
      "learning_rate": 0.0035482333522824424,
      "loss": 0.4278,
      "num_input_tokens_seen": 19888120,
      "step": 30485
    },
    {
      "epoch": 15.980083857442349,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0035438035432902976,
      "loss": 0.4181,
      "num_input_tokens_seen": 19891352,
      "step": 30490
    },
    {
      "epoch": 15.982704402515724,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0035393761308037527,
      "loss": 0.4167,
      "num_input_tokens_seen": 19894872,
      "step": 30495
    },
    {
      "epoch": 15.985324947589099,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0035349511157489713,
      "loss": 0.3277,
      "num_input_tokens_seen": 19898904,
      "step": 30500
    },
    {
      "epoch": 15.987945492662474,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.003530528499051621,
      "loss": 0.4118,
      "num_input_tokens_seen": 19902840,
      "step": 30505
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0035261082816368567,
      "loss": 0.3257,
      "num_input_tokens_seen": 19906168,
      "step": 30510
    },
    {
      "epoch": 15.993186582809225,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0035216904644293366,
      "loss": 0.5112,
      "num_input_tokens_seen": 19908920,
      "step": 30515
    },
    {
      "epoch": 15.9958071278826,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.003517275048353217,
      "loss": 0.4194,
      "num_input_tokens_seen": 19912216,
      "step": 30520
    },
    {
      "epoch": 15.998427672955975,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0035128620343321576,
      "loss": 0.3709,
      "num_input_tokens_seen": 19914904,
      "step": 30525
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.4946008622646332,
      "eval_runtime": 13.6746,
      "eval_samples_per_second": 62.013,
      "eval_steps_per_second": 15.503,
      "num_input_tokens_seen": 19916008,
      "step": 30528
    },
    {
      "epoch": 16.00104821802935,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0035084514232893126,
      "loss": 0.3791,
      "num_input_tokens_seen": 19917032,
      "step": 30530
    },
    {
      "epoch": 16.003668763102727,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.003504043216147323,
      "loss": 0.4696,
      "num_input_tokens_seen": 19919976,
      "step": 30535
    },
    {
      "epoch": 16.0062893081761,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0034996374138283434,
      "loss": 0.3783,
      "num_input_tokens_seen": 19922312,
      "step": 30540
    },
    {
      "epoch": 16.008909853249477,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0034952340172540097,
      "loss": 0.4467,
      "num_input_tokens_seen": 19924392,
      "step": 30545
    },
    {
      "epoch": 16.01153039832285,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0034908330273454658,
      "loss": 0.3572,
      "num_input_tokens_seen": 19927528,
      "step": 30550
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0034864344450233503,
      "loss": 0.3403,
      "num_input_tokens_seen": 19930920,
      "step": 30555
    },
    {
      "epoch": 16.0167714884696,
      "grad_norm": 0.076171875,
      "learning_rate": 0.003482038271207799,
      "loss": 0.2763,
      "num_input_tokens_seen": 19934152,
      "step": 30560
    },
    {
      "epoch": 16.019392033542978,
      "grad_norm": 0.111328125,
      "learning_rate": 0.0034776445068184364,
      "loss": 0.2934,
      "num_input_tokens_seen": 19938216,
      "step": 30565
    },
    {
      "epoch": 16.02201257861635,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.003473253152774385,
      "loss": 0.4304,
      "num_input_tokens_seen": 19942088,
      "step": 30570
    },
    {
      "epoch": 16.02463312368973,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.003468864209994271,
      "loss": 0.6509,
      "num_input_tokens_seen": 19944744,
      "step": 30575
    },
    {
      "epoch": 16.0272536687631,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0034644776793962113,
      "loss": 0.3519,
      "num_input_tokens_seen": 19948360,
      "step": 30580
    },
    {
      "epoch": 16.02987421383648,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0034600935618978143,
      "loss": 0.3313,
      "num_input_tokens_seen": 19951016,
      "step": 30585
    },
    {
      "epoch": 16.032494758909852,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0034557118584161994,
      "loss": 0.3429,
      "num_input_tokens_seen": 19954376,
      "step": 30590
    },
    {
      "epoch": 16.03511530398323,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0034513325698679507,
      "loss": 0.3508,
      "num_input_tokens_seen": 19957864,
      "step": 30595
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 0.06640625,
      "learning_rate": 0.003446955697169174,
      "loss": 0.3331,
      "num_input_tokens_seen": 19961256,
      "step": 30600
    },
    {
      "epoch": 16.04035639412998,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0034425812412354627,
      "loss": 0.3103,
      "num_input_tokens_seen": 19964744,
      "step": 30605
    },
    {
      "epoch": 16.042976939203353,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0034382092029819006,
      "loss": 0.3882,
      "num_input_tokens_seen": 19968232,
      "step": 30610
    },
    {
      "epoch": 16.04559748427673,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0034338395833230745,
      "loss": 0.3128,
      "num_input_tokens_seen": 19971176,
      "step": 30615
    },
    {
      "epoch": 16.048218029350103,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0034294723831730534,
      "loss": 0.3759,
      "num_input_tokens_seen": 19974888,
      "step": 30620
    },
    {
      "epoch": 16.05083857442348,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0034251076034454014,
      "loss": 0.3278,
      "num_input_tokens_seen": 19977864,
      "step": 30625
    },
    {
      "epoch": 16.053459119496857,
      "grad_norm": 0.078125,
      "learning_rate": 0.0034207452450531867,
      "loss": 0.4145,
      "num_input_tokens_seen": 19981640,
      "step": 30630
    },
    {
      "epoch": 16.05607966457023,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0034163853089089623,
      "loss": 0.4172,
      "num_input_tokens_seen": 19985640,
      "step": 30635
    },
    {
      "epoch": 16.058700209643607,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0034120277959247824,
      "loss": 0.2724,
      "num_input_tokens_seen": 19988776,
      "step": 30640
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.003407672707012186,
      "loss": 0.3177,
      "num_input_tokens_seen": 19991976,
      "step": 30645
    },
    {
      "epoch": 16.063941299790358,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0034033200430822023,
      "loss": 0.3403,
      "num_input_tokens_seen": 19995048,
      "step": 30650
    },
    {
      "epoch": 16.06656184486373,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0033989698050453637,
      "loss": 0.423,
      "num_input_tokens_seen": 19997992,
      "step": 30655
    },
    {
      "epoch": 16.069182389937108,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.003394621993811692,
      "loss": 0.293,
      "num_input_tokens_seen": 20000520,
      "step": 30660
    },
    {
      "epoch": 16.07180293501048,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0033902766102906954,
      "loss": 0.4904,
      "num_input_tokens_seen": 20003176,
      "step": 30665
    },
    {
      "epoch": 16.07442348008386,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0033859336553913864,
      "loss": 0.4278,
      "num_input_tokens_seen": 20006664,
      "step": 30670
    },
    {
      "epoch": 16.07704402515723,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.003381593130022256,
      "loss": 0.3414,
      "num_input_tokens_seen": 20009320,
      "step": 30675
    },
    {
      "epoch": 16.07966457023061,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0033772550350912897,
      "loss": 0.27,
      "num_input_tokens_seen": 20011656,
      "step": 30680
    },
    {
      "epoch": 16.082285115303982,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0033729193715059716,
      "loss": 0.3739,
      "num_input_tokens_seen": 20015304,
      "step": 30685
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0033685861401732712,
      "loss": 0.324,
      "num_input_tokens_seen": 20018504,
      "step": 30690
    },
    {
      "epoch": 16.087526205450732,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.003364255341999655,
      "loss": 0.3086,
      "num_input_tokens_seen": 20021096,
      "step": 30695
    },
    {
      "epoch": 16.09014675052411,
      "grad_norm": 0.115234375,
      "learning_rate": 0.003359926977891075,
      "loss": 0.3991,
      "num_input_tokens_seen": 20023912,
      "step": 30700
    },
    {
      "epoch": 16.092767295597483,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00335560104875297,
      "loss": 0.3705,
      "num_input_tokens_seen": 20026792,
      "step": 30705
    },
    {
      "epoch": 16.09538784067086,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0033512775554902802,
      "loss": 0.3455,
      "num_input_tokens_seen": 20028840,
      "step": 30710
    },
    {
      "epoch": 16.098008385744233,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0033469564990074277,
      "loss": 0.3346,
      "num_input_tokens_seen": 20032072,
      "step": 30715
    },
    {
      "epoch": 16.10062893081761,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.003342637880208337,
      "loss": 0.3117,
      "num_input_tokens_seen": 20034824,
      "step": 30720
    },
    {
      "epoch": 16.103249475890987,
      "grad_norm": 0.046875,
      "learning_rate": 0.003338321699996402,
      "loss": 0.2571,
      "num_input_tokens_seen": 20038792,
      "step": 30725
    },
    {
      "epoch": 16.10587002096436,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0033340079592745287,
      "loss": 0.2757,
      "num_input_tokens_seen": 20042024,
      "step": 30730
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0033296966589450947,
      "loss": 0.4316,
      "num_input_tokens_seen": 20045576,
      "step": 30735
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0033253877999099763,
      "loss": 0.2897,
      "num_input_tokens_seen": 20048872,
      "step": 30740
    },
    {
      "epoch": 16.113731656184488,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0033210813830705397,
      "loss": 0.397,
      "num_input_tokens_seen": 20053096,
      "step": 30745
    },
    {
      "epoch": 16.11635220125786,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0033167774093276403,
      "loss": 0.3369,
      "num_input_tokens_seen": 20056296,
      "step": 30750
    },
    {
      "epoch": 16.118972746331238,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0033124758795816155,
      "loss": 0.4035,
      "num_input_tokens_seen": 20059368,
      "step": 30755
    },
    {
      "epoch": 16.12159329140461,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0033081767947323007,
      "loss": 0.2602,
      "num_input_tokens_seen": 20062472,
      "step": 30760
    },
    {
      "epoch": 16.12421383647799,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0033038801556790104,
      "loss": 0.298,
      "num_input_tokens_seen": 20066024,
      "step": 30765
    },
    {
      "epoch": 16.12683438155136,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0032995859633205547,
      "loss": 0.3122,
      "num_input_tokens_seen": 20068968,
      "step": 30770
    },
    {
      "epoch": 16.12945492662474,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.003295294218555234,
      "loss": 0.3467,
      "num_input_tokens_seen": 20071720,
      "step": 30775
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.003291004922280825,
      "loss": 0.5831,
      "num_input_tokens_seen": 20074984,
      "step": 30780
    },
    {
      "epoch": 16.13469601677149,
      "grad_norm": 0.068359375,
      "learning_rate": 0.003286718075394607,
      "loss": 0.4637,
      "num_input_tokens_seen": 20078760,
      "step": 30785
    },
    {
      "epoch": 16.137316561844862,
      "grad_norm": 0.123046875,
      "learning_rate": 0.00328243367879333,
      "loss": 0.3199,
      "num_input_tokens_seen": 20081480,
      "step": 30790
    },
    {
      "epoch": 16.13993710691824,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.003278151733373248,
      "loss": 0.3538,
      "num_input_tokens_seen": 20084072,
      "step": 30795
    },
    {
      "epoch": 16.142557651991613,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0032738722400300958,
      "loss": 0.4764,
      "num_input_tokens_seen": 20088488,
      "step": 30800
    },
    {
      "epoch": 16.14517819706499,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00326959519965909,
      "loss": 0.3612,
      "num_input_tokens_seen": 20092008,
      "step": 30805
    },
    {
      "epoch": 16.147798742138363,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.00326532061315494,
      "loss": 0.3909,
      "num_input_tokens_seen": 20094664,
      "step": 30810
    },
    {
      "epoch": 16.15041928721174,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0032610484814118456,
      "loss": 0.3299,
      "num_input_tokens_seen": 20101640,
      "step": 30815
    },
    {
      "epoch": 16.153039832285117,
      "grad_norm": 0.0859375,
      "learning_rate": 0.003256778805323479,
      "loss": 0.3958,
      "num_input_tokens_seen": 20104840,
      "step": 30820
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 0.052734375,
      "learning_rate": 0.003252511585783014,
      "loss": 0.4305,
      "num_input_tokens_seen": 20108456,
      "step": 30825
    },
    {
      "epoch": 16.158280922431867,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.003248246823683106,
      "loss": 0.4219,
      "num_input_tokens_seen": 20110792,
      "step": 30830
    },
    {
      "epoch": 16.16090146750524,
      "grad_norm": 0.09765625,
      "learning_rate": 0.003243984519915888,
      "loss": 0.2395,
      "num_input_tokens_seen": 20113256,
      "step": 30835
    },
    {
      "epoch": 16.163522012578618,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0032397246753729874,
      "loss": 0.2808,
      "num_input_tokens_seen": 20116424,
      "step": 30840
    },
    {
      "epoch": 16.16614255765199,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0032354672909455217,
      "loss": 0.4083,
      "num_input_tokens_seen": 20120200,
      "step": 30845
    },
    {
      "epoch": 16.168763102725368,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0032312123675240782,
      "loss": 0.3805,
      "num_input_tokens_seen": 20122664,
      "step": 30850
    },
    {
      "epoch": 16.17138364779874,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.003226959905998745,
      "loss": 0.4197,
      "num_input_tokens_seen": 20125224,
      "step": 30855
    },
    {
      "epoch": 16.17400419287212,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0032227099072590806,
      "loss": 0.3661,
      "num_input_tokens_seen": 20128968,
      "step": 30860
    },
    {
      "epoch": 16.17662473794549,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0032184623721941423,
      "loss": 0.3275,
      "num_input_tokens_seen": 20131656,
      "step": 30865
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.003214217301692466,
      "loss": 0.3599,
      "num_input_tokens_seen": 20134696,
      "step": 30870
    },
    {
      "epoch": 16.181865828092242,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.003209974696642074,
      "loss": 0.3881,
      "num_input_tokens_seen": 20137192,
      "step": 30875
    },
    {
      "epoch": 16.18448637316562,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0032057345579304666,
      "loss": 0.2786,
      "num_input_tokens_seen": 20140712,
      "step": 30880
    },
    {
      "epoch": 16.187106918238992,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0032014968864446302,
      "loss": 0.4452,
      "num_input_tokens_seen": 20143496,
      "step": 30885
    },
    {
      "epoch": 16.18972746331237,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0031972616830710414,
      "loss": 0.3081,
      "num_input_tokens_seen": 20148296,
      "step": 30890
    },
    {
      "epoch": 16.192348008385743,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.003193028948695655,
      "loss": 0.3394,
      "num_input_tokens_seen": 20152168,
      "step": 30895
    },
    {
      "epoch": 16.19496855345912,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0031887986842039167,
      "loss": 0.4149,
      "num_input_tokens_seen": 20155784,
      "step": 30900
    },
    {
      "epoch": 16.197589098532493,
      "grad_norm": 0.068359375,
      "learning_rate": 0.003184570890480743,
      "loss": 0.4881,
      "num_input_tokens_seen": 20159112,
      "step": 30905
    },
    {
      "epoch": 16.20020964360587,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.003180345568410538,
      "loss": 0.3112,
      "num_input_tokens_seen": 20162152,
      "step": 30910
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.003176122718877195,
      "loss": 0.3894,
      "num_input_tokens_seen": 20165032,
      "step": 30915
    },
    {
      "epoch": 16.20545073375262,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.003171902342764084,
      "loss": 0.3567,
      "num_input_tokens_seen": 20168136,
      "step": 30920
    },
    {
      "epoch": 16.208071278825997,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0031676844409540606,
      "loss": 0.2921,
      "num_input_tokens_seen": 20171368,
      "step": 30925
    },
    {
      "epoch": 16.21069182389937,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0031634690143294668,
      "loss": 0.2212,
      "num_input_tokens_seen": 20174664,
      "step": 30930
    },
    {
      "epoch": 16.213312368972748,
      "grad_norm": 0.091796875,
      "learning_rate": 0.003159256063772118,
      "loss": 0.512,
      "num_input_tokens_seen": 20177224,
      "step": 30935
    },
    {
      "epoch": 16.21593291404612,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0031550455901633084,
      "loss": 0.309,
      "num_input_tokens_seen": 20180200,
      "step": 30940
    },
    {
      "epoch": 16.218553459119498,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0031508375943838274,
      "loss": 0.2961,
      "num_input_tokens_seen": 20183592,
      "step": 30945
    },
    {
      "epoch": 16.22117400419287,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0031466320773139403,
      "loss": 0.351,
      "num_input_tokens_seen": 20186760,
      "step": 30950
    },
    {
      "epoch": 16.22379454926625,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0031424290398333957,
      "loss": 0.3442,
      "num_input_tokens_seen": 20189864,
      "step": 30955
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.003138228482821417,
      "loss": 0.397,
      "num_input_tokens_seen": 20193096,
      "step": 30960
    },
    {
      "epoch": 16.229035639413,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.00313403040715671,
      "loss": 0.3147,
      "num_input_tokens_seen": 20196136,
      "step": 30965
    },
    {
      "epoch": 16.231656184486372,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0031298348137174673,
      "loss": 0.3884,
      "num_input_tokens_seen": 20199880,
      "step": 30970
    },
    {
      "epoch": 16.23427672955975,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0031256417033813607,
      "loss": 0.3614,
      "num_input_tokens_seen": 20203752,
      "step": 30975
    },
    {
      "epoch": 16.236897274633122,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.003121451077025539,
      "loss": 0.27,
      "num_input_tokens_seen": 20206760,
      "step": 30980
    },
    {
      "epoch": 16.2395178197065,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0031172629355266427,
      "loss": 0.3776,
      "num_input_tokens_seen": 20209096,
      "step": 30985
    },
    {
      "epoch": 16.242138364779873,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.003113077279760766,
      "loss": 0.2662,
      "num_input_tokens_seen": 20212520,
      "step": 30990
    },
    {
      "epoch": 16.24475890985325,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0031088941106035112,
      "loss": 0.2804,
      "num_input_tokens_seen": 20215720,
      "step": 30995
    },
    {
      "epoch": 16.247379454926623,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.003104713428929945,
      "loss": 0.2583,
      "num_input_tokens_seen": 20219048,
      "step": 31000
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0031005352356146226,
      "loss": 0.5071,
      "num_input_tokens_seen": 20222440,
      "step": 31005
    },
    {
      "epoch": 16.252620545073377,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0030963595315315765,
      "loss": 0.383,
      "num_input_tokens_seen": 20225224,
      "step": 31010
    },
    {
      "epoch": 16.25524109014675,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0030921863175543108,
      "loss": 0.3824,
      "num_input_tokens_seen": 20228680,
      "step": 31015
    },
    {
      "epoch": 16.257861635220127,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0030880155945558147,
      "loss": 0.4314,
      "num_input_tokens_seen": 20231176,
      "step": 31020
    },
    {
      "epoch": 16.2604821802935,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0030838473634085557,
      "loss": 0.3161,
      "num_input_tokens_seen": 20233672,
      "step": 31025
    },
    {
      "epoch": 16.263102725366878,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0030796816249844807,
      "loss": 0.2612,
      "num_input_tokens_seen": 20236424,
      "step": 31030
    },
    {
      "epoch": 16.26572327044025,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0030755183801550205,
      "loss": 0.3617,
      "num_input_tokens_seen": 20239336,
      "step": 31035
    },
    {
      "epoch": 16.268343815513628,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0030713576297910694,
      "loss": 0.3813,
      "num_input_tokens_seen": 20242056,
      "step": 31040
    },
    {
      "epoch": 16.270964360587,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0030671993747630175,
      "loss": 0.3098,
      "num_input_tokens_seen": 20245384,
      "step": 31045
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0030630436159407148,
      "loss": 0.3299,
      "num_input_tokens_seen": 20248936,
      "step": 31050
    },
    {
      "epoch": 16.27620545073375,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0030588903541935034,
      "loss": 0.3295,
      "num_input_tokens_seen": 20253416,
      "step": 31055
    },
    {
      "epoch": 16.27882599580713,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0030547395903902006,
      "loss": 0.394,
      "num_input_tokens_seen": 20256968,
      "step": 31060
    },
    {
      "epoch": 16.281446540880502,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0030505913253990984,
      "loss": 0.4993,
      "num_input_tokens_seen": 20259752,
      "step": 31065
    },
    {
      "epoch": 16.28406708595388,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0030464455600879656,
      "loss": 0.3856,
      "num_input_tokens_seen": 20262600,
      "step": 31070
    },
    {
      "epoch": 16.286687631027252,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.003042302295324046,
      "loss": 0.293,
      "num_input_tokens_seen": 20265896,
      "step": 31075
    },
    {
      "epoch": 16.28930817610063,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0030381615319740635,
      "loss": 0.4171,
      "num_input_tokens_seen": 20269224,
      "step": 31080
    },
    {
      "epoch": 16.291928721174003,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0030340232709042208,
      "loss": 0.2793,
      "num_input_tokens_seen": 20272072,
      "step": 31085
    },
    {
      "epoch": 16.29454926624738,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0030298875129802006,
      "loss": 0.3704,
      "num_input_tokens_seen": 20275400,
      "step": 31090
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.003025754259067144,
      "loss": 0.3513,
      "num_input_tokens_seen": 20279560,
      "step": 31095
    },
    {
      "epoch": 16.29979035639413,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0030216235100296923,
      "loss": 0.3804,
      "num_input_tokens_seen": 20282728,
      "step": 31100
    },
    {
      "epoch": 16.302410901467507,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0030174952667319415,
      "loss": 0.3872,
      "num_input_tokens_seen": 20285224,
      "step": 31105
    },
    {
      "epoch": 16.30503144654088,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.003013369530037477,
      "loss": 0.2981,
      "num_input_tokens_seen": 20288552,
      "step": 31110
    },
    {
      "epoch": 16.307651991614257,
      "grad_norm": 0.095703125,
      "learning_rate": 0.003009246300809361,
      "loss": 0.3736,
      "num_input_tokens_seen": 20292776,
      "step": 31115
    },
    {
      "epoch": 16.31027253668763,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.003005125579910116,
      "loss": 0.4909,
      "num_input_tokens_seen": 20295208,
      "step": 31120
    },
    {
      "epoch": 16.312893081761008,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0030010073682017564,
      "loss": 0.3422,
      "num_input_tokens_seen": 20297960,
      "step": 31125
    },
    {
      "epoch": 16.31551362683438,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0029968916665457654,
      "loss": 0.46,
      "num_input_tokens_seen": 20300968,
      "step": 31130
    },
    {
      "epoch": 16.318134171907758,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.002992778475803097,
      "loss": 0.3808,
      "num_input_tokens_seen": 20304648,
      "step": 31135
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.0029886677968341863,
      "loss": 0.4214,
      "num_input_tokens_seen": 20308488,
      "step": 31140
    },
    {
      "epoch": 16.32337526205451,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.002984559630498943,
      "loss": 0.3932,
      "num_input_tokens_seen": 20311560,
      "step": 31145
    },
    {
      "epoch": 16.32599580712788,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.00298045397765674,
      "loss": 0.3354,
      "num_input_tokens_seen": 20315592,
      "step": 31150
    },
    {
      "epoch": 16.32861635220126,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0029763508391664455,
      "loss": 0.41,
      "num_input_tokens_seen": 20319304,
      "step": 31155
    },
    {
      "epoch": 16.331236897274632,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.002972250215886376,
      "loss": 0.3097,
      "num_input_tokens_seen": 20324264,
      "step": 31160
    },
    {
      "epoch": 16.33385744234801,
      "grad_norm": 0.125,
      "learning_rate": 0.002968152108674342,
      "loss": 0.2656,
      "num_input_tokens_seen": 20326696,
      "step": 31165
    },
    {
      "epoch": 16.336477987421382,
      "grad_norm": 0.072265625,
      "learning_rate": 0.002964056518387625,
      "loss": 0.4393,
      "num_input_tokens_seen": 20330248,
      "step": 31170
    },
    {
      "epoch": 16.33909853249476,
      "grad_norm": 0.11328125,
      "learning_rate": 0.002959963445882967,
      "loss": 0.2686,
      "num_input_tokens_seen": 20332552,
      "step": 31175
    },
    {
      "epoch": 16.341719077568133,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.002955872892016597,
      "loss": 0.4158,
      "num_input_tokens_seen": 20335304,
      "step": 31180
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.002951784857644214,
      "loss": 0.4116,
      "num_input_tokens_seen": 20337992,
      "step": 31185
    },
    {
      "epoch": 16.346960167714883,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0029476993436209817,
      "loss": 0.3364,
      "num_input_tokens_seen": 20340680,
      "step": 31190
    },
    {
      "epoch": 16.34958071278826,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0029436163508015506,
      "loss": 0.3274,
      "num_input_tokens_seen": 20343848,
      "step": 31195
    },
    {
      "epoch": 16.352201257861637,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0029395358800400283,
      "loss": 0.2978,
      "num_input_tokens_seen": 20346792,
      "step": 31200
    },
    {
      "epoch": 16.35482180293501,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0029354579321900073,
      "loss": 0.36,
      "num_input_tokens_seen": 20349544,
      "step": 31205
    },
    {
      "epoch": 16.357442348008387,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0029313825081045473,
      "loss": 0.3955,
      "num_input_tokens_seen": 20352904,
      "step": 31210
    },
    {
      "epoch": 16.36006289308176,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.002927309608636186,
      "loss": 0.2772,
      "num_input_tokens_seen": 20356424,
      "step": 31215
    },
    {
      "epoch": 16.362683438155138,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.002923239234636915,
      "loss": 0.3082,
      "num_input_tokens_seen": 20360264,
      "step": 31220
    },
    {
      "epoch": 16.36530398322851,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0029191713869582236,
      "loss": 0.3737,
      "num_input_tokens_seen": 20363848,
      "step": 31225
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 0.09375,
      "learning_rate": 0.0029151060664510476,
      "loss": 0.3061,
      "num_input_tokens_seen": 20366280,
      "step": 31230
    },
    {
      "epoch": 16.37054507337526,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0029110432739658113,
      "loss": 0.2455,
      "num_input_tokens_seen": 20370568,
      "step": 31235
    },
    {
      "epoch": 16.37316561844864,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.002906983010352404,
      "loss": 0.3306,
      "num_input_tokens_seen": 20373928,
      "step": 31240
    },
    {
      "epoch": 16.37578616352201,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.002902925276460189,
      "loss": 0.3416,
      "num_input_tokens_seen": 20376488,
      "step": 31245
    },
    {
      "epoch": 16.37840670859539,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.002898870073137996,
      "loss": 0.3207,
      "num_input_tokens_seen": 20379176,
      "step": 31250
    },
    {
      "epoch": 16.381027253668762,
      "grad_norm": 0.138671875,
      "learning_rate": 0.002894817401234125,
      "loss": 0.4168,
      "num_input_tokens_seen": 20381736,
      "step": 31255
    },
    {
      "epoch": 16.38364779874214,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0028907672615963486,
      "loss": 0.2844,
      "num_input_tokens_seen": 20385064,
      "step": 31260
    },
    {
      "epoch": 16.386268343815512,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0028867196550719115,
      "loss": 0.3742,
      "num_input_tokens_seen": 20388072,
      "step": 31265
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.166015625,
      "learning_rate": 0.002882674582507531,
      "loss": 0.4196,
      "num_input_tokens_seen": 20391080,
      "step": 31270
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0028786320447493855,
      "loss": 0.37,
      "num_input_tokens_seen": 20394024,
      "step": 31275
    },
    {
      "epoch": 16.39412997903564,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.002874592042643125,
      "loss": 0.2953,
      "num_input_tokens_seen": 20397672,
      "step": 31280
    },
    {
      "epoch": 16.396750524109013,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0028705545770338755,
      "loss": 0.4378,
      "num_input_tokens_seen": 20401352,
      "step": 31285
    },
    {
      "epoch": 16.39937106918239,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0028665196487662296,
      "loss": 0.3644,
      "num_input_tokens_seen": 20404008,
      "step": 31290
    },
    {
      "epoch": 16.401991614255767,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0028624872586842454,
      "loss": 0.5247,
      "num_input_tokens_seen": 20407496,
      "step": 31295
    },
    {
      "epoch": 16.40461215932914,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0028584574076314594,
      "loss": 0.3526,
      "num_input_tokens_seen": 20410696,
      "step": 31300
    },
    {
      "epoch": 16.407232704402517,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0028544300964508645,
      "loss": 0.3979,
      "num_input_tokens_seen": 20413736,
      "step": 31305
    },
    {
      "epoch": 16.40985324947589,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0028504053259849264,
      "loss": 0.3069,
      "num_input_tokens_seen": 20417160,
      "step": 31310
    },
    {
      "epoch": 16.412473794549268,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0028463830970755838,
      "loss": 0.3556,
      "num_input_tokens_seen": 20420520,
      "step": 31315
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0028423634105642416,
      "loss": 0.3593,
      "num_input_tokens_seen": 20423272,
      "step": 31320
    },
    {
      "epoch": 16.417714884696018,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0028383462672917767,
      "loss": 0.3369,
      "num_input_tokens_seen": 20426216,
      "step": 31325
    },
    {
      "epoch": 16.42033542976939,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0028343316680985247,
      "loss": 0.3953,
      "num_input_tokens_seen": 20429512,
      "step": 31330
    },
    {
      "epoch": 16.42295597484277,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0028303196138242913,
      "loss": 0.3979,
      "num_input_tokens_seen": 20433064,
      "step": 31335
    },
    {
      "epoch": 16.42557651991614,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0028263101053083556,
      "loss": 0.3756,
      "num_input_tokens_seen": 20436232,
      "step": 31340
    },
    {
      "epoch": 16.42819706498952,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0028223031433894596,
      "loss": 0.3847,
      "num_input_tokens_seen": 20439688,
      "step": 31345
    },
    {
      "epoch": 16.430817610062892,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0028182987289058197,
      "loss": 0.5696,
      "num_input_tokens_seen": 20442888,
      "step": 31350
    },
    {
      "epoch": 16.43343815513627,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0028142968626951106,
      "loss": 0.4082,
      "num_input_tokens_seen": 20445928,
      "step": 31355
    },
    {
      "epoch": 16.436058700209642,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0028102975455944717,
      "loss": 0.2977,
      "num_input_tokens_seen": 20448552,
      "step": 31360
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.00280630077844052,
      "loss": 0.3275,
      "num_input_tokens_seen": 20451304,
      "step": 31365
    },
    {
      "epoch": 16.441299790356393,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.002802306562069329,
      "loss": 0.3968,
      "num_input_tokens_seen": 20454408,
      "step": 31370
    },
    {
      "epoch": 16.44392033542977,
      "grad_norm": 0.06640625,
      "learning_rate": 0.002798314897316449,
      "loss": 0.3206,
      "num_input_tokens_seen": 20457384,
      "step": 31375
    },
    {
      "epoch": 16.446540880503143,
      "grad_norm": 0.087890625,
      "learning_rate": 0.002794325785016892,
      "loss": 0.3266,
      "num_input_tokens_seen": 20460456,
      "step": 31380
    },
    {
      "epoch": 16.44916142557652,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.002790339226005132,
      "loss": 0.3229,
      "num_input_tokens_seen": 20463336,
      "step": 31385
    },
    {
      "epoch": 16.451781970649897,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0027863552211151053,
      "loss": 0.3701,
      "num_input_tokens_seen": 20465928,
      "step": 31390
    },
    {
      "epoch": 16.45440251572327,
      "grad_norm": 0.09375,
      "learning_rate": 0.0027823737711802263,
      "loss": 0.3385,
      "num_input_tokens_seen": 20468584,
      "step": 31395
    },
    {
      "epoch": 16.457023060796647,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0027783948770333687,
      "loss": 0.3671,
      "num_input_tokens_seen": 20471336,
      "step": 31400
    },
    {
      "epoch": 16.45964360587002,
      "grad_norm": 0.14453125,
      "learning_rate": 0.002774418539506875,
      "loss": 0.2994,
      "num_input_tokens_seen": 20474696,
      "step": 31405
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0027704447594325407,
      "loss": 0.3657,
      "num_input_tokens_seen": 20477832,
      "step": 31410
    },
    {
      "epoch": 16.46488469601677,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0027664735376416454,
      "loss": 0.445,
      "num_input_tokens_seen": 20481160,
      "step": 31415
    },
    {
      "epoch": 16.467505241090148,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0027625048749649155,
      "loss": 0.3949,
      "num_input_tokens_seen": 20484648,
      "step": 31420
    },
    {
      "epoch": 16.47012578616352,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0027585387722325523,
      "loss": 0.1951,
      "num_input_tokens_seen": 20486696,
      "step": 31425
    },
    {
      "epoch": 16.4727463312369,
      "grad_norm": 0.09375,
      "learning_rate": 0.0027545752302742246,
      "loss": 0.3373,
      "num_input_tokens_seen": 20489640,
      "step": 31430
    },
    {
      "epoch": 16.47536687631027,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0027506142499190505,
      "loss": 0.4235,
      "num_input_tokens_seen": 20492392,
      "step": 31435
    },
    {
      "epoch": 16.47798742138365,
      "grad_norm": 0.125,
      "learning_rate": 0.0027466558319956314,
      "loss": 0.299,
      "num_input_tokens_seen": 20494984,
      "step": 31440
    },
    {
      "epoch": 16.480607966457022,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0027426999773320147,
      "loss": 0.3454,
      "num_input_tokens_seen": 20497832,
      "step": 31445
    },
    {
      "epoch": 16.4832285115304,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.002738746686755724,
      "loss": 0.301,
      "num_input_tokens_seen": 20500488,
      "step": 31450
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0027347959610937445,
      "loss": 0.3808,
      "num_input_tokens_seen": 20503688,
      "step": 31455
    },
    {
      "epoch": 16.48846960167715,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0027308478011725226,
      "loss": 0.2901,
      "num_input_tokens_seen": 20507816,
      "step": 31460
    },
    {
      "epoch": 16.491090146750523,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0027269022078179636,
      "loss": 0.4783,
      "num_input_tokens_seen": 20510920,
      "step": 31465
    },
    {
      "epoch": 16.4937106918239,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0027229591818554487,
      "loss": 0.3483,
      "num_input_tokens_seen": 20514408,
      "step": 31470
    },
    {
      "epoch": 16.496331236897273,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0027190187241098026,
      "loss": 0.3147,
      "num_input_tokens_seen": 20518472,
      "step": 31475
    },
    {
      "epoch": 16.49895178197065,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0027150808354053334,
      "loss": 0.4454,
      "num_input_tokens_seen": 20521064,
      "step": 31480
    },
    {
      "epoch": 16.501572327044027,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0027111455165658013,
      "loss": 0.402,
      "num_input_tokens_seen": 20523784,
      "step": 31485
    },
    {
      "epoch": 16.5041928721174,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0027072127684144268,
      "loss": 0.395,
      "num_input_tokens_seen": 20527976,
      "step": 31490
    },
    {
      "epoch": 16.506813417190777,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.002703282591773897,
      "loss": 0.463,
      "num_input_tokens_seen": 20531784,
      "step": 31495
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 0.140625,
      "learning_rate": 0.0026993549874663627,
      "loss": 0.4166,
      "num_input_tokens_seen": 20535048,
      "step": 31500
    },
    {
      "epoch": 16.512054507337528,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0026954299563134304,
      "loss": 0.4513,
      "num_input_tokens_seen": 20537672,
      "step": 31505
    },
    {
      "epoch": 16.5146750524109,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.002691507499136175,
      "loss": 0.2905,
      "num_input_tokens_seen": 20540264,
      "step": 31510
    },
    {
      "epoch": 16.517295597484278,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0026875876167551265,
      "loss": 0.3316,
      "num_input_tokens_seen": 20543208,
      "step": 31515
    },
    {
      "epoch": 16.51991614255765,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0026836703099902795,
      "loss": 0.3251,
      "num_input_tokens_seen": 20546632,
      "step": 31520
    },
    {
      "epoch": 16.52253668763103,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.002679755579661093,
      "loss": 0.4501,
      "num_input_tokens_seen": 20549256,
      "step": 31525
    },
    {
      "epoch": 16.5251572327044,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.002675843426586484,
      "loss": 0.4177,
      "num_input_tokens_seen": 20551880,
      "step": 31530
    },
    {
      "epoch": 16.52777777777778,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0026719338515848266,
      "loss": 0.3595,
      "num_input_tokens_seen": 20555368,
      "step": 31535
    },
    {
      "epoch": 16.530398322851152,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0026680268554739655,
      "loss": 0.2503,
      "num_input_tokens_seen": 20557864,
      "step": 31540
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0026641224390711897,
      "loss": 0.3217,
      "num_input_tokens_seen": 20561800,
      "step": 31545
    },
    {
      "epoch": 16.535639412997902,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.002660220603193266,
      "loss": 0.2834,
      "num_input_tokens_seen": 20565128,
      "step": 31550
    },
    {
      "epoch": 16.53825995807128,
      "grad_norm": 0.09765625,
      "learning_rate": 0.002656321348656417,
      "loss": 0.261,
      "num_input_tokens_seen": 20568648,
      "step": 31555
    },
    {
      "epoch": 16.540880503144653,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0026524246762763123,
      "loss": 0.2849,
      "num_input_tokens_seen": 20571816,
      "step": 31560
    },
    {
      "epoch": 16.54350104821803,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.002648530586868103,
      "loss": 0.4615,
      "num_input_tokens_seen": 20576136,
      "step": 31565
    },
    {
      "epoch": 16.546121593291403,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0026446390812463767,
      "loss": 0.4029,
      "num_input_tokens_seen": 20578472,
      "step": 31570
    },
    {
      "epoch": 16.54874213836478,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.002640750160225197,
      "loss": 0.4526,
      "num_input_tokens_seen": 20581800,
      "step": 31575
    },
    {
      "epoch": 16.551362683438157,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.002636863824618082,
      "loss": 0.5317,
      "num_input_tokens_seen": 20584296,
      "step": 31580
    },
    {
      "epoch": 16.55398322851153,
      "grad_norm": 0.115234375,
      "learning_rate": 0.002632980075238011,
      "loss": 0.3402,
      "num_input_tokens_seen": 20587368,
      "step": 31585
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0026290989128974194,
      "loss": 0.3936,
      "num_input_tokens_seen": 20590504,
      "step": 31590
    },
    {
      "epoch": 16.55922431865828,
      "grad_norm": 0.07421875,
      "learning_rate": 0.002625220338408198,
      "loss": 0.3589,
      "num_input_tokens_seen": 20593448,
      "step": 31595
    },
    {
      "epoch": 16.561844863731658,
      "grad_norm": 0.2734375,
      "learning_rate": 0.002621344352581702,
      "loss": 0.3946,
      "num_input_tokens_seen": 20596904,
      "step": 31600
    },
    {
      "epoch": 16.56446540880503,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0026174709562287435,
      "loss": 0.3475,
      "num_input_tokens_seen": 20600040,
      "step": 31605
    },
    {
      "epoch": 16.567085953878408,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0026136001501595937,
      "loss": 0.3804,
      "num_input_tokens_seen": 20603112,
      "step": 31610
    },
    {
      "epoch": 16.56970649895178,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.002609731935183986,
      "loss": 0.3694,
      "num_input_tokens_seen": 20606760,
      "step": 31615
    },
    {
      "epoch": 16.572327044025158,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0026058663121111,
      "loss": 0.3523,
      "num_input_tokens_seen": 20609896,
      "step": 31620
    },
    {
      "epoch": 16.57494758909853,
      "grad_norm": 0.294921875,
      "learning_rate": 0.0026020032817495784,
      "loss": 0.3384,
      "num_input_tokens_seen": 20612712,
      "step": 31625
    },
    {
      "epoch": 16.57756813417191,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0025981428449075275,
      "loss": 0.6007,
      "num_input_tokens_seen": 20616488,
      "step": 31630
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.002594285002392505,
      "loss": 0.4875,
      "num_input_tokens_seen": 20619944,
      "step": 31635
    },
    {
      "epoch": 16.58280922431866,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0025904297550115316,
      "loss": 0.3657,
      "num_input_tokens_seen": 20622952,
      "step": 31640
    },
    {
      "epoch": 16.585429769392032,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0025865771035710776,
      "loss": 0.2613,
      "num_input_tokens_seen": 20626248,
      "step": 31645
    },
    {
      "epoch": 16.58805031446541,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.00258272704887707,
      "loss": 0.2711,
      "num_input_tokens_seen": 20630088,
      "step": 31650
    },
    {
      "epoch": 16.590670859538783,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0025788795917349003,
      "loss": 0.3534,
      "num_input_tokens_seen": 20632840,
      "step": 31655
    },
    {
      "epoch": 16.59329140461216,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.002575034732949413,
      "loss": 0.3557,
      "num_input_tokens_seen": 20635720,
      "step": 31660
    },
    {
      "epoch": 16.595911949685533,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0025711924733249113,
      "loss": 0.2998,
      "num_input_tokens_seen": 20639400,
      "step": 31665
    },
    {
      "epoch": 16.59853249475891,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0025673528136651485,
      "loss": 0.3655,
      "num_input_tokens_seen": 20643656,
      "step": 31670
    },
    {
      "epoch": 16.601153039832283,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0025635157547733334,
      "loss": 0.3131,
      "num_input_tokens_seen": 20646504,
      "step": 31675
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0025596812974521405,
      "loss": 0.4374,
      "num_input_tokens_seen": 20650120,
      "step": 31680
    },
    {
      "epoch": 16.606394129979037,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.002555849442503695,
      "loss": 0.3332,
      "num_input_tokens_seen": 20653064,
      "step": 31685
    },
    {
      "epoch": 16.60901467505241,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0025520201907295735,
      "loss": 0.3768,
      "num_input_tokens_seen": 20657160,
      "step": 31690
    },
    {
      "epoch": 16.611635220125788,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0025481935429308177,
      "loss": 0.2857,
      "num_input_tokens_seen": 20659912,
      "step": 31695
    },
    {
      "epoch": 16.61425576519916,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.002544369499907917,
      "loss": 0.2737,
      "num_input_tokens_seen": 20662312,
      "step": 31700
    },
    {
      "epoch": 16.616876310272538,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0025405480624608096,
      "loss": 0.3671,
      "num_input_tokens_seen": 20666248,
      "step": 31705
    },
    {
      "epoch": 16.61949685534591,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0025367292313889043,
      "loss": 0.3648,
      "num_input_tokens_seen": 20670888,
      "step": 31710
    },
    {
      "epoch": 16.622117400419288,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.002532913007491054,
      "loss": 0.3138,
      "num_input_tokens_seen": 20673320,
      "step": 31715
    },
    {
      "epoch": 16.62473794549266,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0025290993915655755,
      "loss": 0.2787,
      "num_input_tokens_seen": 20676552,
      "step": 31720
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0025252883844102286,
      "loss": 0.314,
      "num_input_tokens_seen": 20678792,
      "step": 31725
    },
    {
      "epoch": 16.629979035639412,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.002521479986822227,
      "loss": 0.2937,
      "num_input_tokens_seen": 20683240,
      "step": 31730
    },
    {
      "epoch": 16.63259958071279,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.002517674199598252,
      "loss": 0.4433,
      "num_input_tokens_seen": 20686536,
      "step": 31735
    },
    {
      "epoch": 16.635220125786162,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.002513871023534427,
      "loss": 0.3918,
      "num_input_tokens_seen": 20689832,
      "step": 31740
    },
    {
      "epoch": 16.63784067085954,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.00251007045942634,
      "loss": 0.4012,
      "num_input_tokens_seen": 20692648,
      "step": 31745
    },
    {
      "epoch": 16.640461215932913,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0025062725080690156,
      "loss": 0.387,
      "num_input_tokens_seen": 20695048,
      "step": 31750
    },
    {
      "epoch": 16.64308176100629,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00250247717025695,
      "loss": 0.3306,
      "num_input_tokens_seen": 20698216,
      "step": 31755
    },
    {
      "epoch": 16.645702306079663,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0024986844467840773,
      "loss": 0.4221,
      "num_input_tokens_seen": 20702024,
      "step": 31760
    },
    {
      "epoch": 16.64832285115304,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0024948943384437952,
      "loss": 0.2548,
      "num_input_tokens_seen": 20705032,
      "step": 31765
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.002491106846028953,
      "loss": 0.2972,
      "num_input_tokens_seen": 20708072,
      "step": 31770
    },
    {
      "epoch": 16.65356394129979,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0024873219703318533,
      "loss": 0.343,
      "num_input_tokens_seen": 20711432,
      "step": 31775
    },
    {
      "epoch": 16.656184486373167,
      "grad_norm": 0.1875,
      "learning_rate": 0.002483539712144241,
      "loss": 0.4258,
      "num_input_tokens_seen": 20715112,
      "step": 31780
    },
    {
      "epoch": 16.65880503144654,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0024797600722573283,
      "loss": 0.3378,
      "num_input_tokens_seen": 20718024,
      "step": 31785
    },
    {
      "epoch": 16.661425576519918,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0024759830514617673,
      "loss": 0.3317,
      "num_input_tokens_seen": 20721192,
      "step": 31790
    },
    {
      "epoch": 16.66404612159329,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.0024722086505476705,
      "loss": 0.2876,
      "num_input_tokens_seen": 20723656,
      "step": 31795
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.083984375,
      "learning_rate": 0.002468436870304603,
      "loss": 0.256,
      "num_input_tokens_seen": 20726088,
      "step": 31800
    },
    {
      "epoch": 16.66928721174004,
      "grad_norm": 0.0703125,
      "learning_rate": 0.002464667711521572,
      "loss": 0.2448,
      "num_input_tokens_seen": 20728840,
      "step": 31805
    },
    {
      "epoch": 16.671907756813418,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.002460901174987047,
      "loss": 0.228,
      "num_input_tokens_seen": 20732424,
      "step": 31810
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0024571372614889393,
      "loss": 0.2695,
      "num_input_tokens_seen": 20738184,
      "step": 31815
    },
    {
      "epoch": 16.67714884696017,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.002453375971814621,
      "loss": 0.2806,
      "num_input_tokens_seen": 20743464,
      "step": 31820
    },
    {
      "epoch": 16.679769392033542,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.002449617306750913,
      "loss": 0.3387,
      "num_input_tokens_seen": 20746536,
      "step": 31825
    },
    {
      "epoch": 16.68238993710692,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0024458612670840785,
      "loss": 0.4239,
      "num_input_tokens_seen": 20749896,
      "step": 31830
    },
    {
      "epoch": 16.685010482180292,
      "grad_norm": 0.083984375,
      "learning_rate": 0.002442107853599842,
      "loss": 0.3375,
      "num_input_tokens_seen": 20753448,
      "step": 31835
    },
    {
      "epoch": 16.68763102725367,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.002438357067083377,
      "loss": 0.4033,
      "num_input_tokens_seen": 20757032,
      "step": 31840
    },
    {
      "epoch": 16.690251572327043,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0024346089083192995,
      "loss": 0.2831,
      "num_input_tokens_seen": 20760136,
      "step": 31845
    },
    {
      "epoch": 16.69287211740042,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0024308633780916853,
      "loss": 0.3887,
      "num_input_tokens_seen": 20763144,
      "step": 31850
    },
    {
      "epoch": 16.695492662473793,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0024271204771840597,
      "loss": 0.4041,
      "num_input_tokens_seen": 20766248,
      "step": 31855
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.002423380206379388,
      "loss": 0.3913,
      "num_input_tokens_seen": 20769160,
      "step": 31860
    },
    {
      "epoch": 16.700733752620543,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0024196425664600955,
      "loss": 0.3253,
      "num_input_tokens_seen": 20773384,
      "step": 31865
    },
    {
      "epoch": 16.70335429769392,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0024159075582080576,
      "loss": 0.4238,
      "num_input_tokens_seen": 20777416,
      "step": 31870
    },
    {
      "epoch": 16.705974842767297,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0024121751824045866,
      "loss": 0.4023,
      "num_input_tokens_seen": 20780808,
      "step": 31875
    },
    {
      "epoch": 16.70859538784067,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.002408445439830462,
      "loss": 0.2444,
      "num_input_tokens_seen": 20784040,
      "step": 31880
    },
    {
      "epoch": 16.711215932914047,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0024047183312658964,
      "loss": 0.3472,
      "num_input_tokens_seen": 20787144,
      "step": 31885
    },
    {
      "epoch": 16.71383647798742,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.002400993857490561,
      "loss": 0.3359,
      "num_input_tokens_seen": 20790408,
      "step": 31890
    },
    {
      "epoch": 16.716457023060798,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0023972720192835736,
      "loss": 0.4058,
      "num_input_tokens_seen": 20793768,
      "step": 31895
    },
    {
      "epoch": 16.71907756813417,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.002393552817423504,
      "loss": 0.3284,
      "num_input_tokens_seen": 20799624,
      "step": 31900
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.002389836252688363,
      "loss": 0.3674,
      "num_input_tokens_seen": 20802152,
      "step": 31905
    },
    {
      "epoch": 16.72431865828092,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0023861223258556104,
      "loss": 0.3678,
      "num_input_tokens_seen": 20804776,
      "step": 31910
    },
    {
      "epoch": 16.7269392033543,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0023824110377021587,
      "loss": 0.3877,
      "num_input_tokens_seen": 20807880,
      "step": 31915
    },
    {
      "epoch": 16.729559748427672,
      "grad_norm": 0.10546875,
      "learning_rate": 0.00237870238900437,
      "loss": 0.3527,
      "num_input_tokens_seen": 20810568,
      "step": 31920
    },
    {
      "epoch": 16.73218029350105,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0023749963805380545,
      "loss": 0.3667,
      "num_input_tokens_seen": 20815368,
      "step": 31925
    },
    {
      "epoch": 16.734800838574422,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0023712930130784576,
      "loss": 0.4514,
      "num_input_tokens_seen": 20818472,
      "step": 31930
    },
    {
      "epoch": 16.7374213836478,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0023675922874002906,
      "loss": 0.342,
      "num_input_tokens_seen": 20822120,
      "step": 31935
    },
    {
      "epoch": 16.740041928721173,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.002363894204277697,
      "loss": 0.3583,
      "num_input_tokens_seen": 20825160,
      "step": 31940
    },
    {
      "epoch": 16.74266247379455,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.002360198764484273,
      "loss": 0.556,
      "num_input_tokens_seen": 20830056,
      "step": 31945
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0023565059687930667,
      "loss": 0.4324,
      "num_input_tokens_seen": 20833032,
      "step": 31950
    },
    {
      "epoch": 16.7479035639413,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0023528158179765696,
      "loss": 0.3554,
      "num_input_tokens_seen": 20836200,
      "step": 31955
    },
    {
      "epoch": 16.750524109014677,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0023491283128067174,
      "loss": 0.359,
      "num_input_tokens_seen": 20839784,
      "step": 31960
    },
    {
      "epoch": 16.75314465408805,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.00234544345405489,
      "loss": 0.3126,
      "num_input_tokens_seen": 20842952,
      "step": 31965
    },
    {
      "epoch": 16.755765199161427,
      "grad_norm": 0.0859375,
      "learning_rate": 0.002341761242491921,
      "loss": 0.3696,
      "num_input_tokens_seen": 20846344,
      "step": 31970
    },
    {
      "epoch": 16.7583857442348,
      "grad_norm": 0.173828125,
      "learning_rate": 0.002338081678888085,
      "loss": 0.3467,
      "num_input_tokens_seen": 20849512,
      "step": 31975
    },
    {
      "epoch": 16.761006289308177,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0023344047640131145,
      "loss": 0.467,
      "num_input_tokens_seen": 20852744,
      "step": 31980
    },
    {
      "epoch": 16.76362683438155,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0023307304986361676,
      "loss": 0.4232,
      "num_input_tokens_seen": 20856392,
      "step": 31985
    },
    {
      "epoch": 16.766247379454928,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0023270588835258587,
      "loss": 0.3637,
      "num_input_tokens_seen": 20859048,
      "step": 31990
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0023233899194502515,
      "loss": 0.3359,
      "num_input_tokens_seen": 20861992,
      "step": 31995
    },
    {
      "epoch": 16.771488469601678,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0023197236071768506,
      "loss": 0.3511,
      "num_input_tokens_seen": 20864904,
      "step": 32000
    },
    {
      "epoch": 16.77410901467505,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0023160599474726074,
      "loss": 0.3573,
      "num_input_tokens_seen": 20867976,
      "step": 32005
    },
    {
      "epoch": 16.77672955974843,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0023123989411039203,
      "loss": 0.366,
      "num_input_tokens_seen": 20871464,
      "step": 32010
    },
    {
      "epoch": 16.779350104821802,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0023087405888366267,
      "loss": 0.4038,
      "num_input_tokens_seen": 20874984,
      "step": 32015
    },
    {
      "epoch": 16.78197064989518,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.002305084891436008,
      "loss": 0.2585,
      "num_input_tokens_seen": 20877832,
      "step": 32020
    },
    {
      "epoch": 16.784591194968552,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0023014318496667996,
      "loss": 0.4791,
      "num_input_tokens_seen": 20881096,
      "step": 32025
    },
    {
      "epoch": 16.78721174004193,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.002297781464293176,
      "loss": 0.3764,
      "num_input_tokens_seen": 20884680,
      "step": 32030
    },
    {
      "epoch": 16.789832285115303,
      "grad_norm": 0.115234375,
      "learning_rate": 0.002294133736078758,
      "loss": 0.4845,
      "num_input_tokens_seen": 20888008,
      "step": 32035
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.002290488665786606,
      "loss": 0.3323,
      "num_input_tokens_seen": 20891240,
      "step": 32040
    },
    {
      "epoch": 16.795073375262053,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.002286846254179225,
      "loss": 0.2908,
      "num_input_tokens_seen": 20894664,
      "step": 32045
    },
    {
      "epoch": 16.79769392033543,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.002283206502018568,
      "loss": 0.4084,
      "num_input_tokens_seen": 20897608,
      "step": 32050
    },
    {
      "epoch": 16.800314465408803,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.00227956941006603,
      "loss": 0.347,
      "num_input_tokens_seen": 20901096,
      "step": 32055
    },
    {
      "epoch": 16.80293501048218,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.002275934979082452,
      "loss": 0.3257,
      "num_input_tokens_seen": 20904520,
      "step": 32060
    },
    {
      "epoch": 16.805555555555557,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0022723032098281096,
      "loss": 0.3273,
      "num_input_tokens_seen": 20907592,
      "step": 32065
    },
    {
      "epoch": 16.80817610062893,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0022686741030627333,
      "loss": 0.3625,
      "num_input_tokens_seen": 20911752,
      "step": 32070
    },
    {
      "epoch": 16.810796645702307,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0022650476595454856,
      "loss": 0.3672,
      "num_input_tokens_seen": 20914728,
      "step": 32075
    },
    {
      "epoch": 16.81341719077568,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.002261423880034978,
      "loss": 0.3463,
      "num_input_tokens_seen": 20917704,
      "step": 32080
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0022578027652892657,
      "loss": 0.3581,
      "num_input_tokens_seen": 20920808,
      "step": 32085
    },
    {
      "epoch": 16.81865828092243,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.002254184316065848,
      "loss": 0.3361,
      "num_input_tokens_seen": 20923528,
      "step": 32090
    },
    {
      "epoch": 16.821278825995808,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0022505685331216593,
      "loss": 0.2961,
      "num_input_tokens_seen": 20926376,
      "step": 32095
    },
    {
      "epoch": 16.82389937106918,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0022469554172130785,
      "loss": 0.3294,
      "num_input_tokens_seen": 20929672,
      "step": 32100
    },
    {
      "epoch": 16.82651991614256,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0022433449690959285,
      "loss": 0.3927,
      "num_input_tokens_seen": 20932616,
      "step": 32105
    },
    {
      "epoch": 16.829140461215932,
      "grad_norm": 0.265625,
      "learning_rate": 0.002239737189525475,
      "loss": 0.2817,
      "num_input_tokens_seen": 20935400,
      "step": 32110
    },
    {
      "epoch": 16.83176100628931,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0022361320792564287,
      "loss": 0.2932,
      "num_input_tokens_seen": 20938824,
      "step": 32115
    },
    {
      "epoch": 16.834381551362682,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.002232529639042929,
      "loss": 0.5183,
      "num_input_tokens_seen": 20941960,
      "step": 32120
    },
    {
      "epoch": 16.83700209643606,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0022289298696385744,
      "loss": 0.3559,
      "num_input_tokens_seen": 20948712,
      "step": 32125
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 0.10546875,
      "learning_rate": 0.002225332771796387,
      "loss": 0.3937,
      "num_input_tokens_seen": 20951592,
      "step": 32130
    },
    {
      "epoch": 16.84224318658281,
      "grad_norm": 0.078125,
      "learning_rate": 0.0022217383462688435,
      "loss": 0.4468,
      "num_input_tokens_seen": 20955880,
      "step": 32135
    },
    {
      "epoch": 16.844863731656183,
      "grad_norm": 0.07421875,
      "learning_rate": 0.002218146593807858,
      "loss": 0.3475,
      "num_input_tokens_seen": 20959112,
      "step": 32140
    },
    {
      "epoch": 16.84748427672956,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00221455751516478,
      "loss": 0.2604,
      "num_input_tokens_seen": 20961992,
      "step": 32145
    },
    {
      "epoch": 16.850104821802937,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.002210971111090405,
      "loss": 0.2777,
      "num_input_tokens_seen": 20964776,
      "step": 32150
    },
    {
      "epoch": 16.85272536687631,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0022073873823349713,
      "loss": 0.3632,
      "num_input_tokens_seen": 20967944,
      "step": 32155
    },
    {
      "epoch": 16.855345911949687,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0022038063296481495,
      "loss": 0.3815,
      "num_input_tokens_seen": 20971976,
      "step": 32160
    },
    {
      "epoch": 16.85796645702306,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.002200227953779055,
      "loss": 0.3387,
      "num_input_tokens_seen": 20975208,
      "step": 32165
    },
    {
      "epoch": 16.860587002096437,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.002196652255476251,
      "loss": 0.3336,
      "num_input_tokens_seen": 20977736,
      "step": 32170
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.002193079235487721,
      "loss": 0.3612,
      "num_input_tokens_seen": 20981704,
      "step": 32175
    },
    {
      "epoch": 16.865828092243188,
      "grad_norm": 0.076171875,
      "learning_rate": 0.002189508894560911,
      "loss": 0.4332,
      "num_input_tokens_seen": 20985992,
      "step": 32180
    },
    {
      "epoch": 16.86844863731656,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.002185941233442685,
      "loss": 0.4262,
      "num_input_tokens_seen": 20989544,
      "step": 32185
    },
    {
      "epoch": 16.871069182389938,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0021823762528793656,
      "loss": 0.4141,
      "num_input_tokens_seen": 20996392,
      "step": 32190
    },
    {
      "epoch": 16.87368972746331,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0021788139536167038,
      "loss": 0.3657,
      "num_input_tokens_seen": 20999400,
      "step": 32195
    },
    {
      "epoch": 16.87631027253669,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0021752543363998877,
      "loss": 0.3026,
      "num_input_tokens_seen": 21001704,
      "step": 32200
    },
    {
      "epoch": 16.878930817610062,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0021716974019735527,
      "loss": 0.4207,
      "num_input_tokens_seen": 21004872,
      "step": 32205
    },
    {
      "epoch": 16.88155136268344,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.002168143151081771,
      "loss": 0.4478,
      "num_input_tokens_seen": 21011656,
      "step": 32210
    },
    {
      "epoch": 16.884171907756812,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0021645915844680456,
      "loss": 0.3426,
      "num_input_tokens_seen": 21014472,
      "step": 32215
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.002161042702875331,
      "loss": 0.3095,
      "num_input_tokens_seen": 21017960,
      "step": 32220
    },
    {
      "epoch": 16.889412997903563,
      "grad_norm": 0.0703125,
      "learning_rate": 0.002157496507046004,
      "loss": 0.3708,
      "num_input_tokens_seen": 21021288,
      "step": 32225
    },
    {
      "epoch": 16.89203354297694,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0021539529977218927,
      "loss": 0.3801,
      "num_input_tokens_seen": 21024872,
      "step": 32230
    },
    {
      "epoch": 16.894654088050313,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0021504121756442582,
      "loss": 0.3363,
      "num_input_tokens_seen": 21028008,
      "step": 32235
    },
    {
      "epoch": 16.89727463312369,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.002146874041553805,
      "loss": 0.2691,
      "num_input_tokens_seen": 21030920,
      "step": 32240
    },
    {
      "epoch": 16.899895178197063,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0021433385961906627,
      "loss": 0.3942,
      "num_input_tokens_seen": 21034248,
      "step": 32245
    },
    {
      "epoch": 16.90251572327044,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.002139805840294414,
      "loss": 0.3487,
      "num_input_tokens_seen": 21037672,
      "step": 32250
    },
    {
      "epoch": 16.905136268343817,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0021362757746040623,
      "loss": 0.3999,
      "num_input_tokens_seen": 21041192,
      "step": 32255
    },
    {
      "epoch": 16.90775681341719,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.002132748399858061,
      "loss": 0.3121,
      "num_input_tokens_seen": 21044488,
      "step": 32260
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0021292237167942966,
      "loss": 0.4008,
      "num_input_tokens_seen": 21047752,
      "step": 32265
    },
    {
      "epoch": 16.91299790356394,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0021257017261500965,
      "loss": 0.3193,
      "num_input_tokens_seen": 21050344,
      "step": 32270
    },
    {
      "epoch": 16.915618448637318,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0021221824286622192,
      "loss": 0.4031,
      "num_input_tokens_seen": 21053288,
      "step": 32275
    },
    {
      "epoch": 16.91823899371069,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0021186658250668543,
      "loss": 0.3563,
      "num_input_tokens_seen": 21056904,
      "step": 32280
    },
    {
      "epoch": 16.920859538784068,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.002115151916099642,
      "loss": 0.2937,
      "num_input_tokens_seen": 21059848,
      "step": 32285
    },
    {
      "epoch": 16.92348008385744,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0021116407024956516,
      "loss": 0.292,
      "num_input_tokens_seen": 21062472,
      "step": 32290
    },
    {
      "epoch": 16.92610062893082,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.00210813218498939,
      "loss": 0.3503,
      "num_input_tokens_seen": 21067752,
      "step": 32295
    },
    {
      "epoch": 16.928721174004192,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0021046263643147984,
      "loss": 0.3403,
      "num_input_tokens_seen": 21070632,
      "step": 32300
    },
    {
      "epoch": 16.93134171907757,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0021011232412052475,
      "loss": 0.412,
      "num_input_tokens_seen": 21074568,
      "step": 32305
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.002097622816393559,
      "loss": 0.3321,
      "num_input_tokens_seen": 21077288,
      "step": 32310
    },
    {
      "epoch": 16.93658280922432,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.002094125090611978,
      "loss": 0.3166,
      "num_input_tokens_seen": 21079912,
      "step": 32315
    },
    {
      "epoch": 16.939203354297693,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0020906300645921892,
      "loss": 0.4633,
      "num_input_tokens_seen": 21084040,
      "step": 32320
    },
    {
      "epoch": 16.94182389937107,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0020871377390653183,
      "loss": 0.3908,
      "num_input_tokens_seen": 21087528,
      "step": 32325
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.002083648114761914,
      "loss": 0.3063,
      "num_input_tokens_seen": 21090888,
      "step": 32330
    },
    {
      "epoch": 16.94706498951782,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0020801611924119623,
      "loss": 0.4096,
      "num_input_tokens_seen": 21094792,
      "step": 32335
    },
    {
      "epoch": 16.949685534591197,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.002076676972744893,
      "loss": 0.3139,
      "num_input_tokens_seen": 21097736,
      "step": 32340
    },
    {
      "epoch": 16.95230607966457,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.002073195456489563,
      "loss": 0.2713,
      "num_input_tokens_seen": 21100776,
      "step": 32345
    },
    {
      "epoch": 16.954926624737947,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00206971664437427,
      "loss": 0.3478,
      "num_input_tokens_seen": 21104040,
      "step": 32350
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.002066240537126739,
      "loss": 0.2798,
      "num_input_tokens_seen": 21106856,
      "step": 32355
    },
    {
      "epoch": 16.960167714884697,
      "grad_norm": 0.146484375,
      "learning_rate": 0.002062767135474129,
      "loss": 0.3613,
      "num_input_tokens_seen": 21109768,
      "step": 32360
    },
    {
      "epoch": 16.96278825995807,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0020592964401430374,
      "loss": 0.3404,
      "num_input_tokens_seen": 21113512,
      "step": 32365
    },
    {
      "epoch": 16.965408805031448,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0020558284518594956,
      "loss": 0.3671,
      "num_input_tokens_seen": 21117544,
      "step": 32370
    },
    {
      "epoch": 16.96802935010482,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.002052363171348969,
      "loss": 0.3505,
      "num_input_tokens_seen": 21121864,
      "step": 32375
    },
    {
      "epoch": 16.970649895178198,
      "grad_norm": 0.0703125,
      "learning_rate": 0.002048900599336354,
      "loss": 0.3714,
      "num_input_tokens_seen": 21124328,
      "step": 32380
    },
    {
      "epoch": 16.97327044025157,
      "grad_norm": 0.10546875,
      "learning_rate": 0.002045440736545975,
      "loss": 0.2842,
      "num_input_tokens_seen": 21127336,
      "step": 32385
    },
    {
      "epoch": 16.97589098532495,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0020419835837016008,
      "loss": 0.2749,
      "num_input_tokens_seen": 21129896,
      "step": 32390
    },
    {
      "epoch": 16.978511530398322,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.002038529141526429,
      "loss": 0.4053,
      "num_input_tokens_seen": 21133512,
      "step": 32395
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 0.07421875,
      "learning_rate": 0.002035077410743088,
      "loss": 0.3306,
      "num_input_tokens_seen": 21135816,
      "step": 32400
    },
    {
      "epoch": 16.983752620545072,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0020316283920736437,
      "loss": 0.2326,
      "num_input_tokens_seen": 21138536,
      "step": 32405
    },
    {
      "epoch": 16.98637316561845,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.002028182086239589,
      "loss": 0.3348,
      "num_input_tokens_seen": 21141160,
      "step": 32410
    },
    {
      "epoch": 16.988993710691823,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0020247384939618496,
      "loss": 0.3219,
      "num_input_tokens_seen": 21143624,
      "step": 32415
    },
    {
      "epoch": 16.9916142557652,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0020212976159607863,
      "loss": 0.4245,
      "num_input_tokens_seen": 21147528,
      "step": 32420
    },
    {
      "epoch": 16.994234800838573,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.002017859452956193,
      "loss": 0.3618,
      "num_input_tokens_seen": 21150376,
      "step": 32425
    },
    {
      "epoch": 16.99685534591195,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0020144240056672977,
      "loss": 0.4245,
      "num_input_tokens_seen": 21155016,
      "step": 32430
    },
    {
      "epoch": 16.999475890985323,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0020109912748127504,
      "loss": 0.1985,
      "num_input_tokens_seen": 21157992,
      "step": 32435
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.5011940598487854,
      "eval_runtime": 13.7623,
      "eval_samples_per_second": 61.618,
      "eval_steps_per_second": 15.404,
      "num_input_tokens_seen": 21158120,
      "step": 32436
    },
    {
      "epoch": 17.0020964360587,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0020075612611106445,
      "loss": 0.2534,
      "num_input_tokens_seen": 21160712,
      "step": 32440
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0020041339652784942,
      "loss": 0.2888,
      "num_input_tokens_seen": 21163528,
      "step": 32445
    },
    {
      "epoch": 17.00733752620545,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0020007093880332534,
      "loss": 0.4204,
      "num_input_tokens_seen": 21167304,
      "step": 32450
    },
    {
      "epoch": 17.009958071278827,
      "grad_norm": 0.06640625,
      "learning_rate": 0.001997287530091308,
      "loss": 0.398,
      "num_input_tokens_seen": 21170792,
      "step": 32455
    },
    {
      "epoch": 17.0125786163522,
      "grad_norm": 0.08984375,
      "learning_rate": 0.001993868392168466,
      "loss": 0.3363,
      "num_input_tokens_seen": 21174408,
      "step": 32460
    },
    {
      "epoch": 17.015199161425578,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0019904519749799764,
      "loss": 0.4082,
      "num_input_tokens_seen": 21178248,
      "step": 32465
    },
    {
      "epoch": 17.01781970649895,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.00198703827924051,
      "loss": 0.235,
      "num_input_tokens_seen": 21181064,
      "step": 32470
    },
    {
      "epoch": 17.020440251572328,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0019836273056641767,
      "loss": 0.3987,
      "num_input_tokens_seen": 21184296,
      "step": 32475
    },
    {
      "epoch": 17.0230607966457,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0019802190549645126,
      "loss": 0.3609,
      "num_input_tokens_seen": 21187176,
      "step": 32480
    },
    {
      "epoch": 17.02568134171908,
      "grad_norm": 0.162109375,
      "learning_rate": 0.001976813527854487,
      "loss": 0.2959,
      "num_input_tokens_seen": 21189800,
      "step": 32485
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0019734107250464926,
      "loss": 0.4361,
      "num_input_tokens_seen": 21196328,
      "step": 32490
    },
    {
      "epoch": 17.03092243186583,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0019700106472523627,
      "loss": 0.3843,
      "num_input_tokens_seen": 21199400,
      "step": 32495
    },
    {
      "epoch": 17.033542976939202,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0019666132951833464,
      "loss": 0.2976,
      "num_input_tokens_seen": 21202248,
      "step": 32500
    },
    {
      "epoch": 17.03616352201258,
      "grad_norm": 0.14453125,
      "learning_rate": 0.001963218669550137,
      "loss": 0.4429,
      "num_input_tokens_seen": 21205448,
      "step": 32505
    },
    {
      "epoch": 17.038784067085953,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.001959826771062854,
      "loss": 0.2986,
      "num_input_tokens_seen": 21209576,
      "step": 32510
    },
    {
      "epoch": 17.04140461215933,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0019564376004310375,
      "loss": 0.413,
      "num_input_tokens_seen": 21212904,
      "step": 32515
    },
    {
      "epoch": 17.044025157232703,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0019530511583636667,
      "loss": 0.3284,
      "num_input_tokens_seen": 21216104,
      "step": 32520
    },
    {
      "epoch": 17.04664570230608,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.001949667445569147,
      "loss": 0.4254,
      "num_input_tokens_seen": 21220104,
      "step": 32525
    },
    {
      "epoch": 17.049266247379453,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0019462864627553106,
      "loss": 0.3836,
      "num_input_tokens_seen": 21222920,
      "step": 32530
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0019429082106294231,
      "loss": 0.3382,
      "num_input_tokens_seen": 21226408,
      "step": 32535
    },
    {
      "epoch": 17.054507337526207,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.001939532689898173,
      "loss": 0.4326,
      "num_input_tokens_seen": 21230184,
      "step": 32540
    },
    {
      "epoch": 17.05712788259958,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0019361599012676816,
      "loss": 0.3308,
      "num_input_tokens_seen": 21233288,
      "step": 32545
    },
    {
      "epoch": 17.059748427672957,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0019327898454435033,
      "loss": 0.3218,
      "num_input_tokens_seen": 21235720,
      "step": 32550
    },
    {
      "epoch": 17.06236897274633,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0019294225231306055,
      "loss": 0.3175,
      "num_input_tokens_seen": 21238408,
      "step": 32555
    },
    {
      "epoch": 17.064989517819708,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0019260579350334,
      "loss": 0.3418,
      "num_input_tokens_seen": 21242632,
      "step": 32560
    },
    {
      "epoch": 17.06761006289308,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.001922696081855722,
      "loss": 0.3343,
      "num_input_tokens_seen": 21246056,
      "step": 32565
    },
    {
      "epoch": 17.070230607966458,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0019193369643008256,
      "loss": 0.4302,
      "num_input_tokens_seen": 21249032,
      "step": 32570
    },
    {
      "epoch": 17.07285115303983,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0019159805830714055,
      "loss": 0.3244,
      "num_input_tokens_seen": 21251816,
      "step": 32575
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.00191262693886958,
      "loss": 0.3173,
      "num_input_tokens_seen": 21256200,
      "step": 32580
    },
    {
      "epoch": 17.078092243186582,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0019092760323968876,
      "loss": 0.3273,
      "num_input_tokens_seen": 21260392,
      "step": 32585
    },
    {
      "epoch": 17.08071278825996,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0019059278643543047,
      "loss": 0.3774,
      "num_input_tokens_seen": 21264296,
      "step": 32590
    },
    {
      "epoch": 17.083333333333332,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0019025824354422243,
      "loss": 0.4133,
      "num_input_tokens_seen": 21267848,
      "step": 32595
    },
    {
      "epoch": 17.08595387840671,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0018992397463604765,
      "loss": 0.3157,
      "num_input_tokens_seen": 21271560,
      "step": 32600
    },
    {
      "epoch": 17.088574423480082,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0018958997978083102,
      "loss": 0.3872,
      "num_input_tokens_seen": 21274760,
      "step": 32605
    },
    {
      "epoch": 17.09119496855346,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0018925625904844118,
      "loss": 0.3059,
      "num_input_tokens_seen": 21277352,
      "step": 32610
    },
    {
      "epoch": 17.093815513626833,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0018892281250868807,
      "loss": 0.4617,
      "num_input_tokens_seen": 21280168,
      "step": 32615
    },
    {
      "epoch": 17.09643605870021,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0018858964023132478,
      "loss": 0.3896,
      "num_input_tokens_seen": 21283688,
      "step": 32620
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0018825674228604738,
      "loss": 0.3636,
      "num_input_tokens_seen": 21286920,
      "step": 32625
    },
    {
      "epoch": 17.10167714884696,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.001879241187424943,
      "loss": 0.3713,
      "num_input_tokens_seen": 21290600,
      "step": 32630
    },
    {
      "epoch": 17.104297693920337,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0018759176967024676,
      "loss": 0.3655,
      "num_input_tokens_seen": 21293032,
      "step": 32635
    },
    {
      "epoch": 17.10691823899371,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0018725969513882844,
      "loss": 0.3681,
      "num_input_tokens_seen": 21295976,
      "step": 32640
    },
    {
      "epoch": 17.109538784067087,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.001869278952177053,
      "loss": 0.3579,
      "num_input_tokens_seen": 21299816,
      "step": 32645
    },
    {
      "epoch": 17.11215932914046,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0018659636997628596,
      "loss": 0.3474,
      "num_input_tokens_seen": 21302696,
      "step": 32650
    },
    {
      "epoch": 17.114779874213838,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00186265119483922,
      "loss": 0.3309,
      "num_input_tokens_seen": 21305288,
      "step": 32655
    },
    {
      "epoch": 17.11740041928721,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.001859341438099069,
      "loss": 0.2926,
      "num_input_tokens_seen": 21308808,
      "step": 32660
    },
    {
      "epoch": 17.120020964360588,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0018560344302347787,
      "loss": 0.315,
      "num_input_tokens_seen": 21312008,
      "step": 32665
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0018527301719381299,
      "loss": 0.3369,
      "num_input_tokens_seen": 21314280,
      "step": 32670
    },
    {
      "epoch": 17.12526205450734,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0018494286639003338,
      "loss": 0.2937,
      "num_input_tokens_seen": 21316808,
      "step": 32675
    },
    {
      "epoch": 17.127882599580712,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0018461299068120323,
      "loss": 0.3437,
      "num_input_tokens_seen": 21320712,
      "step": 32680
    },
    {
      "epoch": 17.13050314465409,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0018428339013632855,
      "loss": 0.3572,
      "num_input_tokens_seen": 21324968,
      "step": 32685
    },
    {
      "epoch": 17.133123689727462,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0018395406482435866,
      "loss": 0.3503,
      "num_input_tokens_seen": 21328360,
      "step": 32690
    },
    {
      "epoch": 17.13574423480084,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0018362501481418397,
      "loss": 0.3586,
      "num_input_tokens_seen": 21331336,
      "step": 32695
    },
    {
      "epoch": 17.138364779874212,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0018329624017463808,
      "loss": 0.488,
      "num_input_tokens_seen": 21334344,
      "step": 32700
    },
    {
      "epoch": 17.14098532494759,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.001829677409744968,
      "loss": 0.4184,
      "num_input_tokens_seen": 21340360,
      "step": 32705
    },
    {
      "epoch": 17.143605870020963,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0018263951728247883,
      "loss": 0.3368,
      "num_input_tokens_seen": 21343176,
      "step": 32710
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0018231156916724443,
      "loss": 0.3526,
      "num_input_tokens_seen": 21346024,
      "step": 32715
    },
    {
      "epoch": 17.148846960167713,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0018198389669739723,
      "loss": 0.3552,
      "num_input_tokens_seen": 21350120,
      "step": 32720
    },
    {
      "epoch": 17.15146750524109,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0018165649994148203,
      "loss": 0.4589,
      "num_input_tokens_seen": 21353608,
      "step": 32725
    },
    {
      "epoch": 17.154088050314467,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0018132937896798622,
      "loss": 0.3645,
      "num_input_tokens_seen": 21358376,
      "step": 32730
    },
    {
      "epoch": 17.15670859538784,
      "grad_norm": 0.123046875,
      "learning_rate": 0.001810025338453401,
      "loss": 0.2894,
      "num_input_tokens_seen": 21361128,
      "step": 32735
    },
    {
      "epoch": 17.159329140461217,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0018067596464191608,
      "loss": 0.3659,
      "num_input_tokens_seen": 21364584,
      "step": 32740
    },
    {
      "epoch": 17.16194968553459,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.001803496714260289,
      "loss": 0.3359,
      "num_input_tokens_seen": 21368168,
      "step": 32745
    },
    {
      "epoch": 17.164570230607968,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0018002365426593512,
      "loss": 0.3017,
      "num_input_tokens_seen": 21371752,
      "step": 32750
    },
    {
      "epoch": 17.16719077568134,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0017969791322983352,
      "loss": 0.4353,
      "num_input_tokens_seen": 21374632,
      "step": 32755
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0017937244838586556,
      "loss": 0.3639,
      "num_input_tokens_seen": 21377608,
      "step": 32760
    },
    {
      "epoch": 17.17243186582809,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00179047259802115,
      "loss": 0.4024,
      "num_input_tokens_seen": 21381192,
      "step": 32765
    },
    {
      "epoch": 17.17505241090147,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.001787223475466077,
      "loss": 0.4633,
      "num_input_tokens_seen": 21384168,
      "step": 32770
    },
    {
      "epoch": 17.177672955974842,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0017839771168731099,
      "loss": 0.3457,
      "num_input_tokens_seen": 21387240,
      "step": 32775
    },
    {
      "epoch": 17.18029350104822,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.001780733522921357,
      "loss": 0.2905,
      "num_input_tokens_seen": 21390216,
      "step": 32780
    },
    {
      "epoch": 17.182914046121592,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.001777492694289336,
      "loss": 0.3821,
      "num_input_tokens_seen": 21393544,
      "step": 32785
    },
    {
      "epoch": 17.18553459119497,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0017742546316549923,
      "loss": 0.3825,
      "num_input_tokens_seen": 21397288,
      "step": 32790
    },
    {
      "epoch": 17.188155136268342,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.001771019335695691,
      "loss": 0.2817,
      "num_input_tokens_seen": 21399912,
      "step": 32795
    },
    {
      "epoch": 17.19077568134172,
      "grad_norm": 0.107421875,
      "learning_rate": 0.001767786807088224,
      "loss": 0.347,
      "num_input_tokens_seen": 21402472,
      "step": 32800
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.001764557046508794,
      "loss": 0.3228,
      "num_input_tokens_seen": 21405352,
      "step": 32805
    },
    {
      "epoch": 17.19601677148847,
      "grad_norm": 0.07421875,
      "learning_rate": 0.001761330054633035,
      "loss": 0.3025,
      "num_input_tokens_seen": 21408008,
      "step": 32810
    },
    {
      "epoch": 17.198637316561843,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0017581058321359893,
      "loss": 0.4617,
      "num_input_tokens_seen": 21411880,
      "step": 32815
    },
    {
      "epoch": 17.20125786163522,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0017548843796921336,
      "loss": 0.3466,
      "num_input_tokens_seen": 21414504,
      "step": 32820
    },
    {
      "epoch": 17.203878406708597,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.001751665697975359,
      "loss": 0.3209,
      "num_input_tokens_seen": 21417608,
      "step": 32825
    },
    {
      "epoch": 17.20649895178197,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0017484497876589739,
      "loss": 0.3877,
      "num_input_tokens_seen": 21420712,
      "step": 32830
    },
    {
      "epoch": 17.209119496855347,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0017452366494157144,
      "loss": 0.4841,
      "num_input_tokens_seen": 21423400,
      "step": 32835
    },
    {
      "epoch": 17.21174004192872,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0017420262839177253,
      "loss": 0.4041,
      "num_input_tokens_seen": 21426888,
      "step": 32840
    },
    {
      "epoch": 17.214360587002098,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0017388186918365837,
      "loss": 0.3786,
      "num_input_tokens_seen": 21429736,
      "step": 32845
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0017356138738432818,
      "loss": 0.3619,
      "num_input_tokens_seen": 21433320,
      "step": 32850
    },
    {
      "epoch": 17.219601677148848,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0017324118306082281,
      "loss": 0.317,
      "num_input_tokens_seen": 21436584,
      "step": 32855
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0017292125628012532,
      "loss": 0.4705,
      "num_input_tokens_seen": 21439528,
      "step": 32860
    },
    {
      "epoch": 17.2248427672956,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0017260160710916139,
      "loss": 0.3027,
      "num_input_tokens_seen": 21442824,
      "step": 32865
    },
    {
      "epoch": 17.22746331236897,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00172282235614797,
      "loss": 0.4039,
      "num_input_tokens_seen": 21446024,
      "step": 32870
    },
    {
      "epoch": 17.23008385744235,
      "grad_norm": 0.14453125,
      "learning_rate": 0.001719631418638417,
      "loss": 0.2764,
      "num_input_tokens_seen": 21448744,
      "step": 32875
    },
    {
      "epoch": 17.232704402515722,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0017164432592304628,
      "loss": 0.3494,
      "num_input_tokens_seen": 21451688,
      "step": 32880
    },
    {
      "epoch": 17.2353249475891,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.001713257878591029,
      "loss": 0.318,
      "num_input_tokens_seen": 21453864,
      "step": 32885
    },
    {
      "epoch": 17.237945492662472,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0017100752773864625,
      "loss": 0.3615,
      "num_input_tokens_seen": 21456936,
      "step": 32890
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0017068954562825338,
      "loss": 0.3165,
      "num_input_tokens_seen": 21460424,
      "step": 32895
    },
    {
      "epoch": 17.243186582809223,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0017037184159444162,
      "loss": 0.3355,
      "num_input_tokens_seen": 21463016,
      "step": 32900
    },
    {
      "epoch": 17.2458071278826,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0017005441570367163,
      "loss": 0.3303,
      "num_input_tokens_seen": 21466632,
      "step": 32905
    },
    {
      "epoch": 17.248427672955973,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0016973726802234478,
      "loss": 0.3073,
      "num_input_tokens_seen": 21469544,
      "step": 32910
    },
    {
      "epoch": 17.25104821802935,
      "grad_norm": 0.09375,
      "learning_rate": 0.0016942039861680502,
      "loss": 0.2833,
      "num_input_tokens_seen": 21472040,
      "step": 32915
    },
    {
      "epoch": 17.253668763102727,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0016910380755333781,
      "loss": 0.3187,
      "num_input_tokens_seen": 21475784,
      "step": 32920
    },
    {
      "epoch": 17.2562893081761,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0016878749489817068,
      "loss": 0.3059,
      "num_input_tokens_seen": 21479656,
      "step": 32925
    },
    {
      "epoch": 17.258909853249477,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0016847146071747236,
      "loss": 0.3022,
      "num_input_tokens_seen": 21482824,
      "step": 32930
    },
    {
      "epoch": 17.26153039832285,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0016815570507735334,
      "loss": 0.3187,
      "num_input_tokens_seen": 21485672,
      "step": 32935
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0016784022804386622,
      "loss": 0.3124,
      "num_input_tokens_seen": 21488616,
      "step": 32940
    },
    {
      "epoch": 17.2667714884696,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0016752502968300541,
      "loss": 0.3186,
      "num_input_tokens_seen": 21490952,
      "step": 32945
    },
    {
      "epoch": 17.269392033542978,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00167210110060707,
      "loss": 0.3617,
      "num_input_tokens_seen": 21493832,
      "step": 32950
    },
    {
      "epoch": 17.27201257861635,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0016689546924284798,
      "loss": 0.3026,
      "num_input_tokens_seen": 21496136,
      "step": 32955
    },
    {
      "epoch": 17.27463312368973,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0016658110729524816,
      "loss": 0.3869,
      "num_input_tokens_seen": 21499432,
      "step": 32960
    },
    {
      "epoch": 17.2772536687631,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0016626702428366796,
      "loss": 0.4947,
      "num_input_tokens_seen": 21502472,
      "step": 32965
    },
    {
      "epoch": 17.27987421383648,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0016595322027381032,
      "loss": 0.2648,
      "num_input_tokens_seen": 21506088,
      "step": 32970
    },
    {
      "epoch": 17.282494758909852,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0016563969533131939,
      "loss": 0.2912,
      "num_input_tokens_seen": 21508904,
      "step": 32975
    },
    {
      "epoch": 17.28511530398323,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0016532644952178138,
      "loss": 0.299,
      "num_input_tokens_seen": 21511624,
      "step": 32980
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.001650134829107232,
      "loss": 0.2494,
      "num_input_tokens_seen": 21514504,
      "step": 32985
    },
    {
      "epoch": 17.29035639412998,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0016470079556361382,
      "loss": 0.4217,
      "num_input_tokens_seen": 21517256,
      "step": 32990
    },
    {
      "epoch": 17.292976939203353,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.001643883875458641,
      "loss": 0.3217,
      "num_input_tokens_seen": 21520648,
      "step": 32995
    },
    {
      "epoch": 17.29559748427673,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0016407625892282639,
      "loss": 0.3184,
      "num_input_tokens_seen": 21523688,
      "step": 33000
    },
    {
      "epoch": 17.298218029350103,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0016376440975979416,
      "loss": 0.395,
      "num_input_tokens_seen": 21527016,
      "step": 33005
    },
    {
      "epoch": 17.30083857442348,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0016345284012200339,
      "loss": 0.3319,
      "num_input_tokens_seen": 21530408,
      "step": 33010
    },
    {
      "epoch": 17.303459119496857,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0016314155007462977,
      "loss": 0.4028,
      "num_input_tokens_seen": 21533288,
      "step": 33015
    },
    {
      "epoch": 17.30607966457023,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0016283053968279221,
      "loss": 0.3355,
      "num_input_tokens_seen": 21536520,
      "step": 33020
    },
    {
      "epoch": 17.308700209643607,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.001625198090115505,
      "loss": 0.3896,
      "num_input_tokens_seen": 21539304,
      "step": 33025
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.0016220935812590598,
      "loss": 0.3091,
      "num_input_tokens_seen": 21542472,
      "step": 33030
    },
    {
      "epoch": 17.313941299790358,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0016189918709080169,
      "loss": 0.3956,
      "num_input_tokens_seen": 21545736,
      "step": 33035
    },
    {
      "epoch": 17.31656184486373,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.001615892959711216,
      "loss": 0.4414,
      "num_input_tokens_seen": 21548744,
      "step": 33040
    },
    {
      "epoch": 17.319182389937108,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0016127968483169114,
      "loss": 0.4363,
      "num_input_tokens_seen": 21551592,
      "step": 33045
    },
    {
      "epoch": 17.32180293501048,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0016097035373727768,
      "loss": 0.345,
      "num_input_tokens_seen": 21554216,
      "step": 33050
    },
    {
      "epoch": 17.32442348008386,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0016066130275258965,
      "loss": 0.2501,
      "num_input_tokens_seen": 21556776,
      "step": 33055
    },
    {
      "epoch": 17.32704402515723,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0016035253194227749,
      "loss": 0.3911,
      "num_input_tokens_seen": 21559976,
      "step": 33060
    },
    {
      "epoch": 17.32966457023061,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0016004404137093219,
      "loss": 0.3528,
      "num_input_tokens_seen": 21562792,
      "step": 33065
    },
    {
      "epoch": 17.332285115303982,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0015973583110308598,
      "loss": 0.4055,
      "num_input_tokens_seen": 21565736,
      "step": 33070
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0015942790120321343,
      "loss": 0.3064,
      "num_input_tokens_seen": 21569864,
      "step": 33075
    },
    {
      "epoch": 17.337526205450732,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0015912025173572986,
      "loss": 0.2279,
      "num_input_tokens_seen": 21572776,
      "step": 33080
    },
    {
      "epoch": 17.34014675052411,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.001588128827649921,
      "loss": 0.536,
      "num_input_tokens_seen": 21576584,
      "step": 33085
    },
    {
      "epoch": 17.342767295597483,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.001585057943552986,
      "loss": 0.3323,
      "num_input_tokens_seen": 21579496,
      "step": 33090
    },
    {
      "epoch": 17.34538784067086,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0015819898657088826,
      "loss": 0.392,
      "num_input_tokens_seen": 21582120,
      "step": 33095
    },
    {
      "epoch": 17.348008385744233,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0015789245947594142,
      "loss": 0.3889,
      "num_input_tokens_seen": 21585224,
      "step": 33100
    },
    {
      "epoch": 17.35062893081761,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0015758621313458065,
      "loss": 0.2715,
      "num_input_tokens_seen": 21588264,
      "step": 33105
    },
    {
      "epoch": 17.353249475890987,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.00157280247610869,
      "loss": 0.4209,
      "num_input_tokens_seen": 21591464,
      "step": 33110
    },
    {
      "epoch": 17.35587002096436,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0015697456296881128,
      "loss": 0.3255,
      "num_input_tokens_seen": 21594024,
      "step": 33115
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0015666915927235285,
      "loss": 0.4112,
      "num_input_tokens_seen": 21596776,
      "step": 33120
    },
    {
      "epoch": 17.36111111111111,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.001563640365853806,
      "loss": 0.2347,
      "num_input_tokens_seen": 21600392,
      "step": 33125
    },
    {
      "epoch": 17.363731656184488,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.001560591949717228,
      "loss": 0.5085,
      "num_input_tokens_seen": 21603880,
      "step": 33130
    },
    {
      "epoch": 17.36635220125786,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0015575463449514898,
      "loss": 0.3387,
      "num_input_tokens_seen": 21606600,
      "step": 33135
    },
    {
      "epoch": 17.368972746331238,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0015545035521936995,
      "loss": 0.4815,
      "num_input_tokens_seen": 21610856,
      "step": 33140
    },
    {
      "epoch": 17.37159329140461,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0015514635720803682,
      "loss": 0.3272,
      "num_input_tokens_seen": 21615016,
      "step": 33145
    },
    {
      "epoch": 17.37421383647799,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0015484264052474323,
      "loss": 0.409,
      "num_input_tokens_seen": 21618920,
      "step": 33150
    },
    {
      "epoch": 17.37683438155136,
      "grad_norm": 0.09375,
      "learning_rate": 0.0015453920523302266,
      "loss": 0.2537,
      "num_input_tokens_seen": 21621960,
      "step": 33155
    },
    {
      "epoch": 17.37945492662474,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0015423605139635037,
      "loss": 0.3156,
      "num_input_tokens_seen": 21624584,
      "step": 33160
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0015393317907814296,
      "loss": 0.3202,
      "num_input_tokens_seen": 21628200,
      "step": 33165
    },
    {
      "epoch": 17.38469601677149,
      "grad_norm": 0.056640625,
      "learning_rate": 0.001536305883417579,
      "loss": 0.2614,
      "num_input_tokens_seen": 21630792,
      "step": 33170
    },
    {
      "epoch": 17.387316561844862,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.001533282792504934,
      "loss": 0.4361,
      "num_input_tokens_seen": 21633480,
      "step": 33175
    },
    {
      "epoch": 17.38993710691824,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0015302625186758938,
      "loss": 0.3617,
      "num_input_tokens_seen": 21636584,
      "step": 33180
    },
    {
      "epoch": 17.392557651991613,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0015272450625622629,
      "loss": 0.3276,
      "num_input_tokens_seen": 21639624,
      "step": 33185
    },
    {
      "epoch": 17.39517819706499,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0015242304247952581,
      "loss": 0.446,
      "num_input_tokens_seen": 21642248,
      "step": 33190
    },
    {
      "epoch": 17.397798742138363,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0015212186060055116,
      "loss": 0.4489,
      "num_input_tokens_seen": 21645576,
      "step": 33195
    },
    {
      "epoch": 17.40041928721174,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0015182096068230577,
      "loss": 0.4349,
      "num_input_tokens_seen": 21648712,
      "step": 33200
    },
    {
      "epoch": 17.403039832285117,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0015152034278773463,
      "loss": 0.3886,
      "num_input_tokens_seen": 21651624,
      "step": 33205
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0015122000697972327,
      "loss": 0.4416,
      "num_input_tokens_seen": 21655624,
      "step": 33210
    },
    {
      "epoch": 17.408280922431867,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0015091995332109874,
      "loss": 0.413,
      "num_input_tokens_seen": 21658632,
      "step": 33215
    },
    {
      "epoch": 17.41090146750524,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.001506201818746292,
      "loss": 0.2863,
      "num_input_tokens_seen": 21661736,
      "step": 33220
    },
    {
      "epoch": 17.413522012578618,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0015032069270302284,
      "loss": 0.4074,
      "num_input_tokens_seen": 21664776,
      "step": 33225
    },
    {
      "epoch": 17.41614255765199,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.001500214858689295,
      "loss": 0.3195,
      "num_input_tokens_seen": 21667752,
      "step": 33230
    },
    {
      "epoch": 17.418763102725368,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0014972256143494028,
      "loss": 0.4871,
      "num_input_tokens_seen": 21672168,
      "step": 33235
    },
    {
      "epoch": 17.42138364779874,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0014942391946358618,
      "loss": 0.3707,
      "num_input_tokens_seen": 21674472,
      "step": 33240
    },
    {
      "epoch": 17.42400419287212,
      "grad_norm": 0.1328125,
      "learning_rate": 0.001491255600173399,
      "loss": 0.4961,
      "num_input_tokens_seen": 21677608,
      "step": 33245
    },
    {
      "epoch": 17.42662473794549,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0014882748315861514,
      "loss": 0.3764,
      "num_input_tokens_seen": 21680552,
      "step": 33250
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0014852968894976575,
      "loss": 0.3989,
      "num_input_tokens_seen": 21682696,
      "step": 33255
    },
    {
      "epoch": 17.431865828092242,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0014823217745308687,
      "loss": 0.3894,
      "num_input_tokens_seen": 21686920,
      "step": 33260
    },
    {
      "epoch": 17.43448637316562,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0014793494873081502,
      "loss": 0.2692,
      "num_input_tokens_seen": 21689416,
      "step": 33265
    },
    {
      "epoch": 17.437106918238992,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.001476380028451265,
      "loss": 0.2444,
      "num_input_tokens_seen": 21693480,
      "step": 33270
    },
    {
      "epoch": 17.43972746331237,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0014734133985813947,
      "loss": 0.3571,
      "num_input_tokens_seen": 21696296,
      "step": 33275
    },
    {
      "epoch": 17.442348008385743,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.001470449598319119,
      "loss": 0.3066,
      "num_input_tokens_seen": 21702088,
      "step": 33280
    },
    {
      "epoch": 17.44496855345912,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.001467488628284434,
      "loss": 0.4009,
      "num_input_tokens_seen": 21705160,
      "step": 33285
    },
    {
      "epoch": 17.447589098532493,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0014645304890967409,
      "loss": 0.4519,
      "num_input_tokens_seen": 21708328,
      "step": 33290
    },
    {
      "epoch": 17.45020964360587,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0014615751813748528,
      "loss": 0.506,
      "num_input_tokens_seen": 21713768,
      "step": 33295
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0014586227057369805,
      "loss": 0.3227,
      "num_input_tokens_seen": 21717192,
      "step": 33300
    },
    {
      "epoch": 17.45545073375262,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0014556730628007479,
      "loss": 0.4205,
      "num_input_tokens_seen": 21721384,
      "step": 33305
    },
    {
      "epoch": 17.458071278825997,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0014527262531831885,
      "loss": 0.3632,
      "num_input_tokens_seen": 21724392,
      "step": 33310
    },
    {
      "epoch": 17.46069182389937,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0014497822775007424,
      "loss": 0.4045,
      "num_input_tokens_seen": 21727208,
      "step": 33315
    },
    {
      "epoch": 17.463312368972748,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0014468411363692567,
      "loss": 0.2817,
      "num_input_tokens_seen": 21729960,
      "step": 33320
    },
    {
      "epoch": 17.46593291404612,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0014439028304039797,
      "loss": 0.3727,
      "num_input_tokens_seen": 21734216,
      "step": 33325
    },
    {
      "epoch": 17.468553459119498,
      "grad_norm": 0.19140625,
      "learning_rate": 0.0014409673602195777,
      "loss": 0.4124,
      "num_input_tokens_seen": 21738152,
      "step": 33330
    },
    {
      "epoch": 17.47117400419287,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0014380347264301113,
      "loss": 0.3939,
      "num_input_tokens_seen": 21741576,
      "step": 33335
    },
    {
      "epoch": 17.47379454926625,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0014351049296490565,
      "loss": 0.2836,
      "num_input_tokens_seen": 21745224,
      "step": 33340
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0014321779704892944,
      "loss": 0.4377,
      "num_input_tokens_seen": 21748424,
      "step": 33345
    },
    {
      "epoch": 17.479035639413,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.001429253849563114,
      "loss": 0.3713,
      "num_input_tokens_seen": 21751336,
      "step": 33350
    },
    {
      "epoch": 17.481656184486372,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.001426332567482204,
      "loss": 0.3011,
      "num_input_tokens_seen": 21754568,
      "step": 33355
    },
    {
      "epoch": 17.48427672955975,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.0014234141248576626,
      "loss": 0.3048,
      "num_input_tokens_seen": 21757832,
      "step": 33360
    },
    {
      "epoch": 17.486897274633122,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.001420498522299996,
      "loss": 0.2928,
      "num_input_tokens_seen": 21760968,
      "step": 33365
    },
    {
      "epoch": 17.4895178197065,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0014175857604191166,
      "loss": 0.3082,
      "num_input_tokens_seen": 21765160,
      "step": 33370
    },
    {
      "epoch": 17.492138364779873,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.0014146758398243403,
      "loss": 0.4813,
      "num_input_tokens_seen": 21768584,
      "step": 33375
    },
    {
      "epoch": 17.49475890985325,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.00141176876112439,
      "loss": 0.3548,
      "num_input_tokens_seen": 21771304,
      "step": 33380
    },
    {
      "epoch": 17.497379454926623,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0014088645249273894,
      "loss": 0.4201,
      "num_input_tokens_seen": 21774568,
      "step": 33385
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0014059631318408742,
      "loss": 0.4537,
      "num_input_tokens_seen": 21777608,
      "step": 33390
    },
    {
      "epoch": 17.502620545073377,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.0014030645824717825,
      "loss": 0.338,
      "num_input_tokens_seen": 21781288,
      "step": 33395
    },
    {
      "epoch": 17.50524109014675,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0014001688774264586,
      "loss": 0.4108,
      "num_input_tokens_seen": 21784168,
      "step": 33400
    },
    {
      "epoch": 17.507861635220127,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0013972760173106536,
      "loss": 0.3368,
      "num_input_tokens_seen": 21786760,
      "step": 33405
    },
    {
      "epoch": 17.5104821802935,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0013943860027295178,
      "loss": 0.3319,
      "num_input_tokens_seen": 21789352,
      "step": 33410
    },
    {
      "epoch": 17.513102725366878,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.001391498834287605,
      "loss": 0.4001,
      "num_input_tokens_seen": 21792296,
      "step": 33415
    },
    {
      "epoch": 17.51572327044025,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0013886145125888833,
      "loss": 0.3236,
      "num_input_tokens_seen": 21794824,
      "step": 33420
    },
    {
      "epoch": 17.518343815513628,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0013857330382367184,
      "loss": 0.3415,
      "num_input_tokens_seen": 21797928,
      "step": 33425
    },
    {
      "epoch": 17.520964360587,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0013828544118338847,
      "loss": 0.561,
      "num_input_tokens_seen": 21801352,
      "step": 33430
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.001379978633982556,
      "loss": 0.339,
      "num_input_tokens_seen": 21805768,
      "step": 33435
    },
    {
      "epoch": 17.52620545073375,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.001377105705284307,
      "loss": 0.2803,
      "num_input_tokens_seen": 21808648,
      "step": 33440
    },
    {
      "epoch": 17.52882599580713,
      "grad_norm": 0.134765625,
      "learning_rate": 0.001374235626340128,
      "loss": 0.3737,
      "num_input_tokens_seen": 21811080,
      "step": 33445
    },
    {
      "epoch": 17.531446540880502,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0013713683977504042,
      "loss": 0.2897,
      "num_input_tokens_seen": 21814568,
      "step": 33450
    },
    {
      "epoch": 17.53406708595388,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0013685040201149316,
      "loss": 0.3355,
      "num_input_tokens_seen": 21817416,
      "step": 33455
    },
    {
      "epoch": 17.536687631027252,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0013656424940328993,
      "loss": 0.3509,
      "num_input_tokens_seen": 21820648,
      "step": 33460
    },
    {
      "epoch": 17.53930817610063,
      "grad_norm": 0.140625,
      "learning_rate": 0.0013627838201029096,
      "loss": 0.3862,
      "num_input_tokens_seen": 21823560,
      "step": 33465
    },
    {
      "epoch": 17.541928721174003,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0013599279989229612,
      "loss": 0.4364,
      "num_input_tokens_seen": 21827208,
      "step": 33470
    },
    {
      "epoch": 17.54454926624738,
      "grad_norm": 0.14453125,
      "learning_rate": 0.001357075031090461,
      "loss": 0.4299,
      "num_input_tokens_seen": 21830120,
      "step": 33475
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.001354224917202217,
      "loss": 0.3608,
      "num_input_tokens_seen": 21833064,
      "step": 33480
    },
    {
      "epoch": 17.54979035639413,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.001351377657854445,
      "loss": 0.3748,
      "num_input_tokens_seen": 21835752,
      "step": 33485
    },
    {
      "epoch": 17.552410901467507,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.001348533253642754,
      "loss": 0.4422,
      "num_input_tokens_seen": 21838664,
      "step": 33490
    },
    {
      "epoch": 17.55503144654088,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0013456917051621613,
      "loss": 0.3835,
      "num_input_tokens_seen": 21842184,
      "step": 33495
    },
    {
      "epoch": 17.557651991614257,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0013428530130070865,
      "loss": 0.3616,
      "num_input_tokens_seen": 21845224,
      "step": 33500
    },
    {
      "epoch": 17.56027253668763,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.001340017177771351,
      "loss": 0.548,
      "num_input_tokens_seen": 21848552,
      "step": 33505
    },
    {
      "epoch": 17.562893081761008,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.0013371842000481854,
      "loss": 0.31,
      "num_input_tokens_seen": 21850856,
      "step": 33510
    },
    {
      "epoch": 17.56551362683438,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0013343540804302073,
      "loss": 0.2463,
      "num_input_tokens_seen": 21856104,
      "step": 33515
    },
    {
      "epoch": 17.568134171907758,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0013315268195094531,
      "loss": 0.2938,
      "num_input_tokens_seen": 21858440,
      "step": 33520
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0013287024178773464,
      "loss": 0.5363,
      "num_input_tokens_seen": 21861064,
      "step": 33525
    },
    {
      "epoch": 17.57337526205451,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0013258808761247248,
      "loss": 0.2107,
      "num_input_tokens_seen": 21863848,
      "step": 33530
    },
    {
      "epoch": 17.57599580712788,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0013230621948418224,
      "loss": 0.3549,
      "num_input_tokens_seen": 21867208,
      "step": 33535
    },
    {
      "epoch": 17.57861635220126,
      "grad_norm": 0.09375,
      "learning_rate": 0.001320246374618273,
      "loss": 0.333,
      "num_input_tokens_seen": 21871944,
      "step": 33540
    },
    {
      "epoch": 17.581236897274632,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.001317433416043115,
      "loss": 0.1826,
      "num_input_tokens_seen": 21874728,
      "step": 33545
    },
    {
      "epoch": 17.58385744234801,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0013146233197047896,
      "loss": 0.3017,
      "num_input_tokens_seen": 21877320,
      "step": 33550
    },
    {
      "epoch": 17.586477987421382,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0013118160861911314,
      "loss": 0.3312,
      "num_input_tokens_seen": 21880488,
      "step": 33555
    },
    {
      "epoch": 17.58909853249476,
      "grad_norm": 0.064453125,
      "learning_rate": 0.001309011716089386,
      "loss": 0.3926,
      "num_input_tokens_seen": 21883080,
      "step": 33560
    },
    {
      "epoch": 17.591719077568133,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0013062102099861955,
      "loss": 0.4223,
      "num_input_tokens_seen": 21886536,
      "step": 33565
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.001303411568467601,
      "loss": 0.3628,
      "num_input_tokens_seen": 21889896,
      "step": 33570
    },
    {
      "epoch": 17.596960167714883,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0013006157921190498,
      "loss": 0.3385,
      "num_input_tokens_seen": 21893256,
      "step": 33575
    },
    {
      "epoch": 17.59958071278826,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0012978228815253817,
      "loss": 0.3295,
      "num_input_tokens_seen": 21896328,
      "step": 33580
    },
    {
      "epoch": 17.602201257861637,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.001295032837270843,
      "loss": 0.4199,
      "num_input_tokens_seen": 21898856,
      "step": 33585
    },
    {
      "epoch": 17.60482180293501,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0012922456599390834,
      "loss": 0.3436,
      "num_input_tokens_seen": 21901992,
      "step": 33590
    },
    {
      "epoch": 17.607442348008387,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0012894613501131435,
      "loss": 0.2795,
      "num_input_tokens_seen": 21904904,
      "step": 33595
    },
    {
      "epoch": 17.61006289308176,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0012866799083754703,
      "loss": 0.3064,
      "num_input_tokens_seen": 21908808,
      "step": 33600
    },
    {
      "epoch": 17.612683438155138,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0012839013353079136,
      "loss": 0.335,
      "num_input_tokens_seen": 21911720,
      "step": 33605
    },
    {
      "epoch": 17.61530398322851,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0012811256314917135,
      "loss": 0.2681,
      "num_input_tokens_seen": 21914024,
      "step": 33610
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0012783527975075203,
      "loss": 0.278,
      "num_input_tokens_seen": 21916648,
      "step": 33615
    },
    {
      "epoch": 17.62054507337526,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0012755828339353736,
      "loss": 0.359,
      "num_input_tokens_seen": 21920136,
      "step": 33620
    },
    {
      "epoch": 17.62316561844864,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0012728157413547231,
      "loss": 0.2832,
      "num_input_tokens_seen": 21924168,
      "step": 33625
    },
    {
      "epoch": 17.62578616352201,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0012700515203444107,
      "loss": 0.3578,
      "num_input_tokens_seen": 21931656,
      "step": 33630
    },
    {
      "epoch": 17.62840670859539,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0012672901714826839,
      "loss": 0.312,
      "num_input_tokens_seen": 21934184,
      "step": 33635
    },
    {
      "epoch": 17.631027253668762,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.001264531695347179,
      "loss": 0.4219,
      "num_input_tokens_seen": 21938856,
      "step": 33640
    },
    {
      "epoch": 17.63364779874214,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0012617760925149428,
      "loss": 0.287,
      "num_input_tokens_seen": 21941736,
      "step": 33645
    },
    {
      "epoch": 17.636268343815512,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0012590233635624126,
      "loss": 0.4117,
      "num_input_tokens_seen": 21945256,
      "step": 33650
    },
    {
      "epoch": 17.63888888888889,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0012562735090654292,
      "loss": 0.3318,
      "num_input_tokens_seen": 21947432,
      "step": 33655
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0012535265295992308,
      "loss": 0.4229,
      "num_input_tokens_seen": 21950664,
      "step": 33660
    },
    {
      "epoch": 17.64412997903564,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0012507824257384581,
      "loss": 0.3847,
      "num_input_tokens_seen": 21953704,
      "step": 33665
    },
    {
      "epoch": 17.646750524109013,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0012480411980571415,
      "loss": 0.2912,
      "num_input_tokens_seen": 21956936,
      "step": 33670
    },
    {
      "epoch": 17.64937106918239,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0012453028471287124,
      "loss": 0.3565,
      "num_input_tokens_seen": 21960200,
      "step": 33675
    },
    {
      "epoch": 17.651991614255767,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0012425673735260061,
      "loss": 0.3221,
      "num_input_tokens_seen": 21962824,
      "step": 33680
    },
    {
      "epoch": 17.65461215932914,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0012398347778212531,
      "loss": 0.3622,
      "num_input_tokens_seen": 21966504,
      "step": 33685
    },
    {
      "epoch": 17.657232704402517,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0012371050605860829,
      "loss": 0.347,
      "num_input_tokens_seen": 21970664,
      "step": 33690
    },
    {
      "epoch": 17.65985324947589,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0012343782223915183,
      "loss": 0.3037,
      "num_input_tokens_seen": 21974312,
      "step": 33695
    },
    {
      "epoch": 17.662473794549268,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0012316542638079802,
      "loss": 0.4432,
      "num_input_tokens_seen": 21977288,
      "step": 33700
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0012289331854052927,
      "loss": 0.315,
      "num_input_tokens_seen": 21979720,
      "step": 33705
    },
    {
      "epoch": 17.667714884696018,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0012262149877526735,
      "loss": 0.3506,
      "num_input_tokens_seen": 21982600,
      "step": 33710
    },
    {
      "epoch": 17.67033542976939,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0012234996714187395,
      "loss": 0.3453,
      "num_input_tokens_seen": 21986184,
      "step": 33715
    },
    {
      "epoch": 17.67295597484277,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0012207872369715067,
      "loss": 0.4235,
      "num_input_tokens_seen": 21989352,
      "step": 33720
    },
    {
      "epoch": 17.67557651991614,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0012180776849783825,
      "loss": 0.3318,
      "num_input_tokens_seen": 21992072,
      "step": 33725
    },
    {
      "epoch": 17.67819706498952,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.0012153710160061704,
      "loss": 0.3965,
      "num_input_tokens_seen": 21996200,
      "step": 33730
    },
    {
      "epoch": 17.680817610062892,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0012126672306210788,
      "loss": 0.3472,
      "num_input_tokens_seen": 21998728,
      "step": 33735
    },
    {
      "epoch": 17.68343815513627,
      "grad_norm": 0.0703125,
      "learning_rate": 0.001209966329388709,
      "loss": 0.3751,
      "num_input_tokens_seen": 22001480,
      "step": 33740
    },
    {
      "epoch": 17.686058700209642,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0012072683128740607,
      "loss": 0.4695,
      "num_input_tokens_seen": 22005800,
      "step": 33745
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0012045731816415256,
      "loss": 0.254,
      "num_input_tokens_seen": 22009096,
      "step": 33750
    },
    {
      "epoch": 17.691299790356393,
      "grad_norm": 0.0703125,
      "learning_rate": 0.001201880936254891,
      "loss": 0.3314,
      "num_input_tokens_seen": 22012072,
      "step": 33755
    },
    {
      "epoch": 17.69392033542977,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0011991915772773482,
      "loss": 0.4468,
      "num_input_tokens_seen": 22015304,
      "step": 33760
    },
    {
      "epoch": 17.696540880503143,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0011965051052714787,
      "loss": 0.3097,
      "num_input_tokens_seen": 22019016,
      "step": 33765
    },
    {
      "epoch": 17.69916142557652,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0011938215207992647,
      "loss": 0.3999,
      "num_input_tokens_seen": 22021992,
      "step": 33770
    },
    {
      "epoch": 17.701781970649897,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0011911408244220806,
      "loss": 0.2875,
      "num_input_tokens_seen": 22024808,
      "step": 33775
    },
    {
      "epoch": 17.70440251572327,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0011884630167006925,
      "loss": 0.3619,
      "num_input_tokens_seen": 22027336,
      "step": 33780
    },
    {
      "epoch": 17.707023060796647,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0011857880981952728,
      "loss": 0.3339,
      "num_input_tokens_seen": 22030440,
      "step": 33785
    },
    {
      "epoch": 17.70964360587002,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0011831160694653802,
      "loss": 0.4687,
      "num_input_tokens_seen": 22033160,
      "step": 33790
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.001180446931069976,
      "loss": 0.205,
      "num_input_tokens_seen": 22035720,
      "step": 33795
    },
    {
      "epoch": 17.71488469601677,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0011777806835674136,
      "loss": 0.4249,
      "num_input_tokens_seen": 22038664,
      "step": 33800
    },
    {
      "epoch": 17.717505241090148,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0011751173275154402,
      "loss": 0.2835,
      "num_input_tokens_seen": 22041512,
      "step": 33805
    },
    {
      "epoch": 17.72012578616352,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0011724568634711968,
      "loss": 0.356,
      "num_input_tokens_seen": 22045320,
      "step": 33810
    },
    {
      "epoch": 17.7227463312369,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0011697992919912248,
      "loss": 0.2648,
      "num_input_tokens_seen": 22048232,
      "step": 33815
    },
    {
      "epoch": 17.72536687631027,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0011671446136314561,
      "loss": 0.3178,
      "num_input_tokens_seen": 22051528,
      "step": 33820
    },
    {
      "epoch": 17.72798742138365,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.001164492828947225,
      "loss": 0.3741,
      "num_input_tokens_seen": 22054248,
      "step": 33825
    },
    {
      "epoch": 17.730607966457022,
      "grad_norm": 0.09375,
      "learning_rate": 0.0011618439384932476,
      "loss": 0.2488,
      "num_input_tokens_seen": 22058600,
      "step": 33830
    },
    {
      "epoch": 17.7332285115304,
      "grad_norm": 0.1171875,
      "learning_rate": 0.001159197942823646,
      "loss": 0.3176,
      "num_input_tokens_seen": 22061896,
      "step": 33835
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0011565548424919286,
      "loss": 0.4769,
      "num_input_tokens_seen": 22065192,
      "step": 33840
    },
    {
      "epoch": 17.73846960167715,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0011539146380510033,
      "loss": 0.3963,
      "num_input_tokens_seen": 22068296,
      "step": 33845
    },
    {
      "epoch": 17.741090146750523,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0011512773300531736,
      "loss": 0.4096,
      "num_input_tokens_seen": 22071976,
      "step": 33850
    },
    {
      "epoch": 17.7437106918239,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0011486429190501295,
      "loss": 0.4253,
      "num_input_tokens_seen": 22075656,
      "step": 33855
    },
    {
      "epoch": 17.746331236897273,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0011460114055929654,
      "loss": 0.306,
      "num_input_tokens_seen": 22079816,
      "step": 33860
    },
    {
      "epoch": 17.74895178197065,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0011433827902321574,
      "loss": 0.3417,
      "num_input_tokens_seen": 22082472,
      "step": 33865
    },
    {
      "epoch": 17.751572327044027,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.001140757073517586,
      "loss": 0.3528,
      "num_input_tokens_seen": 22086024,
      "step": 33870
    },
    {
      "epoch": 17.7541928721174,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0011381342559985198,
      "loss": 0.3983,
      "num_input_tokens_seen": 22090024,
      "step": 33875
    },
    {
      "epoch": 17.756813417190777,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0011355143382236232,
      "loss": 0.2659,
      "num_input_tokens_seen": 22092680,
      "step": 33880
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0011328973207409532,
      "loss": 0.4122,
      "num_input_tokens_seen": 22095112,
      "step": 33885
    },
    {
      "epoch": 17.762054507337528,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0011302832040979598,
      "loss": 0.3993,
      "num_input_tokens_seen": 22098696,
      "step": 33890
    },
    {
      "epoch": 17.7646750524109,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0011276719888414844,
      "loss": 0.3849,
      "num_input_tokens_seen": 22102024,
      "step": 33895
    },
    {
      "epoch": 17.767295597484278,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0011250636755177645,
      "loss": 0.3704,
      "num_input_tokens_seen": 22105064,
      "step": 33900
    },
    {
      "epoch": 17.76991614255765,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.001122458264672434,
      "loss": 0.4462,
      "num_input_tokens_seen": 22108968,
      "step": 33905
    },
    {
      "epoch": 17.77253668763103,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0011198557568505069,
      "loss": 0.33,
      "num_input_tokens_seen": 22113288,
      "step": 33910
    },
    {
      "epoch": 17.7751572327044,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0011172561525964026,
      "loss": 0.4518,
      "num_input_tokens_seen": 22117768,
      "step": 33915
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.001114659452453931,
      "loss": 0.3208,
      "num_input_tokens_seen": 22120936,
      "step": 33920
    },
    {
      "epoch": 17.780398322851152,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0011120656569662878,
      "loss": 0.276,
      "num_input_tokens_seen": 22124616,
      "step": 33925
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0011094747666760706,
      "loss": 0.3226,
      "num_input_tokens_seen": 22128072,
      "step": 33930
    },
    {
      "epoch": 17.785639412997902,
      "grad_norm": 0.14453125,
      "learning_rate": 0.001106886782125256,
      "loss": 0.3419,
      "num_input_tokens_seen": 22131272,
      "step": 33935
    },
    {
      "epoch": 17.78825995807128,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0011043017038552272,
      "loss": 0.2731,
      "num_input_tokens_seen": 22134888,
      "step": 33940
    },
    {
      "epoch": 17.790880503144653,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0011017195324067536,
      "loss": 0.2779,
      "num_input_tokens_seen": 22137480,
      "step": 33945
    },
    {
      "epoch": 17.79350104821803,
      "grad_norm": 0.125,
      "learning_rate": 0.0010991402683199912,
      "loss": 0.3466,
      "num_input_tokens_seen": 22141000,
      "step": 33950
    },
    {
      "epoch": 17.796121593291403,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0010965639121344956,
      "loss": 0.2441,
      "num_input_tokens_seen": 22143752,
      "step": 33955
    },
    {
      "epoch": 17.79874213836478,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0010939904643892117,
      "loss": 0.336,
      "num_input_tokens_seen": 22147272,
      "step": 33960
    },
    {
      "epoch": 17.801362683438157,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0010914199256224742,
      "loss": 0.2258,
      "num_input_tokens_seen": 22151080,
      "step": 33965
    },
    {
      "epoch": 17.80398322851153,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0010888522963720098,
      "loss": 0.2787,
      "num_input_tokens_seen": 22154664,
      "step": 33970
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0010862875771749402,
      "loss": 0.2967,
      "num_input_tokens_seen": 22157608,
      "step": 33975
    },
    {
      "epoch": 17.80922431865828,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0010837257685677715,
      "loss": 0.3866,
      "num_input_tokens_seen": 22160136,
      "step": 33980
    },
    {
      "epoch": 17.811844863731658,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0010811668710864098,
      "loss": 0.3009,
      "num_input_tokens_seen": 22162376,
      "step": 33985
    },
    {
      "epoch": 17.81446540880503,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0010786108852661419,
      "loss": 0.5536,
      "num_input_tokens_seen": 22166376,
      "step": 33990
    },
    {
      "epoch": 17.817085953878408,
      "grad_norm": 0.107421875,
      "learning_rate": 0.001076057811641652,
      "loss": 0.3097,
      "num_input_tokens_seen": 22169128,
      "step": 33995
    },
    {
      "epoch": 17.81970649895178,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.001073507650747016,
      "loss": 0.42,
      "num_input_tokens_seen": 22173544,
      "step": 34000
    },
    {
      "epoch": 17.822327044025158,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0010709604031156987,
      "loss": 0.325,
      "num_input_tokens_seen": 22176712,
      "step": 34005
    },
    {
      "epoch": 17.82494758909853,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.001068416069280556,
      "loss": 0.3171,
      "num_input_tokens_seen": 22179816,
      "step": 34010
    },
    {
      "epoch": 17.82756813417191,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0010658746497738286,
      "loss": 0.3094,
      "num_input_tokens_seen": 22182888,
      "step": 34015
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 0.0625,
      "learning_rate": 0.0010633361451271562,
      "loss": 0.3926,
      "num_input_tokens_seen": 22186472,
      "step": 34020
    },
    {
      "epoch": 17.83280922431866,
      "grad_norm": 0.0859375,
      "learning_rate": 0.001060800555871562,
      "loss": 0.2988,
      "num_input_tokens_seen": 22189320,
      "step": 34025
    },
    {
      "epoch": 17.835429769392032,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0010582678825374675,
      "loss": 0.3648,
      "num_input_tokens_seen": 22192200,
      "step": 34030
    },
    {
      "epoch": 17.83805031446541,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0010557381256546766,
      "loss": 0.2771,
      "num_input_tokens_seen": 22194696,
      "step": 34035
    },
    {
      "epoch": 17.840670859538783,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.0010532112857523878,
      "loss": 0.3027,
      "num_input_tokens_seen": 22197192,
      "step": 34040
    },
    {
      "epoch": 17.84329140461216,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00105068736335918,
      "loss": 0.32,
      "num_input_tokens_seen": 22199976,
      "step": 34045
    },
    {
      "epoch": 17.845911949685533,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0010481663590030342,
      "loss": 0.3616,
      "num_input_tokens_seen": 22203176,
      "step": 34050
    },
    {
      "epoch": 17.84853249475891,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0010456482732113153,
      "loss": 0.2587,
      "num_input_tokens_seen": 22207368,
      "step": 34055
    },
    {
      "epoch": 17.851153039832283,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0010431331065107801,
      "loss": 0.3802,
      "num_input_tokens_seen": 22210376,
      "step": 34060
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0010406208594275711,
      "loss": 0.3327,
      "num_input_tokens_seen": 22214856,
      "step": 34065
    },
    {
      "epoch": 17.856394129979037,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0010381115324872181,
      "loss": 0.3129,
      "num_input_tokens_seen": 22218504,
      "step": 34070
    },
    {
      "epoch": 17.85901467505241,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0010356051262146477,
      "loss": 0.3389,
      "num_input_tokens_seen": 22221384,
      "step": 34075
    },
    {
      "epoch": 17.861635220125788,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0010331016411341703,
      "loss": 0.4304,
      "num_input_tokens_seen": 22225480,
      "step": 34080
    },
    {
      "epoch": 17.86425576519916,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0010306010777694906,
      "loss": 0.3826,
      "num_input_tokens_seen": 22228488,
      "step": 34085
    },
    {
      "epoch": 17.866876310272538,
      "grad_norm": 0.06640625,
      "learning_rate": 0.001028103436643692,
      "loss": 0.3676,
      "num_input_tokens_seen": 22231912,
      "step": 34090
    },
    {
      "epoch": 17.86949685534591,
      "grad_norm": 0.09375,
      "learning_rate": 0.0010256087182792528,
      "loss": 0.3717,
      "num_input_tokens_seen": 22235144,
      "step": 34095
    },
    {
      "epoch": 17.872117400419288,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0010231169231980424,
      "loss": 0.4452,
      "num_input_tokens_seen": 22238088,
      "step": 34100
    },
    {
      "epoch": 17.87473794549266,
      "grad_norm": 0.064453125,
      "learning_rate": 0.001020628051921314,
      "loss": 0.4332,
      "num_input_tokens_seen": 22241800,
      "step": 34105
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0010181421049697125,
      "loss": 0.405,
      "num_input_tokens_seen": 22245512,
      "step": 34110
    },
    {
      "epoch": 17.879979035639412,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0010156590828632704,
      "loss": 0.4048,
      "num_input_tokens_seen": 22248328,
      "step": 34115
    },
    {
      "epoch": 17.88259958071279,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0010131789861214074,
      "loss": 0.4386,
      "num_input_tokens_seen": 22252744,
      "step": 34120
    },
    {
      "epoch": 17.885220125786162,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0010107018152629283,
      "loss": 0.2906,
      "num_input_tokens_seen": 22257544,
      "step": 34125
    },
    {
      "epoch": 17.88784067085954,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0010082275708060318,
      "loss": 0.3536,
      "num_input_tokens_seen": 22260456,
      "step": 34130
    },
    {
      "epoch": 17.890461215932913,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.001005756253268299,
      "loss": 0.3871,
      "num_input_tokens_seen": 22264328,
      "step": 34135
    },
    {
      "epoch": 17.89308176100629,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0010032878631667064,
      "loss": 0.2535,
      "num_input_tokens_seen": 22267080,
      "step": 34140
    },
    {
      "epoch": 17.895702306079663,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0010008224010176108,
      "loss": 0.2634,
      "num_input_tokens_seen": 22270248,
      "step": 34145
    },
    {
      "epoch": 17.89832285115304,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0009983598673367533,
      "loss": 0.3422,
      "num_input_tokens_seen": 22273960,
      "step": 34150
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 0.15625,
      "learning_rate": 0.0009959002626392748,
      "loss": 0.326,
      "num_input_tokens_seen": 22277256,
      "step": 34155
    },
    {
      "epoch": 17.90356394129979,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0009934435874396925,
      "loss": 0.3905,
      "num_input_tokens_seen": 22280424,
      "step": 34160
    },
    {
      "epoch": 17.906184486373167,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0009909898422519197,
      "loss": 0.3043,
      "num_input_tokens_seen": 22284808,
      "step": 34165
    },
    {
      "epoch": 17.90880503144654,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0009885390275892463,
      "loss": 0.3094,
      "num_input_tokens_seen": 22287912,
      "step": 34170
    },
    {
      "epoch": 17.911425576519918,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0009860911439643588,
      "loss": 0.402,
      "num_input_tokens_seen": 22291112,
      "step": 34175
    },
    {
      "epoch": 17.91404612159329,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0009836461918893225,
      "loss": 0.4163,
      "num_input_tokens_seen": 22294280,
      "step": 34180
    },
    {
      "epoch": 17.916666666666668,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0009812041718755965,
      "loss": 0.2493,
      "num_input_tokens_seen": 22297640,
      "step": 34185
    },
    {
      "epoch": 17.91928721174004,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0009787650844340222,
      "loss": 0.3122,
      "num_input_tokens_seen": 22300712,
      "step": 34190
    },
    {
      "epoch": 17.921907756813418,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0009763289300748334,
      "loss": 0.3131,
      "num_input_tokens_seen": 22304072,
      "step": 34195
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0009738957093076389,
      "loss": 0.4873,
      "num_input_tokens_seen": 22307336,
      "step": 34200
    },
    {
      "epoch": 17.92714884696017,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0009714654226414449,
      "loss": 0.3474,
      "num_input_tokens_seen": 22310216,
      "step": 34205
    },
    {
      "epoch": 17.929769392033542,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0009690380705846363,
      "loss": 0.3276,
      "num_input_tokens_seen": 22314216,
      "step": 34210
    },
    {
      "epoch": 17.93238993710692,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0009666136536449904,
      "loss": 0.2742,
      "num_input_tokens_seen": 22317416,
      "step": 34215
    },
    {
      "epoch": 17.935010482180292,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0009641921723296697,
      "loss": 0.3216,
      "num_input_tokens_seen": 22320456,
      "step": 34220
    },
    {
      "epoch": 17.93763102725367,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0009617736271452143,
      "loss": 0.2737,
      "num_input_tokens_seen": 22323048,
      "step": 34225
    },
    {
      "epoch": 17.940251572327043,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0009593580185975625,
      "loss": 0.2651,
      "num_input_tokens_seen": 22326568,
      "step": 34230
    },
    {
      "epoch": 17.94287211740042,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0009569453471920269,
      "loss": 0.3905,
      "num_input_tokens_seen": 22329864,
      "step": 34235
    },
    {
      "epoch": 17.945492662473793,
      "grad_norm": 0.15625,
      "learning_rate": 0.0009545356134333121,
      "loss": 0.5082,
      "num_input_tokens_seen": 22333000,
      "step": 34240
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.0009521288178255099,
      "loss": 0.308,
      "num_input_tokens_seen": 22336328,
      "step": 34245
    },
    {
      "epoch": 17.950733752620543,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0009497249608720909,
      "loss": 0.2763,
      "num_input_tokens_seen": 22338920,
      "step": 34250
    },
    {
      "epoch": 17.95335429769392,
      "grad_norm": 0.125,
      "learning_rate": 0.000947324043075916,
      "loss": 0.5491,
      "num_input_tokens_seen": 22342568,
      "step": 34255
    },
    {
      "epoch": 17.955974842767297,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0009449260649392338,
      "loss": 0.3798,
      "num_input_tokens_seen": 22345512,
      "step": 34260
    },
    {
      "epoch": 17.95859538784067,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.000942531026963666,
      "loss": 0.3943,
      "num_input_tokens_seen": 22348488,
      "step": 34265
    },
    {
      "epoch": 17.961215932914047,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0009401389296502305,
      "loss": 0.3295,
      "num_input_tokens_seen": 22351848,
      "step": 34270
    },
    {
      "epoch": 17.96383647798742,
      "grad_norm": 0.24609375,
      "learning_rate": 0.0009377497734993317,
      "loss": 0.2755,
      "num_input_tokens_seen": 22354856,
      "step": 34275
    },
    {
      "epoch": 17.966457023060798,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0009353635590107467,
      "loss": 0.228,
      "num_input_tokens_seen": 22358152,
      "step": 34280
    },
    {
      "epoch": 17.96907756813417,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0009329802866836461,
      "loss": 0.4043,
      "num_input_tokens_seen": 22361704,
      "step": 34285
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0009305999570165879,
      "loss": 0.4321,
      "num_input_tokens_seen": 22364680,
      "step": 34290
    },
    {
      "epoch": 17.97431865828092,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0009282225705075037,
      "loss": 0.2864,
      "num_input_tokens_seen": 22367752,
      "step": 34295
    },
    {
      "epoch": 17.9769392033543,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0009258481276537222,
      "loss": 0.3297,
      "num_input_tokens_seen": 22370632,
      "step": 34300
    },
    {
      "epoch": 17.979559748427672,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0009234766289519413,
      "loss": 0.3379,
      "num_input_tokens_seen": 22373224,
      "step": 34305
    },
    {
      "epoch": 17.98218029350105,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0009211080748982575,
      "loss": 0.3462,
      "num_input_tokens_seen": 22379816,
      "step": 34310
    },
    {
      "epoch": 17.984800838574422,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.000918742465988146,
      "loss": 0.2748,
      "num_input_tokens_seen": 22383112,
      "step": 34315
    },
    {
      "epoch": 17.9874213836478,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0009163798027164609,
      "loss": 0.4234,
      "num_input_tokens_seen": 22386472,
      "step": 34320
    },
    {
      "epoch": 17.990041928721173,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0009140200855774489,
      "loss": 0.3726,
      "num_input_tokens_seen": 22388968,
      "step": 34325
    },
    {
      "epoch": 17.99266247379455,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0009116633150647329,
      "loss": 0.2874,
      "num_input_tokens_seen": 22392008,
      "step": 34330
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0009093094916713207,
      "loss": 0.2973,
      "num_input_tokens_seen": 22394984,
      "step": 34335
    },
    {
      "epoch": 17.9979035639413,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0009069586158896097,
      "loss": 0.3645,
      "num_input_tokens_seen": 22398376,
      "step": 34340
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.5047969222068787,
      "eval_runtime": 13.7729,
      "eval_samples_per_second": 61.57,
      "eval_steps_per_second": 15.393,
      "num_input_tokens_seen": 22400368,
      "step": 34344
    },
    {
      "epoch": 18.000524109014677,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0009046106882113752,
      "loss": 0.4825,
      "num_input_tokens_seen": 22400816,
      "step": 34345
    },
    {
      "epoch": 18.00314465408805,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0009022657091277742,
      "loss": 0.359,
      "num_input_tokens_seen": 22403888,
      "step": 34350
    },
    {
      "epoch": 18.005765199161427,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.000899923679129354,
      "loss": 0.417,
      "num_input_tokens_seen": 22408208,
      "step": 34355
    },
    {
      "epoch": 18.0083857442348,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0008975845987060348,
      "loss": 0.3425,
      "num_input_tokens_seen": 22411856,
      "step": 34360
    },
    {
      "epoch": 18.011006289308177,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0008952484683471301,
      "loss": 0.3321,
      "num_input_tokens_seen": 22415248,
      "step": 34365
    },
    {
      "epoch": 18.01362683438155,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0008929152885413288,
      "loss": 0.3714,
      "num_input_tokens_seen": 22418416,
      "step": 34370
    },
    {
      "epoch": 18.016247379454928,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0008905850597767107,
      "loss": 0.3373,
      "num_input_tokens_seen": 22421456,
      "step": 34375
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0008882577825407272,
      "loss": 0.2199,
      "num_input_tokens_seen": 22424464,
      "step": 34380
    },
    {
      "epoch": 18.021488469601678,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.000885933457320221,
      "loss": 0.3068,
      "num_input_tokens_seen": 22427152,
      "step": 34385
    },
    {
      "epoch": 18.02410901467505,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0008836120846014111,
      "loss": 0.3041,
      "num_input_tokens_seen": 22431696,
      "step": 34390
    },
    {
      "epoch": 18.02672955974843,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.000881293664869906,
      "loss": 0.4782,
      "num_input_tokens_seen": 22434704,
      "step": 34395
    },
    {
      "epoch": 18.029350104821802,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0008789781986106942,
      "loss": 0.3723,
      "num_input_tokens_seen": 22437776,
      "step": 34400
    },
    {
      "epoch": 18.03197064989518,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0008766656863081417,
      "loss": 0.3124,
      "num_input_tokens_seen": 22440912,
      "step": 34405
    },
    {
      "epoch": 18.034591194968552,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0008743561284459983,
      "loss": 0.3828,
      "num_input_tokens_seen": 22444112,
      "step": 34410
    },
    {
      "epoch": 18.03721174004193,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0008720495255073973,
      "loss": 0.3471,
      "num_input_tokens_seen": 22448080,
      "step": 34415
    },
    {
      "epoch": 18.039832285115303,
      "grad_norm": 0.361328125,
      "learning_rate": 0.0008697458779748562,
      "loss": 0.3514,
      "num_input_tokens_seen": 22450800,
      "step": 34420
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0008674451863302728,
      "loss": 0.4211,
      "num_input_tokens_seen": 22453648,
      "step": 34425
    },
    {
      "epoch": 18.045073375262053,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0008651474510549239,
      "loss": 0.3153,
      "num_input_tokens_seen": 22458480,
      "step": 34430
    },
    {
      "epoch": 18.04769392033543,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0008628526726294699,
      "loss": 0.3671,
      "num_input_tokens_seen": 22461840,
      "step": 34435
    },
    {
      "epoch": 18.050314465408807,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0008605608515339485,
      "loss": 0.267,
      "num_input_tokens_seen": 22465104,
      "step": 34440
    },
    {
      "epoch": 18.05293501048218,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0008582719882477879,
      "loss": 0.3475,
      "num_input_tokens_seen": 22468976,
      "step": 34445
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0008559860832497883,
      "loss": 0.2974,
      "num_input_tokens_seen": 22471888,
      "step": 34450
    },
    {
      "epoch": 18.05817610062893,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0008537031370181392,
      "loss": 0.3787,
      "num_input_tokens_seen": 22474640,
      "step": 34455
    },
    {
      "epoch": 18.060796645702307,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0008514231500304048,
      "loss": 0.4198,
      "num_input_tokens_seen": 22477168,
      "step": 34460
    },
    {
      "epoch": 18.06341719077568,
      "grad_norm": 0.09375,
      "learning_rate": 0.0008491461227635288,
      "loss": 0.4019,
      "num_input_tokens_seen": 22480240,
      "step": 34465
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0008468720556938436,
      "loss": 0.3759,
      "num_input_tokens_seen": 22483344,
      "step": 34470
    },
    {
      "epoch": 18.06865828092243,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0008446009492970568,
      "loss": 0.2726,
      "num_input_tokens_seen": 22485840,
      "step": 34475
    },
    {
      "epoch": 18.071278825995808,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0008423328040482585,
      "loss": 0.3376,
      "num_input_tokens_seen": 22489392,
      "step": 34480
    },
    {
      "epoch": 18.07389937106918,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0008400676204219176,
      "loss": 0.2905,
      "num_input_tokens_seen": 22491696,
      "step": 34485
    },
    {
      "epoch": 18.07651991614256,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0008378053988918882,
      "loss": 0.3216,
      "num_input_tokens_seen": 22494448,
      "step": 34490
    },
    {
      "epoch": 18.079140461215932,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0008355461399313951,
      "loss": 0.3229,
      "num_input_tokens_seen": 22497264,
      "step": 34495
    },
    {
      "epoch": 18.08176100628931,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0008332898440130536,
      "loss": 0.2848,
      "num_input_tokens_seen": 22500080,
      "step": 34500
    },
    {
      "epoch": 18.084381551362682,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0008310365116088547,
      "loss": 0.3647,
      "num_input_tokens_seen": 22503152,
      "step": 34505
    },
    {
      "epoch": 18.08700209643606,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0008287861431901711,
      "loss": 0.4054,
      "num_input_tokens_seen": 22506288,
      "step": 34510
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0008265387392277528,
      "loss": 0.3749,
      "num_input_tokens_seen": 22508496,
      "step": 34515
    },
    {
      "epoch": 18.09224318658281,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.000824294300191729,
      "loss": 0.3949,
      "num_input_tokens_seen": 22512496,
      "step": 34520
    },
    {
      "epoch": 18.094863731656183,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0008220528265516125,
      "loss": 0.344,
      "num_input_tokens_seen": 22515888,
      "step": 34525
    },
    {
      "epoch": 18.09748427672956,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0008198143187762929,
      "loss": 0.3039,
      "num_input_tokens_seen": 22518928,
      "step": 34530
    },
    {
      "epoch": 18.100104821802937,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0008175787773340443,
      "loss": 0.3293,
      "num_input_tokens_seen": 22523120,
      "step": 34535
    },
    {
      "epoch": 18.10272536687631,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0008153462026925123,
      "loss": 0.2634,
      "num_input_tokens_seen": 22525392,
      "step": 34540
    },
    {
      "epoch": 18.105345911949687,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0008131165953187302,
      "loss": 0.307,
      "num_input_tokens_seen": 22527920,
      "step": 34545
    },
    {
      "epoch": 18.10796645702306,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0008108899556791016,
      "loss": 0.3841,
      "num_input_tokens_seen": 22531472,
      "step": 34550
    },
    {
      "epoch": 18.110587002096437,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0008086662842394154,
      "loss": 0.4028,
      "num_input_tokens_seen": 22534672,
      "step": 34555
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0008064455814648414,
      "loss": 0.3838,
      "num_input_tokens_seen": 22537424,
      "step": 34560
    },
    {
      "epoch": 18.115828092243188,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0008042278478199211,
      "loss": 0.4421,
      "num_input_tokens_seen": 22540944,
      "step": 34565
    },
    {
      "epoch": 18.11844863731656,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0008020130837685818,
      "loss": 0.3126,
      "num_input_tokens_seen": 22544368,
      "step": 34570
    },
    {
      "epoch": 18.121069182389938,
      "grad_norm": 0.083984375,
      "learning_rate": 0.000799801289774128,
      "loss": 0.3698,
      "num_input_tokens_seen": 22547568,
      "step": 34575
    },
    {
      "epoch": 18.12368972746331,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0007975924662992378,
      "loss": 0.3381,
      "num_input_tokens_seen": 22551504,
      "step": 34580
    },
    {
      "epoch": 18.12631027253669,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0007953866138059734,
      "loss": 0.3209,
      "num_input_tokens_seen": 22554416,
      "step": 34585
    },
    {
      "epoch": 18.128930817610062,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0007931837327557772,
      "loss": 0.2312,
      "num_input_tokens_seen": 22557232,
      "step": 34590
    },
    {
      "epoch": 18.13155136268344,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0007909838236094624,
      "loss": 0.3184,
      "num_input_tokens_seen": 22560976,
      "step": 34595
    },
    {
      "epoch": 18.134171907756812,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.000788786886827229,
      "loss": 0.4884,
      "num_input_tokens_seen": 22564272,
      "step": 34600
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0007865929228686463,
      "loss": 0.6176,
      "num_input_tokens_seen": 22567280,
      "step": 34605
    },
    {
      "epoch": 18.139412997903563,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0007844019321926688,
      "loss": 0.3573,
      "num_input_tokens_seen": 22571120,
      "step": 34610
    },
    {
      "epoch": 18.14203354297694,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0007822139152576296,
      "loss": 0.3094,
      "num_input_tokens_seen": 22574672,
      "step": 34615
    },
    {
      "epoch": 18.144654088050313,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0007800288725212311,
      "loss": 0.4041,
      "num_input_tokens_seen": 22577616,
      "step": 34620
    },
    {
      "epoch": 18.14727463312369,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0007778468044405645,
      "loss": 0.4002,
      "num_input_tokens_seen": 22579888,
      "step": 34625
    },
    {
      "epoch": 18.149895178197063,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.000775667711472091,
      "loss": 0.3913,
      "num_input_tokens_seen": 22582896,
      "step": 34630
    },
    {
      "epoch": 18.15251572327044,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0007734915940716513,
      "loss": 0.2865,
      "num_input_tokens_seen": 22585616,
      "step": 34635
    },
    {
      "epoch": 18.155136268343817,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0007713184526944694,
      "loss": 0.4406,
      "num_input_tokens_seen": 22588496,
      "step": 34640
    },
    {
      "epoch": 18.15775681341719,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0007691482877951333,
      "loss": 0.3136,
      "num_input_tokens_seen": 22590896,
      "step": 34645
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0007669810998276233,
      "loss": 0.6035,
      "num_input_tokens_seen": 22593104,
      "step": 34650
    },
    {
      "epoch": 18.16299790356394,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0007648168892452866,
      "loss": 0.358,
      "num_input_tokens_seen": 22596656,
      "step": 34655
    },
    {
      "epoch": 18.165618448637318,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0007626556565008563,
      "loss": 0.3327,
      "num_input_tokens_seen": 22599312,
      "step": 34660
    },
    {
      "epoch": 18.16823899371069,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0007604974020464322,
      "loss": 0.2805,
      "num_input_tokens_seen": 22603440,
      "step": 34665
    },
    {
      "epoch": 18.170859538784068,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0007583421263334999,
      "loss": 0.3727,
      "num_input_tokens_seen": 22606512,
      "step": 34670
    },
    {
      "epoch": 18.17348008385744,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0007561898298129154,
      "loss": 0.4269,
      "num_input_tokens_seen": 22609584,
      "step": 34675
    },
    {
      "epoch": 18.17610062893082,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0007540405129349187,
      "loss": 0.3922,
      "num_input_tokens_seen": 22613712,
      "step": 34680
    },
    {
      "epoch": 18.178721174004192,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0007518941761491182,
      "loss": 0.3266,
      "num_input_tokens_seen": 22617008,
      "step": 34685
    },
    {
      "epoch": 18.18134171907757,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0007497508199045066,
      "loss": 0.2492,
      "num_input_tokens_seen": 22620112,
      "step": 34690
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0007476104446494502,
      "loss": 0.2676,
      "num_input_tokens_seen": 22622608,
      "step": 34695
    },
    {
      "epoch": 18.18658280922432,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0007454730508316842,
      "loss": 0.4439,
      "num_input_tokens_seen": 22626416,
      "step": 34700
    },
    {
      "epoch": 18.189203354297693,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0007433386388983343,
      "loss": 0.2636,
      "num_input_tokens_seen": 22629328,
      "step": 34705
    },
    {
      "epoch": 18.19182389937107,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0007412072092958915,
      "loss": 0.4345,
      "num_input_tokens_seen": 22632144,
      "step": 34710
    },
    {
      "epoch": 18.194444444444443,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0007390787624702294,
      "loss": 0.3324,
      "num_input_tokens_seen": 22634992,
      "step": 34715
    },
    {
      "epoch": 18.19706498951782,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0007369532988665933,
      "loss": 0.3424,
      "num_input_tokens_seen": 22638256,
      "step": 34720
    },
    {
      "epoch": 18.199685534591197,
      "grad_norm": 0.09375,
      "learning_rate": 0.0007348308189296026,
      "loss": 0.3583,
      "num_input_tokens_seen": 22641712,
      "step": 34725
    },
    {
      "epoch": 18.20230607966457,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0007327113231032605,
      "loss": 0.3613,
      "num_input_tokens_seen": 22645904,
      "step": 34730
    },
    {
      "epoch": 18.204926624737947,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.000730594811830939,
      "loss": 0.3808,
      "num_input_tokens_seen": 22648944,
      "step": 34735
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.000728481285555389,
      "loss": 0.399,
      "num_input_tokens_seen": 22651888,
      "step": 34740
    },
    {
      "epoch": 18.210167714884697,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0007263707447187383,
      "loss": 0.4015,
      "num_input_tokens_seen": 22655120,
      "step": 34745
    },
    {
      "epoch": 18.21278825995807,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0007242631897624857,
      "loss": 0.6435,
      "num_input_tokens_seen": 22658928,
      "step": 34750
    },
    {
      "epoch": 18.215408805031448,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0007221586211275049,
      "loss": 0.3831,
      "num_input_tokens_seen": 22662288,
      "step": 34755
    },
    {
      "epoch": 18.21802935010482,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0007200570392540506,
      "loss": 0.2808,
      "num_input_tokens_seen": 22664848,
      "step": 34760
    },
    {
      "epoch": 18.220649895178198,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0007179584445817493,
      "loss": 0.2504,
      "num_input_tokens_seen": 22667376,
      "step": 34765
    },
    {
      "epoch": 18.22327044025157,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0007158628375496051,
      "loss": 0.4334,
      "num_input_tokens_seen": 22671440,
      "step": 34770
    },
    {
      "epoch": 18.22589098532495,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0007137702185959937,
      "loss": 0.3297,
      "num_input_tokens_seen": 22675408,
      "step": 34775
    },
    {
      "epoch": 18.228511530398322,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0007116805881586651,
      "loss": 0.2947,
      "num_input_tokens_seen": 22678224,
      "step": 34780
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0007095939466747464,
      "loss": 0.2669,
      "num_input_tokens_seen": 22681712,
      "step": 34785
    },
    {
      "epoch": 18.233752620545072,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0007075102945807415,
      "loss": 0.2293,
      "num_input_tokens_seen": 22685008,
      "step": 34790
    },
    {
      "epoch": 18.23637316561845,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0007054296323125269,
      "loss": 0.3805,
      "num_input_tokens_seen": 22688368,
      "step": 34795
    },
    {
      "epoch": 18.238993710691823,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0007033519603053529,
      "loss": 0.365,
      "num_input_tokens_seen": 22691184,
      "step": 34800
    },
    {
      "epoch": 18.2416142557652,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0007012772789938415,
      "loss": 0.3789,
      "num_input_tokens_seen": 22695024,
      "step": 34805
    },
    {
      "epoch": 18.244234800838573,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0006992055888119957,
      "loss": 0.3339,
      "num_input_tokens_seen": 22699088,
      "step": 34810
    },
    {
      "epoch": 18.24685534591195,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0006971368901931873,
      "loss": 0.3018,
      "num_input_tokens_seen": 22702480,
      "step": 34815
    },
    {
      "epoch": 18.249475890985323,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0006950711835701667,
      "loss": 0.227,
      "num_input_tokens_seen": 22705840,
      "step": 34820
    },
    {
      "epoch": 18.2520964360587,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0006930084693750566,
      "loss": 0.3456,
      "num_input_tokens_seen": 22708752,
      "step": 34825
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0006909487480393534,
      "loss": 0.4201,
      "num_input_tokens_seen": 22711504,
      "step": 34830
    },
    {
      "epoch": 18.25733752620545,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0006888920199939224,
      "loss": 0.4985,
      "num_input_tokens_seen": 22714384,
      "step": 34835
    },
    {
      "epoch": 18.259958071278827,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0006868382856690114,
      "loss": 0.377,
      "num_input_tokens_seen": 22718192,
      "step": 34840
    },
    {
      "epoch": 18.2625786163522,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0006847875454942382,
      "loss": 0.3976,
      "num_input_tokens_seen": 22721616,
      "step": 34845
    },
    {
      "epoch": 18.265199161425578,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0006827397998985945,
      "loss": 0.4157,
      "num_input_tokens_seen": 22724624,
      "step": 34850
    },
    {
      "epoch": 18.26781970649895,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0006806950493104447,
      "loss": 0.4099,
      "num_input_tokens_seen": 22727600,
      "step": 34855
    },
    {
      "epoch": 18.270440251572328,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.000678653294157528,
      "loss": 0.2462,
      "num_input_tokens_seen": 22730608,
      "step": 34860
    },
    {
      "epoch": 18.2730607966457,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0006766145348669544,
      "loss": 0.3461,
      "num_input_tokens_seen": 22733424,
      "step": 34865
    },
    {
      "epoch": 18.27568134171908,
      "grad_norm": 0.0625,
      "learning_rate": 0.0006745787718652097,
      "loss": 0.3419,
      "num_input_tokens_seen": 22737008,
      "step": 34870
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0006725460055781546,
      "loss": 0.335,
      "num_input_tokens_seen": 22740272,
      "step": 34875
    },
    {
      "epoch": 18.28092243186583,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0006705162364310174,
      "loss": 0.3231,
      "num_input_tokens_seen": 22744464,
      "step": 34880
    },
    {
      "epoch": 18.283542976939202,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0006684894648484068,
      "loss": 0.3462,
      "num_input_tokens_seen": 22747152,
      "step": 34885
    },
    {
      "epoch": 18.28616352201258,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0006664656912542954,
      "loss": 0.3457,
      "num_input_tokens_seen": 22751408,
      "step": 34890
    },
    {
      "epoch": 18.288784067085953,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0006644449160720345,
      "loss": 0.3733,
      "num_input_tokens_seen": 22755568,
      "step": 34895
    },
    {
      "epoch": 18.29140461215933,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0006624271397243492,
      "loss": 0.3273,
      "num_input_tokens_seen": 22758576,
      "step": 34900
    },
    {
      "epoch": 18.294025157232703,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0006604123626333369,
      "loss": 0.3862,
      "num_input_tokens_seen": 22762128,
      "step": 34905
    },
    {
      "epoch": 18.29664570230608,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0006584005852204621,
      "loss": 0.353,
      "num_input_tokens_seen": 22764784,
      "step": 34910
    },
    {
      "epoch": 18.299266247379457,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0006563918079065683,
      "loss": 0.3442,
      "num_input_tokens_seen": 22768304,
      "step": 34915
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.000654386031111866,
      "loss": 0.3436,
      "num_input_tokens_seen": 22771632,
      "step": 34920
    },
    {
      "epoch": 18.304507337526207,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0006523832552559428,
      "loss": 0.2784,
      "num_input_tokens_seen": 22774928,
      "step": 34925
    },
    {
      "epoch": 18.30712788259958,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0006503834807577585,
      "loss": 0.2817,
      "num_input_tokens_seen": 22778640,
      "step": 34930
    },
    {
      "epoch": 18.309748427672957,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0006483867080356386,
      "loss": 0.2933,
      "num_input_tokens_seen": 22781936,
      "step": 34935
    },
    {
      "epoch": 18.31236897274633,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0006463929375072874,
      "loss": 0.3042,
      "num_input_tokens_seen": 22784528,
      "step": 34940
    },
    {
      "epoch": 18.314989517819708,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0006444021695897827,
      "loss": 0.273,
      "num_input_tokens_seen": 22790704,
      "step": 34945
    },
    {
      "epoch": 18.31761006289308,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0006424144046995634,
      "loss": 0.3799,
      "num_input_tokens_seen": 22794320,
      "step": 34950
    },
    {
      "epoch": 18.320230607966458,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0006404296432524514,
      "loss": 0.3949,
      "num_input_tokens_seen": 22798544,
      "step": 34955
    },
    {
      "epoch": 18.32285115303983,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.000638447885663635,
      "loss": 0.2873,
      "num_input_tokens_seen": 22801392,
      "step": 34960
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0006364691323476756,
      "loss": 0.3363,
      "num_input_tokens_seen": 22804880,
      "step": 34965
    },
    {
      "epoch": 18.328092243186582,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0006344933837185074,
      "loss": 0.5206,
      "num_input_tokens_seen": 22808112,
      "step": 34970
    },
    {
      "epoch": 18.33071278825996,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0006325206401894312,
      "loss": 0.3825,
      "num_input_tokens_seen": 22814544,
      "step": 34975
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0006305509021731237,
      "loss": 0.281,
      "num_input_tokens_seen": 22817200,
      "step": 34980
    },
    {
      "epoch": 18.33595387840671,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0006285841700816353,
      "loss": 0.269,
      "num_input_tokens_seen": 22820048,
      "step": 34985
    },
    {
      "epoch": 18.338574423480082,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0006266204443263784,
      "loss": 0.2856,
      "num_input_tokens_seen": 22823280,
      "step": 34990
    },
    {
      "epoch": 18.34119496855346,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.000624659725318143,
      "loss": 0.395,
      "num_input_tokens_seen": 22826224,
      "step": 34995
    },
    {
      "epoch": 18.343815513626833,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.000622702013467094,
      "loss": 0.2704,
      "num_input_tokens_seen": 22829360,
      "step": 35000
    },
    {
      "epoch": 18.34643605870021,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0006207473091827558,
      "loss": 0.3323,
      "num_input_tokens_seen": 22832688,
      "step": 35005
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0006187956128740362,
      "loss": 0.3066,
      "num_input_tokens_seen": 22836176,
      "step": 35010
    },
    {
      "epoch": 18.35167714884696,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0006168469249492037,
      "loss": 0.2505,
      "num_input_tokens_seen": 22839088,
      "step": 35015
    },
    {
      "epoch": 18.354297693920337,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.000614901245815902,
      "loss": 0.3401,
      "num_input_tokens_seen": 22842384,
      "step": 35020
    },
    {
      "epoch": 18.35691823899371,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0006129585758811456,
      "loss": 0.3488,
      "num_input_tokens_seen": 22845232,
      "step": 35025
    },
    {
      "epoch": 18.359538784067087,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0006110189155513212,
      "loss": 0.3477,
      "num_input_tokens_seen": 22848016,
      "step": 35030
    },
    {
      "epoch": 18.36215932914046,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0006090822652321809,
      "loss": 0.4473,
      "num_input_tokens_seen": 22850672,
      "step": 35035
    },
    {
      "epoch": 18.364779874213838,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0006071486253288488,
      "loss": 0.2436,
      "num_input_tokens_seen": 22853584,
      "step": 35040
    },
    {
      "epoch": 18.36740041928721,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0006052179962458215,
      "loss": 0.344,
      "num_input_tokens_seen": 22857424,
      "step": 35045
    },
    {
      "epoch": 18.370020964360588,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0006032903783869659,
      "loss": 0.3233,
      "num_input_tokens_seen": 22861456,
      "step": 35050
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 0.08984375,
      "learning_rate": 0.000601365772155516,
      "loss": 0.3274,
      "num_input_tokens_seen": 22864752,
      "step": 35055
    },
    {
      "epoch": 18.37526205450734,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0005994441779540816,
      "loss": 0.3064,
      "num_input_tokens_seen": 22868624,
      "step": 35060
    },
    {
      "epoch": 18.377882599580712,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0005975255961846343,
      "loss": 0.404,
      "num_input_tokens_seen": 22871760,
      "step": 35065
    },
    {
      "epoch": 18.38050314465409,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0005956100272485182,
      "loss": 0.3111,
      "num_input_tokens_seen": 22875248,
      "step": 35070
    },
    {
      "epoch": 18.383123689727462,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0005936974715464494,
      "loss": 0.3974,
      "num_input_tokens_seen": 22878640,
      "step": 35075
    },
    {
      "epoch": 18.38574423480084,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0005917879294785144,
      "loss": 0.3105,
      "num_input_tokens_seen": 22881584,
      "step": 35080
    },
    {
      "epoch": 18.388364779874212,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0005898814014441689,
      "loss": 0.3108,
      "num_input_tokens_seen": 22884368,
      "step": 35085
    },
    {
      "epoch": 18.39098532494759,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0005879778878422353,
      "loss": 0.3956,
      "num_input_tokens_seen": 22886896,
      "step": 35090
    },
    {
      "epoch": 18.393605870020963,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0005860773890709053,
      "loss": 0.4129,
      "num_input_tokens_seen": 22890448,
      "step": 35095
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0005841799055277408,
      "loss": 0.3616,
      "num_input_tokens_seen": 22894224,
      "step": 35100
    },
    {
      "epoch": 18.398846960167713,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0005822854376096775,
      "loss": 0.4386,
      "num_input_tokens_seen": 22897680,
      "step": 35105
    },
    {
      "epoch": 18.40146750524109,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0005803939857130152,
      "loss": 0.3297,
      "num_input_tokens_seen": 22900880,
      "step": 35110
    },
    {
      "epoch": 18.404088050314467,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.000578505550233424,
      "loss": 0.3649,
      "num_input_tokens_seen": 22903984,
      "step": 35115
    },
    {
      "epoch": 18.40670859538784,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0005766201315659413,
      "loss": 0.3846,
      "num_input_tokens_seen": 22907216,
      "step": 35120
    },
    {
      "epoch": 18.409329140461217,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0005747377301049767,
      "loss": 0.4766,
      "num_input_tokens_seen": 22910352,
      "step": 35125
    },
    {
      "epoch": 18.41194968553459,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0005728583462443065,
      "loss": 0.5015,
      "num_input_tokens_seen": 22913776,
      "step": 35130
    },
    {
      "epoch": 18.414570230607968,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0005709819803770765,
      "loss": 0.4899,
      "num_input_tokens_seen": 22917584,
      "step": 35135
    },
    {
      "epoch": 18.41719077568134,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005691086328958029,
      "loss": 0.3327,
      "num_input_tokens_seen": 22920592,
      "step": 35140
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0005672383041923668,
      "loss": 0.3132,
      "num_input_tokens_seen": 22923248,
      "step": 35145
    },
    {
      "epoch": 18.42243186582809,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0005653709946580188,
      "loss": 0.302,
      "num_input_tokens_seen": 22926832,
      "step": 35150
    },
    {
      "epoch": 18.42505241090147,
      "grad_norm": 0.15234375,
      "learning_rate": 0.000563506704683378,
      "loss": 0.3081,
      "num_input_tokens_seen": 22930256,
      "step": 35155
    },
    {
      "epoch": 18.427672955974842,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0005616454346584359,
      "loss": 0.4575,
      "num_input_tokens_seen": 22933552,
      "step": 35160
    },
    {
      "epoch": 18.43029350104822,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0005597871849725494,
      "loss": 0.3121,
      "num_input_tokens_seen": 22936112,
      "step": 35165
    },
    {
      "epoch": 18.432914046121592,
      "grad_norm": 0.1875,
      "learning_rate": 0.0005579319560144408,
      "loss": 0.4452,
      "num_input_tokens_seen": 22939376,
      "step": 35170
    },
    {
      "epoch": 18.43553459119497,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0005560797481722013,
      "loss": 0.3045,
      "num_input_tokens_seen": 22942416,
      "step": 35175
    },
    {
      "epoch": 18.438155136268342,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0005542305618332926,
      "loss": 0.4021,
      "num_input_tokens_seen": 22945520,
      "step": 35180
    },
    {
      "epoch": 18.44077568134172,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0005523843973845455,
      "loss": 0.4145,
      "num_input_tokens_seen": 22948240,
      "step": 35185
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0005505412552121575,
      "loss": 0.4106,
      "num_input_tokens_seen": 22951536,
      "step": 35190
    },
    {
      "epoch": 18.44601677148847,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0005487011357016886,
      "loss": 0.3359,
      "num_input_tokens_seen": 22955280,
      "step": 35195
    },
    {
      "epoch": 18.448637316561843,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0005468640392380741,
      "loss": 0.3774,
      "num_input_tokens_seen": 22958576,
      "step": 35200
    },
    {
      "epoch": 18.45125786163522,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0005450299662056118,
      "loss": 0.3827,
      "num_input_tokens_seen": 22961328,
      "step": 35205
    },
    {
      "epoch": 18.453878406708597,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.000543198916987968,
      "loss": 0.2711,
      "num_input_tokens_seen": 22964880,
      "step": 35210
    },
    {
      "epoch": 18.45649895178197,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0005413708919681798,
      "loss": 0.3339,
      "num_input_tokens_seen": 22968848,
      "step": 35215
    },
    {
      "epoch": 18.459119496855347,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.000539545891528651,
      "loss": 0.4721,
      "num_input_tokens_seen": 22972944,
      "step": 35220
    },
    {
      "epoch": 18.46174004192872,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0005377239160511449,
      "loss": 0.381,
      "num_input_tokens_seen": 22976208,
      "step": 35225
    },
    {
      "epoch": 18.464360587002098,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0005359049659168031,
      "loss": 0.3419,
      "num_input_tokens_seen": 22978832,
      "step": 35230
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0005340890415061261,
      "loss": 0.4098,
      "num_input_tokens_seen": 22982096,
      "step": 35235
    },
    {
      "epoch": 18.469601677148848,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0005322761431989853,
      "loss": 0.5097,
      "num_input_tokens_seen": 22986032,
      "step": 35240
    },
    {
      "epoch": 18.47222222222222,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0005304662713746205,
      "loss": 0.2958,
      "num_input_tokens_seen": 22989168,
      "step": 35245
    },
    {
      "epoch": 18.4748427672956,
      "grad_norm": 0.064453125,
      "learning_rate": 0.000528659426411634,
      "loss": 0.4169,
      "num_input_tokens_seen": 22992784,
      "step": 35250
    },
    {
      "epoch": 18.47746331236897,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0005268556086879982,
      "loss": 0.2712,
      "num_input_tokens_seen": 22995824,
      "step": 35255
    },
    {
      "epoch": 18.48008385744235,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0005250548185810499,
      "loss": 0.4034,
      "num_input_tokens_seen": 22998928,
      "step": 35260
    },
    {
      "epoch": 18.482704402515722,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0005232570564674926,
      "loss": 0.2729,
      "num_input_tokens_seen": 23001424,
      "step": 35265
    },
    {
      "epoch": 18.4853249475891,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0005214623227234039,
      "loss": 0.3035,
      "num_input_tokens_seen": 23004784,
      "step": 35270
    },
    {
      "epoch": 18.487945492662472,
      "grad_norm": 0.109375,
      "learning_rate": 0.000519670617724215,
      "loss": 0.2946,
      "num_input_tokens_seen": 23008048,
      "step": 35275
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0005178819418447311,
      "loss": 0.2567,
      "num_input_tokens_seen": 23010416,
      "step": 35280
    },
    {
      "epoch": 18.493186582809223,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.000516096295459128,
      "loss": 0.3381,
      "num_input_tokens_seen": 23013616,
      "step": 35285
    },
    {
      "epoch": 18.4958071278826,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0005143136789409352,
      "loss": 0.347,
      "num_input_tokens_seen": 23017808,
      "step": 35290
    },
    {
      "epoch": 18.498427672955973,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0005125340926630612,
      "loss": 0.2626,
      "num_input_tokens_seen": 23020720,
      "step": 35295
    },
    {
      "epoch": 18.50104821802935,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0005107575369977729,
      "loss": 0.2302,
      "num_input_tokens_seen": 23024560,
      "step": 35300
    },
    {
      "epoch": 18.503668763102727,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0005089840123167049,
      "loss": 0.3257,
      "num_input_tokens_seen": 23027856,
      "step": 35305
    },
    {
      "epoch": 18.5062893081761,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0005072135189908606,
      "loss": 0.2862,
      "num_input_tokens_seen": 23031024,
      "step": 35310
    },
    {
      "epoch": 18.508909853249477,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0005054460573906067,
      "loss": 0.2751,
      "num_input_tokens_seen": 23034192,
      "step": 35315
    },
    {
      "epoch": 18.51153039832285,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0005036816278856726,
      "loss": 0.3813,
      "num_input_tokens_seen": 23037680,
      "step": 35320
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0005019202308451614,
      "loss": 0.3148,
      "num_input_tokens_seen": 23040784,
      "step": 35325
    },
    {
      "epoch": 18.5167714884696,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0005001618666375335,
      "loss": 0.3402,
      "num_input_tokens_seen": 23043888,
      "step": 35330
    },
    {
      "epoch": 18.519392033542978,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0004984065356306195,
      "loss": 0.3578,
      "num_input_tokens_seen": 23046928,
      "step": 35335
    },
    {
      "epoch": 18.52201257861635,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0004966542381916173,
      "loss": 0.2871,
      "num_input_tokens_seen": 23052016,
      "step": 35340
    },
    {
      "epoch": 18.52463312368973,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0004949049746870837,
      "loss": 0.3188,
      "num_input_tokens_seen": 23055632,
      "step": 35345
    },
    {
      "epoch": 18.5272536687631,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0004931587454829494,
      "loss": 0.3499,
      "num_input_tokens_seen": 23058640,
      "step": 35350
    },
    {
      "epoch": 18.52987421383648,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0004914155509445006,
      "loss": 0.3754,
      "num_input_tokens_seen": 23061488,
      "step": 35355
    },
    {
      "epoch": 18.532494758909852,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0004896753914363955,
      "loss": 0.2664,
      "num_input_tokens_seen": 23064016,
      "step": 35360
    },
    {
      "epoch": 18.53511530398323,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0004879382673226545,
      "loss": 0.3143,
      "num_input_tokens_seen": 23066576,
      "step": 35365
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 0.060546875,
      "learning_rate": 0.00048620417896666875,
      "loss": 0.3956,
      "num_input_tokens_seen": 23069936,
      "step": 35370
    },
    {
      "epoch": 18.54035639412998,
      "grad_norm": 0.083984375,
      "learning_rate": 0.00048447312673118633,
      "loss": 0.6297,
      "num_input_tokens_seen": 23073392,
      "step": 35375
    },
    {
      "epoch": 18.542976939203353,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00048274511097832427,
      "loss": 0.3624,
      "num_input_tokens_seen": 23079568,
      "step": 35380
    },
    {
      "epoch": 18.54559748427673,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004810201320695617,
      "loss": 0.2644,
      "num_input_tokens_seen": 23082480,
      "step": 35385
    },
    {
      "epoch": 18.548218029350103,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.00047929819036574505,
      "loss": 0.2942,
      "num_input_tokens_seen": 23085328,
      "step": 35390
    },
    {
      "epoch": 18.55083857442348,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00047757928622708777,
      "loss": 0.3768,
      "num_input_tokens_seen": 23088208,
      "step": 35395
    },
    {
      "epoch": 18.553459119496857,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.00047586342001316393,
      "loss": 0.2681,
      "num_input_tokens_seen": 23091376,
      "step": 35400
    },
    {
      "epoch": 18.55607966457023,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0004741505920829131,
      "loss": 0.4002,
      "num_input_tokens_seen": 23094832,
      "step": 35405
    },
    {
      "epoch": 18.558700209643607,
      "grad_norm": 0.078125,
      "learning_rate": 0.00047244080279463694,
      "loss": 0.3076,
      "num_input_tokens_seen": 23097808,
      "step": 35410
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.00047073405250600605,
      "loss": 0.4253,
      "num_input_tokens_seen": 23100464,
      "step": 35415
    },
    {
      "epoch": 18.563941299790358,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00046903034157405154,
      "loss": 0.2877,
      "num_input_tokens_seen": 23103760,
      "step": 35420
    },
    {
      "epoch": 18.56656184486373,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0004673296703551732,
      "loss": 0.3468,
      "num_input_tokens_seen": 23106832,
      "step": 35425
    },
    {
      "epoch": 18.569182389937108,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0004656320392051333,
      "loss": 0.4086,
      "num_input_tokens_seen": 23110832,
      "step": 35430
    },
    {
      "epoch": 18.57180293501048,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.00046393744847904924,
      "loss": 0.2344,
      "num_input_tokens_seen": 23113968,
      "step": 35435
    },
    {
      "epoch": 18.57442348008386,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.00046224589853141596,
      "loss": 0.3383,
      "num_input_tokens_seen": 23116944,
      "step": 35440
    },
    {
      "epoch": 18.57704402515723,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.00046055738971608525,
      "loss": 0.4199,
      "num_input_tokens_seen": 23120112,
      "step": 35445
    },
    {
      "epoch": 18.57966457023061,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00045887192238627307,
      "loss": 0.421,
      "num_input_tokens_seen": 23124176,
      "step": 35450
    },
    {
      "epoch": 18.582285115303982,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0004571894968945639,
      "loss": 0.3121,
      "num_input_tokens_seen": 23126832,
      "step": 35455
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.00045551011359289795,
      "loss": 0.3513,
      "num_input_tokens_seen": 23129904,
      "step": 35460
    },
    {
      "epoch": 18.587526205450732,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00045383377283258087,
      "loss": 0.3363,
      "num_input_tokens_seen": 23133200,
      "step": 35465
    },
    {
      "epoch": 18.59014675052411,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.00045216047496428877,
      "loss": 0.3027,
      "num_input_tokens_seen": 23136496,
      "step": 35470
    },
    {
      "epoch": 18.592767295597483,
      "grad_norm": 0.078125,
      "learning_rate": 0.00045049022033805504,
      "loss": 0.2652,
      "num_input_tokens_seen": 23139472,
      "step": 35475
    },
    {
      "epoch": 18.59538784067086,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0004488230093032785,
      "loss": 0.3176,
      "num_input_tokens_seen": 23142384,
      "step": 35480
    },
    {
      "epoch": 18.598008385744233,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.00044715884220872026,
      "loss": 0.3517,
      "num_input_tokens_seen": 23145328,
      "step": 35485
    },
    {
      "epoch": 18.60062893081761,
      "grad_norm": 0.109375,
      "learning_rate": 0.0004454977194025017,
      "loss": 0.3575,
      "num_input_tokens_seen": 23148912,
      "step": 35490
    },
    {
      "epoch": 18.603249475890987,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.00044383964123211495,
      "loss": 0.3561,
      "num_input_tokens_seen": 23151632,
      "step": 35495
    },
    {
      "epoch": 18.60587002096436,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0004421846080444108,
      "loss": 0.3223,
      "num_input_tokens_seen": 23154032,
      "step": 35500
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0004405326201856008,
      "loss": 0.377,
      "num_input_tokens_seen": 23156976,
      "step": 35505
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.12890625,
      "learning_rate": 0.00043888367800126504,
      "loss": 0.3329,
      "num_input_tokens_seen": 23159472,
      "step": 35510
    },
    {
      "epoch": 18.613731656184488,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0004372377818363426,
      "loss": 0.4128,
      "num_input_tokens_seen": 23162672,
      "step": 35515
    },
    {
      "epoch": 18.61635220125786,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0004355949320351332,
      "loss": 0.3785,
      "num_input_tokens_seen": 23165456,
      "step": 35520
    },
    {
      "epoch": 18.618972746331238,
      "grad_norm": 0.119140625,
      "learning_rate": 0.00043395512894130514,
      "loss": 0.3422,
      "num_input_tokens_seen": 23168656,
      "step": 35525
    },
    {
      "epoch": 18.62159329140461,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00043231837289788407,
      "loss": 0.3543,
      "num_input_tokens_seen": 23171856,
      "step": 35530
    },
    {
      "epoch": 18.62421383647799,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.00043068466424726616,
      "loss": 0.3153,
      "num_input_tokens_seen": 23175856,
      "step": 35535
    },
    {
      "epoch": 18.62683438155136,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0004290540033311996,
      "loss": 0.3927,
      "num_input_tokens_seen": 23178928,
      "step": 35540
    },
    {
      "epoch": 18.62945492662474,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.00042742639049080154,
      "loss": 0.4115,
      "num_input_tokens_seen": 23182384,
      "step": 35545
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 0.099609375,
      "learning_rate": 0.000425801826066548,
      "loss": 0.2443,
      "num_input_tokens_seen": 23186192,
      "step": 35550
    },
    {
      "epoch": 18.63469601677149,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0004241803103982822,
      "loss": 0.3794,
      "num_input_tokens_seen": 23191056,
      "step": 35555
    },
    {
      "epoch": 18.637316561844862,
      "grad_norm": 0.111328125,
      "learning_rate": 0.00042256184382520763,
      "loss": 0.3654,
      "num_input_tokens_seen": 23194160,
      "step": 35560
    },
    {
      "epoch": 18.63993710691824,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0004209464266858853,
      "loss": 0.3324,
      "num_input_tokens_seen": 23197392,
      "step": 35565
    },
    {
      "epoch": 18.642557651991613,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00041933405931824486,
      "loss": 0.3234,
      "num_input_tokens_seen": 23201008,
      "step": 35570
    },
    {
      "epoch": 18.64517819706499,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0004177247420595731,
      "loss": 0.3033,
      "num_input_tokens_seen": 23208176,
      "step": 35575
    },
    {
      "epoch": 18.647798742138363,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0004161184752465208,
      "loss": 0.3204,
      "num_input_tokens_seen": 23210768,
      "step": 35580
    },
    {
      "epoch": 18.65041928721174,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0004145152592151041,
      "loss": 0.5507,
      "num_input_tokens_seen": 23214480,
      "step": 35585
    },
    {
      "epoch": 18.653039832285117,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.00041291509430069483,
      "loss": 0.3226,
      "num_input_tokens_seen": 23217296,
      "step": 35590
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0004113179808380285,
      "loss": 0.2881,
      "num_input_tokens_seen": 23220272,
      "step": 35595
    },
    {
      "epoch": 18.658280922431867,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0004097239191612062,
      "loss": 0.2962,
      "num_input_tokens_seen": 23224912,
      "step": 35600
    },
    {
      "epoch": 18.66090146750524,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00040813290960368286,
      "loss": 0.2746,
      "num_input_tokens_seen": 23227600,
      "step": 35605
    },
    {
      "epoch": 18.663522012578618,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00040654495249828224,
      "loss": 0.4783,
      "num_input_tokens_seen": 23230576,
      "step": 35610
    },
    {
      "epoch": 18.66614255765199,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.00040496004817718864,
      "loss": 0.4227,
      "num_input_tokens_seen": 23233360,
      "step": 35615
    },
    {
      "epoch": 18.668763102725368,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0004033781969719419,
      "loss": 0.5332,
      "num_input_tokens_seen": 23237328,
      "step": 35620
    },
    {
      "epoch": 18.67138364779874,
      "grad_norm": 0.083984375,
      "learning_rate": 0.00040179939921345054,
      "loss": 0.3122,
      "num_input_tokens_seen": 23240592,
      "step": 35625
    },
    {
      "epoch": 18.67400419287212,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00040022365523197876,
      "loss": 0.3332,
      "num_input_tokens_seen": 23243472,
      "step": 35630
    },
    {
      "epoch": 18.67662473794549,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00039865096535715626,
      "loss": 0.3027,
      "num_input_tokens_seen": 23246032,
      "step": 35635
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0003970813299179715,
      "loss": 0.2786,
      "num_input_tokens_seen": 23248304,
      "step": 35640
    },
    {
      "epoch": 18.681865828092242,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.00039551474924277185,
      "loss": 0.3557,
      "num_input_tokens_seen": 23251632,
      "step": 35645
    },
    {
      "epoch": 18.68448637316562,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.00039395122365927016,
      "loss": 0.332,
      "num_input_tokens_seen": 23255216,
      "step": 35650
    },
    {
      "epoch": 18.687106918238992,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0003923907534945381,
      "loss": 0.4087,
      "num_input_tokens_seen": 23258736,
      "step": 35655
    },
    {
      "epoch": 18.68972746331237,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0003908333390750079,
      "loss": 0.292,
      "num_input_tokens_seen": 23261872,
      "step": 35660
    },
    {
      "epoch": 18.692348008385743,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.00038927898072647237,
      "loss": 0.4485,
      "num_input_tokens_seen": 23265008,
      "step": 35665
    },
    {
      "epoch": 18.69496855345912,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.00038772767877408797,
      "loss": 0.3324,
      "num_input_tokens_seen": 23268912,
      "step": 35670
    },
    {
      "epoch": 18.697589098532493,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.00038617943354236683,
      "loss": 0.1882,
      "num_input_tokens_seen": 23271344,
      "step": 35675
    },
    {
      "epoch": 18.70020964360587,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0003846342453551832,
      "loss": 0.3565,
      "num_input_tokens_seen": 23274960,
      "step": 35680
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 0.111328125,
      "learning_rate": 0.00038309211453577683,
      "loss": 0.3851,
      "num_input_tokens_seen": 23278160,
      "step": 35685
    },
    {
      "epoch": 18.70545073375262,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0003815530414067414,
      "loss": 0.3899,
      "num_input_tokens_seen": 23282192,
      "step": 35690
    },
    {
      "epoch": 18.708071278825997,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0003800170262900326,
      "loss": 0.3633,
      "num_input_tokens_seen": 23285648,
      "step": 35695
    },
    {
      "epoch": 18.71069182389937,
      "grad_norm": 0.2177734375,
      "learning_rate": 0.00037848406950696854,
      "loss": 0.534,
      "num_input_tokens_seen": 23288304,
      "step": 35700
    },
    {
      "epoch": 18.713312368972748,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.00037695417137822594,
      "loss": 0.3416,
      "num_input_tokens_seen": 23291408,
      "step": 35705
    },
    {
      "epoch": 18.71593291404612,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0003754273322238438,
      "loss": 0.3352,
      "num_input_tokens_seen": 23294448,
      "step": 35710
    },
    {
      "epoch": 18.718553459119498,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.00037390355236321504,
      "loss": 0.3686,
      "num_input_tokens_seen": 23297232,
      "step": 35715
    },
    {
      "epoch": 18.72117400419287,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.00037238283211510294,
      "loss": 0.4048,
      "num_input_tokens_seen": 23301264,
      "step": 35720
    },
    {
      "epoch": 18.72379454926625,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.00037086517179761965,
      "loss": 0.3828,
      "num_input_tokens_seen": 23304720,
      "step": 35725
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 0.046875,
      "learning_rate": 0.00036935057172824304,
      "loss": 0.3827,
      "num_input_tokens_seen": 23308592,
      "step": 35730
    },
    {
      "epoch": 18.729035639413,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0003678390322238129,
      "loss": 0.7767,
      "num_input_tokens_seen": 23312080,
      "step": 35735
    },
    {
      "epoch": 18.731656184486372,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0003663305536005262,
      "loss": 0.5679,
      "num_input_tokens_seen": 23314800,
      "step": 35740
    },
    {
      "epoch": 18.73427672955975,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0003648251361739374,
      "loss": 0.4555,
      "num_input_tokens_seen": 23317840,
      "step": 35745
    },
    {
      "epoch": 18.736897274633122,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0003633227802589628,
      "loss": 0.3151,
      "num_input_tokens_seen": 23321200,
      "step": 35750
    },
    {
      "epoch": 18.7395178197065,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0003618234861698777,
      "loss": 0.3962,
      "num_input_tokens_seen": 23323824,
      "step": 35755
    },
    {
      "epoch": 18.742138364779873,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00036032725422031783,
      "loss": 0.3554,
      "num_input_tokens_seen": 23326928,
      "step": 35760
    },
    {
      "epoch": 18.74475890985325,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0003588340847232796,
      "loss": 0.4516,
      "num_input_tokens_seen": 23330576,
      "step": 35765
    },
    {
      "epoch": 18.747379454926623,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.00035734397799111635,
      "loss": 0.4119,
      "num_input_tokens_seen": 23334832,
      "step": 35770
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00035585693433554057,
      "loss": 0.2795,
      "num_input_tokens_seen": 23337488,
      "step": 35775
    },
    {
      "epoch": 18.752620545073377,
      "grad_norm": 0.076171875,
      "learning_rate": 0.00035437295406762323,
      "loss": 0.2809,
      "num_input_tokens_seen": 23341296,
      "step": 35780
    },
    {
      "epoch": 18.75524109014675,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.00035289203749779783,
      "loss": 0.4165,
      "num_input_tokens_seen": 23348144,
      "step": 35785
    },
    {
      "epoch": 18.757861635220127,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00035141418493585486,
      "loss": 0.3471,
      "num_input_tokens_seen": 23352080,
      "step": 35790
    },
    {
      "epoch": 18.7604821802935,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0003499393966909453,
      "loss": 0.2228,
      "num_input_tokens_seen": 23355536,
      "step": 35795
    },
    {
      "epoch": 18.763102725366878,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0003484676730715791,
      "loss": 0.2622,
      "num_input_tokens_seen": 23358160,
      "step": 35800
    },
    {
      "epoch": 18.76572327044025,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.00034699901438561994,
      "loss": 0.454,
      "num_input_tokens_seen": 23361296,
      "step": 35805
    },
    {
      "epoch": 18.768343815513628,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00034553342094029545,
      "loss": 0.3453,
      "num_input_tokens_seen": 23365264,
      "step": 35810
    },
    {
      "epoch": 18.770964360587,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0003440708930421937,
      "loss": 0.2631,
      "num_input_tokens_seen": 23368208,
      "step": 35815
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 0.087890625,
      "learning_rate": 0.00034261143099725665,
      "loss": 0.3726,
      "num_input_tokens_seen": 23371248,
      "step": 35820
    },
    {
      "epoch": 18.77620545073375,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.00034115503511079003,
      "loss": 0.2797,
      "num_input_tokens_seen": 23374960,
      "step": 35825
    },
    {
      "epoch": 18.77882599580713,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0003397017056874535,
      "loss": 0.2842,
      "num_input_tokens_seen": 23378832,
      "step": 35830
    },
    {
      "epoch": 18.781446540880502,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0003382514430312638,
      "loss": 0.4303,
      "num_input_tokens_seen": 23382032,
      "step": 35835
    },
    {
      "epoch": 18.78406708595388,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0003368042474456034,
      "loss": 0.3422,
      "num_input_tokens_seen": 23385456,
      "step": 35840
    },
    {
      "epoch": 18.786687631027252,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.00033536011923320674,
      "loss": 0.3792,
      "num_input_tokens_seen": 23388592,
      "step": 35845
    },
    {
      "epoch": 18.78930817610063,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0003339190586961721,
      "loss": 0.3162,
      "num_input_tokens_seen": 23391824,
      "step": 35850
    },
    {
      "epoch": 18.791928721174003,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0003324810661359517,
      "loss": 0.3436,
      "num_input_tokens_seen": 23394896,
      "step": 35855
    },
    {
      "epoch": 18.79454926624738,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0003310461418533533,
      "loss": 0.3124,
      "num_input_tokens_seen": 23399888,
      "step": 35860
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0003296142861485518,
      "loss": 0.3246,
      "num_input_tokens_seen": 23402832,
      "step": 35865
    },
    {
      "epoch": 18.79979035639413,
      "grad_norm": 0.078125,
      "learning_rate": 0.00032818549932107086,
      "loss": 0.4719,
      "num_input_tokens_seen": 23405968,
      "step": 35870
    },
    {
      "epoch": 18.802410901467507,
      "grad_norm": 0.109375,
      "learning_rate": 0.00032675978166980146,
      "loss": 0.4647,
      "num_input_tokens_seen": 23408656,
      "step": 35875
    },
    {
      "epoch": 18.80503144654088,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0003253371334929833,
      "loss": 0.466,
      "num_input_tokens_seen": 23411952,
      "step": 35880
    },
    {
      "epoch": 18.807651991614257,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.00032391755508822,
      "loss": 0.3773,
      "num_input_tokens_seen": 23414896,
      "step": 35885
    },
    {
      "epoch": 18.81027253668763,
      "grad_norm": 0.111328125,
      "learning_rate": 0.00032250104675246913,
      "loss": 0.4504,
      "num_input_tokens_seen": 23418224,
      "step": 35890
    },
    {
      "epoch": 18.812893081761008,
      "grad_norm": 0.125,
      "learning_rate": 0.0003210876087820502,
      "loss": 0.3078,
      "num_input_tokens_seen": 23421008,
      "step": 35895
    },
    {
      "epoch": 18.81551362683438,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00031967724147263674,
      "loss": 0.3043,
      "num_input_tokens_seen": 23423792,
      "step": 35900
    },
    {
      "epoch": 18.818134171907758,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0003182699451192644,
      "loss": 0.3964,
      "num_input_tokens_seen": 23426320,
      "step": 35905
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0003168657200163211,
      "loss": 0.4204,
      "num_input_tokens_seen": 23429840,
      "step": 35910
    },
    {
      "epoch": 18.82337526205451,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0003154645664575534,
      "loss": 0.2733,
      "num_input_tokens_seen": 23432720,
      "step": 35915
    },
    {
      "epoch": 18.82599580712788,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.00031406648473607024,
      "loss": 0.2924,
      "num_input_tokens_seen": 23436592,
      "step": 35920
    },
    {
      "epoch": 18.82861635220126,
      "grad_norm": 0.130859375,
      "learning_rate": 0.000312671475144331,
      "loss": 0.2708,
      "num_input_tokens_seen": 23439856,
      "step": 35925
    },
    {
      "epoch": 18.831236897274632,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.00031127953797415895,
      "loss": 0.2849,
      "num_input_tokens_seen": 23442832,
      "step": 35930
    },
    {
      "epoch": 18.83385744234801,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0003098906735167278,
      "loss": 0.2815,
      "num_input_tokens_seen": 23446416,
      "step": 35935
    },
    {
      "epoch": 18.836477987421382,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0003085048820625752,
      "loss": 0.3226,
      "num_input_tokens_seen": 23449104,
      "step": 35940
    },
    {
      "epoch": 18.83909853249476,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0003071221639015925,
      "loss": 0.3997,
      "num_input_tokens_seen": 23452272,
      "step": 35945
    },
    {
      "epoch": 18.841719077568133,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0003057425193230251,
      "loss": 0.3527,
      "num_input_tokens_seen": 23455088,
      "step": 35950
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0003043659486154854,
      "loss": 0.4055,
      "num_input_tokens_seen": 23458704,
      "step": 35955
    },
    {
      "epoch": 18.846960167714883,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0003029924520669297,
      "loss": 0.3721,
      "num_input_tokens_seen": 23462704,
      "step": 35960
    },
    {
      "epoch": 18.84958071278826,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.00030162202996468154,
      "loss": 0.2772,
      "num_input_tokens_seen": 23465328,
      "step": 35965
    },
    {
      "epoch": 18.852201257861637,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0003002546825954183,
      "loss": 0.3989,
      "num_input_tokens_seen": 23468912,
      "step": 35970
    },
    {
      "epoch": 18.85482180293501,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0002988904102451711,
      "loss": 0.3289,
      "num_input_tokens_seen": 23471984,
      "step": 35975
    },
    {
      "epoch": 18.857442348008387,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0002975292131993301,
      "loss": 0.2884,
      "num_input_tokens_seen": 23475888,
      "step": 35980
    },
    {
      "epoch": 18.86006289308176,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0002961710917426441,
      "loss": 0.3265,
      "num_input_tokens_seen": 23478704,
      "step": 35985
    },
    {
      "epoch": 18.862683438155138,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0002948160461592142,
      "loss": 0.3095,
      "num_input_tokens_seen": 23481136,
      "step": 35990
    },
    {
      "epoch": 18.86530398322851,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0002934640767325036,
      "loss": 0.2695,
      "num_input_tokens_seen": 23483408,
      "step": 35995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.00029211518374532616,
      "loss": 0.3577,
      "num_input_tokens_seen": 23487568,
      "step": 36000
    },
    {
      "epoch": 18.87054507337526,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.00029076936747985446,
      "loss": 0.4181,
      "num_input_tokens_seen": 23490512,
      "step": 36005
    },
    {
      "epoch": 18.87316561844864,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.00028942662821762166,
      "loss": 0.4142,
      "num_input_tokens_seen": 23493680,
      "step": 36010
    },
    {
      "epoch": 18.87578616352201,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0002880869662395097,
      "loss": 0.3458,
      "num_input_tokens_seen": 23497680,
      "step": 36015
    },
    {
      "epoch": 18.87840670859539,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.00028675038182576274,
      "loss": 0.3471,
      "num_input_tokens_seen": 23500464,
      "step": 36020
    },
    {
      "epoch": 18.881027253668762,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0002854168752559788,
      "loss": 0.2943,
      "num_input_tokens_seen": 23503184,
      "step": 36025
    },
    {
      "epoch": 18.88364779874214,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00028408644680910975,
      "loss": 0.3752,
      "num_input_tokens_seen": 23506064,
      "step": 36030
    },
    {
      "epoch": 18.886268343815512,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0002827590967634696,
      "loss": 0.3729,
      "num_input_tokens_seen": 23508848,
      "step": 36035
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.130859375,
      "learning_rate": 0.00028143482539672303,
      "loss": 0.4292,
      "num_input_tokens_seen": 23512432,
      "step": 36040
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 0.296875,
      "learning_rate": 0.00028011363298589164,
      "loss": 0.4191,
      "num_input_tokens_seen": 23515312,
      "step": 36045
    },
    {
      "epoch": 18.89412997903564,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.00027879551980735604,
      "loss": 0.4163,
      "num_input_tokens_seen": 23519024,
      "step": 36050
    },
    {
      "epoch": 18.896750524109013,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00027748048613684907,
      "loss": 0.272,
      "num_input_tokens_seen": 23522032,
      "step": 36055
    },
    {
      "epoch": 18.89937106918239,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00027616853224946057,
      "loss": 0.4364,
      "num_input_tokens_seen": 23524912,
      "step": 36060
    },
    {
      "epoch": 18.901991614255767,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0002748596584196394,
      "loss": 0.241,
      "num_input_tokens_seen": 23527984,
      "step": 36065
    },
    {
      "epoch": 18.90461215932914,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0002735538649211816,
      "loss": 0.4071,
      "num_input_tokens_seen": 23531792,
      "step": 36070
    },
    {
      "epoch": 18.907232704402517,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0002722511520272469,
      "loss": 0.3422,
      "num_input_tokens_seen": 23535504,
      "step": 36075
    },
    {
      "epoch": 18.90985324947589,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00027095152001034903,
      "loss": 0.3416,
      "num_input_tokens_seen": 23538864,
      "step": 36080
    },
    {
      "epoch": 18.912473794549268,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.00026965496914235554,
      "loss": 0.3937,
      "num_input_tokens_seen": 23542736,
      "step": 36085
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0002683614996944894,
      "loss": 0.4136,
      "num_input_tokens_seen": 23547504,
      "step": 36090
    },
    {
      "epoch": 18.917714884696018,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0002670711119373309,
      "loss": 0.3228,
      "num_input_tokens_seen": 23550224,
      "step": 36095
    },
    {
      "epoch": 18.92033542976939,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00026578380614081243,
      "loss": 0.41,
      "num_input_tokens_seen": 23553488,
      "step": 36100
    },
    {
      "epoch": 18.92295597484277,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0002644995825742252,
      "loss": 0.4443,
      "num_input_tokens_seen": 23556816,
      "step": 36105
    },
    {
      "epoch": 18.92557651991614,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0002632184415062144,
      "loss": 0.2323,
      "num_input_tokens_seen": 23559280,
      "step": 36110
    },
    {
      "epoch": 18.92819706498952,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0002619403832047806,
      "loss": 0.4367,
      "num_input_tokens_seen": 23562416,
      "step": 36115
    },
    {
      "epoch": 18.930817610062892,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.000260665407937275,
      "loss": 0.289,
      "num_input_tokens_seen": 23566032,
      "step": 36120
    },
    {
      "epoch": 18.93343815513627,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.00025939351597041246,
      "loss": 0.4043,
      "num_input_tokens_seen": 23568880,
      "step": 36125
    },
    {
      "epoch": 18.936058700209642,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0002581247075702569,
      "loss": 0.3029,
      "num_input_tokens_seen": 23571216,
      "step": 36130
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.00025685898300222774,
      "loss": 0.4551,
      "num_input_tokens_seen": 23575216,
      "step": 36135
    },
    {
      "epoch": 18.941299790356393,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00025559634253110143,
      "loss": 0.3332,
      "num_input_tokens_seen": 23578576,
      "step": 36140
    },
    {
      "epoch": 18.94392033542977,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.00025433678642100663,
      "loss": 0.3199,
      "num_input_tokens_seen": 23580848,
      "step": 36145
    },
    {
      "epoch": 18.946540880503143,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0002530803149354294,
      "loss": 0.3807,
      "num_input_tokens_seen": 23583440,
      "step": 36150
    },
    {
      "epoch": 18.94916142557652,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.00025182692833720764,
      "loss": 0.2632,
      "num_input_tokens_seen": 23586640,
      "step": 36155
    },
    {
      "epoch": 18.951781970649897,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00025057662688853676,
      "loss": 0.3199,
      "num_input_tokens_seen": 23589968,
      "step": 36160
    },
    {
      "epoch": 18.95440251572327,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.00024932941085096584,
      "loss": 0.2903,
      "num_input_tokens_seen": 23592368,
      "step": 36165
    },
    {
      "epoch": 18.957023060796647,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0002480852804853978,
      "loss": 0.3171,
      "num_input_tokens_seen": 23595248,
      "step": 36170
    },
    {
      "epoch": 18.95964360587002,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.00024684423605208784,
      "loss": 0.3702,
      "num_input_tokens_seen": 23598384,
      "step": 36175
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0002456062778106532,
      "loss": 0.4235,
      "num_input_tokens_seen": 23602128,
      "step": 36180
    },
    {
      "epoch": 18.96488469601677,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0002443714060200569,
      "loss": 0.2744,
      "num_input_tokens_seen": 23605232,
      "step": 36185
    },
    {
      "epoch": 18.967505241090148,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.00024313962093862218,
      "loss": 0.288,
      "num_input_tokens_seen": 23608272,
      "step": 36190
    },
    {
      "epoch": 18.97012578616352,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0002419109228240246,
      "loss": 0.3041,
      "num_input_tokens_seen": 23610576,
      "step": 36195
    },
    {
      "epoch": 18.9727463312369,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.00024068531193329024,
      "loss": 0.3863,
      "num_input_tokens_seen": 23614000,
      "step": 36200
    },
    {
      "epoch": 18.97536687631027,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.00023946278852280732,
      "loss": 0.2961,
      "num_input_tokens_seen": 23617264,
      "step": 36205
    },
    {
      "epoch": 18.97798742138365,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.00023824335284831132,
      "loss": 0.379,
      "num_input_tokens_seen": 23620240,
      "step": 36210
    },
    {
      "epoch": 18.980607966457022,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.00023702700516489315,
      "loss": 0.3114,
      "num_input_tokens_seen": 23623216,
      "step": 36215
    },
    {
      "epoch": 18.9832285115304,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0002358137457270043,
      "loss": 0.3173,
      "num_input_tokens_seen": 23626064,
      "step": 36220
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.00023460357478844007,
      "loss": 0.3942,
      "num_input_tokens_seen": 23629360,
      "step": 36225
    },
    {
      "epoch": 18.98846960167715,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00023339649260235295,
      "loss": 0.3829,
      "num_input_tokens_seen": 23631792,
      "step": 36230
    },
    {
      "epoch": 18.991090146750523,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00023219249942125597,
      "loss": 0.3088,
      "num_input_tokens_seen": 23636400,
      "step": 36235
    },
    {
      "epoch": 18.9937106918239,
      "grad_norm": 0.111328125,
      "learning_rate": 0.00023099159549700598,
      "loss": 0.3298,
      "num_input_tokens_seen": 23639056,
      "step": 36240
    },
    {
      "epoch": 18.996331236897273,
      "grad_norm": 0.099609375,
      "learning_rate": 0.00022979378108082204,
      "loss": 0.2882,
      "num_input_tokens_seen": 23642672,
      "step": 36245
    },
    {
      "epoch": 18.99895178197065,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.00022859905642327036,
      "loss": 0.3265,
      "num_input_tokens_seen": 23644848,
      "step": 36250
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.5051435828208923,
      "eval_runtime": 13.6899,
      "eval_samples_per_second": 61.944,
      "eval_steps_per_second": 15.486,
      "num_input_tokens_seen": 23645440,
      "step": 36252
    },
    {
      "epoch": 19.001572327044027,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.00022740742177427775,
      "loss": 0.2365,
      "num_input_tokens_seen": 23647040,
      "step": 36255
    },
    {
      "epoch": 19.0041928721174,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.00022621887738311474,
      "loss": 0.3468,
      "num_input_tokens_seen": 23649760,
      "step": 36260
    },
    {
      "epoch": 19.006813417190777,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0002250334234984158,
      "loss": 0.2705,
      "num_input_tokens_seen": 23652672,
      "step": 36265
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0002238510603681626,
      "loss": 0.3259,
      "num_input_tokens_seen": 23655584,
      "step": 36270
    },
    {
      "epoch": 19.012054507337528,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00022267178823969224,
      "loss": 0.399,
      "num_input_tokens_seen": 23658496,
      "step": 36275
    },
    {
      "epoch": 19.0146750524109,
      "grad_norm": 0.107421875,
      "learning_rate": 0.00022149560735969576,
      "loss": 0.2765,
      "num_input_tokens_seen": 23660928,
      "step": 36280
    },
    {
      "epoch": 19.017295597484278,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.00022032251797421464,
      "loss": 0.5208,
      "num_input_tokens_seen": 23664448,
      "step": 36285
    },
    {
      "epoch": 19.01991614255765,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00021915252032864927,
      "loss": 0.4726,
      "num_input_tokens_seen": 23668128,
      "step": 36290
    },
    {
      "epoch": 19.02253668763103,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00021798561466774557,
      "loss": 0.3987,
      "num_input_tokens_seen": 23672128,
      "step": 36295
    },
    {
      "epoch": 19.0251572327044,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00021682180123561157,
      "loss": 0.3225,
      "num_input_tokens_seen": 23675392,
      "step": 36300
    },
    {
      "epoch": 19.02777777777778,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0002156610802756992,
      "loss": 0.4139,
      "num_input_tokens_seen": 23679136,
      "step": 36305
    },
    {
      "epoch": 19.030398322851152,
      "grad_norm": 0.216796875,
      "learning_rate": 0.00021450345203082255,
      "loss": 0.3481,
      "num_input_tokens_seen": 23681920,
      "step": 36310
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0002133489167431396,
      "loss": 0.3098,
      "num_input_tokens_seen": 23685824,
      "step": 36315
    },
    {
      "epoch": 19.035639412997902,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.00021219747465416882,
      "loss": 0.388,
      "num_input_tokens_seen": 23689344,
      "step": 36320
    },
    {
      "epoch": 19.03825995807128,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0002110491260047792,
      "loss": 0.2769,
      "num_input_tokens_seen": 23692384,
      "step": 36325
    },
    {
      "epoch": 19.040880503144653,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.00020990387103519026,
      "loss": 0.3326,
      "num_input_tokens_seen": 23695936,
      "step": 36330
    },
    {
      "epoch": 19.04350104821803,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.000208761709984977,
      "loss": 0.3707,
      "num_input_tokens_seen": 23698848,
      "step": 36335
    },
    {
      "epoch": 19.046121593291403,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00020762264309306833,
      "loss": 0.318,
      "num_input_tokens_seen": 23702464,
      "step": 36340
    },
    {
      "epoch": 19.04874213836478,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0002064866705977403,
      "loss": 0.2838,
      "num_input_tokens_seen": 23705344,
      "step": 36345
    },
    {
      "epoch": 19.051362683438157,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.00020535379273662778,
      "loss": 0.4086,
      "num_input_tokens_seen": 23708992,
      "step": 36350
    },
    {
      "epoch": 19.05398322851153,
      "grad_norm": 0.1015625,
      "learning_rate": 0.00020422400974671627,
      "loss": 0.2665,
      "num_input_tokens_seen": 23711808,
      "step": 36355
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0002030973218643417,
      "loss": 0.4451,
      "num_input_tokens_seen": 23714720,
      "step": 36360
    },
    {
      "epoch": 19.05922431865828,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0002019737293251972,
      "loss": 0.4347,
      "num_input_tokens_seen": 23717984,
      "step": 36365
    },
    {
      "epoch": 19.061844863731658,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00020085323236432306,
      "loss": 0.3838,
      "num_input_tokens_seen": 23721184,
      "step": 36370
    },
    {
      "epoch": 19.06446540880503,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00019973583121611347,
      "loss": 0.3786,
      "num_input_tokens_seen": 23724384,
      "step": 36375
    },
    {
      "epoch": 19.067085953878408,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.00019862152611431982,
      "loss": 0.4004,
      "num_input_tokens_seen": 23727680,
      "step": 36380
    },
    {
      "epoch": 19.06970649895178,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00019751031729203894,
      "loss": 0.435,
      "num_input_tokens_seen": 23730464,
      "step": 36385
    },
    {
      "epoch": 19.072327044025158,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00019640220498172323,
      "loss": 0.2532,
      "num_input_tokens_seen": 23733888,
      "step": 36390
    },
    {
      "epoch": 19.07494758909853,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.00019529718941517892,
      "loss": 0.349,
      "num_input_tokens_seen": 23736736,
      "step": 36395
    },
    {
      "epoch": 19.07756813417191,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00019419527082356112,
      "loss": 0.3662,
      "num_input_tokens_seen": 23739872,
      "step": 36400
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.00019309644943738046,
      "loss": 0.2929,
      "num_input_tokens_seen": 23742592,
      "step": 36405
    },
    {
      "epoch": 19.08280922431866,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0001920007254864947,
      "loss": 0.5001,
      "num_input_tokens_seen": 23747296,
      "step": 36410
    },
    {
      "epoch": 19.085429769392032,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00019090809920011885,
      "loss": 0.3687,
      "num_input_tokens_seen": 23751040,
      "step": 36415
    },
    {
      "epoch": 19.08805031446541,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.00018981857080681842,
      "loss": 0.3779,
      "num_input_tokens_seen": 23754976,
      "step": 36420
    },
    {
      "epoch": 19.090670859538783,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.00018873214053450937,
      "loss": 0.4262,
      "num_input_tokens_seen": 23758464,
      "step": 36425
    },
    {
      "epoch": 19.09329140461216,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00018764880861045995,
      "loss": 0.4282,
      "num_input_tokens_seen": 23762880,
      "step": 36430
    },
    {
      "epoch": 19.095911949685533,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0001865685752612922,
      "loss": 0.4675,
      "num_input_tokens_seen": 23765664,
      "step": 36435
    },
    {
      "epoch": 19.09853249475891,
      "grad_norm": 0.087890625,
      "learning_rate": 0.00018549144071297707,
      "loss": 0.5233,
      "num_input_tokens_seen": 23770016,
      "step": 36440
    },
    {
      "epoch": 19.101153039832287,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00018441740519084093,
      "loss": 0.3819,
      "num_input_tokens_seen": 23773408,
      "step": 36445
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001833464689195574,
      "loss": 0.2993,
      "num_input_tokens_seen": 23776672,
      "step": 36450
    },
    {
      "epoch": 19.106394129979037,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.00018227863212315565,
      "loss": 0.2942,
      "num_input_tokens_seen": 23779264,
      "step": 36455
    },
    {
      "epoch": 19.10901467505241,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00018121389502501528,
      "loss": 0.4014,
      "num_input_tokens_seen": 23783104,
      "step": 36460
    },
    {
      "epoch": 19.111635220125788,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001801522578478648,
      "loss": 0.2956,
      "num_input_tokens_seen": 23785856,
      "step": 36465
    },
    {
      "epoch": 19.11425576519916,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.00017909372081378994,
      "loss": 0.2917,
      "num_input_tokens_seen": 23789344,
      "step": 36470
    },
    {
      "epoch": 19.116876310272538,
      "grad_norm": 0.0625,
      "learning_rate": 0.00017803828414422184,
      "loss": 0.3717,
      "num_input_tokens_seen": 23793056,
      "step": 36475
    },
    {
      "epoch": 19.11949685534591,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.00017698594805994728,
      "loss": 0.3673,
      "num_input_tokens_seen": 23795552,
      "step": 36480
    },
    {
      "epoch": 19.122117400419288,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00017593671278110345,
      "loss": 0.2975,
      "num_input_tokens_seen": 23797920,
      "step": 36485
    },
    {
      "epoch": 19.12473794549266,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00017489057852717482,
      "loss": 0.4129,
      "num_input_tokens_seen": 23800864,
      "step": 36490
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.00017384754551700465,
      "loss": 0.3497,
      "num_input_tokens_seen": 23804480,
      "step": 36495
    },
    {
      "epoch": 19.129979035639412,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0001728076139687834,
      "loss": 0.3152,
      "num_input_tokens_seen": 23807296,
      "step": 36500
    },
    {
      "epoch": 19.13259958071279,
      "grad_norm": 0.091796875,
      "learning_rate": 0.00017177078410005042,
      "loss": 0.2615,
      "num_input_tokens_seen": 23810432,
      "step": 36505
    },
    {
      "epoch": 19.135220125786162,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00017073705612770217,
      "loss": 0.2705,
      "num_input_tokens_seen": 23813088,
      "step": 36510
    },
    {
      "epoch": 19.13784067085954,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.00016970643026797738,
      "loss": 0.5165,
      "num_input_tokens_seen": 23815968,
      "step": 36515
    },
    {
      "epoch": 19.140461215932913,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.00016867890673647522,
      "loss": 0.3448,
      "num_input_tokens_seen": 23819392,
      "step": 36520
    },
    {
      "epoch": 19.14308176100629,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0001676544857481421,
      "loss": 0.4555,
      "num_input_tokens_seen": 23822144,
      "step": 36525
    },
    {
      "epoch": 19.145702306079663,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0001666331675172733,
      "loss": 0.3683,
      "num_input_tokens_seen": 23826144,
      "step": 36530
    },
    {
      "epoch": 19.14832285115304,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.00016561495225751955,
      "loss": 0.3306,
      "num_input_tokens_seen": 23829536,
      "step": 36535
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00016459984018187712,
      "loss": 0.2753,
      "num_input_tokens_seen": 23833408,
      "step": 36540
    },
    {
      "epoch": 19.15356394129979,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.00016358783150269784,
      "loss": 0.2842,
      "num_input_tokens_seen": 23835808,
      "step": 36545
    },
    {
      "epoch": 19.156184486373167,
      "grad_norm": 0.080078125,
      "learning_rate": 0.00016257892643167904,
      "loss": 0.4337,
      "num_input_tokens_seen": 23838624,
      "step": 36550
    },
    {
      "epoch": 19.15880503144654,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00016157312517987686,
      "loss": 0.2375,
      "num_input_tokens_seen": 23841312,
      "step": 36555
    },
    {
      "epoch": 19.161425576519918,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00016057042795769138,
      "loss": 0.3184,
      "num_input_tokens_seen": 23844416,
      "step": 36560
    },
    {
      "epoch": 19.16404612159329,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.00015957083497487477,
      "loss": 0.3,
      "num_input_tokens_seen": 23847744,
      "step": 36565
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0001585743464405298,
      "loss": 0.4375,
      "num_input_tokens_seen": 23850144,
      "step": 36570
    },
    {
      "epoch": 19.16928721174004,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00015758096256311138,
      "loss": 0.4783,
      "num_input_tokens_seen": 23853952,
      "step": 36575
    },
    {
      "epoch": 19.171907756813418,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00015659068355042494,
      "loss": 0.3622,
      "num_input_tokens_seen": 23857376,
      "step": 36580
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.00015560350960962642,
      "loss": 0.3882,
      "num_input_tokens_seen": 23861056,
      "step": 36585
    },
    {
      "epoch": 19.17714884696017,
      "grad_norm": 0.09375,
      "learning_rate": 0.00015461944094721734,
      "loss": 0.2643,
      "num_input_tokens_seen": 23863968,
      "step": 36590
    },
    {
      "epoch": 19.179769392033542,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00015363847776905803,
      "loss": 0.3194,
      "num_input_tokens_seen": 23866912,
      "step": 36595
    },
    {
      "epoch": 19.18238993710692,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.00015266062028035265,
      "loss": 0.2568,
      "num_input_tokens_seen": 23869856,
      "step": 36600
    },
    {
      "epoch": 19.185010482180292,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0001516858686856576,
      "loss": 0.3573,
      "num_input_tokens_seen": 23873088,
      "step": 36605
    },
    {
      "epoch": 19.18763102725367,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0001507142231888797,
      "loss": 0.3171,
      "num_input_tokens_seen": 23876160,
      "step": 36610
    },
    {
      "epoch": 19.190251572327043,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.00014974568399327815,
      "loss": 0.408,
      "num_input_tokens_seen": 23878880,
      "step": 36615
    },
    {
      "epoch": 19.19287211740042,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.00014878025130145745,
      "loss": 0.3264,
      "num_input_tokens_seen": 23882912,
      "step": 36620
    },
    {
      "epoch": 19.195492662473793,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00014781792531537774,
      "loss": 0.3213,
      "num_input_tokens_seen": 23886560,
      "step": 36625
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.00014685870623634467,
      "loss": 0.4321,
      "num_input_tokens_seen": 23889728,
      "step": 36630
    },
    {
      "epoch": 19.200733752620547,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.00014590259426501773,
      "loss": 0.3463,
      "num_input_tokens_seen": 23893408,
      "step": 36635
    },
    {
      "epoch": 19.20335429769392,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00014494958960140357,
      "loss": 0.2983,
      "num_input_tokens_seen": 23896960,
      "step": 36640
    },
    {
      "epoch": 19.205974842767297,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.00014399969244485776,
      "loss": 0.3187,
      "num_input_tokens_seen": 23899936,
      "step": 36645
    },
    {
      "epoch": 19.20859538784067,
      "grad_norm": 0.08984375,
      "learning_rate": 0.000143052902994093,
      "loss": 0.3114,
      "num_input_tokens_seen": 23903360,
      "step": 36650
    },
    {
      "epoch": 19.211215932914047,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00014210922144716087,
      "loss": 0.3034,
      "num_input_tokens_seen": 23906784,
      "step": 36655
    },
    {
      "epoch": 19.21383647798742,
      "grad_norm": 0.07421875,
      "learning_rate": 0.00014116864800147344,
      "loss": 0.2456,
      "num_input_tokens_seen": 23910240,
      "step": 36660
    },
    {
      "epoch": 19.216457023060798,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.00014023118285378666,
      "loss": 0.3645,
      "num_input_tokens_seen": 23913888,
      "step": 36665
    },
    {
      "epoch": 19.21907756813417,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.000139296826200207,
      "loss": 0.2959,
      "num_input_tokens_seen": 23917312,
      "step": 36670
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00013836557823618977,
      "loss": 0.324,
      "num_input_tokens_seen": 23920032,
      "step": 36675
    },
    {
      "epoch": 19.22431865828092,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0001374374391565458,
      "loss": 0.4286,
      "num_input_tokens_seen": 23924000,
      "step": 36680
    },
    {
      "epoch": 19.2269392033543,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0001365124091554265,
      "loss": 0.2715,
      "num_input_tokens_seen": 23928128,
      "step": 36685
    },
    {
      "epoch": 19.229559748427672,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00013559048842634036,
      "loss": 0.3675,
      "num_input_tokens_seen": 23931712,
      "step": 36690
    },
    {
      "epoch": 19.23218029350105,
      "grad_norm": 0.09375,
      "learning_rate": 0.00013467167716214146,
      "loss": 0.3942,
      "num_input_tokens_seen": 23934912,
      "step": 36695
    },
    {
      "epoch": 19.234800838574422,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00013375597555503603,
      "loss": 0.4383,
      "num_input_tokens_seen": 23938560,
      "step": 36700
    },
    {
      "epoch": 19.2374213836478,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001328433837965759,
      "loss": 0.3645,
      "num_input_tokens_seen": 23941440,
      "step": 36705
    },
    {
      "epoch": 19.240041928721173,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001319339020776683,
      "loss": 0.3605,
      "num_input_tokens_seen": 23944736,
      "step": 36710
    },
    {
      "epoch": 19.24266247379455,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.00013102753058856276,
      "loss": 0.5775,
      "num_input_tokens_seen": 23949376,
      "step": 36715
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 0.0625,
      "learning_rate": 0.00013012426951886425,
      "loss": 0.3852,
      "num_input_tokens_seen": 23952640,
      "step": 36720
    },
    {
      "epoch": 19.2479035639413,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.00012922411905752496,
      "loss": 0.35,
      "num_input_tokens_seen": 23956992,
      "step": 36725
    },
    {
      "epoch": 19.250524109014677,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.00012832707939284426,
      "loss": 0.285,
      "num_input_tokens_seen": 23960192,
      "step": 36730
    },
    {
      "epoch": 19.25314465408805,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00012743315071247374,
      "loss": 0.2608,
      "num_input_tokens_seen": 23963040,
      "step": 36735
    },
    {
      "epoch": 19.255765199161427,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00012654233320341212,
      "loss": 0.3351,
      "num_input_tokens_seen": 23965664,
      "step": 36740
    },
    {
      "epoch": 19.2583857442348,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.00012565462705201036,
      "loss": 0.3432,
      "num_input_tokens_seen": 23968992,
      "step": 36745
    },
    {
      "epoch": 19.261006289308177,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0001247700324439649,
      "loss": 0.6684,
      "num_input_tokens_seen": 23972032,
      "step": 36750
    },
    {
      "epoch": 19.26362683438155,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.00012388854956432106,
      "loss": 0.4734,
      "num_input_tokens_seen": 23975040,
      "step": 36755
    },
    {
      "epoch": 19.266247379454928,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.00012301017859747964,
      "loss": 0.349,
      "num_input_tokens_seen": 23977408,
      "step": 36760
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 0.119140625,
      "learning_rate": 0.00012213491972718203,
      "loss": 0.3756,
      "num_input_tokens_seen": 23980032,
      "step": 36765
    },
    {
      "epoch": 19.271488469601678,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.00012126277313652345,
      "loss": 0.3008,
      "num_input_tokens_seen": 23982720,
      "step": 36770
    },
    {
      "epoch": 19.27410901467505,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00012039373900794792,
      "loss": 0.3855,
      "num_input_tokens_seen": 23985984,
      "step": 36775
    },
    {
      "epoch": 19.27672955974843,
      "grad_norm": 0.07421875,
      "learning_rate": 0.00011952781752324503,
      "loss": 0.4569,
      "num_input_tokens_seen": 23989504,
      "step": 36780
    },
    {
      "epoch": 19.279350104821802,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.00011866500886355823,
      "loss": 0.3659,
      "num_input_tokens_seen": 23992160,
      "step": 36785
    },
    {
      "epoch": 19.28197064989518,
      "grad_norm": 0.083984375,
      "learning_rate": 0.00011780531320937647,
      "loss": 0.4701,
      "num_input_tokens_seen": 23998816,
      "step": 36790
    },
    {
      "epoch": 19.284591194968552,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00011694873074053924,
      "loss": 0.3802,
      "num_input_tokens_seen": 24003040,
      "step": 36795
    },
    {
      "epoch": 19.28721174004193,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00011609526163623151,
      "loss": 0.3874,
      "num_input_tokens_seen": 24006272,
      "step": 36800
    },
    {
      "epoch": 19.289832285115303,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00011524490607499049,
      "loss": 0.3074,
      "num_input_tokens_seen": 24009792,
      "step": 36805
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 0.107421875,
      "learning_rate": 0.00011439766423470054,
      "loss": 0.2849,
      "num_input_tokens_seen": 24012960,
      "step": 36810
    },
    {
      "epoch": 19.295073375262053,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001135535362925949,
      "loss": 0.4834,
      "num_input_tokens_seen": 24016416,
      "step": 36815
    },
    {
      "epoch": 19.29769392033543,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.00011271252242525731,
      "loss": 0.279,
      "num_input_tokens_seen": 24019520,
      "step": 36820
    },
    {
      "epoch": 19.300314465408803,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00011187462280861704,
      "loss": 0.3225,
      "num_input_tokens_seen": 24023168,
      "step": 36825
    },
    {
      "epoch": 19.30293501048218,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.00011103983761795222,
      "loss": 0.3544,
      "num_input_tokens_seen": 24026368,
      "step": 36830
    },
    {
      "epoch": 19.305555555555557,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00011020816702788982,
      "loss": 0.279,
      "num_input_tokens_seen": 24028960,
      "step": 36835
    },
    {
      "epoch": 19.30817610062893,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00010937961121240902,
      "loss": 0.3526,
      "num_input_tokens_seen": 24031392,
      "step": 36840
    },
    {
      "epoch": 19.310796645702307,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.00010855417034483117,
      "loss": 0.3307,
      "num_input_tokens_seen": 24034368,
      "step": 36845
    },
    {
      "epoch": 19.31341719077568,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00010773184459783147,
      "loss": 0.3917,
      "num_input_tokens_seen": 24037472,
      "step": 36850
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.00010691263414343066,
      "loss": 0.3854,
      "num_input_tokens_seen": 24040672,
      "step": 36855
    },
    {
      "epoch": 19.31865828092243,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00010609653915299499,
      "loss": 0.3382,
      "num_input_tokens_seen": 24044192,
      "step": 36860
    },
    {
      "epoch": 19.321278825995808,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.00010528355979724624,
      "loss": 0.3471,
      "num_input_tokens_seen": 24047520,
      "step": 36865
    },
    {
      "epoch": 19.32389937106918,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.00010447369624624836,
      "loss": 0.4496,
      "num_input_tokens_seen": 24050944,
      "step": 36870
    },
    {
      "epoch": 19.32651991614256,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.00010366694866941583,
      "loss": 0.3089,
      "num_input_tokens_seen": 24053440,
      "step": 36875
    },
    {
      "epoch": 19.329140461215932,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00010286331723551201,
      "loss": 0.3597,
      "num_input_tokens_seen": 24056480,
      "step": 36880
    },
    {
      "epoch": 19.33176100628931,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.00010206280211264573,
      "loss": 0.2987,
      "num_input_tokens_seen": 24060768,
      "step": 36885
    },
    {
      "epoch": 19.334381551362682,
      "grad_norm": 0.09765625,
      "learning_rate": 0.00010126540346827806,
      "loss": 0.3869,
      "num_input_tokens_seen": 24063200,
      "step": 36890
    },
    {
      "epoch": 19.33700209643606,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00010047112146921222,
      "loss": 0.4748,
      "num_input_tokens_seen": 24066272,
      "step": 36895
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 0.08447265625,
      "learning_rate": 9.967995628160697e-05,
      "loss": 0.3343,
      "num_input_tokens_seen": 24068960,
      "step": 36900
    },
    {
      "epoch": 19.34224318658281,
      "grad_norm": 0.10400390625,
      "learning_rate": 9.889190807096159e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 24071936,
      "step": 36905
    },
    {
      "epoch": 19.344863731656183,
      "grad_norm": 0.171875,
      "learning_rate": 9.810697700212922e-05,
      "loss": 0.3352,
      "num_input_tokens_seen": 24074752,
      "step": 36910
    },
    {
      "epoch": 19.34748427672956,
      "grad_norm": 0.05712890625,
      "learning_rate": 9.732516323930684e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 24078240,
      "step": 36915
    },
    {
      "epoch": 19.350104821802937,
      "grad_norm": 0.09423828125,
      "learning_rate": 9.654646694604197e-05,
      "loss": 0.3664,
      "num_input_tokens_seen": 24081920,
      "step": 36920
    },
    {
      "epoch": 19.35272536687631,
      "grad_norm": 0.11181640625,
      "learning_rate": 9.577088828522761e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 24085952,
      "step": 36925
    },
    {
      "epoch": 19.355345911949687,
      "grad_norm": 0.11279296875,
      "learning_rate": 9.499842741910902e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 24088672,
      "step": 36930
    },
    {
      "epoch": 19.35796645702306,
      "grad_norm": 0.060302734375,
      "learning_rate": 9.422908450927358e-05,
      "loss": 0.3992,
      "num_input_tokens_seen": 24092064,
      "step": 36935
    },
    {
      "epoch": 19.360587002096437,
      "grad_norm": 0.068359375,
      "learning_rate": 9.346285971665924e-05,
      "loss": 0.426,
      "num_input_tokens_seen": 24095328,
      "step": 36940
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 0.0966796875,
      "learning_rate": 9.26997532015511e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 24098560,
      "step": 36945
    },
    {
      "epoch": 19.365828092243188,
      "grad_norm": 0.06494140625,
      "learning_rate": 9.193976512358314e-05,
      "loss": 0.3394,
      "num_input_tokens_seen": 24101696,
      "step": 36950
    },
    {
      "epoch": 19.36844863731656,
      "grad_norm": 0.11572265625,
      "learning_rate": 9.11828956417382e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 24104448,
      "step": 36955
    },
    {
      "epoch": 19.371069182389938,
      "grad_norm": 0.0712890625,
      "learning_rate": 9.042914491433961e-05,
      "loss": 0.3537,
      "num_input_tokens_seen": 24107520,
      "step": 36960
    },
    {
      "epoch": 19.37368972746331,
      "grad_norm": 0.11083984375,
      "learning_rate": 8.96785130990696e-05,
      "loss": 0.3021,
      "num_input_tokens_seen": 24110304,
      "step": 36965
    },
    {
      "epoch": 19.37631027253669,
      "grad_norm": 0.11669921875,
      "learning_rate": 8.893100035294754e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 24113792,
      "step": 36970
    },
    {
      "epoch": 19.378930817610062,
      "grad_norm": 0.1513671875,
      "learning_rate": 8.818660683234502e-05,
      "loss": 0.3435,
      "num_input_tokens_seen": 24116224,
      "step": 36975
    },
    {
      "epoch": 19.38155136268344,
      "grad_norm": 0.0673828125,
      "learning_rate": 8.744533269298248e-05,
      "loss": 0.2968,
      "num_input_tokens_seen": 24120480,
      "step": 36980
    },
    {
      "epoch": 19.384171907756812,
      "grad_norm": 0.2578125,
      "learning_rate": 8.670717808992423e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 24123424,
      "step": 36985
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 0.1123046875,
      "learning_rate": 8.597214317758339e-05,
      "loss": 0.324,
      "num_input_tokens_seen": 24126528,
      "step": 36990
    },
    {
      "epoch": 19.389412997903563,
      "grad_norm": 0.0751953125,
      "learning_rate": 8.524022810972365e-05,
      "loss": 0.3048,
      "num_input_tokens_seen": 24131360,
      "step": 36995
    },
    {
      "epoch": 19.39203354297694,
      "grad_norm": 0.12353515625,
      "learning_rate": 8.451143303945085e-05,
      "loss": 0.3525,
      "num_input_tokens_seen": 24133920,
      "step": 37000
    },
    {
      "epoch": 19.394654088050313,
      "grad_norm": 0.07861328125,
      "learning_rate": 8.37857581192214e-05,
      "loss": 0.3161,
      "num_input_tokens_seen": 24137536,
      "step": 37005
    },
    {
      "epoch": 19.39727463312369,
      "grad_norm": 0.1240234375,
      "learning_rate": 8.306320350083885e-05,
      "loss": 0.3412,
      "num_input_tokens_seen": 24140448,
      "step": 37010
    },
    {
      "epoch": 19.399895178197063,
      "grad_norm": 0.11083984375,
      "learning_rate": 8.234376933545229e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 24143488,
      "step": 37015
    },
    {
      "epoch": 19.40251572327044,
      "grad_norm": 0.2060546875,
      "learning_rate": 8.162745577355968e-05,
      "loss": 0.3108,
      "num_input_tokens_seen": 24146944,
      "step": 37020
    },
    {
      "epoch": 19.405136268343817,
      "grad_norm": 0.06640625,
      "learning_rate": 8.091426296500614e-05,
      "loss": 0.3655,
      "num_input_tokens_seen": 24150944,
      "step": 37025
    },
    {
      "epoch": 19.40775681341719,
      "grad_norm": 0.1796875,
      "learning_rate": 8.020419105898068e-05,
      "loss": 0.3122,
      "num_input_tokens_seen": 24153504,
      "step": 37030
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 0.07958984375,
      "learning_rate": 7.949724020402615e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 24156480,
      "step": 37035
    },
    {
      "epoch": 19.41299790356394,
      "grad_norm": 0.0986328125,
      "learning_rate": 7.879341054802757e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 24160192,
      "step": 37040
    },
    {
      "epoch": 19.415618448637318,
      "grad_norm": 0.10205078125,
      "learning_rate": 7.809270223821552e-05,
      "loss": 0.3572,
      "num_input_tokens_seen": 24162560,
      "step": 37045
    },
    {
      "epoch": 19.41823899371069,
      "grad_norm": 0.0703125,
      "learning_rate": 7.739511542117438e-05,
      "loss": 0.3478,
      "num_input_tokens_seen": 24165568,
      "step": 37050
    },
    {
      "epoch": 19.420859538784068,
      "grad_norm": 0.06982421875,
      "learning_rate": 7.670065024282746e-05,
      "loss": 0.2709,
      "num_input_tokens_seen": 24168608,
      "step": 37055
    },
    {
      "epoch": 19.42348008385744,
      "grad_norm": 0.0703125,
      "learning_rate": 7.60093068484502e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 24171552,
      "step": 37060
    },
    {
      "epoch": 19.42610062893082,
      "grad_norm": 0.1240234375,
      "learning_rate": 7.532108538266358e-05,
      "loss": 0.3296,
      "num_input_tokens_seen": 24174336,
      "step": 37065
    },
    {
      "epoch": 19.428721174004192,
      "grad_norm": 0.05517578125,
      "learning_rate": 7.463598598943743e-05,
      "loss": 0.2656,
      "num_input_tokens_seen": 24178400,
      "step": 37070
    },
    {
      "epoch": 19.43134171907757,
      "grad_norm": 0.1005859375,
      "learning_rate": 7.395400881208546e-05,
      "loss": 0.2953,
      "num_input_tokens_seen": 24181312,
      "step": 37075
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 0.0771484375,
      "learning_rate": 7.327515399326855e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 24184544,
      "step": 37080
    },
    {
      "epoch": 19.43658280922432,
      "grad_norm": 0.1123046875,
      "learning_rate": 7.25994216749981e-05,
      "loss": 0.2252,
      "num_input_tokens_seen": 24186720,
      "step": 37085
    },
    {
      "epoch": 19.439203354297693,
      "grad_norm": 0.1064453125,
      "learning_rate": 7.192681199862604e-05,
      "loss": 0.3679,
      "num_input_tokens_seen": 24189984,
      "step": 37090
    },
    {
      "epoch": 19.44182389937107,
      "grad_norm": 0.09326171875,
      "learning_rate": 7.125732510485649e-05,
      "loss": 0.3121,
      "num_input_tokens_seen": 24192448,
      "step": 37095
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.076171875,
      "learning_rate": 7.059096113373908e-05,
      "loss": 0.3138,
      "num_input_tokens_seen": 24196128,
      "step": 37100
    },
    {
      "epoch": 19.44706498951782,
      "grad_norm": 0.0732421875,
      "learning_rate": 6.992772022467064e-05,
      "loss": 0.2906,
      "num_input_tokens_seen": 24199104,
      "step": 37105
    },
    {
      "epoch": 19.449685534591197,
      "grad_norm": 0.07861328125,
      "learning_rate": 6.926760251638852e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 24202176,
      "step": 37110
    },
    {
      "epoch": 19.45230607966457,
      "grad_norm": 0.09326171875,
      "learning_rate": 6.861060814698727e-05,
      "loss": 0.4136,
      "num_input_tokens_seen": 24205824,
      "step": 37115
    },
    {
      "epoch": 19.454926624737947,
      "grad_norm": 0.07080078125,
      "learning_rate": 6.795673725390027e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 24208640,
      "step": 37120
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 0.025390625,
      "learning_rate": 6.73059899739098e-05,
      "loss": 0.2207,
      "num_input_tokens_seen": 24215456,
      "step": 37125
    },
    {
      "epoch": 19.460167714884697,
      "grad_norm": 0.0751953125,
      "learning_rate": 6.66583664431436e-05,
      "loss": 0.3527,
      "num_input_tokens_seen": 24218752,
      "step": 37130
    },
    {
      "epoch": 19.46278825995807,
      "grad_norm": 0.08251953125,
      "learning_rate": 6.601386679708165e-05,
      "loss": 0.3489,
      "num_input_tokens_seen": 24221632,
      "step": 37135
    },
    {
      "epoch": 19.465408805031448,
      "grad_norm": 0.095703125,
      "learning_rate": 6.537249117054111e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 24225824,
      "step": 37140
    },
    {
      "epoch": 19.46802935010482,
      "grad_norm": 0.10595703125,
      "learning_rate": 6.473423969769132e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 24228960,
      "step": 37145
    },
    {
      "epoch": 19.470649895178198,
      "grad_norm": 0.0966796875,
      "learning_rate": 6.40991125120488e-05,
      "loss": 0.3307,
      "num_input_tokens_seen": 24232352,
      "step": 37150
    },
    {
      "epoch": 19.47327044025157,
      "grad_norm": 0.07275390625,
      "learning_rate": 6.346710974647563e-05,
      "loss": 0.3069,
      "num_input_tokens_seen": 24235456,
      "step": 37155
    },
    {
      "epoch": 19.47589098532495,
      "grad_norm": 0.1083984375,
      "learning_rate": 6.283823153317602e-05,
      "loss": 0.3466,
      "num_input_tokens_seen": 24238016,
      "step": 37160
    },
    {
      "epoch": 19.478511530398322,
      "grad_norm": 0.0888671875,
      "learning_rate": 6.22124780037081e-05,
      "loss": 0.3183,
      "num_input_tokens_seen": 24240672,
      "step": 37165
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 0.09375,
      "learning_rate": 6.158984928896882e-05,
      "loss": 0.2497,
      "num_input_tokens_seen": 24243520,
      "step": 37170
    },
    {
      "epoch": 19.483752620545072,
      "grad_norm": 0.1005859375,
      "learning_rate": 6.097034551920732e-05,
      "loss": 0.3556,
      "num_input_tokens_seen": 24246784,
      "step": 37175
    },
    {
      "epoch": 19.48637316561845,
      "grad_norm": 0.0859375,
      "learning_rate": 6.0353966824016615e-05,
      "loss": 0.5668,
      "num_input_tokens_seen": 24250848,
      "step": 37180
    },
    {
      "epoch": 19.488993710691823,
      "grad_norm": 0.06396484375,
      "learning_rate": 5.974071333233355e-05,
      "loss": 0.3795,
      "num_input_tokens_seen": 24254208,
      "step": 37185
    },
    {
      "epoch": 19.4916142557652,
      "grad_norm": 0.09814453125,
      "learning_rate": 5.9130585172448845e-05,
      "loss": 0.3942,
      "num_input_tokens_seen": 24260896,
      "step": 37190
    },
    {
      "epoch": 19.494234800838573,
      "grad_norm": 0.20703125,
      "learning_rate": 5.852358247199041e-05,
      "loss": 0.3397,
      "num_input_tokens_seen": 24263232,
      "step": 37195
    },
    {
      "epoch": 19.49685534591195,
      "grad_norm": 0.06591796875,
      "learning_rate": 5.7919705357935e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 24266784,
      "step": 37200
    },
    {
      "epoch": 19.499475890985323,
      "grad_norm": 0.0712890625,
      "learning_rate": 5.731895395660991e-05,
      "loss": 0.3032,
      "num_input_tokens_seen": 24270112,
      "step": 37205
    },
    {
      "epoch": 19.5020964360587,
      "grad_norm": 0.07958984375,
      "learning_rate": 5.672132839368626e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 24273184,
      "step": 37210
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 0.09912109375,
      "learning_rate": 5.61268287941774e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 24276224,
      "step": 37215
    },
    {
      "epoch": 19.50733752620545,
      "grad_norm": 0.0703125,
      "learning_rate": 5.553545528244719e-05,
      "loss": 0.3081,
      "num_input_tokens_seen": 24281024,
      "step": 37220
    },
    {
      "epoch": 19.509958071278827,
      "grad_norm": 0.08251953125,
      "learning_rate": 5.494720798220498e-05,
      "loss": 0.4594,
      "num_input_tokens_seen": 24283840,
      "step": 37225
    },
    {
      "epoch": 19.5125786163522,
      "grad_norm": 0.0869140625,
      "learning_rate": 5.436208701650402e-05,
      "loss": 0.2152,
      "num_input_tokens_seen": 24286240,
      "step": 37230
    },
    {
      "epoch": 19.515199161425578,
      "grad_norm": 0.1240234375,
      "learning_rate": 5.3780092507744734e-05,
      "loss": 0.3039,
      "num_input_tokens_seen": 24288992,
      "step": 37235
    },
    {
      "epoch": 19.51781970649895,
      "grad_norm": 0.0859375,
      "learning_rate": 5.3201224577676396e-05,
      "loss": 0.3487,
      "num_input_tokens_seen": 24292288,
      "step": 37240
    },
    {
      "epoch": 19.520440251572328,
      "grad_norm": 0.0703125,
      "learning_rate": 5.262548334738881e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 24295584,
      "step": 37245
    },
    {
      "epoch": 19.5230607966457,
      "grad_norm": 0.076171875,
      "learning_rate": 5.2052868937322306e-05,
      "loss": 0.3106,
      "num_input_tokens_seen": 24298400,
      "step": 37250
    },
    {
      "epoch": 19.52568134171908,
      "grad_norm": 0.083984375,
      "learning_rate": 5.148338146725939e-05,
      "loss": 0.2784,
      "num_input_tokens_seen": 24301280,
      "step": 37255
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 0.08984375,
      "learning_rate": 5.091702105633144e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 24303968,
      "step": 37260
    },
    {
      "epoch": 19.53092243186583,
      "grad_norm": 0.07568359375,
      "learning_rate": 5.035378782301369e-05,
      "loss": 0.3719,
      "num_input_tokens_seen": 24307456,
      "step": 37265
    },
    {
      "epoch": 19.533542976939202,
      "grad_norm": 0.125,
      "learning_rate": 4.979368188513189e-05,
      "loss": 0.318,
      "num_input_tokens_seen": 24310912,
      "step": 37270
    },
    {
      "epoch": 19.53616352201258,
      "grad_norm": 0.07958984375,
      "learning_rate": 4.9236703359848975e-05,
      "loss": 0.3363,
      "num_input_tokens_seen": 24313728,
      "step": 37275
    },
    {
      "epoch": 19.538784067085953,
      "grad_norm": 0.11279296875,
      "learning_rate": 4.8682852363680084e-05,
      "loss": 0.3284,
      "num_input_tokens_seen": 24316928,
      "step": 37280
    },
    {
      "epoch": 19.54140461215933,
      "grad_norm": 0.109375,
      "learning_rate": 4.813212901248587e-05,
      "loss": 0.2913,
      "num_input_tokens_seen": 24319552,
      "step": 37285
    },
    {
      "epoch": 19.544025157232703,
      "grad_norm": 0.10400390625,
      "learning_rate": 4.7584533421469176e-05,
      "loss": 0.3282,
      "num_input_tokens_seen": 24321984,
      "step": 37290
    },
    {
      "epoch": 19.54664570230608,
      "grad_norm": 0.1455078125,
      "learning_rate": 4.704006570518171e-05,
      "loss": 0.4085,
      "num_input_tokens_seen": 24324928,
      "step": 37295
    },
    {
      "epoch": 19.549266247379457,
      "grad_norm": 0.06494140625,
      "learning_rate": 4.6498725977520695e-05,
      "loss": 0.275,
      "num_input_tokens_seen": 24328064,
      "step": 37300
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 0.1328125,
      "learning_rate": 4.596051435172887e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 24331136,
      "step": 37305
    },
    {
      "epoch": 19.554507337526207,
      "grad_norm": 0.11376953125,
      "learning_rate": 4.542543094039119e-05,
      "loss": 0.3482,
      "num_input_tokens_seen": 24334112,
      "step": 37310
    },
    {
      "epoch": 19.55712788259958,
      "grad_norm": 0.04443359375,
      "learning_rate": 4.489347585544312e-05,
      "loss": 0.3035,
      "num_input_tokens_seen": 24338048,
      "step": 37315
    },
    {
      "epoch": 19.559748427672957,
      "grad_norm": 0.058349609375,
      "learning_rate": 4.43646492081623e-05,
      "loss": 0.4792,
      "num_input_tokens_seen": 24341632,
      "step": 37320
    },
    {
      "epoch": 19.56236897274633,
      "grad_norm": 0.0869140625,
      "learning_rate": 4.383895110917524e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 24345088,
      "step": 37325
    },
    {
      "epoch": 19.564989517819708,
      "grad_norm": 0.0546875,
      "learning_rate": 4.331638166845064e-05,
      "loss": 0.3431,
      "num_input_tokens_seen": 24348736,
      "step": 37330
    },
    {
      "epoch": 19.56761006289308,
      "grad_norm": 0.07373046875,
      "learning_rate": 4.279694099530273e-05,
      "loss": 0.2642,
      "num_input_tokens_seen": 24351776,
      "step": 37335
    },
    {
      "epoch": 19.570230607966458,
      "grad_norm": 0.07177734375,
      "learning_rate": 4.2280629198394594e-05,
      "loss": 0.3019,
      "num_input_tokens_seen": 24355296,
      "step": 37340
    },
    {
      "epoch": 19.57285115303983,
      "grad_norm": 0.078125,
      "learning_rate": 4.1767446385733155e-05,
      "loss": 0.3716,
      "num_input_tokens_seen": 24359168,
      "step": 37345
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 0.09228515625,
      "learning_rate": 4.1257392664669214e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 24362720,
      "step": 37350
    },
    {
      "epoch": 19.578092243186582,
      "grad_norm": 0.0654296875,
      "learning_rate": 4.075046814189909e-05,
      "loss": 0.3098,
      "num_input_tokens_seen": 24365824,
      "step": 37355
    },
    {
      "epoch": 19.58071278825996,
      "grad_norm": 0.10107421875,
      "learning_rate": 4.0246672923466285e-05,
      "loss": 0.2952,
      "num_input_tokens_seen": 24368928,
      "step": 37360
    },
    {
      "epoch": 19.583333333333332,
      "grad_norm": 0.0986328125,
      "learning_rate": 3.974600711476151e-05,
      "loss": 0.3084,
      "num_input_tokens_seen": 24372192,
      "step": 37365
    },
    {
      "epoch": 19.58595387840671,
      "grad_norm": 0.12060546875,
      "learning_rate": 3.9248470820515964e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 24374912,
      "step": 37370
    },
    {
      "epoch": 19.588574423480082,
      "grad_norm": 0.083984375,
      "learning_rate": 3.875406414480975e-05,
      "loss": 0.3409,
      "num_input_tokens_seen": 24378176,
      "step": 37375
    },
    {
      "epoch": 19.59119496855346,
      "grad_norm": 0.103515625,
      "learning_rate": 3.826278719106513e-05,
      "loss": 0.318,
      "num_input_tokens_seen": 24381760,
      "step": 37380
    },
    {
      "epoch": 19.593815513626833,
      "grad_norm": 0.06689453125,
      "learning_rate": 3.777464006205322e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 24386144,
      "step": 37385
    },
    {
      "epoch": 19.59643605870021,
      "grad_norm": 0.056640625,
      "learning_rate": 3.728962285988901e-05,
      "loss": 0.3417,
      "num_input_tokens_seen": 24389600,
      "step": 37390
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 0.1591796875,
      "learning_rate": 3.680773568603135e-05,
      "loss": 0.3107,
      "num_input_tokens_seen": 24392160,
      "step": 37395
    },
    {
      "epoch": 19.60167714884696,
      "grad_norm": 0.080078125,
      "learning_rate": 3.6328978641286256e-05,
      "loss": 0.3493,
      "num_input_tokens_seen": 24395584,
      "step": 37400
    },
    {
      "epoch": 19.604297693920337,
      "grad_norm": 0.185546875,
      "learning_rate": 3.585335182580529e-05,
      "loss": 0.2662,
      "num_input_tokens_seen": 24397696,
      "step": 37405
    },
    {
      "epoch": 19.60691823899371,
      "grad_norm": 0.068359375,
      "learning_rate": 3.538085533908219e-05,
      "loss": 0.3451,
      "num_input_tokens_seen": 24401696,
      "step": 37410
    },
    {
      "epoch": 19.609538784067087,
      "grad_norm": 0.1005859375,
      "learning_rate": 3.4911489279957904e-05,
      "loss": 0.3365,
      "num_input_tokens_seen": 24404416,
      "step": 37415
    },
    {
      "epoch": 19.61215932914046,
      "grad_norm": 0.09765625,
      "learning_rate": 3.444525374662222e-05,
      "loss": 0.3043,
      "num_input_tokens_seen": 24407008,
      "step": 37420
    },
    {
      "epoch": 19.614779874213838,
      "grad_norm": 0.0634765625,
      "learning_rate": 3.398214883660044e-05,
      "loss": 0.2344,
      "num_input_tokens_seen": 24410208,
      "step": 37425
    },
    {
      "epoch": 19.61740041928721,
      "grad_norm": 0.043212890625,
      "learning_rate": 3.3522174646773424e-05,
      "loss": 0.2603,
      "num_input_tokens_seen": 24415168,
      "step": 37430
    },
    {
      "epoch": 19.620020964360588,
      "grad_norm": 0.0947265625,
      "learning_rate": 3.30653312733592e-05,
      "loss": 0.2999,
      "num_input_tokens_seen": 24417984,
      "step": 37435
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 0.12109375,
      "learning_rate": 3.261161881192798e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 24421376,
      "step": 37440
    },
    {
      "epoch": 19.62526205450734,
      "grad_norm": 0.044677734375,
      "learning_rate": 3.2161037357387175e-05,
      "loss": 0.227,
      "num_input_tokens_seen": 24424544,
      "step": 37445
    },
    {
      "epoch": 19.627882599580712,
      "grad_norm": 0.080078125,
      "learning_rate": 3.17135870039964e-05,
      "loss": 0.2707,
      "num_input_tokens_seen": 24427488,
      "step": 37450
    },
    {
      "epoch": 19.63050314465409,
      "grad_norm": 0.09765625,
      "learning_rate": 3.126926784535577e-05,
      "loss": 0.3712,
      "num_input_tokens_seen": 24430336,
      "step": 37455
    },
    {
      "epoch": 19.633123689727462,
      "grad_norm": 0.1015625,
      "learning_rate": 3.0828079974412614e-05,
      "loss": 0.3549,
      "num_input_tokens_seen": 24435136,
      "step": 37460
    },
    {
      "epoch": 19.63574423480084,
      "grad_norm": 0.11572265625,
      "learning_rate": 3.039002348345643e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 24438848,
      "step": 37465
    },
    {
      "epoch": 19.638364779874212,
      "grad_norm": 0.0791015625,
      "learning_rate": 2.9955098464125583e-05,
      "loss": 0.3552,
      "num_input_tokens_seen": 24441248,
      "step": 37470
    },
    {
      "epoch": 19.64098532494759,
      "grad_norm": 0.1220703125,
      "learning_rate": 2.9523305007402298e-05,
      "loss": 0.3102,
      "num_input_tokens_seen": 24443840,
      "step": 37475
    },
    {
      "epoch": 19.643605870020963,
      "grad_norm": 0.08056640625,
      "learning_rate": 2.9094643203609328e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 24447200,
      "step": 37480
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 0.314453125,
      "learning_rate": 2.86691131424216e-05,
      "loss": 0.4368,
      "num_input_tokens_seen": 24450368,
      "step": 37485
    },
    {
      "epoch": 19.648846960167717,
      "grad_norm": 0.06689453125,
      "learning_rate": 2.8246714912851243e-05,
      "loss": 0.2384,
      "num_input_tokens_seen": 24452992,
      "step": 37490
    },
    {
      "epoch": 19.65146750524109,
      "grad_norm": 0.07275390625,
      "learning_rate": 2.7827448603262583e-05,
      "loss": 0.3628,
      "num_input_tokens_seen": 24455264,
      "step": 37495
    },
    {
      "epoch": 19.654088050314467,
      "grad_norm": 0.0732421875,
      "learning_rate": 2.7411314301360456e-05,
      "loss": 0.4146,
      "num_input_tokens_seen": 24458304,
      "step": 37500
    },
    {
      "epoch": 19.65670859538784,
      "grad_norm": 0.09716796875,
      "learning_rate": 2.6998312094193565e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 24461824,
      "step": 37505
    },
    {
      "epoch": 19.659329140461217,
      "grad_norm": 0.09521484375,
      "learning_rate": 2.6588442068161134e-05,
      "loss": 0.3339,
      "num_input_tokens_seen": 24464736,
      "step": 37510
    },
    {
      "epoch": 19.66194968553459,
      "grad_norm": 0.11572265625,
      "learning_rate": 2.6181704308999575e-05,
      "loss": 0.2908,
      "num_input_tokens_seen": 24467264,
      "step": 37515
    },
    {
      "epoch": 19.664570230607968,
      "grad_norm": 0.0859375,
      "learning_rate": 2.5778098901794167e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 24470912,
      "step": 37520
    },
    {
      "epoch": 19.66719077568134,
      "grad_norm": 0.062255859375,
      "learning_rate": 2.5377625930977366e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 24474304,
      "step": 37525
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 0.1357421875,
      "learning_rate": 2.4980285480320496e-05,
      "loss": 0.3041,
      "num_input_tokens_seen": 24477280,
      "step": 37530
    },
    {
      "epoch": 19.67243186582809,
      "grad_norm": 0.1064453125,
      "learning_rate": 2.4586077632943735e-05,
      "loss": 0.2934,
      "num_input_tokens_seen": 24479456,
      "step": 37535
    },
    {
      "epoch": 19.67505241090147,
      "grad_norm": 0.162109375,
      "learning_rate": 2.4195002471312788e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 24482624,
      "step": 37540
    },
    {
      "epoch": 19.677672955974842,
      "grad_norm": 0.1376953125,
      "learning_rate": 2.3807060077232212e-05,
      "loss": 0.314,
      "num_input_tokens_seen": 24485792,
      "step": 37545
    },
    {
      "epoch": 19.68029350104822,
      "grad_norm": 0.10400390625,
      "learning_rate": 2.342225053185709e-05,
      "loss": 0.3067,
      "num_input_tokens_seen": 24488576,
      "step": 37550
    },
    {
      "epoch": 19.682914046121592,
      "grad_norm": 0.11865234375,
      "learning_rate": 2.3040573915686367e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 24491136,
      "step": 37555
    },
    {
      "epoch": 19.68553459119497,
      "grad_norm": 0.0556640625,
      "learning_rate": 2.2662030308561177e-05,
      "loss": 0.5662,
      "num_input_tokens_seen": 24494528,
      "step": 37560
    },
    {
      "epoch": 19.688155136268342,
      "grad_norm": 0.051513671875,
      "learning_rate": 2.2286619789669836e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 24498816,
      "step": 37565
    },
    {
      "epoch": 19.69077568134172,
      "grad_norm": 0.08056640625,
      "learning_rate": 2.19143424375412e-05,
      "loss": 0.3113,
      "num_input_tokens_seen": 24502368,
      "step": 37570
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 0.11376953125,
      "learning_rate": 2.154519833005297e-05,
      "loss": 0.3368,
      "num_input_tokens_seen": 24505440,
      "step": 37575
    },
    {
      "epoch": 19.69601677148847,
      "grad_norm": 0.0703125,
      "learning_rate": 2.1179187544426713e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 24508512,
      "step": 37580
    },
    {
      "epoch": 19.698637316561843,
      "grad_norm": 0.060546875,
      "learning_rate": 2.0816310157227845e-05,
      "loss": 0.3157,
      "num_input_tokens_seen": 24511360,
      "step": 37585
    },
    {
      "epoch": 19.70125786163522,
      "grad_norm": 0.06884765625,
      "learning_rate": 2.0456566244365648e-05,
      "loss": 0.36,
      "num_input_tokens_seen": 24514048,
      "step": 37590
    },
    {
      "epoch": 19.703878406708597,
      "grad_norm": 0.208984375,
      "learning_rate": 2.009995588109159e-05,
      "loss": 0.2846,
      "num_input_tokens_seen": 24517088,
      "step": 37595
    },
    {
      "epoch": 19.70649895178197,
      "grad_norm": 0.11181640625,
      "learning_rate": 1.9746479142009333e-05,
      "loss": 0.3984,
      "num_input_tokens_seen": 24519808,
      "step": 37600
    },
    {
      "epoch": 19.709119496855347,
      "grad_norm": 0.1416015625,
      "learning_rate": 1.9396136101058058e-05,
      "loss": 0.411,
      "num_input_tokens_seen": 24524320,
      "step": 37605
    },
    {
      "epoch": 19.71174004192872,
      "grad_norm": 0.0830078125,
      "learning_rate": 1.9048926831529145e-05,
      "loss": 0.3819,
      "num_input_tokens_seen": 24528192,
      "step": 37610
    },
    {
      "epoch": 19.714360587002098,
      "grad_norm": 0.1123046875,
      "learning_rate": 1.870485140605116e-05,
      "loss": 0.4084,
      "num_input_tokens_seen": 24530880,
      "step": 37615
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 0.08251953125,
      "learning_rate": 1.8363909896604856e-05,
      "loss": 0.3686,
      "num_input_tokens_seen": 24533952,
      "step": 37620
    },
    {
      "epoch": 19.719601677148848,
      "grad_norm": 0.10595703125,
      "learning_rate": 1.8026102374506526e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 24537152,
      "step": 37625
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.2001953125,
      "learning_rate": 1.7691428910426297e-05,
      "loss": 0.2477,
      "num_input_tokens_seen": 24539776,
      "step": 37630
    },
    {
      "epoch": 19.7248427672956,
      "grad_norm": 0.09375,
      "learning_rate": 1.7359889574369846e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 24542944,
      "step": 37635
    },
    {
      "epoch": 19.72746331236897,
      "grad_norm": 0.1005859375,
      "learning_rate": 1.703148443569502e-05,
      "loss": 0.3167,
      "num_input_tokens_seen": 24545504,
      "step": 37640
    },
    {
      "epoch": 19.73008385744235,
      "grad_norm": 0.107421875,
      "learning_rate": 1.6706213563098537e-05,
      "loss": 0.3543,
      "num_input_tokens_seen": 24548736,
      "step": 37645
    },
    {
      "epoch": 19.732704402515722,
      "grad_norm": 0.1142578125,
      "learning_rate": 1.63840770246243e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 24552000,
      "step": 37650
    },
    {
      "epoch": 19.7353249475891,
      "grad_norm": 0.07275390625,
      "learning_rate": 1.6065074887658404e-05,
      "loss": 0.3156,
      "num_input_tokens_seen": 24555968,
      "step": 37655
    },
    {
      "epoch": 19.737945492662472,
      "grad_norm": 0.08154296875,
      "learning_rate": 1.574920721893247e-05,
      "loss": 0.3902,
      "num_input_tokens_seen": 24559104,
      "step": 37660
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 0.154296875,
      "learning_rate": 1.543647408452531e-05,
      "loss": 0.3581,
      "num_input_tokens_seen": 24561824,
      "step": 37665
    },
    {
      "epoch": 19.743186582809223,
      "grad_norm": 0.333984375,
      "learning_rate": 1.5126875549852924e-05,
      "loss": 0.337,
      "num_input_tokens_seen": 24564864,
      "step": 37670
    },
    {
      "epoch": 19.7458071278826,
      "grad_norm": 0.037841796875,
      "learning_rate": 1.4820411679681844e-05,
      "loss": 0.3355,
      "num_input_tokens_seen": 24568224,
      "step": 37675
    },
    {
      "epoch": 19.748427672955973,
      "grad_norm": 0.047119140625,
      "learning_rate": 1.4517082538120785e-05,
      "loss": 0.3598,
      "num_input_tokens_seen": 24573312,
      "step": 37680
    },
    {
      "epoch": 19.75104821802935,
      "grad_norm": 0.06298828125,
      "learning_rate": 1.4216888188622323e-05,
      "loss": 0.3332,
      "num_input_tokens_seen": 24577248,
      "step": 37685
    },
    {
      "epoch": 19.753668763102727,
      "grad_norm": 0.06640625,
      "learning_rate": 1.391982869398456e-05,
      "loss": 0.3283,
      "num_input_tokens_seen": 24580544,
      "step": 37690
    },
    {
      "epoch": 19.7562893081761,
      "grad_norm": 0.0830078125,
      "learning_rate": 1.3625904116347787e-05,
      "loss": 0.3274,
      "num_input_tokens_seen": 24582944,
      "step": 37695
    },
    {
      "epoch": 19.758909853249477,
      "grad_norm": 0.06201171875,
      "learning_rate": 1.3335114517199487e-05,
      "loss": 0.4442,
      "num_input_tokens_seen": 24585728,
      "step": 37700
    },
    {
      "epoch": 19.76153039832285,
      "grad_norm": 0.08837890625,
      "learning_rate": 1.3047459957367669e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 24588640,
      "step": 37705
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 0.07275390625,
      "learning_rate": 1.276294049702753e-05,
      "loss": 0.3834,
      "num_input_tokens_seen": 24592064,
      "step": 37710
    },
    {
      "epoch": 19.7667714884696,
      "grad_norm": 0.0634765625,
      "learning_rate": 1.2481556195694798e-05,
      "loss": 0.3214,
      "num_input_tokens_seen": 24598944,
      "step": 37715
    },
    {
      "epoch": 19.769392033542978,
      "grad_norm": 0.10888671875,
      "learning_rate": 1.2203307112235717e-05,
      "loss": 0.3307,
      "num_input_tokens_seen": 24601152,
      "step": 37720
    },
    {
      "epoch": 19.77201257861635,
      "grad_norm": 0.0751953125,
      "learning_rate": 1.1928193304855394e-05,
      "loss": 0.3294,
      "num_input_tokens_seen": 24603904,
      "step": 37725
    },
    {
      "epoch": 19.77463312368973,
      "grad_norm": 0.1103515625,
      "learning_rate": 1.1656214831102795e-05,
      "loss": 0.2865,
      "num_input_tokens_seen": 24606240,
      "step": 37730
    },
    {
      "epoch": 19.7772536687631,
      "grad_norm": 0.1064453125,
      "learning_rate": 1.1387371747874076e-05,
      "loss": 0.3484,
      "num_input_tokens_seen": 24609088,
      "step": 37735
    },
    {
      "epoch": 19.77987421383648,
      "grad_norm": 0.119140625,
      "learning_rate": 1.1121664111409246e-05,
      "loss": 0.3273,
      "num_input_tokens_seen": 24612320,
      "step": 37740
    },
    {
      "epoch": 19.782494758909852,
      "grad_norm": 0.06298828125,
      "learning_rate": 1.0859091977288848e-05,
      "loss": 0.3438,
      "num_input_tokens_seen": 24617664,
      "step": 37745
    },
    {
      "epoch": 19.78511530398323,
      "grad_norm": 0.07470703125,
      "learning_rate": 1.059965540044061e-05,
      "loss": 0.3181,
      "num_input_tokens_seen": 24622304,
      "step": 37750
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 0.10107421875,
      "learning_rate": 1.0343354435137785e-05,
      "loss": 0.3384,
      "num_input_tokens_seen": 24625088,
      "step": 37755
    },
    {
      "epoch": 19.79035639412998,
      "grad_norm": 0.061767578125,
      "learning_rate": 1.0090189134994153e-05,
      "loss": 0.3343,
      "num_input_tokens_seen": 24628544,
      "step": 37760
    },
    {
      "epoch": 19.792976939203353,
      "grad_norm": 0.158203125,
      "learning_rate": 9.84015955296902e-06,
      "loss": 0.4856,
      "num_input_tokens_seen": 24631776,
      "step": 37765
    },
    {
      "epoch": 19.79559748427673,
      "grad_norm": 0.109375,
      "learning_rate": 9.593265741365542e-06,
      "loss": 0.3414,
      "num_input_tokens_seen": 24634976,
      "step": 37770
    },
    {
      "epoch": 19.798218029350103,
      "grad_norm": 0.08984375,
      "learning_rate": 9.349507751830744e-06,
      "loss": 0.4015,
      "num_input_tokens_seen": 24637696,
      "step": 37775
    },
    {
      "epoch": 19.80083857442348,
      "grad_norm": 0.0927734375,
      "learning_rate": 9.108885635357167e-06,
      "loss": 0.2934,
      "num_input_tokens_seen": 24640864,
      "step": 37780
    },
    {
      "epoch": 19.803459119496857,
      "grad_norm": 0.041015625,
      "learning_rate": 8.871399442277882e-06,
      "loss": 0.2317,
      "num_input_tokens_seen": 24646784,
      "step": 37785
    },
    {
      "epoch": 19.80607966457023,
      "grad_norm": 0.0888671875,
      "learning_rate": 8.637049222276483e-06,
      "loss": 0.3278,
      "num_input_tokens_seen": 24650048,
      "step": 37790
    },
    {
      "epoch": 19.808700209643607,
      "grad_norm": 0.11767578125,
      "learning_rate": 8.405835024372088e-06,
      "loss": 0.3487,
      "num_input_tokens_seen": 24653728,
      "step": 37795
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 0.08935546875,
      "learning_rate": 8.177756896934341e-06,
      "loss": 0.2688,
      "num_input_tokens_seen": 24656288,
      "step": 37800
    },
    {
      "epoch": 19.813941299790358,
      "grad_norm": 0.06689453125,
      "learning_rate": 7.952814887675075e-06,
      "loss": 0.3353,
      "num_input_tokens_seen": 24659744,
      "step": 37805
    },
    {
      "epoch": 19.81656184486373,
      "grad_norm": 0.054443359375,
      "learning_rate": 7.731009043648318e-06,
      "loss": 0.4301,
      "num_input_tokens_seen": 24663488,
      "step": 37810
    },
    {
      "epoch": 19.819182389937108,
      "grad_norm": 0.0859375,
      "learning_rate": 7.512339411251956e-06,
      "loss": 0.2849,
      "num_input_tokens_seen": 24666272,
      "step": 37815
    },
    {
      "epoch": 19.82180293501048,
      "grad_norm": 0.05126953125,
      "learning_rate": 7.296806036232728e-06,
      "loss": 0.2042,
      "num_input_tokens_seen": 24670400,
      "step": 37820
    },
    {
      "epoch": 19.82442348008386,
      "grad_norm": 0.1318359375,
      "learning_rate": 7.084408963674571e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 24673280,
      "step": 37825
    },
    {
      "epoch": 19.82704402515723,
      "grad_norm": 0.07568359375,
      "learning_rate": 6.875148238010276e-06,
      "loss": 0.3961,
      "num_input_tokens_seen": 24676320,
      "step": 37830
    },
    {
      "epoch": 19.82966457023061,
      "grad_norm": 0.1708984375,
      "learning_rate": 6.669023903014826e-06,
      "loss": 0.2537,
      "num_input_tokens_seen": 24680960,
      "step": 37835
    },
    {
      "epoch": 19.832285115303982,
      "grad_norm": 0.0966796875,
      "learning_rate": 6.466036001807062e-06,
      "loss": 0.3922,
      "num_input_tokens_seen": 24685792,
      "step": 37840
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 0.08984375,
      "learning_rate": 6.266184576848022e-06,
      "loss": 0.3101,
      "num_input_tokens_seen": 24688448,
      "step": 37845
    },
    {
      "epoch": 19.837526205450732,
      "grad_norm": 0.0859375,
      "learning_rate": 6.069469669945926e-06,
      "loss": 0.2727,
      "num_input_tokens_seen": 24691552,
      "step": 37850
    },
    {
      "epoch": 19.84014675052411,
      "grad_norm": 0.12109375,
      "learning_rate": 5.875891322251192e-06,
      "loss": 0.329,
      "num_input_tokens_seen": 24694368,
      "step": 37855
    },
    {
      "epoch": 19.842767295597483,
      "grad_norm": 0.12158203125,
      "learning_rate": 5.685449574258095e-06,
      "loss": 0.3586,
      "num_input_tokens_seen": 24696608,
      "step": 37860
    },
    {
      "epoch": 19.84538784067086,
      "grad_norm": 0.07958984375,
      "learning_rate": 5.498144465804766e-06,
      "loss": 0.3334,
      "num_input_tokens_seen": 24699872,
      "step": 37865
    },
    {
      "epoch": 19.848008385744233,
      "grad_norm": 0.0771484375,
      "learning_rate": 5.313976036073198e-06,
      "loss": 0.3214,
      "num_input_tokens_seen": 24704608,
      "step": 37870
    },
    {
      "epoch": 19.85062893081761,
      "grad_norm": 0.09326171875,
      "learning_rate": 5.132944323589239e-06,
      "loss": 0.3695,
      "num_input_tokens_seen": 24708544,
      "step": 37875
    },
    {
      "epoch": 19.853249475890987,
      "grad_norm": 0.09814453125,
      "learning_rate": 4.955049366224262e-06,
      "loss": 0.3442,
      "num_input_tokens_seen": 24710880,
      "step": 37880
    },
    {
      "epoch": 19.85587002096436,
      "grad_norm": 0.08447265625,
      "learning_rate": 4.7802912011885025e-06,
      "loss": 0.3152,
      "num_input_tokens_seen": 24715008,
      "step": 37885
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 0.126953125,
      "learning_rate": 4.608669865042714e-06,
      "loss": 0.3553,
      "num_input_tokens_seen": 24718240,
      "step": 37890
    },
    {
      "epoch": 19.86111111111111,
      "grad_norm": 0.1416015625,
      "learning_rate": 4.4401853936848474e-06,
      "loss": 0.2841,
      "num_input_tokens_seen": 24720800,
      "step": 37895
    },
    {
      "epoch": 19.863731656184488,
      "grad_norm": 0.068359375,
      "learning_rate": 4.274837822363375e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 24723392,
      "step": 37900
    },
    {
      "epoch": 19.86635220125786,
      "grad_norm": 0.09619140625,
      "learning_rate": 4.112627185665629e-06,
      "loss": 0.3354,
      "num_input_tokens_seen": 24725792,
      "step": 37905
    },
    {
      "epoch": 19.868972746331238,
      "grad_norm": 0.1640625,
      "learning_rate": 3.953553517522801e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 24728672,
      "step": 37910
    },
    {
      "epoch": 19.87159329140461,
      "grad_norm": 0.1103515625,
      "learning_rate": 3.7976168512149356e-06,
      "loss": 0.3918,
      "num_input_tokens_seen": 24732384,
      "step": 37915
    },
    {
      "epoch": 19.87421383647799,
      "grad_norm": 0.25390625,
      "learning_rate": 3.64481721935761e-06,
      "loss": 0.3355,
      "num_input_tokens_seen": 24735744,
      "step": 37920
    },
    {
      "epoch": 19.87683438155136,
      "grad_norm": 0.138671875,
      "learning_rate": 3.4951546539185862e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 24738496,
      "step": 37925
    },
    {
      "epoch": 19.87945492662474,
      "grad_norm": 0.056640625,
      "learning_rate": 3.3486291862028227e-06,
      "loss": 0.3101,
      "num_input_tokens_seen": 24742080,
      "step": 37930
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 0.0830078125,
      "learning_rate": 3.2052408468641324e-06,
      "loss": 0.3147,
      "num_input_tokens_seen": 24744512,
      "step": 37935
    },
    {
      "epoch": 19.88469601677149,
      "grad_norm": 0.056640625,
      "learning_rate": 3.06498966589519e-06,
      "loss": 0.2893,
      "num_input_tokens_seen": 24748224,
      "step": 37940
    },
    {
      "epoch": 19.887316561844862,
      "grad_norm": 0.11865234375,
      "learning_rate": 2.9278756726375255e-06,
      "loss": 0.4524,
      "num_input_tokens_seen": 24751232,
      "step": 37945
    },
    {
      "epoch": 19.88993710691824,
      "grad_norm": 0.051513671875,
      "learning_rate": 2.7938988957715292e-06,
      "loss": 0.3027,
      "num_input_tokens_seen": 24754208,
      "step": 37950
    },
    {
      "epoch": 19.892557651991613,
      "grad_norm": 0.09521484375,
      "learning_rate": 2.6630593633264474e-06,
      "loss": 0.3431,
      "num_input_tokens_seen": 24757440,
      "step": 37955
    },
    {
      "epoch": 19.89517819706499,
      "grad_norm": 0.1328125,
      "learning_rate": 2.5353571026687227e-06,
      "loss": 0.3344,
      "num_input_tokens_seen": 24759808,
      "step": 37960
    },
    {
      "epoch": 19.897798742138363,
      "grad_norm": 0.109375,
      "learning_rate": 2.4107921405153164e-06,
      "loss": 0.2909,
      "num_input_tokens_seen": 24761952,
      "step": 37965
    },
    {
      "epoch": 19.90041928721174,
      "grad_norm": 0.03955078125,
      "learning_rate": 2.289364502922053e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 24765280,
      "step": 37970
    },
    {
      "epoch": 19.903039832285117,
      "grad_norm": 0.12255859375,
      "learning_rate": 2.171074215291946e-06,
      "loss": 0.3704,
      "num_input_tokens_seen": 24768320,
      "step": 37975
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 0.2314453125,
      "learning_rate": 2.055921302368535e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 24770912,
      "step": 37980
    },
    {
      "epoch": 19.908280922431867,
      "grad_norm": 0.099609375,
      "learning_rate": 1.9439057882392194e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 24774176,
      "step": 37985
    },
    {
      "epoch": 19.91090146750524,
      "grad_norm": 0.095703125,
      "learning_rate": 1.8350276963402522e-06,
      "loss": 0.3548,
      "num_input_tokens_seen": 24777120,
      "step": 37990
    },
    {
      "epoch": 19.913522012578618,
      "grad_norm": 0.1005859375,
      "learning_rate": 1.7292870494434176e-06,
      "loss": 0.3099,
      "num_input_tokens_seen": 24779712,
      "step": 37995
    },
    {
      "epoch": 19.91614255765199,
      "grad_norm": 0.09716796875,
      "learning_rate": 1.6266838696710196e-06,
      "loss": 0.4107,
      "num_input_tokens_seen": 24783200,
      "step": 38000
    },
    {
      "epoch": 19.918763102725368,
      "grad_norm": 0.0849609375,
      "learning_rate": 1.5272181784858895e-06,
      "loss": 0.4382,
      "num_input_tokens_seen": 24786592,
      "step": 38005
    },
    {
      "epoch": 19.92138364779874,
      "grad_norm": 0.103515625,
      "learning_rate": 1.4308899966963827e-06,
      "loss": 0.284,
      "num_input_tokens_seen": 24789664,
      "step": 38010
    },
    {
      "epoch": 19.92400419287212,
      "grad_norm": 0.11767578125,
      "learning_rate": 1.3376993444497165e-06,
      "loss": 0.3699,
      "num_input_tokens_seen": 24792288,
      "step": 38015
    },
    {
      "epoch": 19.92662473794549,
      "grad_norm": 0.07421875,
      "learning_rate": 1.2476462412436273e-06,
      "loss": 0.3045,
      "num_input_tokens_seen": 24796000,
      "step": 38020
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 0.10498046875,
      "learning_rate": 1.1607307059163796e-06,
      "loss": 0.322,
      "num_input_tokens_seen": 24798976,
      "step": 38025
    },
    {
      "epoch": 19.931865828092242,
      "grad_norm": 0.11474609375,
      "learning_rate": 1.076952756646765e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 24801760,
      "step": 38030
    },
    {
      "epoch": 19.93448637316562,
      "grad_norm": 0.08203125,
      "learning_rate": 9.9631241096243e-07,
      "loss": 0.3477,
      "num_input_tokens_seen": 24805376,
      "step": 38035
    },
    {
      "epoch": 19.937106918238992,
      "grad_norm": 0.11376953125,
      "learning_rate": 9.188096857315475e-07,
      "loss": 0.4111,
      "num_input_tokens_seen": 24809440,
      "step": 38040
    },
    {
      "epoch": 19.93972746331237,
      "grad_norm": 0.10888671875,
      "learning_rate": 8.444445971678149e-07,
      "loss": 0.375,
      "num_input_tokens_seen": 24813472,
      "step": 38045
    },
    {
      "epoch": 19.942348008385743,
      "grad_norm": 0.0791015625,
      "learning_rate": 7.732171608271221e-07,
      "loss": 0.3797,
      "num_input_tokens_seen": 24816608,
      "step": 38050
    },
    {
      "epoch": 19.94496855345912,
      "grad_norm": 0.10009765625,
      "learning_rate": 7.051273916075517e-07,
      "loss": 0.4294,
      "num_input_tokens_seen": 24819776,
      "step": 38055
    },
    {
      "epoch": 19.947589098532493,
      "grad_norm": 0.10302734375,
      "learning_rate": 6.401753037560409e-07,
      "loss": 0.315,
      "num_input_tokens_seen": 24822560,
      "step": 38060
    },
    {
      "epoch": 19.95020964360587,
      "grad_norm": 0.08642578125,
      "learning_rate": 5.783609108567233e-07,
      "loss": 0.2375,
      "num_input_tokens_seen": 24825600,
      "step": 38065
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 0.1015625,
      "learning_rate": 5.196842258425871e-07,
      "loss": 0.3903,
      "num_input_tokens_seen": 24828224,
      "step": 38070
    },
    {
      "epoch": 19.95545073375262,
      "grad_norm": 0.08447265625,
      "learning_rate": 4.641452609871477e-07,
      "loss": 0.3331,
      "num_input_tokens_seen": 24830848,
      "step": 38075
    },
    {
      "epoch": 19.958071278825997,
      "grad_norm": 0.1044921875,
      "learning_rate": 4.117440279094442e-07,
      "loss": 0.3554,
      "num_input_tokens_seen": 24834304,
      "step": 38080
    },
    {
      "epoch": 19.96069182389937,
      "grad_norm": 0.06884765625,
      "learning_rate": 3.624805375690432e-07,
      "loss": 0.2598,
      "num_input_tokens_seen": 24838112,
      "step": 38085
    },
    {
      "epoch": 19.963312368972748,
      "grad_norm": 0.08544921875,
      "learning_rate": 3.1635480027436545e-07,
      "loss": 0.3303,
      "num_input_tokens_seen": 24842240,
      "step": 38090
    },
    {
      "epoch": 19.96593291404612,
      "grad_norm": 0.07568359375,
      "learning_rate": 2.7336682567269397e-07,
      "loss": 0.2916,
      "num_input_tokens_seen": 24845824,
      "step": 38095
    },
    {
      "epoch": 19.968553459119498,
      "grad_norm": 0.10595703125,
      "learning_rate": 2.335166227551699e-07,
      "loss": 0.3944,
      "num_input_tokens_seen": 24848704,
      "step": 38100
    },
    {
      "epoch": 19.97117400419287,
      "grad_norm": 0.083984375,
      "learning_rate": 1.9680419986178866e-07,
      "loss": 0.2321,
      "num_input_tokens_seen": 24853120,
      "step": 38105
    },
    {
      "epoch": 19.97379454926625,
      "grad_norm": 0.07177734375,
      "learning_rate": 1.6322956466807703e-07,
      "loss": 0.4259,
      "num_input_tokens_seen": 24856800,
      "step": 38110
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 0.11669921875,
      "learning_rate": 1.327927242017468e-07,
      "loss": 0.4084,
      "num_input_tokens_seen": 24859392,
      "step": 38115
    },
    {
      "epoch": 19.979035639413,
      "grad_norm": 0.0830078125,
      "learning_rate": 1.0549368482604126e-07,
      "loss": 0.3776,
      "num_input_tokens_seen": 24863328,
      "step": 38120
    },
    {
      "epoch": 19.981656184486372,
      "grad_norm": 0.06689453125,
      "learning_rate": 8.133245225305784e-08,
      "loss": 0.3241,
      "num_input_tokens_seen": 24866496,
      "step": 38125
    },
    {
      "epoch": 19.98427672955975,
      "grad_norm": 0.10986328125,
      "learning_rate": 6.030903153875222e-08,
      "loss": 0.3356,
      "num_input_tokens_seen": 24871328,
      "step": 38130
    },
    {
      "epoch": 19.986897274633122,
      "grad_norm": 0.1015625,
      "learning_rate": 4.242342707794222e-08,
      "loss": 0.3015,
      "num_input_tokens_seen": 24874336,
      "step": 38135
    },
    {
      "epoch": 19.9895178197065,
      "grad_norm": 0.09619140625,
      "learning_rate": 2.767564261429989e-08,
      "loss": 0.3034,
      "num_input_tokens_seen": 24877024,
      "step": 38140
    },
    {
      "epoch": 19.992138364779873,
      "grad_norm": 0.1474609375,
      "learning_rate": 1.6065681232024784e-08,
      "loss": 0.3531,
      "num_input_tokens_seen": 24880096,
      "step": 38145
    },
    {
      "epoch": 19.99475890985325,
      "grad_norm": 0.0908203125,
      "learning_rate": 7.593545359174669e-09,
      "loss": 0.3312,
      "num_input_tokens_seen": 24883200,
      "step": 38150
    },
    {
      "epoch": 19.997379454926623,
      "grad_norm": 0.17578125,
      "learning_rate": 2.259236769330819e-09,
      "loss": 0.3271,
      "num_input_tokens_seen": 24885664,
      "step": 38155
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.02099609375,
      "learning_rate": 6.275657826737557e-11,
      "loss": 0.2443,
      "num_input_tokens_seen": 24887720,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.5051561594009399,
      "eval_runtime": 13.698,
      "eval_samples_per_second": 61.907,
      "eval_steps_per_second": 15.477,
      "num_input_tokens_seen": 24887720,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 24887720,
      "step": 38160,
      "total_flos": 1.120682783146967e+18,
      "train_loss": 0.42400622701651147,
      "train_runtime": 5899.8605,
      "train_samples_per_second": 25.862,
      "train_steps_per_second": 6.468
    }
  ],
  "logging_steps": 5,
  "max_steps": 38160,
  "num_input_tokens_seen": 24887720,
  "num_train_epochs": 20,
  "save_steps": 1908,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.120682783146967e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}