{
  "best_global_step": 1876,
  "best_metric": 0.5987069010734558,
  "best_model_checkpoint": "saves/p-tuning/llama-3-8b-instruct/train_conala_1754652181/checkpoint-1876",
  "epoch": 10.0,
  "eval_steps": 268,
  "global_step": 5360,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 100.36551666259766,
      "learning_rate": 3.7313432835820895e-07,
      "loss": 4.1856,
      "num_input_tokens_seen": 1216,
      "step": 5
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 154.37203979492188,
      "learning_rate": 8.395522388059702e-07,
      "loss": 3.6219,
      "num_input_tokens_seen": 2528,
      "step": 10
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 42.87506103515625,
      "learning_rate": 1.3059701492537314e-06,
      "loss": 2.7272,
      "num_input_tokens_seen": 4160,
      "step": 15
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 53.51836013793945,
      "learning_rate": 1.7723880597014925e-06,
      "loss": 2.3651,
      "num_input_tokens_seen": 5504,
      "step": 20
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 54.76260757446289,
      "learning_rate": 2.238805970149254e-06,
      "loss": 2.301,
      "num_input_tokens_seen": 6912,
      "step": 25
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 36.159603118896484,
      "learning_rate": 2.705223880597015e-06,
      "loss": 1.8801,
      "num_input_tokens_seen": 8544,
      "step": 30
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 88.50357818603516,
      "learning_rate": 3.171641791044776e-06,
      "loss": 2.4136,
      "num_input_tokens_seen": 9696,
      "step": 35
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 51.878326416015625,
      "learning_rate": 3.6380597014925376e-06,
      "loss": 2.0391,
      "num_input_tokens_seen": 10976,
      "step": 40
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 44.85438537597656,
      "learning_rate": 4.1044776119402985e-06,
      "loss": 1.8727,
      "num_input_tokens_seen": 12320,
      "step": 45
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 26.25723648071289,
      "learning_rate": 4.57089552238806e-06,
      "loss": 1.3923,
      "num_input_tokens_seen": 13920,
      "step": 50
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 29.260101318359375,
      "learning_rate": 5.037313432835821e-06,
      "loss": 1.265,
      "num_input_tokens_seen": 15456,
      "step": 55
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 111.94584655761719,
      "learning_rate": 5.503731343283583e-06,
      "loss": 1.3524,
      "num_input_tokens_seen": 16864,
      "step": 60
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 123.01996612548828,
      "learning_rate": 5.970149253731343e-06,
      "loss": 1.3948,
      "num_input_tokens_seen": 18336,
      "step": 65
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 40.37369918823242,
      "learning_rate": 6.436567164179105e-06,
      "loss": 1.0559,
      "num_input_tokens_seen": 19712,
      "step": 70
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 32.19388198852539,
      "learning_rate": 6.902985074626867e-06,
      "loss": 1.1292,
      "num_input_tokens_seen": 21152,
      "step": 75
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 24.646595001220703,
      "learning_rate": 7.369402985074628e-06,
      "loss": 1.005,
      "num_input_tokens_seen": 22560,
      "step": 80
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 20.98089599609375,
      "learning_rate": 7.835820895522389e-06,
      "loss": 1.1229,
      "num_input_tokens_seen": 23968,
      "step": 85
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 22.397136688232422,
      "learning_rate": 8.30223880597015e-06,
      "loss": 0.7712,
      "num_input_tokens_seen": 25600,
      "step": 90
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 31.489791870117188,
      "learning_rate": 8.768656716417911e-06,
      "loss": 0.9389,
      "num_input_tokens_seen": 26912,
      "step": 95
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 18.812299728393555,
      "learning_rate": 9.235074626865672e-06,
      "loss": 0.989,
      "num_input_tokens_seen": 28512,
      "step": 100
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 146.33045959472656,
      "learning_rate": 9.701492537313434e-06,
      "loss": 1.1056,
      "num_input_tokens_seen": 29888,
      "step": 105
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 51.52705764770508,
      "learning_rate": 1.0167910447761195e-05,
      "loss": 1.1815,
      "num_input_tokens_seen": 31296,
      "step": 110
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 7.95074462890625,
      "learning_rate": 1.0634328358208955e-05,
      "loss": 0.9649,
      "num_input_tokens_seen": 32736,
      "step": 115
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 74.2563705444336,
      "learning_rate": 1.1100746268656717e-05,
      "loss": 1.2157,
      "num_input_tokens_seen": 34240,
      "step": 120
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 158.36537170410156,
      "learning_rate": 1.1567164179104478e-05,
      "loss": 1.1574,
      "num_input_tokens_seen": 35648,
      "step": 125
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 12.419695854187012,
      "learning_rate": 1.203358208955224e-05,
      "loss": 1.0702,
      "num_input_tokens_seen": 36992,
      "step": 130
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 15.341970443725586,
      "learning_rate": 1.25e-05,
      "loss": 0.7606,
      "num_input_tokens_seen": 38528,
      "step": 135
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 20.809242248535156,
      "learning_rate": 1.2966417910447761e-05,
      "loss": 1.0,
      "num_input_tokens_seen": 39840,
      "step": 140
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 15.06401538848877,
      "learning_rate": 1.3432835820895523e-05,
      "loss": 0.7942,
      "num_input_tokens_seen": 41216,
      "step": 145
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 19.199960708618164,
      "learning_rate": 1.3899253731343284e-05,
      "loss": 0.8215,
      "num_input_tokens_seen": 42624,
      "step": 150
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 9.108075141906738,
      "learning_rate": 1.4365671641791046e-05,
      "loss": 0.6431,
      "num_input_tokens_seen": 44192,
      "step": 155
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 7.107698440551758,
      "learning_rate": 1.4832089552238807e-05,
      "loss": 0.9962,
      "num_input_tokens_seen": 45504,
      "step": 160
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 16.22126007080078,
      "learning_rate": 1.529850746268657e-05,
      "loss": 0.6303,
      "num_input_tokens_seen": 47360,
      "step": 165
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 15.512596130371094,
      "learning_rate": 1.5764925373134328e-05,
      "loss": 0.8965,
      "num_input_tokens_seen": 48640,
      "step": 170
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 16.319429397583008,
      "learning_rate": 1.623134328358209e-05,
      "loss": 1.0441,
      "num_input_tokens_seen": 50016,
      "step": 175
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 12.855195999145508,
      "learning_rate": 1.6697761194029852e-05,
      "loss": 1.0999,
      "num_input_tokens_seen": 51296,
      "step": 180
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 16.59507179260254,
      "learning_rate": 1.716417910447761e-05,
      "loss": 1.1909,
      "num_input_tokens_seen": 52576,
      "step": 185
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 6.062872886657715,
      "learning_rate": 1.7630597014925373e-05,
      "loss": 0.617,
      "num_input_tokens_seen": 54080,
      "step": 190
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 15.93188190460205,
      "learning_rate": 1.8097014925373135e-05,
      "loss": 1.1725,
      "num_input_tokens_seen": 55392,
      "step": 195
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 18.400327682495117,
      "learning_rate": 1.8563432835820898e-05,
      "loss": 0.9259,
      "num_input_tokens_seen": 56864,
      "step": 200
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 5.059667110443115,
      "learning_rate": 1.9029850746268656e-05,
      "loss": 0.8611,
      "num_input_tokens_seen": 58368,
      "step": 205
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 26.0755615234375,
      "learning_rate": 1.949626865671642e-05,
      "loss": 1.2025,
      "num_input_tokens_seen": 59744,
      "step": 210
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 21.821815490722656,
      "learning_rate": 1.996268656716418e-05,
      "loss": 0.8922,
      "num_input_tokens_seen": 61152,
      "step": 215
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 18.10831069946289,
      "learning_rate": 2.042910447761194e-05,
      "loss": 0.8431,
      "num_input_tokens_seen": 62496,
      "step": 220
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 14.457990646362305,
      "learning_rate": 2.0895522388059702e-05,
      "loss": 0.7331,
      "num_input_tokens_seen": 63808,
      "step": 225
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 16.29780387878418,
      "learning_rate": 2.1361940298507464e-05,
      "loss": 0.7996,
      "num_input_tokens_seen": 65088,
      "step": 230
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 17.378437042236328,
      "learning_rate": 2.1828358208955223e-05,
      "loss": 0.8701,
      "num_input_tokens_seen": 66496,
      "step": 235
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 6.193378448486328,
      "learning_rate": 2.2294776119402985e-05,
      "loss": 0.7778,
      "num_input_tokens_seen": 67872,
      "step": 240
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 5.975141525268555,
      "learning_rate": 2.2761194029850747e-05,
      "loss": 1.0044,
      "num_input_tokens_seen": 69216,
      "step": 245
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 4.410932540893555,
      "learning_rate": 2.3227611940298506e-05,
      "loss": 0.9899,
      "num_input_tokens_seen": 70816,
      "step": 250
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 6.7089643478393555,
      "learning_rate": 2.369402985074627e-05,
      "loss": 0.693,
      "num_input_tokens_seen": 72224,
      "step": 255
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 10.53079891204834,
      "learning_rate": 2.416044776119403e-05,
      "loss": 1.1661,
      "num_input_tokens_seen": 73600,
      "step": 260
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 7.824636936187744,
      "learning_rate": 2.4626865671641793e-05,
      "loss": 0.7295,
      "num_input_tokens_seen": 75168,
      "step": 265
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.7194088101387024,
      "eval_runtime": 4.5719,
      "eval_samples_per_second": 52.057,
      "eval_steps_per_second": 13.124,
      "num_input_tokens_seen": 75936,
      "step": 268
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 6.349889755249023,
      "learning_rate": 2.5093283582089555e-05,
      "loss": 0.8866,
      "num_input_tokens_seen": 76544,
      "step": 270
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 4.774112701416016,
      "learning_rate": 2.5559701492537314e-05,
      "loss": 0.739,
      "num_input_tokens_seen": 78080,
      "step": 275
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 5.692472457885742,
      "learning_rate": 2.6026119402985076e-05,
      "loss": 0.6549,
      "num_input_tokens_seen": 79520,
      "step": 280
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 4.53741979598999,
      "learning_rate": 2.6492537313432835e-05,
      "loss": 0.9459,
      "num_input_tokens_seen": 80992,
      "step": 285
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 5.167804718017578,
      "learning_rate": 2.69589552238806e-05,
      "loss": 0.7845,
      "num_input_tokens_seen": 82496,
      "step": 290
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 3.988497495651245,
      "learning_rate": 2.742537313432836e-05,
      "loss": 0.6354,
      "num_input_tokens_seen": 83712,
      "step": 295
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 4.841502666473389,
      "learning_rate": 2.789179104477612e-05,
      "loss": 0.7602,
      "num_input_tokens_seen": 85120,
      "step": 300
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 6.440793991088867,
      "learning_rate": 2.835820895522388e-05,
      "loss": 0.7904,
      "num_input_tokens_seen": 86560,
      "step": 305
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 7.7904372215271,
      "learning_rate": 2.8824626865671646e-05,
      "loss": 0.6207,
      "num_input_tokens_seen": 87936,
      "step": 310
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 6.175988674163818,
      "learning_rate": 2.92910447761194e-05,
      "loss": 0.757,
      "num_input_tokens_seen": 89312,
      "step": 315
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 7.592132091522217,
      "learning_rate": 2.9757462686567167e-05,
      "loss": 0.8405,
      "num_input_tokens_seen": 90560,
      "step": 320
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 4.181941509246826,
      "learning_rate": 3.0223880597014926e-05,
      "loss": 0.8746,
      "num_input_tokens_seen": 91936,
      "step": 325
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 4.667055130004883,
      "learning_rate": 3.069029850746269e-05,
      "loss": 0.6,
      "num_input_tokens_seen": 93344,
      "step": 330
    },
    {
      "epoch": 0.625,
      "grad_norm": 3.164530038833618,
      "learning_rate": 3.115671641791045e-05,
      "loss": 0.8976,
      "num_input_tokens_seen": 94560,
      "step": 335
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 3.0226097106933594,
      "learning_rate": 3.162313432835821e-05,
      "loss": 0.6209,
      "num_input_tokens_seen": 96032,
      "step": 340
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 3.280397891998291,
      "learning_rate": 3.208955223880597e-05,
      "loss": 0.8654,
      "num_input_tokens_seen": 97504,
      "step": 345
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 4.144200801849365,
      "learning_rate": 3.2555970149253733e-05,
      "loss": 0.9549,
      "num_input_tokens_seen": 98720,
      "step": 350
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 5.324551582336426,
      "learning_rate": 3.302238805970149e-05,
      "loss": 0.8041,
      "num_input_tokens_seen": 100064,
      "step": 355
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 4.421558380126953,
      "learning_rate": 3.348880597014926e-05,
      "loss": 0.6841,
      "num_input_tokens_seen": 101536,
      "step": 360
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 3.013723850250244,
      "learning_rate": 3.395522388059701e-05,
      "loss": 0.7948,
      "num_input_tokens_seen": 103072,
      "step": 365
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 4.454566478729248,
      "learning_rate": 3.4421641791044776e-05,
      "loss": 0.9889,
      "num_input_tokens_seen": 104416,
      "step": 370
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 3.346811532974243,
      "learning_rate": 3.488805970149254e-05,
      "loss": 0.7354,
      "num_input_tokens_seen": 105952,
      "step": 375
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 4.115444183349609,
      "learning_rate": 3.53544776119403e-05,
      "loss": 0.7968,
      "num_input_tokens_seen": 107520,
      "step": 380
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 2.6957967281341553,
      "learning_rate": 3.582089552238806e-05,
      "loss": 0.9186,
      "num_input_tokens_seen": 109248,
      "step": 385
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 6.732184410095215,
      "learning_rate": 3.6287313432835824e-05,
      "loss": 0.8202,
      "num_input_tokens_seen": 110752,
      "step": 390
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 3.8247575759887695,
      "learning_rate": 3.675373134328358e-05,
      "loss": 0.6186,
      "num_input_tokens_seen": 112320,
      "step": 395
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 2.5585551261901855,
      "learning_rate": 3.722014925373135e-05,
      "loss": 0.911,
      "num_input_tokens_seen": 113792,
      "step": 400
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 4.387516498565674,
      "learning_rate": 3.7686567164179104e-05,
      "loss": 0.6569,
      "num_input_tokens_seen": 115328,
      "step": 405
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 4.66252326965332,
      "learning_rate": 3.8152985074626867e-05,
      "loss": 0.6056,
      "num_input_tokens_seen": 116800,
      "step": 410
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 2.222792387008667,
      "learning_rate": 3.861940298507463e-05,
      "loss": 0.6315,
      "num_input_tokens_seen": 118592,
      "step": 415
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 12.575801849365234,
      "learning_rate": 3.908582089552239e-05,
      "loss": 0.6381,
      "num_input_tokens_seen": 119936,
      "step": 420
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 3.243664026260376,
      "learning_rate": 3.9552238805970146e-05,
      "loss": 0.9194,
      "num_input_tokens_seen": 121280,
      "step": 425
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 2.1990668773651123,
      "learning_rate": 4.0018656716417915e-05,
      "loss": 0.6753,
      "num_input_tokens_seen": 122944,
      "step": 430
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 3.342103958129883,
      "learning_rate": 4.048507462686567e-05,
      "loss": 0.6586,
      "num_input_tokens_seen": 124320,
      "step": 435
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 2.6422414779663086,
      "learning_rate": 4.095149253731343e-05,
      "loss": 0.818,
      "num_input_tokens_seen": 125888,
      "step": 440
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 1.9941433668136597,
      "learning_rate": 4.1417910447761195e-05,
      "loss": 0.7356,
      "num_input_tokens_seen": 127296,
      "step": 445
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 6.570806503295898,
      "learning_rate": 4.188432835820896e-05,
      "loss": 0.7253,
      "num_input_tokens_seen": 128512,
      "step": 450
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 2.254293918609619,
      "learning_rate": 4.235074626865671e-05,
      "loss": 0.6539,
      "num_input_tokens_seen": 130048,
      "step": 455
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 3.071747064590454,
      "learning_rate": 4.281716417910448e-05,
      "loss": 0.6185,
      "num_input_tokens_seen": 131424,
      "step": 460
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 1.909755825996399,
      "learning_rate": 4.328358208955224e-05,
      "loss": 0.5762,
      "num_input_tokens_seen": 132704,
      "step": 465
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 6.39342737197876,
      "learning_rate": 4.375e-05,
      "loss": 0.8091,
      "num_input_tokens_seen": 134048,
      "step": 470
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 1.3354735374450684,
      "learning_rate": 4.421641791044777e-05,
      "loss": 0.6196,
      "num_input_tokens_seen": 135392,
      "step": 475
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 2.911376953125,
      "learning_rate": 4.4682835820895524e-05,
      "loss": 0.9791,
      "num_input_tokens_seen": 136832,
      "step": 480
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 3.955244541168213,
      "learning_rate": 4.5149253731343286e-05,
      "loss": 0.8659,
      "num_input_tokens_seen": 138336,
      "step": 485
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 2.3309555053710938,
      "learning_rate": 4.561567164179105e-05,
      "loss": 0.5936,
      "num_input_tokens_seen": 139840,
      "step": 490
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 4.691043853759766,
      "learning_rate": 4.608208955223881e-05,
      "loss": 0.6507,
      "num_input_tokens_seen": 141344,
      "step": 495
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 2.383185386657715,
      "learning_rate": 4.6548507462686566e-05,
      "loss": 0.6802,
      "num_input_tokens_seen": 142848,
      "step": 500
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 1.5926017761230469,
      "learning_rate": 4.7014925373134335e-05,
      "loss": 0.6252,
      "num_input_tokens_seen": 144384,
      "step": 505
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 3.094663619995117,
      "learning_rate": 4.748134328358209e-05,
      "loss": 0.6142,
      "num_input_tokens_seen": 145760,
      "step": 510
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 1.5612682104110718,
      "learning_rate": 4.794776119402985e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 147168,
      "step": 515
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 1.528640627861023,
      "learning_rate": 4.8414179104477615e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 148480,
      "step": 520
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 1.8824248313903809,
      "learning_rate": 4.888059701492538e-05,
      "loss": 0.6769,
      "num_input_tokens_seen": 149920,
      "step": 525
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 2.4537553787231445,
      "learning_rate": 4.934701492537314e-05,
      "loss": 0.9858,
      "num_input_tokens_seen": 151264,
      "step": 530
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 1.8029956817626953,
      "learning_rate": 4.98134328358209e-05,
      "loss": 0.9109,
      "num_input_tokens_seen": 152608,
      "step": 535
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.6631675958633423,
      "eval_runtime": 4.5705,
      "eval_samples_per_second": 52.073,
      "eval_steps_per_second": 13.128,
      "num_input_tokens_seen": 152672,
      "step": 536
    },
    {
      "epoch": 1.007462686567164,
      "grad_norm": 1.1932291984558105,
      "learning_rate": 4.999995228691131e-05,
      "loss": 0.787,
      "num_input_tokens_seen": 153760,
      "step": 540
    },
    {
      "epoch": 1.0167910447761195,
      "grad_norm": 2.202826499938965,
      "learning_rate": 4.999966070758437e-05,
      "loss": 0.8016,
      "num_input_tokens_seen": 155104,
      "step": 545
    },
    {
      "epoch": 1.0261194029850746,
      "grad_norm": 2.127169609069824,
      "learning_rate": 4.999910405928983e-05,
      "loss": 0.9582,
      "num_input_tokens_seen": 156416,
      "step": 550
    },
    {
      "epoch": 1.0354477611940298,
      "grad_norm": 2.5556702613830566,
      "learning_rate": 4.9998282347929784e-05,
      "loss": 0.856,
      "num_input_tokens_seen": 157888,
      "step": 555
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 3.145081043243408,
      "learning_rate": 4.999719558221674e-05,
      "loss": 0.6926,
      "num_input_tokens_seen": 159392,
      "step": 560
    },
    {
      "epoch": 1.0541044776119404,
      "grad_norm": 3.977997064590454,
      "learning_rate": 4.999584377367359e-05,
      "loss": 0.8115,
      "num_input_tokens_seen": 160864,
      "step": 565
    },
    {
      "epoch": 1.0634328358208955,
      "grad_norm": 2.5034961700439453,
      "learning_rate": 4.9994226936633415e-05,
      "loss": 0.8986,
      "num_input_tokens_seen": 162240,
      "step": 570
    },
    {
      "epoch": 1.0727611940298507,
      "grad_norm": 1.4600768089294434,
      "learning_rate": 4.999234508823938e-05,
      "loss": 0.7446,
      "num_input_tokens_seen": 163616,
      "step": 575
    },
    {
      "epoch": 1.0820895522388059,
      "grad_norm": 7.328063488006592,
      "learning_rate": 4.999019824844455e-05,
      "loss": 1.0083,
      "num_input_tokens_seen": 165152,
      "step": 580
    },
    {
      "epoch": 1.0914179104477613,
      "grad_norm": 1.6883888244628906,
      "learning_rate": 4.998778644001165e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 166560,
      "step": 585
    },
    {
      "epoch": 1.1007462686567164,
      "grad_norm": 1.8932539224624634,
      "learning_rate": 4.9985109688512854e-05,
      "loss": 0.6506,
      "num_input_tokens_seen": 167904,
      "step": 590
    },
    {
      "epoch": 1.1100746268656716,
      "grad_norm": 2.2898571491241455,
      "learning_rate": 4.998216802232949e-05,
      "loss": 0.7991,
      "num_input_tokens_seen": 169280,
      "step": 595
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 2.3847858905792236,
      "learning_rate": 4.9978961472651774e-05,
      "loss": 0.6094,
      "num_input_tokens_seen": 170656,
      "step": 600
    },
    {
      "epoch": 1.1287313432835822,
      "grad_norm": 1.958328366279602,
      "learning_rate": 4.997549007347842e-05,
      "loss": 0.6961,
      "num_input_tokens_seen": 172352,
      "step": 605
    },
    {
      "epoch": 1.1380597014925373,
      "grad_norm": 2.0680088996887207,
      "learning_rate": 4.9971753861616354e-05,
      "loss": 0.5458,
      "num_input_tokens_seen": 173888,
      "step": 610
    },
    {
      "epoch": 1.1473880597014925,
      "grad_norm": 1.2166048288345337,
      "learning_rate": 4.9967752876680254e-05,
      "loss": 0.9095,
      "num_input_tokens_seen": 175040,
      "step": 615
    },
    {
      "epoch": 1.1567164179104479,
      "grad_norm": 1.9659861326217651,
      "learning_rate": 4.996348716109217e-05,
      "loss": 0.6201,
      "num_input_tokens_seen": 176480,
      "step": 620
    },
    {
      "epoch": 1.166044776119403,
      "grad_norm": 1.7151298522949219,
      "learning_rate": 4.9958956760081085e-05,
      "loss": 0.5944,
      "num_input_tokens_seen": 177952,
      "step": 625
    },
    {
      "epoch": 1.1753731343283582,
      "grad_norm": 2.7638487815856934,
      "learning_rate": 4.995416172168239e-05,
      "loss": 0.9196,
      "num_input_tokens_seen": 179552,
      "step": 630
    },
    {
      "epoch": 1.1847014925373134,
      "grad_norm": 1.9493519067764282,
      "learning_rate": 4.994910209673741e-05,
      "loss": 0.7352,
      "num_input_tokens_seen": 181120,
      "step": 635
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 1.1021276712417603,
      "learning_rate": 4.9943777938892855e-05,
      "loss": 0.6201,
      "num_input_tokens_seen": 182560,
      "step": 640
    },
    {
      "epoch": 1.203358208955224,
      "grad_norm": 1.6290100812911987,
      "learning_rate": 4.993818930460026e-05,
      "loss": 0.6135,
      "num_input_tokens_seen": 184064,
      "step": 645
    },
    {
      "epoch": 1.212686567164179,
      "grad_norm": 0.8198497891426086,
      "learning_rate": 4.9932336253115354e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 185600,
      "step": 650
    },
    {
      "epoch": 1.2220149253731343,
      "grad_norm": 1.588991403579712,
      "learning_rate": 4.9926218846497486e-05,
      "loss": 0.6244,
      "num_input_tokens_seen": 186848,
      "step": 655
    },
    {
      "epoch": 1.2313432835820897,
      "grad_norm": 1.2917518615722656,
      "learning_rate": 4.991983714960892e-05,
      "loss": 0.5736,
      "num_input_tokens_seen": 188224,
      "step": 660
    },
    {
      "epoch": 1.2406716417910448,
      "grad_norm": 1.2897745370864868,
      "learning_rate": 4.9913191230114156e-05,
      "loss": 0.5902,
      "num_input_tokens_seen": 189696,
      "step": 665
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.5458823442459106,
      "learning_rate": 4.990628115847924e-05,
      "loss": 0.7197,
      "num_input_tokens_seen": 191360,
      "step": 670
    },
    {
      "epoch": 1.2593283582089552,
      "grad_norm": 1.6638667583465576,
      "learning_rate": 4.9899107007971004e-05,
      "loss": 0.6034,
      "num_input_tokens_seen": 192800,
      "step": 675
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 2.221782684326172,
      "learning_rate": 4.989166885465624e-05,
      "loss": 0.9335,
      "num_input_tokens_seen": 194208,
      "step": 680
    },
    {
      "epoch": 1.2779850746268657,
      "grad_norm": 2.3120720386505127,
      "learning_rate": 4.988396677740097e-05,
      "loss": 0.7154,
      "num_input_tokens_seen": 195552,
      "step": 685
    },
    {
      "epoch": 1.287313432835821,
      "grad_norm": 0.9551522731781006,
      "learning_rate": 4.9876000857869583e-05,
      "loss": 0.546,
      "num_input_tokens_seen": 196832,
      "step": 690
    },
    {
      "epoch": 1.296641791044776,
      "grad_norm": 1.3219599723815918,
      "learning_rate": 4.986777118052393e-05,
      "loss": 0.6161,
      "num_input_tokens_seen": 198112,
      "step": 695
    },
    {
      "epoch": 1.3059701492537314,
      "grad_norm": 1.8073735237121582,
      "learning_rate": 4.9859277832622494e-05,
      "loss": 0.7514,
      "num_input_tokens_seen": 199680,
      "step": 700
    },
    {
      "epoch": 1.3152985074626866,
      "grad_norm": 1.543784260749817,
      "learning_rate": 4.98505209042194e-05,
      "loss": 0.7255,
      "num_input_tokens_seen": 200960,
      "step": 705
    },
    {
      "epoch": 1.3246268656716418,
      "grad_norm": 1.5509998798370361,
      "learning_rate": 4.9841500488163526e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 202368,
      "step": 710
    },
    {
      "epoch": 1.333955223880597,
      "grad_norm": 1.4129292964935303,
      "learning_rate": 4.983221668009744e-05,
      "loss": 0.6942,
      "num_input_tokens_seen": 203744,
      "step": 715
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 1.8519418239593506,
      "learning_rate": 4.982266957845648e-05,
      "loss": 0.6477,
      "num_input_tokens_seen": 205056,
      "step": 720
    },
    {
      "epoch": 1.3526119402985075,
      "grad_norm": 1.606263279914856,
      "learning_rate": 4.981285928446762e-05,
      "loss": 0.6991,
      "num_input_tokens_seen": 206432,
      "step": 725
    },
    {
      "epoch": 1.3619402985074627,
      "grad_norm": 1.5719319581985474,
      "learning_rate": 4.9802785902148455e-05,
      "loss": 0.9099,
      "num_input_tokens_seen": 207712,
      "step": 730
    },
    {
      "epoch": 1.3712686567164178,
      "grad_norm": 0.8573477864265442,
      "learning_rate": 4.979244953830608e-05,
      "loss": 0.6248,
      "num_input_tokens_seen": 208992,
      "step": 735
    },
    {
      "epoch": 1.3805970149253732,
      "grad_norm": 1.813030481338501,
      "learning_rate": 4.9781850302535945e-05,
      "loss": 0.8095,
      "num_input_tokens_seen": 210400,
      "step": 740
    },
    {
      "epoch": 1.3899253731343284,
      "grad_norm": 2.0639684200286865,
      "learning_rate": 4.9770988307220736e-05,
      "loss": 0.7118,
      "num_input_tokens_seen": 211936,
      "step": 745
    },
    {
      "epoch": 1.3992537313432836,
      "grad_norm": 1.5449402332305908,
      "learning_rate": 4.97598636675291e-05,
      "loss": 0.6724,
      "num_input_tokens_seen": 213376,
      "step": 750
    },
    {
      "epoch": 1.4085820895522387,
      "grad_norm": 2.0232129096984863,
      "learning_rate": 4.974847650141453e-05,
      "loss": 0.8811,
      "num_input_tokens_seen": 214784,
      "step": 755
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 2.43575382232666,
      "learning_rate": 4.973682692961403e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 216224,
      "step": 760
    },
    {
      "epoch": 1.4272388059701493,
      "grad_norm": 2.0031626224517822,
      "learning_rate": 4.972491507564688e-05,
      "loss": 0.857,
      "num_input_tokens_seen": 217600,
      "step": 765
    },
    {
      "epoch": 1.4365671641791045,
      "grad_norm": 1.3304368257522583,
      "learning_rate": 4.971274106581331e-05,
      "loss": 0.7182,
      "num_input_tokens_seen": 219040,
      "step": 770
    },
    {
      "epoch": 1.4458955223880596,
      "grad_norm": 4.4156646728515625,
      "learning_rate": 4.970030502919315e-05,
      "loss": 0.8641,
      "num_input_tokens_seen": 220480,
      "step": 775
    },
    {
      "epoch": 1.455223880597015,
      "grad_norm": 1.0709304809570312,
      "learning_rate": 4.9687607097644495e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 222208,
      "step": 780
    },
    {
      "epoch": 1.4645522388059702,
      "grad_norm": 1.6765780448913574,
      "learning_rate": 4.967464740580227e-05,
      "loss": 0.5145,
      "num_input_tokens_seen": 223648,
      "step": 785
    },
    {
      "epoch": 1.4738805970149254,
      "grad_norm": 1.2266260385513306,
      "learning_rate": 4.9661426091076834e-05,
      "loss": 0.5974,
      "num_input_tokens_seen": 225152,
      "step": 790
    },
    {
      "epoch": 1.4832089552238805,
      "grad_norm": 1.4691288471221924,
      "learning_rate": 4.9647943293652486e-05,
      "loss": 0.689,
      "num_input_tokens_seen": 226528,
      "step": 795
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 2.961132287979126,
      "learning_rate": 4.963419915648603e-05,
      "loss": 0.6458,
      "num_input_tokens_seen": 228000,
      "step": 800
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.6388996839523315,
      "eval_runtime": 4.5898,
      "eval_samples_per_second": 51.854,
      "eval_steps_per_second": 13.072,
      "num_input_tokens_seen": 229344,
      "step": 804
    },
    {
      "epoch": 1.501865671641791,
      "grad_norm": 1.2205604314804077,
      "learning_rate": 4.962019382530521e-05,
      "loss": 0.488,
      "num_input_tokens_seen": 229600,
      "step": 805
    },
    {
      "epoch": 1.5111940298507462,
      "grad_norm": 0.8918901681900024,
      "learning_rate": 4.960592744860717e-05,
      "loss": 0.6591,
      "num_input_tokens_seen": 230912,
      "step": 810
    },
    {
      "epoch": 1.5205223880597014,
      "grad_norm": 1.3232108354568481,
      "learning_rate": 4.9591400177656935e-05,
      "loss": 0.5742,
      "num_input_tokens_seen": 232256,
      "step": 815
    },
    {
      "epoch": 1.5298507462686568,
      "grad_norm": 1.5085234642028809,
      "learning_rate": 4.957661216648573e-05,
      "loss": 0.4847,
      "num_input_tokens_seen": 233728,
      "step": 820
    },
    {
      "epoch": 1.539179104477612,
      "grad_norm": 1.1434234380722046,
      "learning_rate": 4.95615635718894e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 235072,
      "step": 825
    },
    {
      "epoch": 1.5485074626865671,
      "grad_norm": 1.1906145811080933,
      "learning_rate": 4.954625455342674e-05,
      "loss": 0.7332,
      "num_input_tokens_seen": 236512,
      "step": 830
    },
    {
      "epoch": 1.5578358208955225,
      "grad_norm": 1.3204833269119263,
      "learning_rate": 4.953068527341777e-05,
      "loss": 0.5138,
      "num_input_tokens_seen": 237952,
      "step": 835
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 1.231200098991394,
      "learning_rate": 4.9514855896942046e-05,
      "loss": 0.6462,
      "num_input_tokens_seen": 239328,
      "step": 840
    },
    {
      "epoch": 1.5764925373134329,
      "grad_norm": 1.188966155052185,
      "learning_rate": 4.949876659183692e-05,
      "loss": 0.6347,
      "num_input_tokens_seen": 240736,
      "step": 845
    },
    {
      "epoch": 1.585820895522388,
      "grad_norm": 1.532457947731018,
      "learning_rate": 4.948241752869571e-05,
      "loss": 0.6066,
      "num_input_tokens_seen": 241984,
      "step": 850
    },
    {
      "epoch": 1.5951492537313432,
      "grad_norm": 1.5619561672210693,
      "learning_rate": 4.946580888086595e-05,
      "loss": 0.7548,
      "num_input_tokens_seen": 243296,
      "step": 855
    },
    {
      "epoch": 1.6044776119402986,
      "grad_norm": 1.5813361406326294,
      "learning_rate": 4.9448940824447515e-05,
      "loss": 0.5763,
      "num_input_tokens_seen": 244960,
      "step": 860
    },
    {
      "epoch": 1.6138059701492538,
      "grad_norm": 2.172302007675171,
      "learning_rate": 4.943181353829076e-05,
      "loss": 0.7208,
      "num_input_tokens_seen": 246304,
      "step": 865
    },
    {
      "epoch": 1.623134328358209,
      "grad_norm": 3.6299030780792236,
      "learning_rate": 4.941442720399464e-05,
      "loss": 0.7044,
      "num_input_tokens_seen": 247680,
      "step": 870
    },
    {
      "epoch": 1.6324626865671643,
      "grad_norm": 1.2730520963668823,
      "learning_rate": 4.939678200590475e-05,
      "loss": 0.6817,
      "num_input_tokens_seen": 249152,
      "step": 875
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 1.2772308588027954,
      "learning_rate": 4.937887813111142e-05,
      "loss": 0.6536,
      "num_input_tokens_seen": 250496,
      "step": 880
    },
    {
      "epoch": 1.6511194029850746,
      "grad_norm": 1.3259509801864624,
      "learning_rate": 4.936071576944769e-05,
      "loss": 0.6178,
      "num_input_tokens_seen": 251872,
      "step": 885
    },
    {
      "epoch": 1.6604477611940298,
      "grad_norm": 1.3362723588943481,
      "learning_rate": 4.93422951134873e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 253440,
      "step": 890
    },
    {
      "epoch": 1.669776119402985,
      "grad_norm": 1.3205769062042236,
      "learning_rate": 4.932361635854268e-05,
      "loss": 0.5977,
      "num_input_tokens_seen": 254720,
      "step": 895
    },
    {
      "epoch": 1.6791044776119404,
      "grad_norm": 4.008857727050781,
      "learning_rate": 4.9304679702662854e-05,
      "loss": 0.8268,
      "num_input_tokens_seen": 256320,
      "step": 900
    },
    {
      "epoch": 1.6884328358208955,
      "grad_norm": 1.8653584718704224,
      "learning_rate": 4.9285485346631334e-05,
      "loss": 0.6047,
      "num_input_tokens_seen": 257600,
      "step": 905
    },
    {
      "epoch": 1.6977611940298507,
      "grad_norm": 1.1812169551849365,
      "learning_rate": 4.9266033493964e-05,
      "loss": 0.4307,
      "num_input_tokens_seen": 259008,
      "step": 910
    },
    {
      "epoch": 1.707089552238806,
      "grad_norm": 1.4658585786819458,
      "learning_rate": 4.924632435090696e-05,
      "loss": 0.4938,
      "num_input_tokens_seen": 260448,
      "step": 915
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 1.910929560661316,
      "learning_rate": 4.922635812643434e-05,
      "loss": 0.7848,
      "num_input_tokens_seen": 261888,
      "step": 920
    },
    {
      "epoch": 1.7257462686567164,
      "grad_norm": 1.0371445417404175,
      "learning_rate": 4.920613503224608e-05,
      "loss": 0.6437,
      "num_input_tokens_seen": 263328,
      "step": 925
    },
    {
      "epoch": 1.7350746268656716,
      "grad_norm": 1.303602933883667,
      "learning_rate": 4.9185655282765655e-05,
      "loss": 0.5388,
      "num_input_tokens_seen": 264832,
      "step": 930
    },
    {
      "epoch": 1.7444029850746268,
      "grad_norm": 1.1543775796890259,
      "learning_rate": 4.916491909513787e-05,
      "loss": 0.9968,
      "num_input_tokens_seen": 266368,
      "step": 935
    },
    {
      "epoch": 1.7537313432835822,
      "grad_norm": 1.099280595779419,
      "learning_rate": 4.914392668922651e-05,
      "loss": 0.8145,
      "num_input_tokens_seen": 267904,
      "step": 940
    },
    {
      "epoch": 1.7630597014925373,
      "grad_norm": 1.097183108329773,
      "learning_rate": 4.912267828761199e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 269504,
      "step": 945
    },
    {
      "epoch": 1.7723880597014925,
      "grad_norm": 1.1116201877593994,
      "learning_rate": 4.910117411558906e-05,
      "loss": 0.4948,
      "num_input_tokens_seen": 270912,
      "step": 950
    },
    {
      "epoch": 1.7817164179104479,
      "grad_norm": 0.7362908124923706,
      "learning_rate": 4.907941440116436e-05,
      "loss": 0.5829,
      "num_input_tokens_seen": 272384,
      "step": 955
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 1.39719820022583,
      "learning_rate": 4.905739937505401e-05,
      "loss": 0.6633,
      "num_input_tokens_seen": 273760,
      "step": 960
    },
    {
      "epoch": 1.8003731343283582,
      "grad_norm": 0.9451549053192139,
      "learning_rate": 4.9035129270681196e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 275424,
      "step": 965
    },
    {
      "epoch": 1.8097014925373134,
      "grad_norm": 3.1764304637908936,
      "learning_rate": 4.901260432417367e-05,
      "loss": 0.8598,
      "num_input_tokens_seen": 276832,
      "step": 970
    },
    {
      "epoch": 1.8190298507462686,
      "grad_norm": 1.4072598218917847,
      "learning_rate": 4.8989824774361236e-05,
      "loss": 0.8787,
      "num_input_tokens_seen": 278336,
      "step": 975
    },
    {
      "epoch": 1.828358208955224,
      "grad_norm": 1.3940120935440063,
      "learning_rate": 4.896679086277325e-05,
      "loss": 0.6351,
      "num_input_tokens_seen": 279712,
      "step": 980
    },
    {
      "epoch": 1.837686567164179,
      "grad_norm": 1.6442984342575073,
      "learning_rate": 4.8943502833636026e-05,
      "loss": 0.6894,
      "num_input_tokens_seen": 281152,
      "step": 985
    },
    {
      "epoch": 1.8470149253731343,
      "grad_norm": 3.0127034187316895,
      "learning_rate": 4.891996093387028e-05,
      "loss": 0.7872,
      "num_input_tokens_seen": 282560,
      "step": 990
    },
    {
      "epoch": 1.8563432835820897,
      "grad_norm": 1.2732431888580322,
      "learning_rate": 4.889616541308847e-05,
      "loss": 0.5937,
      "num_input_tokens_seen": 283968,
      "step": 995
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 1.5963600873947144,
      "learning_rate": 4.8872116523592196e-05,
      "loss": 0.6134,
      "num_input_tokens_seen": 285440,
      "step": 1000
    },
    {
      "epoch": 1.875,
      "grad_norm": 1.0796291828155518,
      "learning_rate": 4.8847814520369475e-05,
      "loss": 0.5015,
      "num_input_tokens_seen": 286944,
      "step": 1005
    },
    {
      "epoch": 1.8843283582089554,
      "grad_norm": 1.0519851446151733,
      "learning_rate": 4.8823259661092104e-05,
      "loss": 0.6325,
      "num_input_tokens_seen": 288320,
      "step": 1010
    },
    {
      "epoch": 1.8936567164179103,
      "grad_norm": 1.9387621879577637,
      "learning_rate": 4.879845220611284e-05,
      "loss": 0.5384,
      "num_input_tokens_seen": 289568,
      "step": 1015
    },
    {
      "epoch": 1.9029850746268657,
      "grad_norm": 1.5977712869644165,
      "learning_rate": 4.877339241846273e-05,
      "loss": 0.8395,
      "num_input_tokens_seen": 290976,
      "step": 1020
    },
    {
      "epoch": 1.912313432835821,
      "grad_norm": 1.65835440158844,
      "learning_rate": 4.874808056384825e-05,
      "loss": 0.7479,
      "num_input_tokens_seen": 292576,
      "step": 1025
    },
    {
      "epoch": 1.921641791044776,
      "grad_norm": 1.360770344734192,
      "learning_rate": 4.872251691064854e-05,
      "loss": 0.6412,
      "num_input_tokens_seen": 293952,
      "step": 1030
    },
    {
      "epoch": 1.9309701492537314,
      "grad_norm": 1.352001667022705,
      "learning_rate": 4.869670172991252e-05,
      "loss": 0.5933,
      "num_input_tokens_seen": 295488,
      "step": 1035
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 1.4881118535995483,
      "learning_rate": 4.8670635295356035e-05,
      "loss": 0.6816,
      "num_input_tokens_seen": 296768,
      "step": 1040
    },
    {
      "epoch": 1.9496268656716418,
      "grad_norm": 0.9687556028366089,
      "learning_rate": 4.8644317883358956e-05,
      "loss": 0.5105,
      "num_input_tokens_seen": 298240,
      "step": 1045
    },
    {
      "epoch": 1.9589552238805972,
      "grad_norm": 0.8841202855110168,
      "learning_rate": 4.861774977296223e-05,
      "loss": 0.5589,
      "num_input_tokens_seen": 299712,
      "step": 1050
    },
    {
      "epoch": 1.9682835820895521,
      "grad_norm": 1.7201465368270874,
      "learning_rate": 4.8590931245864954e-05,
      "loss": 0.6308,
      "num_input_tokens_seen": 300864,
      "step": 1055
    },
    {
      "epoch": 1.9776119402985075,
      "grad_norm": 1.226158618927002,
      "learning_rate": 4.856386258642135e-05,
      "loss": 0.6345,
      "num_input_tokens_seen": 302336,
      "step": 1060
    },
    {
      "epoch": 1.9869402985074627,
      "grad_norm": 1.062418818473816,
      "learning_rate": 4.8536544081637787e-05,
      "loss": 0.6448,
      "num_input_tokens_seen": 303584,
      "step": 1065
    },
    {
      "epoch": 1.9962686567164178,
      "grad_norm": 1.3856908082962036,
      "learning_rate": 4.8508976021169705e-05,
      "loss": 0.6543,
      "num_input_tokens_seen": 304960,
      "step": 1070
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6024589538574219,
      "eval_runtime": 4.5711,
      "eval_samples_per_second": 52.066,
      "eval_steps_per_second": 13.126,
      "num_input_tokens_seen": 305288,
      "step": 1072
    },
    {
      "epoch": 2.0055970149253732,
      "grad_norm": 1.7049146890640259,
      "learning_rate": 4.8481158697318564e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 306152,
      "step": 1075
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 1.2097922563552856,
      "learning_rate": 4.845309240502874e-05,
      "loss": 0.7647,
      "num_input_tokens_seen": 307464,
      "step": 1080
    },
    {
      "epoch": 2.0242537313432836,
      "grad_norm": 0.5486364364624023,
      "learning_rate": 4.8424777441884405e-05,
      "loss": 0.6772,
      "num_input_tokens_seen": 308968,
      "step": 1085
    },
    {
      "epoch": 2.033582089552239,
      "grad_norm": 1.4574507474899292,
      "learning_rate": 4.839621410810634e-05,
      "loss": 0.6623,
      "num_input_tokens_seen": 310440,
      "step": 1090
    },
    {
      "epoch": 2.042910447761194,
      "grad_norm": 1.5538582801818848,
      "learning_rate": 4.8367402706548805e-05,
      "loss": 0.6824,
      "num_input_tokens_seen": 312008,
      "step": 1095
    },
    {
      "epoch": 2.0522388059701493,
      "grad_norm": 1.5001722574234009,
      "learning_rate": 4.8338343542696275e-05,
      "loss": 0.767,
      "num_input_tokens_seen": 313416,
      "step": 1100
    },
    {
      "epoch": 2.0615671641791047,
      "grad_norm": 1.1931625604629517,
      "learning_rate": 4.830903692466024e-05,
      "loss": 0.7107,
      "num_input_tokens_seen": 314728,
      "step": 1105
    },
    {
      "epoch": 2.0708955223880596,
      "grad_norm": 1.4544624090194702,
      "learning_rate": 4.82794831631759e-05,
      "loss": 0.7076,
      "num_input_tokens_seen": 316072,
      "step": 1110
    },
    {
      "epoch": 2.080223880597015,
      "grad_norm": 2.525104522705078,
      "learning_rate": 4.8249682571598945e-05,
      "loss": 0.6707,
      "num_input_tokens_seen": 317384,
      "step": 1115
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 1.1973071098327637,
      "learning_rate": 4.821963546590211e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 318696,
      "step": 1120
    },
    {
      "epoch": 2.0988805970149254,
      "grad_norm": 3.119307518005371,
      "learning_rate": 4.8189342164671944e-05,
      "loss": 0.5901,
      "num_input_tokens_seen": 320008,
      "step": 1125
    },
    {
      "epoch": 2.1082089552238807,
      "grad_norm": 1.5846446752548218,
      "learning_rate": 4.815880298910537e-05,
      "loss": 0.53,
      "num_input_tokens_seen": 321480,
      "step": 1130
    },
    {
      "epoch": 2.1175373134328357,
      "grad_norm": 1.6190916299819946,
      "learning_rate": 4.8128018263006305e-05,
      "loss": 0.4959,
      "num_input_tokens_seen": 323016,
      "step": 1135
    },
    {
      "epoch": 2.126865671641791,
      "grad_norm": 1.514156460762024,
      "learning_rate": 4.8096988312782174e-05,
      "loss": 0.4515,
      "num_input_tokens_seen": 324456,
      "step": 1140
    },
    {
      "epoch": 2.1361940298507465,
      "grad_norm": 0.8373875021934509,
      "learning_rate": 4.806571346744053e-05,
      "loss": 0.3586,
      "num_input_tokens_seen": 326120,
      "step": 1145
    },
    {
      "epoch": 2.1455223880597014,
      "grad_norm": 1.5115240812301636,
      "learning_rate": 4.803419405858553e-05,
      "loss": 0.73,
      "num_input_tokens_seen": 327624,
      "step": 1150
    },
    {
      "epoch": 2.154850746268657,
      "grad_norm": 3.2437195777893066,
      "learning_rate": 4.8002430420414356e-05,
      "loss": 0.5212,
      "num_input_tokens_seen": 329160,
      "step": 1155
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 3.2139501571655273,
      "learning_rate": 4.79704228897138e-05,
      "loss": 0.9175,
      "num_input_tokens_seen": 330472,
      "step": 1160
    },
    {
      "epoch": 2.173507462686567,
      "grad_norm": 1.7413250207901,
      "learning_rate": 4.79381718058566e-05,
      "loss": 0.4151,
      "num_input_tokens_seen": 332008,
      "step": 1165
    },
    {
      "epoch": 2.1828358208955225,
      "grad_norm": 1.1538379192352295,
      "learning_rate": 4.790567751079783e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 333384,
      "step": 1170
    },
    {
      "epoch": 2.1921641791044775,
      "grad_norm": 2.0296406745910645,
      "learning_rate": 4.787294034907135e-05,
      "loss": 0.6664,
      "num_input_tokens_seen": 334696,
      "step": 1175
    },
    {
      "epoch": 2.201492537313433,
      "grad_norm": 1.394704818725586,
      "learning_rate": 4.78399606677861e-05,
      "loss": 0.5698,
      "num_input_tokens_seen": 336232,
      "step": 1180
    },
    {
      "epoch": 2.2108208955223883,
      "grad_norm": 1.2067760229110718,
      "learning_rate": 4.780673881662242e-05,
      "loss": 0.5549,
      "num_input_tokens_seen": 337512,
      "step": 1185
    },
    {
      "epoch": 2.220149253731343,
      "grad_norm": 0.7276526689529419,
      "learning_rate": 4.777327514782837e-05,
      "loss": 0.6195,
      "num_input_tokens_seen": 339080,
      "step": 1190
    },
    {
      "epoch": 2.2294776119402986,
      "grad_norm": 0.6832820773124695,
      "learning_rate": 4.773957001621597e-05,
      "loss": 0.7226,
      "num_input_tokens_seen": 340456,
      "step": 1195
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 0.5864564776420593,
      "learning_rate": 4.7705623779157435e-05,
      "loss": 0.474,
      "num_input_tokens_seen": 341864,
      "step": 1200
    },
    {
      "epoch": 2.248134328358209,
      "grad_norm": 1.5740203857421875,
      "learning_rate": 4.7671436796581426e-05,
      "loss": 0.6855,
      "num_input_tokens_seen": 343240,
      "step": 1205
    },
    {
      "epoch": 2.2574626865671643,
      "grad_norm": 1.1795896291732788,
      "learning_rate": 4.7637009430969194e-05,
      "loss": 0.6987,
      "num_input_tokens_seen": 344808,
      "step": 1210
    },
    {
      "epoch": 2.2667910447761193,
      "grad_norm": 1.8666998147964478,
      "learning_rate": 4.760234204735072e-05,
      "loss": 0.5404,
      "num_input_tokens_seen": 346152,
      "step": 1215
    },
    {
      "epoch": 2.2761194029850746,
      "grad_norm": 1.111821174621582,
      "learning_rate": 4.756743501330091e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 347528,
      "step": 1220
    },
    {
      "epoch": 2.28544776119403,
      "grad_norm": 1.3123027086257935,
      "learning_rate": 4.753228869893566e-05,
      "loss": 0.5621,
      "num_input_tokens_seen": 348776,
      "step": 1225
    },
    {
      "epoch": 2.294776119402985,
      "grad_norm": 1.3569413423538208,
      "learning_rate": 4.7496903476907885e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 350088,
      "step": 1230
    },
    {
      "epoch": 2.3041044776119404,
      "grad_norm": 1.2350666522979736,
      "learning_rate": 4.746127972240367e-05,
      "loss": 0.6737,
      "num_input_tokens_seen": 351528,
      "step": 1235
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 1.0653327703475952,
      "learning_rate": 4.742541781313822e-05,
      "loss": 0.7182,
      "num_input_tokens_seen": 352840,
      "step": 1240
    },
    {
      "epoch": 2.3227611940298507,
      "grad_norm": 0.983496367931366,
      "learning_rate": 4.738931812935186e-05,
      "loss": 0.6179,
      "num_input_tokens_seen": 354152,
      "step": 1245
    },
    {
      "epoch": 2.332089552238806,
      "grad_norm": 1.4893721342086792,
      "learning_rate": 4.735298105380601e-05,
      "loss": 0.7028,
      "num_input_tokens_seen": 355496,
      "step": 1250
    },
    {
      "epoch": 2.341417910447761,
      "grad_norm": 0.8644038438796997,
      "learning_rate": 4.7316406971779145e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 357096,
      "step": 1255
    },
    {
      "epoch": 2.3507462686567164,
      "grad_norm": 1.8214662075042725,
      "learning_rate": 4.7279596271062716e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 358312,
      "step": 1260
    },
    {
      "epoch": 2.360074626865672,
      "grad_norm": 1.3043785095214844,
      "learning_rate": 4.724254934195697e-05,
      "loss": 0.796,
      "num_input_tokens_seen": 359752,
      "step": 1265
    },
    {
      "epoch": 2.3694029850746268,
      "grad_norm": 1.4253908395767212,
      "learning_rate": 4.720526657726691e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 361128,
      "step": 1270
    },
    {
      "epoch": 2.378731343283582,
      "grad_norm": 2.104628086090088,
      "learning_rate": 4.716774837229804e-05,
      "loss": 0.514,
      "num_input_tokens_seen": 362600,
      "step": 1275
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 1.254268765449524,
      "learning_rate": 4.712999512485225e-05,
      "loss": 0.6867,
      "num_input_tokens_seen": 364072,
      "step": 1280
    },
    {
      "epoch": 2.3973880597014925,
      "grad_norm": 2.871674060821533,
      "learning_rate": 4.709200723522353e-05,
      "loss": 0.7446,
      "num_input_tokens_seen": 365576,
      "step": 1285
    },
    {
      "epoch": 2.406716417910448,
      "grad_norm": 1.5138633251190186,
      "learning_rate": 4.7053785106193793e-05,
      "loss": 0.5512,
      "num_input_tokens_seen": 367208,
      "step": 1290
    },
    {
      "epoch": 2.416044776119403,
      "grad_norm": 1.2380428314208984,
      "learning_rate": 4.701532914302853e-05,
      "loss": 0.7371,
      "num_input_tokens_seen": 368680,
      "step": 1295
    },
    {
      "epoch": 2.425373134328358,
      "grad_norm": 0.9417859315872192,
      "learning_rate": 4.697663975347258e-05,
      "loss": 0.7131,
      "num_input_tokens_seen": 370408,
      "step": 1300
    },
    {
      "epoch": 2.4347014925373136,
      "grad_norm": 1.5790616273880005,
      "learning_rate": 4.693771734774578e-05,
      "loss": 0.4707,
      "num_input_tokens_seen": 371784,
      "step": 1305
    },
    {
      "epoch": 2.4440298507462686,
      "grad_norm": 1.7903887033462524,
      "learning_rate": 4.6898562338538606e-05,
      "loss": 0.5913,
      "num_input_tokens_seen": 373256,
      "step": 1310
    },
    {
      "epoch": 2.453358208955224,
      "grad_norm": 0.6431801915168762,
      "learning_rate": 4.6859175141007796e-05,
      "loss": 0.4984,
      "num_input_tokens_seen": 374760,
      "step": 1315
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 1.290738821029663,
      "learning_rate": 4.6819556172771974e-05,
      "loss": 0.5889,
      "num_input_tokens_seen": 376360,
      "step": 1320
    },
    {
      "epoch": 2.4720149253731343,
      "grad_norm": 1.147175908088684,
      "learning_rate": 4.6779705853907205e-05,
      "loss": 0.5453,
      "num_input_tokens_seen": 377960,
      "step": 1325
    },
    {
      "epoch": 2.4813432835820897,
      "grad_norm": 1.5055599212646484,
      "learning_rate": 4.673962460694254e-05,
      "loss": 0.8192,
      "num_input_tokens_seen": 379208,
      "step": 1330
    },
    {
      "epoch": 2.4906716417910446,
      "grad_norm": 1.395520567893982,
      "learning_rate": 4.669931285685553e-05,
      "loss": 0.6262,
      "num_input_tokens_seen": 380744,
      "step": 1335
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.1887707710266113,
      "learning_rate": 4.6658771031067734e-05,
      "loss": 0.6009,
      "num_input_tokens_seen": 382120,
      "step": 1340
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.5999053120613098,
      "eval_runtime": 4.5761,
      "eval_samples_per_second": 52.009,
      "eval_steps_per_second": 13.112,
      "num_input_tokens_seen": 382120,
      "step": 1340
    },
    {
      "epoch": 2.5093283582089554,
      "grad_norm": 1.782746434211731,
      "learning_rate": 4.6617999559440187e-05,
      "loss": 0.7068,
      "num_input_tokens_seen": 383400,
      "step": 1345
    },
    {
      "epoch": 2.5186567164179103,
      "grad_norm": 1.1211272478103638,
      "learning_rate": 4.657699887426884e-05,
      "loss": 0.6096,
      "num_input_tokens_seen": 384680,
      "step": 1350
    },
    {
      "epoch": 2.5279850746268657,
      "grad_norm": 1.4074512720108032,
      "learning_rate": 4.653576941027995e-05,
      "loss": 0.652,
      "num_input_tokens_seen": 385992,
      "step": 1355
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 0.8100829720497131,
      "learning_rate": 4.649431160462552e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 387368,
      "step": 1360
    },
    {
      "epoch": 2.546641791044776,
      "grad_norm": 1.991472601890564,
      "learning_rate": 4.645262589687861e-05,
      "loss": 0.7804,
      "num_input_tokens_seen": 388776,
      "step": 1365
    },
    {
      "epoch": 2.5559701492537314,
      "grad_norm": 1.2398386001586914,
      "learning_rate": 4.6410712729028734e-05,
      "loss": 0.608,
      "num_input_tokens_seen": 390440,
      "step": 1370
    },
    {
      "epoch": 2.5652985074626864,
      "grad_norm": 1.5492669343948364,
      "learning_rate": 4.636857254547712e-05,
      "loss": 0.6654,
      "num_input_tokens_seen": 391720,
      "step": 1375
    },
    {
      "epoch": 2.574626865671642,
      "grad_norm": 1.3636995553970337,
      "learning_rate": 4.632620579303203e-05,
      "loss": 0.6797,
      "num_input_tokens_seen": 393032,
      "step": 1380
    },
    {
      "epoch": 2.583955223880597,
      "grad_norm": 0.7820605635643005,
      "learning_rate": 4.628361292090403e-05,
      "loss": 0.5658,
      "num_input_tokens_seen": 394312,
      "step": 1385
    },
    {
      "epoch": 2.593283582089552,
      "grad_norm": 1.408697247505188,
      "learning_rate": 4.624079438070117e-05,
      "loss": 0.7981,
      "num_input_tokens_seen": 395656,
      "step": 1390
    },
    {
      "epoch": 2.6026119402985075,
      "grad_norm": 0.6349939703941345,
      "learning_rate": 4.6197750626424277e-05,
      "loss": 0.6402,
      "num_input_tokens_seen": 397288,
      "step": 1395
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 0.922044038772583,
      "learning_rate": 4.615448211446208e-05,
      "loss": 0.6525,
      "num_input_tokens_seen": 398600,
      "step": 1400
    },
    {
      "epoch": 2.621268656716418,
      "grad_norm": 1.6789649724960327,
      "learning_rate": 4.6110989303586396e-05,
      "loss": 0.7038,
      "num_input_tokens_seen": 400168,
      "step": 1405
    },
    {
      "epoch": 2.6305970149253732,
      "grad_norm": 2.0882503986358643,
      "learning_rate": 4.606727265494727e-05,
      "loss": 0.787,
      "num_input_tokens_seen": 401544,
      "step": 1410
    },
    {
      "epoch": 2.6399253731343286,
      "grad_norm": 1.2559574842453003,
      "learning_rate": 4.602333263206806e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 402984,
      "step": 1415
    },
    {
      "epoch": 2.6492537313432836,
      "grad_norm": 1.3488932847976685,
      "learning_rate": 4.597916970084056e-05,
      "loss": 0.7587,
      "num_input_tokens_seen": 404328,
      "step": 1420
    },
    {
      "epoch": 2.658582089552239,
      "grad_norm": 1.4329274892807007,
      "learning_rate": 4.593478432952002e-05,
      "loss": 0.5994,
      "num_input_tokens_seen": 405608,
      "step": 1425
    },
    {
      "epoch": 2.667910447761194,
      "grad_norm": 1.2357596158981323,
      "learning_rate": 4.5890176988720205e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 407208,
      "step": 1430
    },
    {
      "epoch": 2.6772388059701493,
      "grad_norm": 0.7859914898872375,
      "learning_rate": 4.584534815140842e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 408776,
      "step": 1435
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 1.0628080368041992,
      "learning_rate": 4.5800298292900446e-05,
      "loss": 0.6064,
      "num_input_tokens_seen": 410120,
      "step": 1440
    },
    {
      "epoch": 2.6958955223880596,
      "grad_norm": 1.056727409362793,
      "learning_rate": 4.575502789085555e-05,
      "loss": 0.4442,
      "num_input_tokens_seen": 411624,
      "step": 1445
    },
    {
      "epoch": 2.705223880597015,
      "grad_norm": 1.0404438972473145,
      "learning_rate": 4.57095374252714e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 413096,
      "step": 1450
    },
    {
      "epoch": 2.71455223880597,
      "grad_norm": 2.5030760765075684,
      "learning_rate": 4.5663827378478975e-05,
      "loss": 0.8002,
      "num_input_tokens_seen": 414472,
      "step": 1455
    },
    {
      "epoch": 2.7238805970149254,
      "grad_norm": 0.8684210181236267,
      "learning_rate": 4.561789823513743e-05,
      "loss": 0.6204,
      "num_input_tokens_seen": 416072,
      "step": 1460
    },
    {
      "epoch": 2.7332089552238807,
      "grad_norm": 1.1791789531707764,
      "learning_rate": 4.5571750482229016e-05,
      "loss": 0.7093,
      "num_input_tokens_seen": 417640,
      "step": 1465
    },
    {
      "epoch": 2.7425373134328357,
      "grad_norm": 1.25052011013031,
      "learning_rate": 4.552538460905386e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 419048,
      "step": 1470
    },
    {
      "epoch": 2.751865671641791,
      "grad_norm": 0.8737344145774841,
      "learning_rate": 4.54788011072248e-05,
      "loss": 0.5445,
      "num_input_tokens_seen": 420520,
      "step": 1475
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 1.8435871601104736,
      "learning_rate": 4.543200047066216e-05,
      "loss": 0.6661,
      "num_input_tokens_seen": 421960,
      "step": 1480
    },
    {
      "epoch": 2.7705223880597014,
      "grad_norm": 1.6196060180664062,
      "learning_rate": 4.538498319558854e-05,
      "loss": 0.6095,
      "num_input_tokens_seen": 423304,
      "step": 1485
    },
    {
      "epoch": 2.779850746268657,
      "grad_norm": 1.1326524019241333,
      "learning_rate": 4.5337749780523526e-05,
      "loss": 0.6792,
      "num_input_tokens_seen": 424552,
      "step": 1490
    },
    {
      "epoch": 2.789179104477612,
      "grad_norm": 0.9968588352203369,
      "learning_rate": 4.5290300726278415e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 425896,
      "step": 1495
    },
    {
      "epoch": 2.798507462686567,
      "grad_norm": 0.992361307144165,
      "learning_rate": 4.5242636535950913e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 427208,
      "step": 1500
    },
    {
      "epoch": 2.8078358208955225,
      "grad_norm": 1.4601582288742065,
      "learning_rate": 4.519475771491978e-05,
      "loss": 0.5523,
      "num_input_tokens_seen": 428392,
      "step": 1505
    },
    {
      "epoch": 2.8171641791044775,
      "grad_norm": 1.4612449407577515,
      "learning_rate": 4.5146664770839495e-05,
      "loss": 0.5631,
      "num_input_tokens_seen": 429928,
      "step": 1510
    },
    {
      "epoch": 2.826492537313433,
      "grad_norm": 1.7660143375396729,
      "learning_rate": 4.5098358213634876e-05,
      "loss": 0.6328,
      "num_input_tokens_seen": 431464,
      "step": 1515
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 2.104743003845215,
      "learning_rate": 4.504983855549562e-05,
      "loss": 0.519,
      "num_input_tokens_seen": 432968,
      "step": 1520
    },
    {
      "epoch": 2.845149253731343,
      "grad_norm": 1.5445313453674316,
      "learning_rate": 4.5001106310870946e-05,
      "loss": 0.6846,
      "num_input_tokens_seen": 434472,
      "step": 1525
    },
    {
      "epoch": 2.8544776119402986,
      "grad_norm": 2.0449059009552,
      "learning_rate": 4.49521619964641e-05,
      "loss": 0.5764,
      "num_input_tokens_seen": 435944,
      "step": 1530
    },
    {
      "epoch": 2.8638059701492535,
      "grad_norm": 1.006453514099121,
      "learning_rate": 4.4903006131226874e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 437608,
      "step": 1535
    },
    {
      "epoch": 2.873134328358209,
      "grad_norm": 2.133375883102417,
      "learning_rate": 4.485363923635413e-05,
      "loss": 0.6713,
      "num_input_tokens_seen": 439176,
      "step": 1540
    },
    {
      "epoch": 2.8824626865671643,
      "grad_norm": 1.2206082344055176,
      "learning_rate": 4.480406183527823e-05,
      "loss": 1.0375,
      "num_input_tokens_seen": 440520,
      "step": 1545
    },
    {
      "epoch": 2.8917910447761193,
      "grad_norm": 1.645415186882019,
      "learning_rate": 4.475427445366355e-05,
      "loss": 0.7037,
      "num_input_tokens_seen": 441896,
      "step": 1550
    },
    {
      "epoch": 2.9011194029850746,
      "grad_norm": 1.7647379636764526,
      "learning_rate": 4.4704277619400834e-05,
      "loss": 0.8418,
      "num_input_tokens_seen": 443176,
      "step": 1555
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 0.9811164736747742,
      "learning_rate": 4.4654071862601654e-05,
      "loss": 0.5749,
      "num_input_tokens_seen": 444680,
      "step": 1560
    },
    {
      "epoch": 2.919776119402985,
      "grad_norm": 0.9422204494476318,
      "learning_rate": 4.460365771559275e-05,
      "loss": 0.6281,
      "num_input_tokens_seen": 446024,
      "step": 1565
    },
    {
      "epoch": 2.9291044776119404,
      "grad_norm": 1.1950550079345703,
      "learning_rate": 4.455303571291042e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 447656,
      "step": 1570
    },
    {
      "epoch": 2.9384328358208958,
      "grad_norm": 0.9624826312065125,
      "learning_rate": 4.4502206391294824e-05,
      "loss": 0.6184,
      "num_input_tokens_seen": 448904,
      "step": 1575
    },
    {
      "epoch": 2.9477611940298507,
      "grad_norm": 1.001377820968628,
      "learning_rate": 4.445117028968431e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 450216,
      "step": 1580
    },
    {
      "epoch": 2.957089552238806,
      "grad_norm": 1.2497371435165405,
      "learning_rate": 4.439992794920969e-05,
      "loss": 0.5975,
      "num_input_tokens_seen": 451592,
      "step": 1585
    },
    {
      "epoch": 2.966417910447761,
      "grad_norm": 0.9689121246337891,
      "learning_rate": 4.434847991318851e-05,
      "loss": 0.8465,
      "num_input_tokens_seen": 453032,
      "step": 1590
    },
    {
      "epoch": 2.9757462686567164,
      "grad_norm": 1.1159669160842896,
      "learning_rate": 4.4296826727119296e-05,
      "loss": 0.5455,
      "num_input_tokens_seen": 454408,
      "step": 1595
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 1.3040260076522827,
      "learning_rate": 4.424496893867573e-05,
      "loss": 0.674,
      "num_input_tokens_seen": 455912,
      "step": 1600
    },
    {
      "epoch": 2.9944029850746268,
      "grad_norm": 1.2749884128570557,
      "learning_rate": 4.419290709770091e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 457256,
      "step": 1605
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5991371273994446,
      "eval_runtime": 4.5684,
      "eval_samples_per_second": 52.097,
      "eval_steps_per_second": 13.134,
      "num_input_tokens_seen": 457952,
      "step": 1608
    },
    {
      "epoch": 3.003731343283582,
      "grad_norm": 1.3173999786376953,
      "learning_rate": 4.414064175620146e-05,
      "loss": 0.6624,
      "num_input_tokens_seen": 458464,
      "step": 1610
    },
    {
      "epoch": 3.013059701492537,
      "grad_norm": 0.7166537642478943,
      "learning_rate": 4.408817346834169e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 459872,
      "step": 1615
    },
    {
      "epoch": 3.0223880597014925,
      "grad_norm": 1.2299184799194336,
      "learning_rate": 4.4035502790437764e-05,
      "loss": 0.3844,
      "num_input_tokens_seen": 461408,
      "step": 1620
    },
    {
      "epoch": 3.031716417910448,
      "grad_norm": 0.9069246053695679,
      "learning_rate": 4.398263028095175e-05,
      "loss": 0.643,
      "num_input_tokens_seen": 462624,
      "step": 1625
    },
    {
      "epoch": 3.041044776119403,
      "grad_norm": 1.481077790260315,
      "learning_rate": 4.392955650048571e-05,
      "loss": 0.5278,
      "num_input_tokens_seen": 464096,
      "step": 1630
    },
    {
      "epoch": 3.050373134328358,
      "grad_norm": 3.1422011852264404,
      "learning_rate": 4.387628201177577e-05,
      "loss": 0.6399,
      "num_input_tokens_seen": 465408,
      "step": 1635
    },
    {
      "epoch": 3.0597014925373136,
      "grad_norm": 2.379854679107666,
      "learning_rate": 4.382280737968614e-05,
      "loss": 0.6913,
      "num_input_tokens_seen": 466880,
      "step": 1640
    },
    {
      "epoch": 3.0690298507462686,
      "grad_norm": 0.8965510129928589,
      "learning_rate": 4.3769133171203144e-05,
      "loss": 0.3852,
      "num_input_tokens_seen": 468416,
      "step": 1645
    },
    {
      "epoch": 3.078358208955224,
      "grad_norm": 1.4239137172698975,
      "learning_rate": 4.371525995542918e-05,
      "loss": 0.5092,
      "num_input_tokens_seen": 469952,
      "step": 1650
    },
    {
      "epoch": 3.0876865671641793,
      "grad_norm": 1.249794602394104,
      "learning_rate": 4.366118830357672e-05,
      "loss": 0.5958,
      "num_input_tokens_seen": 471264,
      "step": 1655
    },
    {
      "epoch": 3.0970149253731343,
      "grad_norm": 0.9955365657806396,
      "learning_rate": 4.3606918788962205e-05,
      "loss": 0.7403,
      "num_input_tokens_seen": 472640,
      "step": 1660
    },
    {
      "epoch": 3.1063432835820897,
      "grad_norm": 0.6966485977172852,
      "learning_rate": 4.355245198700003e-05,
      "loss": 0.6524,
      "num_input_tokens_seen": 473984,
      "step": 1665
    },
    {
      "epoch": 3.1156716417910446,
      "grad_norm": 2.00040864944458,
      "learning_rate": 4.3497788475196376e-05,
      "loss": 0.6698,
      "num_input_tokens_seen": 475488,
      "step": 1670
    },
    {
      "epoch": 3.125,
      "grad_norm": 1.651235818862915,
      "learning_rate": 4.3442928833143145e-05,
      "loss": 0.5796,
      "num_input_tokens_seen": 476864,
      "step": 1675
    },
    {
      "epoch": 3.1343283582089554,
      "grad_norm": 1.4639793634414673,
      "learning_rate": 4.338787364251177e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 478272,
      "step": 1680
    },
    {
      "epoch": 3.1436567164179103,
      "grad_norm": 1.6414200067520142,
      "learning_rate": 4.3332623487047084e-05,
      "loss": 0.6368,
      "num_input_tokens_seen": 479648,
      "step": 1685
    },
    {
      "epoch": 3.1529850746268657,
      "grad_norm": 0.9184785485267639,
      "learning_rate": 4.32771789525611e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 481152,
      "step": 1690
    },
    {
      "epoch": 3.1623134328358207,
      "grad_norm": 1.4448952674865723,
      "learning_rate": 4.3221540626926824e-05,
      "loss": 0.6188,
      "num_input_tokens_seen": 482624,
      "step": 1695
    },
    {
      "epoch": 3.171641791044776,
      "grad_norm": 0.775035560131073,
      "learning_rate": 4.3165709100071986e-05,
      "loss": 0.5135,
      "num_input_tokens_seen": 484032,
      "step": 1700
    },
    {
      "epoch": 3.1809701492537314,
      "grad_norm": 1.3805443048477173,
      "learning_rate": 4.310968496397284e-05,
      "loss": 0.6214,
      "num_input_tokens_seen": 485440,
      "step": 1705
    },
    {
      "epoch": 3.1902985074626864,
      "grad_norm": 0.9740598797798157,
      "learning_rate": 4.305346881264785e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 486848,
      "step": 1710
    },
    {
      "epoch": 3.199626865671642,
      "grad_norm": 1.109499216079712,
      "learning_rate": 4.299706124215138e-05,
      "loss": 0.58,
      "num_input_tokens_seen": 488192,
      "step": 1715
    },
    {
      "epoch": 3.208955223880597,
      "grad_norm": 1.719560146331787,
      "learning_rate": 4.294046285056742e-05,
      "loss": 0.67,
      "num_input_tokens_seen": 489696,
      "step": 1720
    },
    {
      "epoch": 3.218283582089552,
      "grad_norm": 1.3051543235778809,
      "learning_rate": 4.288367423800319e-05,
      "loss": 0.8217,
      "num_input_tokens_seen": 491008,
      "step": 1725
    },
    {
      "epoch": 3.2276119402985075,
      "grad_norm": 1.5748628377914429,
      "learning_rate": 4.2826696006582825e-05,
      "loss": 0.5825,
      "num_input_tokens_seen": 492576,
      "step": 1730
    },
    {
      "epoch": 3.236940298507463,
      "grad_norm": 1.371764898300171,
      "learning_rate": 4.276952876044096e-05,
      "loss": 0.8118,
      "num_input_tokens_seen": 493824,
      "step": 1735
    },
    {
      "epoch": 3.246268656716418,
      "grad_norm": 1.7391024827957153,
      "learning_rate": 4.2712173105716346e-05,
      "loss": 0.5449,
      "num_input_tokens_seen": 495200,
      "step": 1740
    },
    {
      "epoch": 3.2555970149253732,
      "grad_norm": 1.294609785079956,
      "learning_rate": 4.265462965054539e-05,
      "loss": 0.7559,
      "num_input_tokens_seen": 496576,
      "step": 1745
    },
    {
      "epoch": 3.264925373134328,
      "grad_norm": 0.9215909838676453,
      "learning_rate": 4.259689900505576e-05,
      "loss": 0.6842,
      "num_input_tokens_seen": 498112,
      "step": 1750
    },
    {
      "epoch": 3.2742537313432836,
      "grad_norm": 0.7093052268028259,
      "learning_rate": 4.253898178135985e-05,
      "loss": 0.5425,
      "num_input_tokens_seen": 499552,
      "step": 1755
    },
    {
      "epoch": 3.283582089552239,
      "grad_norm": 1.284893274307251,
      "learning_rate": 4.2480878593548344e-05,
      "loss": 0.3423,
      "num_input_tokens_seen": 500960,
      "step": 1760
    },
    {
      "epoch": 3.292910447761194,
      "grad_norm": 2.1534950733184814,
      "learning_rate": 4.24225900576837e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 502208,
      "step": 1765
    },
    {
      "epoch": 3.3022388059701493,
      "grad_norm": 1.300766944885254,
      "learning_rate": 4.236411679179357e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 503808,
      "step": 1770
    },
    {
      "epoch": 3.3115671641791042,
      "grad_norm": 0.9290071725845337,
      "learning_rate": 4.230545941586431e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 505248,
      "step": 1775
    },
    {
      "epoch": 3.3208955223880596,
      "grad_norm": 1.1970524787902832,
      "learning_rate": 4.224661855183435e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 507072,
      "step": 1780
    },
    {
      "epoch": 3.330223880597015,
      "grad_norm": 2.128944158554077,
      "learning_rate": 4.218759482358765e-05,
      "loss": 0.6523,
      "num_input_tokens_seen": 508576,
      "step": 1785
    },
    {
      "epoch": 3.33955223880597,
      "grad_norm": 1.0806620121002197,
      "learning_rate": 4.212838885694705e-05,
      "loss": 0.7859,
      "num_input_tokens_seen": 509824,
      "step": 1790
    },
    {
      "epoch": 3.3488805970149254,
      "grad_norm": 1.3884180784225464,
      "learning_rate": 4.206900127966764e-05,
      "loss": 0.4836,
      "num_input_tokens_seen": 511392,
      "step": 1795
    },
    {
      "epoch": 3.3582089552238807,
      "grad_norm": 1.3276972770690918,
      "learning_rate": 4.200943272143013e-05,
      "loss": 0.5972,
      "num_input_tokens_seen": 512672,
      "step": 1800
    },
    {
      "epoch": 3.3675373134328357,
      "grad_norm": 1.5149116516113281,
      "learning_rate": 4.194968381383414e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 514144,
      "step": 1805
    },
    {
      "epoch": 3.376865671641791,
      "grad_norm": 1.401867151260376,
      "learning_rate": 4.188975519039151e-05,
      "loss": 0.5632,
      "num_input_tokens_seen": 515584,
      "step": 1810
    },
    {
      "epoch": 3.3861940298507465,
      "grad_norm": 1.3054430484771729,
      "learning_rate": 4.1829647486519596e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 516928,
      "step": 1815
    },
    {
      "epoch": 3.3955223880597014,
      "grad_norm": 1.7890267372131348,
      "learning_rate": 4.176936133953454e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 518432,
      "step": 1820
    },
    {
      "epoch": 3.404850746268657,
      "grad_norm": 1.3304798603057861,
      "learning_rate": 4.170889738864448e-05,
      "loss": 0.5269,
      "num_input_tokens_seen": 519904,
      "step": 1825
    },
    {
      "epoch": 3.4141791044776117,
      "grad_norm": 1.520775318145752,
      "learning_rate": 4.16482562749428e-05,
      "loss": 0.5409,
      "num_input_tokens_seen": 521472,
      "step": 1830
    },
    {
      "epoch": 3.423507462686567,
      "grad_norm": 1.7513511180877686,
      "learning_rate": 4.158743864140131e-05,
      "loss": 0.6476,
      "num_input_tokens_seen": 522848,
      "step": 1835
    },
    {
      "epoch": 3.4328358208955225,
      "grad_norm": 2.0569376945495605,
      "learning_rate": 4.152644513286348e-05,
      "loss": 0.6094,
      "num_input_tokens_seen": 524160,
      "step": 1840
    },
    {
      "epoch": 3.4421641791044775,
      "grad_norm": 1.427960753440857,
      "learning_rate": 4.146527639603751e-05,
      "loss": 0.6179,
      "num_input_tokens_seen": 525504,
      "step": 1845
    },
    {
      "epoch": 3.451492537313433,
      "grad_norm": 1.1078107357025146,
      "learning_rate": 4.1403933079489585e-05,
      "loss": 0.4016,
      "num_input_tokens_seen": 526816,
      "step": 1850
    },
    {
      "epoch": 3.4608208955223883,
      "grad_norm": 2.032392740249634,
      "learning_rate": 4.1342415833636904e-05,
      "loss": 0.8031,
      "num_input_tokens_seen": 528224,
      "step": 1855
    },
    {
      "epoch": 3.470149253731343,
      "grad_norm": 1.0600571632385254,
      "learning_rate": 4.128072531074084e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 529504,
      "step": 1860
    },
    {
      "epoch": 3.4794776119402986,
      "grad_norm": 1.62373685836792,
      "learning_rate": 4.121886216489998e-05,
      "loss": 0.804,
      "num_input_tokens_seen": 531072,
      "step": 1865
    },
    {
      "epoch": 3.4888059701492535,
      "grad_norm": 0.878396213054657,
      "learning_rate": 4.115682705204326e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 532832,
      "step": 1870
    },
    {
      "epoch": 3.498134328358209,
      "grad_norm": 1.1084483861923218,
      "learning_rate": 4.109462062992293e-05,
      "loss": 0.5311,
      "num_input_tokens_seen": 534272,
      "step": 1875
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.5987069010734558,
      "eval_runtime": 4.5678,
      "eval_samples_per_second": 52.104,
      "eval_steps_per_second": 13.135,
      "num_input_tokens_seen": 534688,
      "step": 1876
    },
    {
      "epoch": 3.5074626865671643,
      "grad_norm": 1.2245299816131592,
      "learning_rate": 4.103224355810761e-05,
      "loss": 0.6022,
      "num_input_tokens_seen": 535872,
      "step": 1880
    },
    {
      "epoch": 3.5167910447761193,
      "grad_norm": 1.758050560951233,
      "learning_rate": 4.096969649797534e-05,
      "loss": 0.5914,
      "num_input_tokens_seen": 537376,
      "step": 1885
    },
    {
      "epoch": 3.5261194029850746,
      "grad_norm": 1.045783519744873,
      "learning_rate": 4.0906980112706494e-05,
      "loss": 0.8767,
      "num_input_tokens_seen": 538880,
      "step": 1890
    },
    {
      "epoch": 3.53544776119403,
      "grad_norm": 0.9992062449455261,
      "learning_rate": 4.08440950672768e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 540512,
      "step": 1895
    },
    {
      "epoch": 3.544776119402985,
      "grad_norm": 1.7374255657196045,
      "learning_rate": 4.078104202845027e-05,
      "loss": 0.6087,
      "num_input_tokens_seen": 541920,
      "step": 1900
    },
    {
      "epoch": 3.5541044776119404,
      "grad_norm": 2.1052746772766113,
      "learning_rate": 4.071782166477213e-05,
      "loss": 0.635,
      "num_input_tokens_seen": 543328,
      "step": 1905
    },
    {
      "epoch": 3.5634328358208958,
      "grad_norm": 1.1379703283309937,
      "learning_rate": 4.065443464656174e-05,
      "loss": 0.5881,
      "num_input_tokens_seen": 544800,
      "step": 1910
    },
    {
      "epoch": 3.5727611940298507,
      "grad_norm": 1.3880736827850342,
      "learning_rate": 4.0590881645905475e-05,
      "loss": 0.6933,
      "num_input_tokens_seen": 546208,
      "step": 1915
    },
    {
      "epoch": 3.582089552238806,
      "grad_norm": 1.554692268371582,
      "learning_rate": 4.052716333664963e-05,
      "loss": 0.4754,
      "num_input_tokens_seen": 547616,
      "step": 1920
    },
    {
      "epoch": 3.591417910447761,
      "grad_norm": 2.0506961345672607,
      "learning_rate": 4.046328039439321e-05,
      "loss": 0.631,
      "num_input_tokens_seen": 549088,
      "step": 1925
    },
    {
      "epoch": 3.6007462686567164,
      "grad_norm": 1.3062870502471924,
      "learning_rate": 4.039923349648084e-05,
      "loss": 0.5158,
      "num_input_tokens_seen": 550464,
      "step": 1930
    },
    {
      "epoch": 3.6100746268656714,
      "grad_norm": 3.3200204372406006,
      "learning_rate": 4.0335023321995545e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 552032,
      "step": 1935
    },
    {
      "epoch": 3.6194029850746268,
      "grad_norm": 1.2594958543777466,
      "learning_rate": 4.0270650551751546e-05,
      "loss": 0.7315,
      "num_input_tokens_seen": 553344,
      "step": 1940
    },
    {
      "epoch": 3.628731343283582,
      "grad_norm": 1.695493459701538,
      "learning_rate": 4.020611586828705e-05,
      "loss": 0.6374,
      "num_input_tokens_seen": 554592,
      "step": 1945
    },
    {
      "epoch": 3.638059701492537,
      "grad_norm": 1.279910922050476,
      "learning_rate": 4.0141419955857044e-05,
      "loss": 0.5249,
      "num_input_tokens_seen": 556064,
      "step": 1950
    },
    {
      "epoch": 3.6473880597014925,
      "grad_norm": 0.8155736327171326,
      "learning_rate": 4.007656350042595e-05,
      "loss": 0.4287,
      "num_input_tokens_seen": 557536,
      "step": 1955
    },
    {
      "epoch": 3.656716417910448,
      "grad_norm": 1.965759515762329,
      "learning_rate": 4.001154718966048e-05,
      "loss": 0.5472,
      "num_input_tokens_seen": 558976,
      "step": 1960
    },
    {
      "epoch": 3.666044776119403,
      "grad_norm": 1.2898179292678833,
      "learning_rate": 3.994637171292223e-05,
      "loss": 0.5616,
      "num_input_tokens_seen": 560192,
      "step": 1965
    },
    {
      "epoch": 3.675373134328358,
      "grad_norm": 2.2137973308563232,
      "learning_rate": 3.988103776126042e-05,
      "loss": 0.5813,
      "num_input_tokens_seen": 561472,
      "step": 1970
    },
    {
      "epoch": 3.6847014925373136,
      "grad_norm": 2.0516977310180664,
      "learning_rate": 3.9815546027404603e-05,
      "loss": 0.7628,
      "num_input_tokens_seen": 562912,
      "step": 1975
    },
    {
      "epoch": 3.6940298507462686,
      "grad_norm": 2.0701968669891357,
      "learning_rate": 3.974989720575724e-05,
      "loss": 0.5456,
      "num_input_tokens_seen": 564288,
      "step": 1980
    },
    {
      "epoch": 3.703358208955224,
      "grad_norm": 1.4374306201934814,
      "learning_rate": 3.9684091992386393e-05,
      "loss": 0.6171,
      "num_input_tokens_seen": 565664,
      "step": 1985
    },
    {
      "epoch": 3.7126865671641793,
      "grad_norm": 1.5016531944274902,
      "learning_rate": 3.961813108501833e-05,
      "loss": 0.7046,
      "num_input_tokens_seen": 567040,
      "step": 1990
    },
    {
      "epoch": 3.7220149253731343,
      "grad_norm": 1.6128549575805664,
      "learning_rate": 3.9552015183030136e-05,
      "loss": 0.6095,
      "num_input_tokens_seen": 568320,
      "step": 1995
    },
    {
      "epoch": 3.7313432835820897,
      "grad_norm": 1.2565516233444214,
      "learning_rate": 3.9485744987442266e-05,
      "loss": 0.7175,
      "num_input_tokens_seen": 569696,
      "step": 2000
    },
    {
      "epoch": 3.7406716417910446,
      "grad_norm": 1.697272539138794,
      "learning_rate": 3.9419321200911155e-05,
      "loss": 0.5668,
      "num_input_tokens_seen": 571296,
      "step": 2005
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.9241491556167603,
      "learning_rate": 3.9352744527721754e-05,
      "loss": 0.7156,
      "num_input_tokens_seen": 572768,
      "step": 2010
    },
    {
      "epoch": 3.7593283582089554,
      "grad_norm": 1.5093331336975098,
      "learning_rate": 3.928601567378003e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 574080,
      "step": 2015
    },
    {
      "epoch": 3.7686567164179103,
      "grad_norm": 2.693692922592163,
      "learning_rate": 3.921913534660552e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 575456,
      "step": 2020
    },
    {
      "epoch": 3.7779850746268657,
      "grad_norm": 1.2875691652297974,
      "learning_rate": 3.915210425532383e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 576864,
      "step": 2025
    },
    {
      "epoch": 3.7873134328358207,
      "grad_norm": 1.6616896390914917,
      "learning_rate": 3.908492311065909e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 578304,
      "step": 2030
    },
    {
      "epoch": 3.796641791044776,
      "grad_norm": 0.9011557698249817,
      "learning_rate": 3.901759262492643e-05,
      "loss": 0.2518,
      "num_input_tokens_seen": 579744,
      "step": 2035
    },
    {
      "epoch": 3.8059701492537314,
      "grad_norm": 1.8044768571853638,
      "learning_rate": 3.895011351202443e-05,
      "loss": 0.6119,
      "num_input_tokens_seen": 581088,
      "step": 2040
    },
    {
      "epoch": 3.8152985074626864,
      "grad_norm": 2.3962883949279785,
      "learning_rate": 3.888248648742756e-05,
      "loss": 0.671,
      "num_input_tokens_seen": 582560,
      "step": 2045
    },
    {
      "epoch": 3.824626865671642,
      "grad_norm": 1.1989905834197998,
      "learning_rate": 3.881471226817858e-05,
      "loss": 0.528,
      "num_input_tokens_seen": 584000,
      "step": 2050
    },
    {
      "epoch": 3.833955223880597,
      "grad_norm": 2.6569294929504395,
      "learning_rate": 3.874679157288092e-05,
      "loss": 0.8116,
      "num_input_tokens_seen": 585248,
      "step": 2055
    },
    {
      "epoch": 3.843283582089552,
      "grad_norm": 1.2410632371902466,
      "learning_rate": 3.86787251216911e-05,
      "loss": 0.7641,
      "num_input_tokens_seen": 586720,
      "step": 2060
    },
    {
      "epoch": 3.8526119402985075,
      "grad_norm": 1.9396847486495972,
      "learning_rate": 3.8610513636311073e-05,
      "loss": 0.5573,
      "num_input_tokens_seen": 588352,
      "step": 2065
    },
    {
      "epoch": 3.861940298507463,
      "grad_norm": 1.6793707609176636,
      "learning_rate": 3.854215783998058e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 589632,
      "step": 2070
    },
    {
      "epoch": 3.871268656716418,
      "grad_norm": 0.6740831732749939,
      "learning_rate": 3.8473658457469466e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 591072,
      "step": 2075
    },
    {
      "epoch": 3.8805970149253732,
      "grad_norm": 1.2159323692321777,
      "learning_rate": 3.840501621507003e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 592544,
      "step": 2080
    },
    {
      "epoch": 3.8899253731343286,
      "grad_norm": 1.9220870733261108,
      "learning_rate": 3.833623184058926e-05,
      "loss": 0.5559,
      "num_input_tokens_seen": 594112,
      "step": 2085
    },
    {
      "epoch": 3.8992537313432836,
      "grad_norm": 1.4277399778366089,
      "learning_rate": 3.826730606334119e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 595744,
      "step": 2090
    },
    {
      "epoch": 3.908582089552239,
      "grad_norm": 2.3160860538482666,
      "learning_rate": 3.819823961413912e-05,
      "loss": 0.6836,
      "num_input_tokens_seen": 597184,
      "step": 2095
    },
    {
      "epoch": 3.917910447761194,
      "grad_norm": 1.28302001953125,
      "learning_rate": 3.812903322528789e-05,
      "loss": 0.5824,
      "num_input_tokens_seen": 598816,
      "step": 2100
    },
    {
      "epoch": 3.9272388059701493,
      "grad_norm": 1.9368464946746826,
      "learning_rate": 3.805968763057609e-05,
      "loss": 0.7782,
      "num_input_tokens_seen": 600192,
      "step": 2105
    },
    {
      "epoch": 3.9365671641791042,
      "grad_norm": 1.3876724243164062,
      "learning_rate": 3.7990203565268314e-05,
      "loss": 0.4846,
      "num_input_tokens_seen": 601472,
      "step": 2110
    },
    {
      "epoch": 3.9458955223880596,
      "grad_norm": 1.3265646696090698,
      "learning_rate": 3.792058176609734e-05,
      "loss": 0.5604,
      "num_input_tokens_seen": 602880,
      "step": 2115
    },
    {
      "epoch": 3.955223880597015,
      "grad_norm": 1.9544258117675781,
      "learning_rate": 3.785082297125631e-05,
      "loss": 0.745,
      "num_input_tokens_seen": 604416,
      "step": 2120
    },
    {
      "epoch": 3.96455223880597,
      "grad_norm": 1.5964033603668213,
      "learning_rate": 3.7780927920390964e-05,
      "loss": 0.6856,
      "num_input_tokens_seen": 605696,
      "step": 2125
    },
    {
      "epoch": 3.9738805970149254,
      "grad_norm": 1.018444299697876,
      "learning_rate": 3.771089735459168e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 607040,
      "step": 2130
    },
    {
      "epoch": 3.9832089552238807,
      "grad_norm": 1.6007020473480225,
      "learning_rate": 3.7640732016385745e-05,
      "loss": 0.6189,
      "num_input_tokens_seen": 608608,
      "step": 2135
    },
    {
      "epoch": 3.9925373134328357,
      "grad_norm": 1.5353299379348755,
      "learning_rate": 3.757043264972941e-05,
      "loss": 0.3989,
      "num_input_tokens_seen": 609984,
      "step": 2140
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.6024206280708313,
      "eval_runtime": 4.5817,
      "eval_samples_per_second": 51.946,
      "eval_steps_per_second": 13.096,
      "num_input_tokens_seen": 610944,
      "step": 2144
    },
    {
      "epoch": 4.001865671641791,
      "grad_norm": 1.361750841140747,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.3484,
      "num_input_tokens_seen": 611200,
      "step": 2145
    },
    {
      "epoch": 4.0111940298507465,
      "grad_norm": 1.5583475828170776,
      "learning_rate": 3.742943481398805e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 612512,
      "step": 2150
    },
    {
      "epoch": 4.020522388059701,
      "grad_norm": 0.9856475591659546,
      "learning_rate": 3.7358737839889356e-05,
      "loss": 0.5583,
      "num_input_tokens_seen": 613920,
      "step": 2155
    },
    {
      "epoch": 4.029850746268656,
      "grad_norm": 1.8877625465393066,
      "learning_rate": 3.728790982729705e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 615264,
      "step": 2160
    },
    {
      "epoch": 4.039179104477612,
      "grad_norm": 0.9619392156600952,
      "learning_rate": 3.721695152719364e-05,
      "loss": 0.8709,
      "num_input_tokens_seen": 616576,
      "step": 2165
    },
    {
      "epoch": 4.048507462686567,
      "grad_norm": 1.738877534866333,
      "learning_rate": 3.7145863691943076e-05,
      "loss": 0.5265,
      "num_input_tokens_seen": 617984,
      "step": 2170
    },
    {
      "epoch": 4.057835820895522,
      "grad_norm": 1.7362086772918701,
      "learning_rate": 3.707464707528275e-05,
      "loss": 0.3869,
      "num_input_tokens_seen": 619520,
      "step": 2175
    },
    {
      "epoch": 4.067164179104478,
      "grad_norm": 1.4337236881256104,
      "learning_rate": 3.700330243231552e-05,
      "loss": 0.7082,
      "num_input_tokens_seen": 620864,
      "step": 2180
    },
    {
      "epoch": 4.076492537313433,
      "grad_norm": 1.6963222026824951,
      "learning_rate": 3.6931830519501685e-05,
      "loss": 0.6452,
      "num_input_tokens_seen": 622272,
      "step": 2185
    },
    {
      "epoch": 4.085820895522388,
      "grad_norm": 1.5358551740646362,
      "learning_rate": 3.686023209465096e-05,
      "loss": 0.4004,
      "num_input_tokens_seen": 623840,
      "step": 2190
    },
    {
      "epoch": 4.095149253731344,
      "grad_norm": 1.9799845218658447,
      "learning_rate": 3.678850791691448e-05,
      "loss": 0.6623,
      "num_input_tokens_seen": 625216,
      "step": 2195
    },
    {
      "epoch": 4.104477611940299,
      "grad_norm": 1.1511236429214478,
      "learning_rate": 3.671665874677673e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 626592,
      "step": 2200
    },
    {
      "epoch": 4.1138059701492535,
      "grad_norm": 1.58088219165802,
      "learning_rate": 3.664468534604745e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 627968,
      "step": 2205
    },
    {
      "epoch": 4.123134328358209,
      "grad_norm": 1.774718165397644,
      "learning_rate": 3.65725884778536e-05,
      "loss": 0.6441,
      "num_input_tokens_seen": 629248,
      "step": 2210
    },
    {
      "epoch": 4.132462686567164,
      "grad_norm": 1.3666579723358154,
      "learning_rate": 3.650036890663124e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 631040,
      "step": 2215
    },
    {
      "epoch": 4.141791044776119,
      "grad_norm": 1.8252735137939453,
      "learning_rate": 3.642802739811747e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 632320,
      "step": 2220
    },
    {
      "epoch": 4.151119402985074,
      "grad_norm": 1.6163835525512695,
      "learning_rate": 3.635556471934224e-05,
      "loss": 0.5887,
      "num_input_tokens_seen": 633792,
      "step": 2225
    },
    {
      "epoch": 4.16044776119403,
      "grad_norm": 1.203169345855713,
      "learning_rate": 3.628298163862029e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 635328,
      "step": 2230
    },
    {
      "epoch": 4.169776119402985,
      "grad_norm": 1.9701123237609863,
      "learning_rate": 3.621027892554295e-05,
      "loss": 0.5175,
      "num_input_tokens_seen": 636864,
      "step": 2235
    },
    {
      "epoch": 4.17910447761194,
      "grad_norm": 1.4490993022918701,
      "learning_rate": 3.613745735096999e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 638112,
      "step": 2240
    },
    {
      "epoch": 4.188432835820896,
      "grad_norm": 1.7466325759887695,
      "learning_rate": 3.606451768702151e-05,
      "loss": 0.5062,
      "num_input_tokens_seen": 639456,
      "step": 2245
    },
    {
      "epoch": 4.197761194029851,
      "grad_norm": 1.7205357551574707,
      "learning_rate": 3.599146070706964e-05,
      "loss": 0.6739,
      "num_input_tokens_seen": 640736,
      "step": 2250
    },
    {
      "epoch": 4.207089552238806,
      "grad_norm": 2.4796130657196045,
      "learning_rate": 3.5918287185730414e-05,
      "loss": 0.5191,
      "num_input_tokens_seen": 642208,
      "step": 2255
    },
    {
      "epoch": 4.2164179104477615,
      "grad_norm": 1.612862467765808,
      "learning_rate": 3.5844997898855566e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 643616,
      "step": 2260
    },
    {
      "epoch": 4.225746268656716,
      "grad_norm": 1.061316967010498,
      "learning_rate": 3.5771593623524265e-05,
      "loss": 0.6123,
      "num_input_tokens_seen": 645024,
      "step": 2265
    },
    {
      "epoch": 4.235074626865671,
      "grad_norm": 1.5158660411834717,
      "learning_rate": 3.569807513803488e-05,
      "loss": 0.6397,
      "num_input_tokens_seen": 646464,
      "step": 2270
    },
    {
      "epoch": 4.244402985074627,
      "grad_norm": 2.4303832054138184,
      "learning_rate": 3.5624443221896776e-05,
      "loss": 0.5074,
      "num_input_tokens_seen": 647968,
      "step": 2275
    },
    {
      "epoch": 4.253731343283582,
      "grad_norm": 1.6570976972579956,
      "learning_rate": 3.555069865582197e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 649312,
      "step": 2280
    },
    {
      "epoch": 4.263059701492537,
      "grad_norm": 1.6410709619522095,
      "learning_rate": 3.547684222171692e-05,
      "loss": 0.7953,
      "num_input_tokens_seen": 650656,
      "step": 2285
    },
    {
      "epoch": 4.272388059701493,
      "grad_norm": 2.3512117862701416,
      "learning_rate": 3.54028747026742e-05,
      "loss": 0.6101,
      "num_input_tokens_seen": 652032,
      "step": 2290
    },
    {
      "epoch": 4.281716417910448,
      "grad_norm": 1.7227802276611328,
      "learning_rate": 3.532879688296421e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 653504,
      "step": 2295
    },
    {
      "epoch": 4.291044776119403,
      "grad_norm": 1.7758419513702393,
      "learning_rate": 3.5254609548026865e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 654912,
      "step": 2300
    },
    {
      "epoch": 4.300373134328359,
      "grad_norm": 2.092322587966919,
      "learning_rate": 3.518031348446324e-05,
      "loss": 0.6688,
      "num_input_tokens_seen": 656256,
      "step": 2305
    },
    {
      "epoch": 4.309701492537314,
      "grad_norm": 1.4643824100494385,
      "learning_rate": 3.5105909480027276e-05,
      "loss": 0.4369,
      "num_input_tokens_seen": 657792,
      "step": 2310
    },
    {
      "epoch": 4.3190298507462686,
      "grad_norm": 0.8759129643440247,
      "learning_rate": 3.5031398323617366e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 659264,
      "step": 2315
    },
    {
      "epoch": 4.3283582089552235,
      "grad_norm": 2.0400428771972656,
      "learning_rate": 3.4956780805268066e-05,
      "loss": 0.7356,
      "num_input_tokens_seen": 660640,
      "step": 2320
    },
    {
      "epoch": 4.337686567164179,
      "grad_norm": 1.0295913219451904,
      "learning_rate": 3.4882057716141635e-05,
      "loss": 0.6875,
      "num_input_tokens_seen": 662016,
      "step": 2325
    },
    {
      "epoch": 4.347014925373134,
      "grad_norm": 1.7252097129821777,
      "learning_rate": 3.480722984851972e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 663392,
      "step": 2330
    },
    {
      "epoch": 4.356343283582089,
      "grad_norm": 2.8781917095184326,
      "learning_rate": 3.473229799579492e-05,
      "loss": 0.6541,
      "num_input_tokens_seen": 664928,
      "step": 2335
    },
    {
      "epoch": 4.365671641791045,
      "grad_norm": 1.4185543060302734,
      "learning_rate": 3.465726295246236e-05,
      "loss": 0.6864,
      "num_input_tokens_seen": 666176,
      "step": 2340
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.7663853168487549,
      "learning_rate": 3.45821255141113e-05,
      "loss": 0.5149,
      "num_input_tokens_seen": 667552,
      "step": 2345
    },
    {
      "epoch": 4.384328358208955,
      "grad_norm": 1.348618507385254,
      "learning_rate": 3.450688647741668e-05,
      "loss": 0.5285,
      "num_input_tokens_seen": 669056,
      "step": 2350
    },
    {
      "epoch": 4.393656716417911,
      "grad_norm": 1.047650694847107,
      "learning_rate": 3.443154664013067e-05,
      "loss": 0.5163,
      "num_input_tokens_seen": 670592,
      "step": 2355
    },
    {
      "epoch": 4.402985074626866,
      "grad_norm": 0.8788366913795471,
      "learning_rate": 3.4356106801074245e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 672096,
      "step": 2360
    },
    {
      "epoch": 4.412313432835821,
      "grad_norm": 1.7129604816436768,
      "learning_rate": 3.4280567760128656e-05,
      "loss": 0.8765,
      "num_input_tokens_seen": 673568,
      "step": 2365
    },
    {
      "epoch": 4.4216417910447765,
      "grad_norm": 0.8176696300506592,
      "learning_rate": 3.4204930318227016e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 674976,
      "step": 2370
    },
    {
      "epoch": 4.4309701492537314,
      "grad_norm": 1.6699620485305786,
      "learning_rate": 3.4129195277345724e-05,
      "loss": 0.6621,
      "num_input_tokens_seen": 676288,
      "step": 2375
    },
    {
      "epoch": 4.440298507462686,
      "grad_norm": 1.5446295738220215,
      "learning_rate": 3.405336344049607e-05,
      "loss": 0.4044,
      "num_input_tokens_seen": 677888,
      "step": 2380
    },
    {
      "epoch": 4.449626865671641,
      "grad_norm": 1.567613959312439,
      "learning_rate": 3.397743561171562e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 679328,
      "step": 2385
    },
    {
      "epoch": 4.458955223880597,
      "grad_norm": 1.7097142934799194,
      "learning_rate": 3.390141259605975e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 680736,
      "step": 2390
    },
    {
      "epoch": 4.468283582089552,
      "grad_norm": 0.8298039436340332,
      "learning_rate": 3.3825295199593084e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 682304,
      "step": 2395
    },
    {
      "epoch": 4.477611940298507,
      "grad_norm": 1.5298432111740112,
      "learning_rate": 3.3749084229380976e-05,
      "loss": 0.644,
      "num_input_tokens_seen": 683936,
      "step": 2400
    },
    {
      "epoch": 4.486940298507463,
      "grad_norm": 1.1645063161849976,
      "learning_rate": 3.367278049348093e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 685312,
      "step": 2405
    },
    {
      "epoch": 4.496268656716418,
      "grad_norm": 1.344023585319519,
      "learning_rate": 3.3596384800934e-05,
      "loss": 0.4024,
      "num_input_tokens_seen": 686752,
      "step": 2410
    },
    {
      "epoch": 4.5,
      "eval_loss": 0.6074901819229126,
      "eval_runtime": 4.5782,
      "eval_samples_per_second": 51.985,
      "eval_steps_per_second": 13.106,
      "num_input_tokens_seen": 687328,
      "step": 2412
    },
    {
      "epoch": 4.505597014925373,
      "grad_norm": 1.4670190811157227,
      "learning_rate": 3.351989796175628e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 688096,
      "step": 2415
    },
    {
      "epoch": 4.514925373134329,
      "grad_norm": 1.435750126838684,
      "learning_rate": 3.3443320786930275e-05,
      "loss": 0.6047,
      "num_input_tokens_seen": 689664,
      "step": 2420
    },
    {
      "epoch": 4.524253731343284,
      "grad_norm": 2.409498453140259,
      "learning_rate": 3.3366654088396326e-05,
      "loss": 0.8501,
      "num_input_tokens_seen": 690912,
      "step": 2425
    },
    {
      "epoch": 4.5335820895522385,
      "grad_norm": 1.8768709897994995,
      "learning_rate": 3.328989867904396e-05,
      "loss": 0.3952,
      "num_input_tokens_seen": 692320,
      "step": 2430
    },
    {
      "epoch": 4.542910447761194,
      "grad_norm": 1.4717427492141724,
      "learning_rate": 3.3213055372703305e-05,
      "loss": 0.457,
      "num_input_tokens_seen": 693568,
      "step": 2435
    },
    {
      "epoch": 4.552238805970149,
      "grad_norm": 1.3881869316101074,
      "learning_rate": 3.313612498413646e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 695168,
      "step": 2440
    },
    {
      "epoch": 4.561567164179104,
      "grad_norm": 2.0043697357177734,
      "learning_rate": 3.305910832902884e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 696384,
      "step": 2445
    },
    {
      "epoch": 4.57089552238806,
      "grad_norm": 0.9538807272911072,
      "learning_rate": 3.298200622398054e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 697792,
      "step": 2450
    },
    {
      "epoch": 4.580223880597015,
      "grad_norm": 1.473875641822815,
      "learning_rate": 3.290481948649767e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 699232,
      "step": 2455
    },
    {
      "epoch": 4.58955223880597,
      "grad_norm": 0.9104630351066589,
      "learning_rate": 3.282754893498369e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 700704,
      "step": 2460
    },
    {
      "epoch": 4.598880597014926,
      "grad_norm": 2.381241798400879,
      "learning_rate": 3.275019538873071e-05,
      "loss": 0.7093,
      "num_input_tokens_seen": 702016,
      "step": 2465
    },
    {
      "epoch": 4.608208955223881,
      "grad_norm": 1.6714354753494263,
      "learning_rate": 3.267275966791088e-05,
      "loss": 0.6003,
      "num_input_tokens_seen": 703456,
      "step": 2470
    },
    {
      "epoch": 4.617537313432836,
      "grad_norm": 2.9158918857574463,
      "learning_rate": 3.259524259356759e-05,
      "loss": 0.5657,
      "num_input_tokens_seen": 704896,
      "step": 2475
    },
    {
      "epoch": 4.6268656716417915,
      "grad_norm": 1.8916445970535278,
      "learning_rate": 3.251764498760683e-05,
      "loss": 0.5218,
      "num_input_tokens_seen": 706272,
      "step": 2480
    },
    {
      "epoch": 4.6361940298507465,
      "grad_norm": 1.530457615852356,
      "learning_rate": 3.243996767278846e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 707744,
      "step": 2485
    },
    {
      "epoch": 4.645522388059701,
      "grad_norm": 1.1385972499847412,
      "learning_rate": 3.2362211472717484e-05,
      "loss": 0.6437,
      "num_input_tokens_seen": 709120,
      "step": 2490
    },
    {
      "epoch": 4.654850746268656,
      "grad_norm": 1.269961953163147,
      "learning_rate": 3.228437721183531e-05,
      "loss": 0.5472,
      "num_input_tokens_seen": 710592,
      "step": 2495
    },
    {
      "epoch": 4.664179104477612,
      "grad_norm": 1.5249689817428589,
      "learning_rate": 3.220646571541105e-05,
      "loss": 0.7148,
      "num_input_tokens_seen": 712000,
      "step": 2500
    },
    {
      "epoch": 4.673507462686567,
      "grad_norm": 3.2751009464263916,
      "learning_rate": 3.2128477809532684e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 713408,
      "step": 2505
    },
    {
      "epoch": 4.682835820895522,
      "grad_norm": 1.3634408712387085,
      "learning_rate": 3.2050414321098385e-05,
      "loss": 0.5955,
      "num_input_tokens_seen": 714688,
      "step": 2510
    },
    {
      "epoch": 4.692164179104478,
      "grad_norm": 2.4926040172576904,
      "learning_rate": 3.197227607780774e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 716128,
      "step": 2515
    },
    {
      "epoch": 4.701492537313433,
      "grad_norm": 2.2082693576812744,
      "learning_rate": 3.1894063908152916e-05,
      "loss": 0.6304,
      "num_input_tokens_seen": 717632,
      "step": 2520
    },
    {
      "epoch": 4.710820895522388,
      "grad_norm": 2.0415198802948,
      "learning_rate": 3.181577864140992e-05,
      "loss": 0.6188,
      "num_input_tokens_seen": 719264,
      "step": 2525
    },
    {
      "epoch": 4.720149253731344,
      "grad_norm": 1.4403802156448364,
      "learning_rate": 3.173742110762984e-05,
      "loss": 0.5871,
      "num_input_tokens_seen": 720832,
      "step": 2530
    },
    {
      "epoch": 4.729477611940299,
      "grad_norm": 1.8611282110214233,
      "learning_rate": 3.165899213762995e-05,
      "loss": 0.612,
      "num_input_tokens_seen": 722240,
      "step": 2535
    },
    {
      "epoch": 4.7388059701492535,
      "grad_norm": 1.445170283317566,
      "learning_rate": 3.158049256298499e-05,
      "loss": 0.7647,
      "num_input_tokens_seen": 723552,
      "step": 2540
    },
    {
      "epoch": 4.7481343283582085,
      "grad_norm": 2.182626724243164,
      "learning_rate": 3.15019232160183e-05,
      "loss": 0.6878,
      "num_input_tokens_seen": 724864,
      "step": 2545
    },
    {
      "epoch": 4.757462686567164,
      "grad_norm": 1.8532062768936157,
      "learning_rate": 3.142328492979301e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 726176,
      "step": 2550
    },
    {
      "epoch": 4.766791044776119,
      "grad_norm": 1.493152379989624,
      "learning_rate": 3.134457853810322e-05,
      "loss": 0.5863,
      "num_input_tokens_seen": 727680,
      "step": 2555
    },
    {
      "epoch": 4.776119402985074,
      "grad_norm": 0.9484604597091675,
      "learning_rate": 3.126580487546513e-05,
      "loss": 0.6949,
      "num_input_tokens_seen": 729024,
      "step": 2560
    },
    {
      "epoch": 4.78544776119403,
      "grad_norm": 1.9140299558639526,
      "learning_rate": 3.1186964777108215e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 730304,
      "step": 2565
    },
    {
      "epoch": 4.794776119402985,
      "grad_norm": 2.299058675765991,
      "learning_rate": 3.110805907896637e-05,
      "loss": 0.4843,
      "num_input_tokens_seen": 731776,
      "step": 2570
    },
    {
      "epoch": 4.80410447761194,
      "grad_norm": 0.8986464738845825,
      "learning_rate": 3.1029088617669e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 733152,
      "step": 2575
    },
    {
      "epoch": 4.813432835820896,
      "grad_norm": 2.075289726257324,
      "learning_rate": 3.0950054230532235e-05,
      "loss": 0.7283,
      "num_input_tokens_seen": 734432,
      "step": 2580
    },
    {
      "epoch": 4.822761194029851,
      "grad_norm": 1.595855712890625,
      "learning_rate": 3.0870956755549976e-05,
      "loss": 0.6463,
      "num_input_tokens_seen": 735808,
      "step": 2585
    },
    {
      "epoch": 4.832089552238806,
      "grad_norm": 1.7077938318252563,
      "learning_rate": 3.079179703138505e-05,
      "loss": 0.604,
      "num_input_tokens_seen": 737088,
      "step": 2590
    },
    {
      "epoch": 4.8414179104477615,
      "grad_norm": 1.4575996398925781,
      "learning_rate": 3.0712575897360304e-05,
      "loss": 0.4881,
      "num_input_tokens_seen": 738528,
      "step": 2595
    },
    {
      "epoch": 4.850746268656716,
      "grad_norm": 1.2425379753112793,
      "learning_rate": 3.0633294193449695e-05,
      "loss": 0.5754,
      "num_input_tokens_seen": 740128,
      "step": 2600
    },
    {
      "epoch": 4.860074626865671,
      "grad_norm": 2.409543991088867,
      "learning_rate": 3.0553952760269426e-05,
      "loss": 0.5891,
      "num_input_tokens_seen": 741760,
      "step": 2605
    },
    {
      "epoch": 4.869402985074627,
      "grad_norm": 2.1821789741516113,
      "learning_rate": 3.0474552439068978e-05,
      "loss": 0.6176,
      "num_input_tokens_seen": 743136,
      "step": 2610
    },
    {
      "epoch": 4.878731343283582,
      "grad_norm": 1.2381558418273926,
      "learning_rate": 3.039509407172222e-05,
      "loss": 0.3686,
      "num_input_tokens_seen": 744384,
      "step": 2615
    },
    {
      "epoch": 4.888059701492537,
      "grad_norm": 1.3593685626983643,
      "learning_rate": 3.0315578500718476e-05,
      "loss": 0.5985,
      "num_input_tokens_seen": 745888,
      "step": 2620
    },
    {
      "epoch": 4.897388059701493,
      "grad_norm": 1.941843867301941,
      "learning_rate": 3.0236006569153617e-05,
      "loss": 0.6793,
      "num_input_tokens_seen": 747360,
      "step": 2625
    },
    {
      "epoch": 4.906716417910448,
      "grad_norm": 2.097285747528076,
      "learning_rate": 3.0156379120721068e-05,
      "loss": 0.8462,
      "num_input_tokens_seen": 748576,
      "step": 2630
    },
    {
      "epoch": 4.916044776119403,
      "grad_norm": 1.4640358686447144,
      "learning_rate": 3.0076696999702913e-05,
      "loss": 0.4057,
      "num_input_tokens_seen": 749792,
      "step": 2635
    },
    {
      "epoch": 4.925373134328359,
      "grad_norm": 1.5823386907577515,
      "learning_rate": 2.9996961050960932e-05,
      "loss": 0.4217,
      "num_input_tokens_seen": 751200,
      "step": 2640
    },
    {
      "epoch": 4.934701492537314,
      "grad_norm": 1.6937531232833862,
      "learning_rate": 2.9917172119927606e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 752608,
      "step": 2645
    },
    {
      "epoch": 4.9440298507462686,
      "grad_norm": 1.498105525970459,
      "learning_rate": 2.9837331052597224e-05,
      "loss": 0.5351,
      "num_input_tokens_seen": 754144,
      "step": 2650
    },
    {
      "epoch": 4.9533582089552235,
      "grad_norm": 1.8841726779937744,
      "learning_rate": 2.9757438695516816e-05,
      "loss": 0.5143,
      "num_input_tokens_seen": 755584,
      "step": 2655
    },
    {
      "epoch": 4.962686567164179,
      "grad_norm": 1.3817956447601318,
      "learning_rate": 2.9677495895777286e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 757152,
      "step": 2660
    },
    {
      "epoch": 4.972014925373134,
      "grad_norm": 1.7007989883422852,
      "learning_rate": 2.9597503501004343e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 758528,
      "step": 2665
    },
    {
      "epoch": 4.981343283582089,
      "grad_norm": 1.9800549745559692,
      "learning_rate": 2.9517462359349553e-05,
      "loss": 0.6581,
      "num_input_tokens_seen": 759904,
      "step": 2670
    },
    {
      "epoch": 4.990671641791045,
      "grad_norm": 1.4242937564849854,
      "learning_rate": 2.943737331948136e-05,
      "loss": 0.54,
      "num_input_tokens_seen": 761344,
      "step": 2675
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.3624165058135986,
      "learning_rate": 2.9357237230576045e-05,
      "loss": 0.8237,
      "num_input_tokens_seen": 762440,
      "step": 2680
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.5997107028961182,
      "eval_runtime": 4.5677,
      "eval_samples_per_second": 52.104,
      "eval_steps_per_second": 13.136,
      "num_input_tokens_seen": 762440,
      "step": 2680
    },
    {
      "epoch": 5.009328358208955,
      "grad_norm": 1.2942463159561157,
      "learning_rate": 2.927705494230875e-05,
      "loss": 0.5786,
      "num_input_tokens_seen": 763656,
      "step": 2685
    },
    {
      "epoch": 5.018656716417911,
      "grad_norm": 1.856074571609497,
      "learning_rate": 2.9196827304844483e-05,
      "loss": 0.6471,
      "num_input_tokens_seen": 764968,
      "step": 2690
    },
    {
      "epoch": 5.027985074626866,
      "grad_norm": 0.861467719078064,
      "learning_rate": 2.911655516882905e-05,
      "loss": 0.3831,
      "num_input_tokens_seen": 766440,
      "step": 2695
    },
    {
      "epoch": 5.037313432835821,
      "grad_norm": 1.7158617973327637,
      "learning_rate": 2.9036239385380098e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 767784,
      "step": 2700
    },
    {
      "epoch": 5.0466417910447765,
      "grad_norm": 2.1512506008148193,
      "learning_rate": 2.8955880806078068e-05,
      "loss": 0.59,
      "num_input_tokens_seen": 769128,
      "step": 2705
    },
    {
      "epoch": 5.0559701492537314,
      "grad_norm": 0.9823271036148071,
      "learning_rate": 2.8875480282957133e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 770536,
      "step": 2710
    },
    {
      "epoch": 5.065298507462686,
      "grad_norm": 2.973296642303467,
      "learning_rate": 2.8795038668496222e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 772040,
      "step": 2715
    },
    {
      "epoch": 5.074626865671641,
      "grad_norm": 2.0613298416137695,
      "learning_rate": 2.8714556815609926e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 773480,
      "step": 2720
    },
    {
      "epoch": 5.083955223880597,
      "grad_norm": 1.8094143867492676,
      "learning_rate": 2.8634035577639505e-05,
      "loss": 0.5133,
      "num_input_tokens_seen": 774888,
      "step": 2725
    },
    {
      "epoch": 5.093283582089552,
      "grad_norm": 2.9016542434692383,
      "learning_rate": 2.8553475808343798e-05,
      "loss": 0.6089,
      "num_input_tokens_seen": 776296,
      "step": 2730
    },
    {
      "epoch": 5.102611940298507,
      "grad_norm": 1.173405647277832,
      "learning_rate": 2.847287836189019e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 777928,
      "step": 2735
    },
    {
      "epoch": 5.111940298507463,
      "grad_norm": 1.099913477897644,
      "learning_rate": 2.839224409284556e-05,
      "loss": 0.4895,
      "num_input_tokens_seen": 779304,
      "step": 2740
    },
    {
      "epoch": 5.121268656716418,
      "grad_norm": 2.125192403793335,
      "learning_rate": 2.83115738561672e-05,
      "loss": 0.5916,
      "num_input_tokens_seen": 780456,
      "step": 2745
    },
    {
      "epoch": 5.130597014925373,
      "grad_norm": 2.2636284828186035,
      "learning_rate": 2.8230868507193785e-05,
      "loss": 0.5908,
      "num_input_tokens_seen": 781800,
      "step": 2750
    },
    {
      "epoch": 5.139925373134329,
      "grad_norm": 1.27336585521698,
      "learning_rate": 2.8150128901636262e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 783176,
      "step": 2755
    },
    {
      "epoch": 5.149253731343284,
      "grad_norm": 1.4693082571029663,
      "learning_rate": 2.8069355895568805e-05,
      "loss": 0.5037,
      "num_input_tokens_seen": 784488,
      "step": 2760
    },
    {
      "epoch": 5.1585820895522385,
      "grad_norm": 1.7354995012283325,
      "learning_rate": 2.7988550345419733e-05,
      "loss": 0.4295,
      "num_input_tokens_seen": 785928,
      "step": 2765
    },
    {
      "epoch": 5.167910447761194,
      "grad_norm": 1.460009217262268,
      "learning_rate": 2.790771310796243e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 787368,
      "step": 2770
    },
    {
      "epoch": 5.177238805970149,
      "grad_norm": 2.7251977920532227,
      "learning_rate": 2.7826845040306238e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 788744,
      "step": 2775
    },
    {
      "epoch": 5.186567164179104,
      "grad_norm": 1.1219333410263062,
      "learning_rate": 2.7745946999887406e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 790216,
      "step": 2780
    },
    {
      "epoch": 5.19589552238806,
      "grad_norm": 1.7065500020980835,
      "learning_rate": 2.766501984445999e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 791656,
      "step": 2785
    },
    {
      "epoch": 5.205223880597015,
      "grad_norm": 1.9253047704696655,
      "learning_rate": 2.7584064432086716e-05,
      "loss": 0.5117,
      "num_input_tokens_seen": 793288,
      "step": 2790
    },
    {
      "epoch": 5.21455223880597,
      "grad_norm": 1.573366403579712,
      "learning_rate": 2.750308162112995e-05,
      "loss": 0.4868,
      "num_input_tokens_seen": 794984,
      "step": 2795
    },
    {
      "epoch": 5.223880597014926,
      "grad_norm": 1.3211243152618408,
      "learning_rate": 2.7422072270242548e-05,
      "loss": 0.5504,
      "num_input_tokens_seen": 796328,
      "step": 2800
    },
    {
      "epoch": 5.233208955223881,
      "grad_norm": 2.41324782371521,
      "learning_rate": 2.7341037238358774e-05,
      "loss": 0.7952,
      "num_input_tokens_seen": 797704,
      "step": 2805
    },
    {
      "epoch": 5.242537313432836,
      "grad_norm": 2.3553287982940674,
      "learning_rate": 2.7259977384685163e-05,
      "loss": 0.4316,
      "num_input_tokens_seen": 798984,
      "step": 2810
    },
    {
      "epoch": 5.251865671641791,
      "grad_norm": 1.6046924591064453,
      "learning_rate": 2.717889356869146e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 800392,
      "step": 2815
    },
    {
      "epoch": 5.2611940298507465,
      "grad_norm": 1.816957950592041,
      "learning_rate": 2.7097786650101458e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 801832,
      "step": 2820
    },
    {
      "epoch": 5.270522388059701,
      "grad_norm": 1.8928248882293701,
      "learning_rate": 2.7016657488883928e-05,
      "loss": 0.7359,
      "num_input_tokens_seen": 803272,
      "step": 2825
    },
    {
      "epoch": 5.279850746268656,
      "grad_norm": 2.4919962882995605,
      "learning_rate": 2.693550694524346e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 804872,
      "step": 2830
    },
    {
      "epoch": 5.289179104477612,
      "grad_norm": 1.5413204431533813,
      "learning_rate": 2.685433587961136e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 806312,
      "step": 2835
    },
    {
      "epoch": 5.298507462686567,
      "grad_norm": 1.1456393003463745,
      "learning_rate": 2.6773145152636536e-05,
      "loss": 0.4171,
      "num_input_tokens_seen": 807944,
      "step": 2840
    },
    {
      "epoch": 5.307835820895522,
      "grad_norm": 2.876997947692871,
      "learning_rate": 2.6691935625176357e-05,
      "loss": 0.5842,
      "num_input_tokens_seen": 809320,
      "step": 2845
    },
    {
      "epoch": 5.317164179104478,
      "grad_norm": 1.3236156702041626,
      "learning_rate": 2.6610708158287535e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 810952,
      "step": 2850
    },
    {
      "epoch": 5.326492537313433,
      "grad_norm": 1.7661559581756592,
      "learning_rate": 2.6529463613216986e-05,
      "loss": 0.4181,
      "num_input_tokens_seen": 812264,
      "step": 2855
    },
    {
      "epoch": 5.335820895522388,
      "grad_norm": 2.199836492538452,
      "learning_rate": 2.644820285139271e-05,
      "loss": 0.6136,
      "num_input_tokens_seen": 813704,
      "step": 2860
    },
    {
      "epoch": 5.345149253731344,
      "grad_norm": 2.0607194900512695,
      "learning_rate": 2.636692673441465e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 815272,
      "step": 2865
    },
    {
      "epoch": 5.354477611940299,
      "grad_norm": 1.781184434890747,
      "learning_rate": 2.628563612404556e-05,
      "loss": 0.581,
      "num_input_tokens_seen": 816904,
      "step": 2870
    },
    {
      "epoch": 5.3638059701492535,
      "grad_norm": 1.3373324871063232,
      "learning_rate": 2.6204331882201864e-05,
      "loss": 0.6485,
      "num_input_tokens_seen": 818440,
      "step": 2875
    },
    {
      "epoch": 5.373134328358209,
      "grad_norm": 2.280576467514038,
      "learning_rate": 2.6123014870944517e-05,
      "loss": 0.6548,
      "num_input_tokens_seen": 819752,
      "step": 2880
    },
    {
      "epoch": 5.382462686567164,
      "grad_norm": 2.9549825191497803,
      "learning_rate": 2.6041685952469875e-05,
      "loss": 0.7016,
      "num_input_tokens_seen": 820936,
      "step": 2885
    },
    {
      "epoch": 5.391791044776119,
      "grad_norm": 3.4383857250213623,
      "learning_rate": 2.596034598910053e-05,
      "loss": 0.5732,
      "num_input_tokens_seen": 822408,
      "step": 2890
    },
    {
      "epoch": 5.401119402985074,
      "grad_norm": 1.6611424684524536,
      "learning_rate": 2.5878995843276204e-05,
      "loss": 0.6594,
      "num_input_tokens_seen": 823784,
      "step": 2895
    },
    {
      "epoch": 5.41044776119403,
      "grad_norm": 2.0401363372802734,
      "learning_rate": 2.5797636377544564e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 825064,
      "step": 2900
    },
    {
      "epoch": 5.419776119402985,
      "grad_norm": 1.7741297483444214,
      "learning_rate": 2.5716268454552095e-05,
      "loss": 0.5565,
      "num_input_tokens_seen": 826600,
      "step": 2905
    },
    {
      "epoch": 5.42910447761194,
      "grad_norm": 1.6536259651184082,
      "learning_rate": 2.563489293703496e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 828104,
      "step": 2910
    },
    {
      "epoch": 5.438432835820896,
      "grad_norm": 2.2615702152252197,
      "learning_rate": 2.5553510687809855e-05,
      "loss": 0.5959,
      "num_input_tokens_seen": 829640,
      "step": 2915
    },
    {
      "epoch": 5.447761194029851,
      "grad_norm": 1.718976616859436,
      "learning_rate": 2.547212256976484e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 831112,
      "step": 2920
    },
    {
      "epoch": 5.457089552238806,
      "grad_norm": 2.1744861602783203,
      "learning_rate": 2.5390729445850198e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 832744,
      "step": 2925
    },
    {
      "epoch": 5.4664179104477615,
      "grad_norm": 1.3987616300582886,
      "learning_rate": 2.53093321790693e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 834248,
      "step": 2930
    },
    {
      "epoch": 5.475746268656716,
      "grad_norm": 1.3535387516021729,
      "learning_rate": 2.5227931632469437e-05,
      "loss": 0.6761,
      "num_input_tokens_seen": 835752,
      "step": 2935
    },
    {
      "epoch": 5.485074626865671,
      "grad_norm": 1.0129848718643188,
      "learning_rate": 2.5146528669132664e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 837064,
      "step": 2940
    },
    {
      "epoch": 5.494402985074627,
      "grad_norm": 2.2536509037017822,
      "learning_rate": 2.506512415216669e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 838440,
      "step": 2945
    },
    {
      "epoch": 5.5,
      "eval_loss": 0.6064518094062805,
      "eval_runtime": 4.5785,
      "eval_samples_per_second": 51.982,
      "eval_steps_per_second": 13.105,
      "num_input_tokens_seen": 839656,
      "step": 2948
    },
    {
      "epoch": 5.503731343283582,
      "grad_norm": 1.7913509607315063,
      "learning_rate": 2.4983718944695672e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 840232,
      "step": 2950
    },
    {
      "epoch": 5.513059701492537,
      "grad_norm": 1.927219271659851,
      "learning_rate": 2.49023139098511e-05,
      "loss": 0.4897,
      "num_input_tokens_seen": 841576,
      "step": 2955
    },
    {
      "epoch": 5.522388059701493,
      "grad_norm": 2.661860942840576,
      "learning_rate": 2.4820909910762628e-05,
      "loss": 0.6269,
      "num_input_tokens_seen": 842984,
      "step": 2960
    },
    {
      "epoch": 5.531716417910448,
      "grad_norm": 1.848433494567871,
      "learning_rate": 2.473950781054893e-05,
      "loss": 0.4699,
      "num_input_tokens_seen": 844232,
      "step": 2965
    },
    {
      "epoch": 5.541044776119403,
      "grad_norm": 2.5200202465057373,
      "learning_rate": 2.4658108472308545e-05,
      "loss": 0.6574,
      "num_input_tokens_seen": 845576,
      "step": 2970
    },
    {
      "epoch": 5.550373134328359,
      "grad_norm": 1.4073548316955566,
      "learning_rate": 2.4576712759110728e-05,
      "loss": 0.4742,
      "num_input_tokens_seen": 846984,
      "step": 2975
    },
    {
      "epoch": 5.559701492537314,
      "grad_norm": 2.6013896465301514,
      "learning_rate": 2.4495321533986297e-05,
      "loss": 0.5868,
      "num_input_tokens_seen": 848328,
      "step": 2980
    },
    {
      "epoch": 5.5690298507462686,
      "grad_norm": 2.357034206390381,
      "learning_rate": 2.441393565991849e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 849800,
      "step": 2985
    },
    {
      "epoch": 5.5783582089552235,
      "grad_norm": 1.8608379364013672,
      "learning_rate": 2.43325559998338e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 851176,
      "step": 2990
    },
    {
      "epoch": 5.587686567164179,
      "grad_norm": 1.3976454734802246,
      "learning_rate": 2.425118341659284e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 852776,
      "step": 2995
    },
    {
      "epoch": 5.597014925373134,
      "grad_norm": 1.4467735290527344,
      "learning_rate": 2.416981877298118e-05,
      "loss": 0.6054,
      "num_input_tokens_seen": 854184,
      "step": 3000
    },
    {
      "epoch": 5.606343283582089,
      "grad_norm": 1.4580317735671997,
      "learning_rate": 2.4088462931700214e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 855496,
      "step": 3005
    },
    {
      "epoch": 5.615671641791045,
      "grad_norm": 1.6334571838378906,
      "learning_rate": 2.4007116755357995e-05,
      "loss": 0.665,
      "num_input_tokens_seen": 856840,
      "step": 3010
    },
    {
      "epoch": 5.625,
      "grad_norm": 4.509402751922607,
      "learning_rate": 2.392578110646012e-05,
      "loss": 0.4688,
      "num_input_tokens_seen": 858184,
      "step": 3015
    },
    {
      "epoch": 5.634328358208955,
      "grad_norm": 3.650466203689575,
      "learning_rate": 2.384445684740055e-05,
      "loss": 0.5398,
      "num_input_tokens_seen": 859624,
      "step": 3020
    },
    {
      "epoch": 5.643656716417911,
      "grad_norm": 1.482712745666504,
      "learning_rate": 2.376314484045248e-05,
      "loss": 0.6066,
      "num_input_tokens_seen": 860968,
      "step": 3025
    },
    {
      "epoch": 5.652985074626866,
      "grad_norm": 1.8888198137283325,
      "learning_rate": 2.3681845947759206e-05,
      "loss": 0.6276,
      "num_input_tokens_seen": 862472,
      "step": 3030
    },
    {
      "epoch": 5.662313432835821,
      "grad_norm": 2.44982647895813,
      "learning_rate": 2.3600561031324958e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 863880,
      "step": 3035
    },
    {
      "epoch": 5.6716417910447765,
      "grad_norm": 1.58240807056427,
      "learning_rate": 2.3519290953005784e-05,
      "loss": 0.5124,
      "num_input_tokens_seen": 865288,
      "step": 3040
    },
    {
      "epoch": 5.6809701492537314,
      "grad_norm": 1.0377243757247925,
      "learning_rate": 2.3438036574500432e-05,
      "loss": 0.3272,
      "num_input_tokens_seen": 866824,
      "step": 3045
    },
    {
      "epoch": 5.690298507462686,
      "grad_norm": 1.284062147140503,
      "learning_rate": 2.3356798757341155e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 868104,
      "step": 3050
    },
    {
      "epoch": 5.699626865671641,
      "grad_norm": 2.1024458408355713,
      "learning_rate": 2.327557836288461e-05,
      "loss": 0.3573,
      "num_input_tokens_seen": 869512,
      "step": 3055
    },
    {
      "epoch": 5.708955223880597,
      "grad_norm": 2.8066346645355225,
      "learning_rate": 2.3194376252302723e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 871048,
      "step": 3060
    },
    {
      "epoch": 5.718283582089552,
      "grad_norm": 1.4103471040725708,
      "learning_rate": 2.311319328657358e-05,
      "loss": 0.6122,
      "num_input_tokens_seen": 872424,
      "step": 3065
    },
    {
      "epoch": 5.727611940298507,
      "grad_norm": 1.7628196477890015,
      "learning_rate": 2.3032030326472235e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 873992,
      "step": 3070
    },
    {
      "epoch": 5.736940298507463,
      "grad_norm": 1.2805485725402832,
      "learning_rate": 2.2950888232561672e-05,
      "loss": 0.3543,
      "num_input_tokens_seen": 875368,
      "step": 3075
    },
    {
      "epoch": 5.746268656716418,
      "grad_norm": 2.103849411010742,
      "learning_rate": 2.2869767865183612e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 876808,
      "step": 3080
    },
    {
      "epoch": 5.755597014925373,
      "grad_norm": 2.2846083641052246,
      "learning_rate": 2.2788670084449394e-05,
      "loss": 0.6798,
      "num_input_tokens_seen": 878152,
      "step": 3085
    },
    {
      "epoch": 5.764925373134329,
      "grad_norm": 1.7905958890914917,
      "learning_rate": 2.2707595750230894e-05,
      "loss": 0.4615,
      "num_input_tokens_seen": 879496,
      "step": 3090
    },
    {
      "epoch": 5.774253731343284,
      "grad_norm": 2.518063545227051,
      "learning_rate": 2.2626545722151384e-05,
      "loss": 0.765,
      "num_input_tokens_seen": 880904,
      "step": 3095
    },
    {
      "epoch": 5.7835820895522385,
      "grad_norm": 2.478855848312378,
      "learning_rate": 2.254552085957641e-05,
      "loss": 0.644,
      "num_input_tokens_seen": 882152,
      "step": 3100
    },
    {
      "epoch": 5.792910447761194,
      "grad_norm": 3.0239367485046387,
      "learning_rate": 2.246452202160471e-05,
      "loss": 0.5638,
      "num_input_tokens_seen": 883400,
      "step": 3105
    },
    {
      "epoch": 5.802238805970149,
      "grad_norm": 1.8436428308486938,
      "learning_rate": 2.2383550067059063e-05,
      "loss": 0.363,
      "num_input_tokens_seen": 884968,
      "step": 3110
    },
    {
      "epoch": 5.811567164179104,
      "grad_norm": 3.2736122608184814,
      "learning_rate": 2.2302605854477228e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 886312,
      "step": 3115
    },
    {
      "epoch": 5.82089552238806,
      "grad_norm": 1.4033927917480469,
      "learning_rate": 2.22216902421028e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 887752,
      "step": 3120
    },
    {
      "epoch": 5.830223880597015,
      "grad_norm": 3.0840935707092285,
      "learning_rate": 2.2140804087876132e-05,
      "loss": 0.5508,
      "num_input_tokens_seen": 889096,
      "step": 3125
    },
    {
      "epoch": 5.83955223880597,
      "grad_norm": 1.9307622909545898,
      "learning_rate": 2.2059948249425243e-05,
      "loss": 0.5277,
      "num_input_tokens_seen": 890408,
      "step": 3130
    },
    {
      "epoch": 5.848880597014926,
      "grad_norm": 1.9568920135498047,
      "learning_rate": 2.197912358405672e-05,
      "loss": 0.3063,
      "num_input_tokens_seen": 892072,
      "step": 3135
    },
    {
      "epoch": 5.858208955223881,
      "grad_norm": 2.353299379348755,
      "learning_rate": 2.189833094874661e-05,
      "loss": 0.6174,
      "num_input_tokens_seen": 893608,
      "step": 3140
    },
    {
      "epoch": 5.867537313432836,
      "grad_norm": 3.459669351577759,
      "learning_rate": 2.1817571200131358e-05,
      "loss": 0.6511,
      "num_input_tokens_seen": 894920,
      "step": 3145
    },
    {
      "epoch": 5.8768656716417915,
      "grad_norm": 1.924635648727417,
      "learning_rate": 2.173684519449872e-05,
      "loss": 0.488,
      "num_input_tokens_seen": 896328,
      "step": 3150
    },
    {
      "epoch": 5.8861940298507465,
      "grad_norm": 1.1924333572387695,
      "learning_rate": 2.1656153787778645e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 897896,
      "step": 3155
    },
    {
      "epoch": 5.895522388059701,
      "grad_norm": 1.045536994934082,
      "learning_rate": 2.157549783553429e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 899272,
      "step": 3160
    },
    {
      "epoch": 5.904850746268656,
      "grad_norm": 1.9812746047973633,
      "learning_rate": 2.1494878192952855e-05,
      "loss": 0.3643,
      "num_input_tokens_seen": 900584,
      "step": 3165
    },
    {
      "epoch": 5.914179104477612,
      "grad_norm": 1.337859034538269,
      "learning_rate": 2.141429571483655e-05,
      "loss": 0.4565,
      "num_input_tokens_seen": 901992,
      "step": 3170
    },
    {
      "epoch": 5.923507462686567,
      "grad_norm": 1.733487606048584,
      "learning_rate": 2.1333751255593552e-05,
      "loss": 0.7281,
      "num_input_tokens_seen": 903304,
      "step": 3175
    },
    {
      "epoch": 5.932835820895522,
      "grad_norm": 3.2020914554595947,
      "learning_rate": 2.1253245669228905e-05,
      "loss": 0.4534,
      "num_input_tokens_seen": 904712,
      "step": 3180
    },
    {
      "epoch": 5.942164179104478,
      "grad_norm": 1.3351143598556519,
      "learning_rate": 2.1172779809335496e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 906120,
      "step": 3185
    },
    {
      "epoch": 5.951492537313433,
      "grad_norm": 1.8775660991668701,
      "learning_rate": 2.109235452908502e-05,
      "loss": 0.6197,
      "num_input_tokens_seen": 907464,
      "step": 3190
    },
    {
      "epoch": 5.960820895522388,
      "grad_norm": 1.8199220895767212,
      "learning_rate": 2.101197068121888e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 909000,
      "step": 3195
    },
    {
      "epoch": 5.970149253731344,
      "grad_norm": 1.4227691888809204,
      "learning_rate": 2.0931629118039185e-05,
      "loss": 0.4629,
      "num_input_tokens_seen": 910280,
      "step": 3200
    },
    {
      "epoch": 5.979477611940299,
      "grad_norm": 1.748199462890625,
      "learning_rate": 2.0851330691399713e-05,
      "loss": 0.6795,
      "num_input_tokens_seen": 911816,
      "step": 3205
    },
    {
      "epoch": 5.9888059701492535,
      "grad_norm": 2.593017101287842,
      "learning_rate": 2.0771076252696846e-05,
      "loss": 0.6185,
      "num_input_tokens_seen": 913320,
      "step": 3210
    },
    {
      "epoch": 5.9981343283582085,
      "grad_norm": 2.1509287357330322,
      "learning_rate": 2.0690866652860584e-05,
      "loss": 0.5024,
      "num_input_tokens_seen": 914856,
      "step": 3215
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.6061959862709045,
      "eval_runtime": 4.5708,
      "eval_samples_per_second": 52.07,
      "eval_steps_per_second": 13.127,
      "num_input_tokens_seen": 914920,
      "step": 3216
    },
    {
      "epoch": 6.007462686567164,
      "grad_norm": 1.9272371530532837,
      "learning_rate": 2.0610702742345517e-05,
      "loss": 0.4233,
      "num_input_tokens_seen": 916136,
      "step": 3220
    },
    {
      "epoch": 6.016791044776119,
      "grad_norm": 0.773199737071991,
      "learning_rate": 2.053058537112177e-05,
      "loss": 0.3792,
      "num_input_tokens_seen": 917544,
      "step": 3225
    },
    {
      "epoch": 6.026119402985074,
      "grad_norm": 1.7177700996398926,
      "learning_rate": 2.0450515388666022e-05,
      "loss": 0.6412,
      "num_input_tokens_seen": 919080,
      "step": 3230
    },
    {
      "epoch": 6.03544776119403,
      "grad_norm": 2.3316898345947266,
      "learning_rate": 2.0370493643952507e-05,
      "loss": 0.5139,
      "num_input_tokens_seen": 920840,
      "step": 3235
    },
    {
      "epoch": 6.044776119402985,
      "grad_norm": 1.0655263662338257,
      "learning_rate": 2.0290520985443985e-05,
      "loss": 0.3291,
      "num_input_tokens_seen": 922248,
      "step": 3240
    },
    {
      "epoch": 6.05410447761194,
      "grad_norm": 2.6309804916381836,
      "learning_rate": 2.0210598261082765e-05,
      "loss": 0.6032,
      "num_input_tokens_seen": 923912,
      "step": 3245
    },
    {
      "epoch": 6.063432835820896,
      "grad_norm": 1.9413639307022095,
      "learning_rate": 2.013072631828171e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 925352,
      "step": 3250
    },
    {
      "epoch": 6.072761194029851,
      "grad_norm": 2.293771505355835,
      "learning_rate": 2.005090600391526e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 926856,
      "step": 3255
    },
    {
      "epoch": 6.082089552238806,
      "grad_norm": 1.7775822877883911,
      "learning_rate": 1.9971138164310422e-05,
      "loss": 0.7555,
      "num_input_tokens_seen": 928168,
      "step": 3260
    },
    {
      "epoch": 6.0914179104477615,
      "grad_norm": 2.4788622856140137,
      "learning_rate": 1.9891423645237834e-05,
      "loss": 0.6329,
      "num_input_tokens_seen": 929544,
      "step": 3265
    },
    {
      "epoch": 6.100746268656716,
      "grad_norm": 1.5743602514266968,
      "learning_rate": 1.9811763291902774e-05,
      "loss": 0.5973,
      "num_input_tokens_seen": 930792,
      "step": 3270
    },
    {
      "epoch": 6.110074626865671,
      "grad_norm": 2.0222790241241455,
      "learning_rate": 1.9732157948936204e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 932072,
      "step": 3275
    },
    {
      "epoch": 6.119402985074627,
      "grad_norm": 1.9080196619033813,
      "learning_rate": 1.9652608460385843e-05,
      "loss": 0.5727,
      "num_input_tokens_seen": 933384,
      "step": 3280
    },
    {
      "epoch": 6.128731343283582,
      "grad_norm": 1.4677373170852661,
      "learning_rate": 1.957311566970716e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 934888,
      "step": 3285
    },
    {
      "epoch": 6.138059701492537,
      "grad_norm": 1.6466163396835327,
      "learning_rate": 1.949368041975447e-05,
      "loss": 0.3252,
      "num_input_tokens_seen": 936328,
      "step": 3290
    },
    {
      "epoch": 6.147388059701493,
      "grad_norm": 2.237051248550415,
      "learning_rate": 1.9414303552771997e-05,
      "loss": 0.5443,
      "num_input_tokens_seen": 937864,
      "step": 3295
    },
    {
      "epoch": 6.156716417910448,
      "grad_norm": 3.4381797313690186,
      "learning_rate": 1.9334985910384923e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 939464,
      "step": 3300
    },
    {
      "epoch": 6.166044776119403,
      "grad_norm": 2.380587100982666,
      "learning_rate": 1.9255728333590475e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 940968,
      "step": 3305
    },
    {
      "epoch": 6.175373134328359,
      "grad_norm": 1.474696397781372,
      "learning_rate": 1.9176531662749044e-05,
      "loss": 0.6863,
      "num_input_tokens_seen": 942152,
      "step": 3310
    },
    {
      "epoch": 6.184701492537314,
      "grad_norm": 1.2008843421936035,
      "learning_rate": 1.9097396737575206e-05,
      "loss": 0.4581,
      "num_input_tokens_seen": 943720,
      "step": 3315
    },
    {
      "epoch": 6.1940298507462686,
      "grad_norm": 1.3337721824645996,
      "learning_rate": 1.9018324397128866e-05,
      "loss": 0.619,
      "num_input_tokens_seen": 944904,
      "step": 3320
    },
    {
      "epoch": 6.2033582089552235,
      "grad_norm": 1.9344011545181274,
      "learning_rate": 1.893931547980635e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 946344,
      "step": 3325
    },
    {
      "epoch": 6.212686567164179,
      "grad_norm": 1.4552773237228394,
      "learning_rate": 1.8860370823331513e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 947816,
      "step": 3330
    },
    {
      "epoch": 6.222014925373134,
      "grad_norm": 2.124227285385132,
      "learning_rate": 1.878149126474685e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 949320,
      "step": 3335
    },
    {
      "epoch": 6.231343283582089,
      "grad_norm": 1.597843050956726,
      "learning_rate": 1.870267764040465e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 950664,
      "step": 3340
    },
    {
      "epoch": 6.240671641791045,
      "grad_norm": 2.6282832622528076,
      "learning_rate": 1.8623930785958092e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 952328,
      "step": 3345
    },
    {
      "epoch": 6.25,
      "grad_norm": 1.7921161651611328,
      "learning_rate": 1.8545251536352402e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 953736,
      "step": 3350
    },
    {
      "epoch": 6.259328358208955,
      "grad_norm": 1.298344612121582,
      "learning_rate": 1.8466640725816e-05,
      "loss": 0.3357,
      "num_input_tokens_seen": 955272,
      "step": 3355
    },
    {
      "epoch": 6.268656716417911,
      "grad_norm": 1.9250441789627075,
      "learning_rate": 1.838809918785166e-05,
      "loss": 0.2602,
      "num_input_tokens_seen": 956808,
      "step": 3360
    },
    {
      "epoch": 6.277985074626866,
      "grad_norm": 2.4121971130371094,
      "learning_rate": 1.8309627755227643e-05,
      "loss": 0.738,
      "num_input_tokens_seen": 958152,
      "step": 3365
    },
    {
      "epoch": 6.287313432835821,
      "grad_norm": 1.4859696626663208,
      "learning_rate": 1.8231227259968926e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 959560,
      "step": 3370
    },
    {
      "epoch": 6.2966417910447765,
      "grad_norm": 2.501481294631958,
      "learning_rate": 1.8152898533348317e-05,
      "loss": 0.5118,
      "num_input_tokens_seen": 960776,
      "step": 3375
    },
    {
      "epoch": 6.3059701492537314,
      "grad_norm": 1.3837236166000366,
      "learning_rate": 1.8074642405877673e-05,
      "loss": 0.3681,
      "num_input_tokens_seen": 962280,
      "step": 3380
    },
    {
      "epoch": 6.315298507462686,
      "grad_norm": 2.2088353633880615,
      "learning_rate": 1.799645970729909e-05,
      "loss": 0.4548,
      "num_input_tokens_seen": 963752,
      "step": 3385
    },
    {
      "epoch": 6.324626865671641,
      "grad_norm": 1.6698986291885376,
      "learning_rate": 1.7918351266576112e-05,
      "loss": 0.3278,
      "num_input_tokens_seen": 965096,
      "step": 3390
    },
    {
      "epoch": 6.333955223880597,
      "grad_norm": 1.4233766794204712,
      "learning_rate": 1.784031791188492e-05,
      "loss": 0.5437,
      "num_input_tokens_seen": 966600,
      "step": 3395
    },
    {
      "epoch": 6.343283582089552,
      "grad_norm": 2.5529987812042236,
      "learning_rate": 1.776236047060558e-05,
      "loss": 0.4456,
      "num_input_tokens_seen": 967976,
      "step": 3400
    },
    {
      "epoch": 6.352611940298507,
      "grad_norm": 1.8620471954345703,
      "learning_rate": 1.768447976931326e-05,
      "loss": 0.584,
      "num_input_tokens_seen": 969256,
      "step": 3405
    },
    {
      "epoch": 6.361940298507463,
      "grad_norm": 1.4414796829223633,
      "learning_rate": 1.760667663376943e-05,
      "loss": 0.6452,
      "num_input_tokens_seen": 970696,
      "step": 3410
    },
    {
      "epoch": 6.371268656716418,
      "grad_norm": 2.0666308403015137,
      "learning_rate": 1.7528951888913177e-05,
      "loss": 0.5942,
      "num_input_tokens_seen": 972072,
      "step": 3415
    },
    {
      "epoch": 6.380597014925373,
      "grad_norm": 2.1074230670928955,
      "learning_rate": 1.7451306358852386e-05,
      "loss": 0.4943,
      "num_input_tokens_seen": 973736,
      "step": 3420
    },
    {
      "epoch": 6.389925373134329,
      "grad_norm": 2.282416820526123,
      "learning_rate": 1.737374086685504e-05,
      "loss": 0.5938,
      "num_input_tokens_seen": 975144,
      "step": 3425
    },
    {
      "epoch": 6.399253731343284,
      "grad_norm": 2.493809461593628,
      "learning_rate": 1.729625623534053e-05,
      "loss": 0.6691,
      "num_input_tokens_seen": 976712,
      "step": 3430
    },
    {
      "epoch": 6.4085820895522385,
      "grad_norm": 1.4737935066223145,
      "learning_rate": 1.721885328587083e-05,
      "loss": 0.6786,
      "num_input_tokens_seen": 978088,
      "step": 3435
    },
    {
      "epoch": 6.417910447761194,
      "grad_norm": 1.562461495399475,
      "learning_rate": 1.714153283914189e-05,
      "loss": 0.4772,
      "num_input_tokens_seen": 979560,
      "step": 3440
    },
    {
      "epoch": 6.427238805970149,
      "grad_norm": 1.566023588180542,
      "learning_rate": 1.706429571497486e-05,
      "loss": 0.5876,
      "num_input_tokens_seen": 981000,
      "step": 3445
    },
    {
      "epoch": 6.436567164179104,
      "grad_norm": 1.5297110080718994,
      "learning_rate": 1.698714273230746e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 982568,
      "step": 3450
    },
    {
      "epoch": 6.44589552238806,
      "grad_norm": 1.953328013420105,
      "learning_rate": 1.6910074709185235e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 984040,
      "step": 3455
    },
    {
      "epoch": 6.455223880597015,
      "grad_norm": 2.411365270614624,
      "learning_rate": 1.6833092462752942e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 985384,
      "step": 3460
    },
    {
      "epoch": 6.46455223880597,
      "grad_norm": 2.568527936935425,
      "learning_rate": 1.6756196809245838e-05,
      "loss": 0.3819,
      "num_input_tokens_seen": 986696,
      "step": 3465
    },
    {
      "epoch": 6.473880597014926,
      "grad_norm": 1.503892421722412,
      "learning_rate": 1.667938856398105e-05,
      "loss": 0.3315,
      "num_input_tokens_seen": 988136,
      "step": 3470
    },
    {
      "epoch": 6.483208955223881,
      "grad_norm": 1.0345922708511353,
      "learning_rate": 1.6602668541348916e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 989608,
      "step": 3475
    },
    {
      "epoch": 6.492537313432836,
      "grad_norm": 4.134031295776367,
      "learning_rate": 1.6526037554804367e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 990984,
      "step": 3480
    },
    {
      "epoch": 6.5,
      "eval_loss": 0.618507981300354,
      "eval_runtime": 4.5965,
      "eval_samples_per_second": 51.779,
      "eval_steps_per_second": 13.053,
      "num_input_tokens_seen": 992104,
      "step": 3484
    },
    {
      "epoch": 6.5018656716417915,
      "grad_norm": 1.612874984741211,
      "learning_rate": 1.6449496416858284e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 992360,
      "step": 3485
    },
    {
      "epoch": 6.5111940298507465,
      "grad_norm": 2.3995234966278076,
      "learning_rate": 1.637304593906891e-05,
      "loss": 0.679,
      "num_input_tokens_seen": 993864,
      "step": 3490
    },
    {
      "epoch": 6.520522388059701,
      "grad_norm": 2.02327036857605,
      "learning_rate": 1.6296686932033203e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 995176,
      "step": 3495
    },
    {
      "epoch": 6.529850746268656,
      "grad_norm": 2.9396958351135254,
      "learning_rate": 1.6220420205378282e-05,
      "loss": 0.5061,
      "num_input_tokens_seen": 996680,
      "step": 3500
    },
    {
      "epoch": 6.539179104477612,
      "grad_norm": 1.8367266654968262,
      "learning_rate": 1.614424656775283e-05,
      "loss": 0.4287,
      "num_input_tokens_seen": 998024,
      "step": 3505
    },
    {
      "epoch": 6.548507462686567,
      "grad_norm": 2.9913573265075684,
      "learning_rate": 1.6068166826818492e-05,
      "loss": 0.5131,
      "num_input_tokens_seen": 999432,
      "step": 3510
    },
    {
      "epoch": 6.557835820895522,
      "grad_norm": 2.3710031509399414,
      "learning_rate": 1.5992181789241353e-05,
      "loss": 0.5706,
      "num_input_tokens_seen": 1000712,
      "step": 3515
    },
    {
      "epoch": 6.567164179104478,
      "grad_norm": 2.325901985168457,
      "learning_rate": 1.591629226068338e-05,
      "loss": 0.5728,
      "num_input_tokens_seen": 1002056,
      "step": 3520
    },
    {
      "epoch": 6.576492537313433,
      "grad_norm": 2.011854648590088,
      "learning_rate": 1.5840499045793843e-05,
      "loss": 0.5342,
      "num_input_tokens_seen": 1003464,
      "step": 3525
    },
    {
      "epoch": 6.585820895522388,
      "grad_norm": 1.2709379196166992,
      "learning_rate": 1.5764802948200825e-05,
      "loss": 0.2772,
      "num_input_tokens_seen": 1004808,
      "step": 3530
    },
    {
      "epoch": 6.595149253731344,
      "grad_norm": 1.4927457571029663,
      "learning_rate": 1.5689204770502666e-05,
      "loss": 0.3178,
      "num_input_tokens_seen": 1006120,
      "step": 3535
    },
    {
      "epoch": 6.604477611940299,
      "grad_norm": 1.9912543296813965,
      "learning_rate": 1.5613705314259492e-05,
      "loss": 0.7316,
      "num_input_tokens_seen": 1007624,
      "step": 3540
    },
    {
      "epoch": 6.6138059701492535,
      "grad_norm": 1.9771320819854736,
      "learning_rate": 1.553830537998466e-05,
      "loss": 0.661,
      "num_input_tokens_seen": 1009000,
      "step": 3545
    },
    {
      "epoch": 6.6231343283582085,
      "grad_norm": 2.6609065532684326,
      "learning_rate": 1.5463005767136356e-05,
      "loss": 0.5081,
      "num_input_tokens_seen": 1010376,
      "step": 3550
    },
    {
      "epoch": 6.632462686567164,
      "grad_norm": 2.0714924335479736,
      "learning_rate": 1.5387807274109038e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 1011816,
      "step": 3555
    },
    {
      "epoch": 6.641791044776119,
      "grad_norm": 1.76613187789917,
      "learning_rate": 1.531271069822499e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 1013416,
      "step": 3560
    },
    {
      "epoch": 6.651119402985074,
      "grad_norm": 1.8337589502334595,
      "learning_rate": 1.5237716835725906e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 1014728,
      "step": 3565
    },
    {
      "epoch": 6.66044776119403,
      "grad_norm": 2.056192636489868,
      "learning_rate": 1.5162826481764398e-05,
      "loss": 0.5888,
      "num_input_tokens_seen": 1016136,
      "step": 3570
    },
    {
      "epoch": 6.669776119402985,
      "grad_norm": 2.0060765743255615,
      "learning_rate": 1.5088040430395589e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 1017416,
      "step": 3575
    },
    {
      "epoch": 6.67910447761194,
      "grad_norm": 2.455496072769165,
      "learning_rate": 1.5013359474568712e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 1019048,
      "step": 3580
    },
    {
      "epoch": 6.688432835820896,
      "grad_norm": 2.891009569168091,
      "learning_rate": 1.493878440611866e-05,
      "loss": 0.5087,
      "num_input_tokens_seen": 1020552,
      "step": 3585
    },
    {
      "epoch": 6.697761194029851,
      "grad_norm": 2.3145925998687744,
      "learning_rate": 1.4864316015757623e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 1021896,
      "step": 3590
    },
    {
      "epoch": 6.707089552238806,
      "grad_norm": 1.4066121578216553,
      "learning_rate": 1.478995509306669e-05,
      "loss": 0.4593,
      "num_input_tokens_seen": 1023240,
      "step": 3595
    },
    {
      "epoch": 6.7164179104477615,
      "grad_norm": 2.0292818546295166,
      "learning_rate": 1.4715702426487482e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 1024712,
      "step": 3600
    },
    {
      "epoch": 6.725746268656716,
      "grad_norm": 1.4507927894592285,
      "learning_rate": 1.4641558803313781e-05,
      "loss": 0.3098,
      "num_input_tokens_seen": 1026216,
      "step": 3605
    },
    {
      "epoch": 6.735074626865671,
      "grad_norm": 1.6267207860946655,
      "learning_rate": 1.4567525009683219e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 1027592,
      "step": 3610
    },
    {
      "epoch": 6.744402985074627,
      "grad_norm": 1.7667746543884277,
      "learning_rate": 1.4493601830568887e-05,
      "loss": 0.3181,
      "num_input_tokens_seen": 1029064,
      "step": 3615
    },
    {
      "epoch": 6.753731343283582,
      "grad_norm": 1.2118422985076904,
      "learning_rate": 1.4419790049771068e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 1030632,
      "step": 3620
    },
    {
      "epoch": 6.763059701492537,
      "grad_norm": 2.6723885536193848,
      "learning_rate": 1.434609044990886e-05,
      "loss": 0.6535,
      "num_input_tokens_seen": 1032072,
      "step": 3625
    },
    {
      "epoch": 6.772388059701493,
      "grad_norm": 1.7491955757141113,
      "learning_rate": 1.4272503812411974e-05,
      "loss": 0.5742,
      "num_input_tokens_seen": 1033480,
      "step": 3630
    },
    {
      "epoch": 6.781716417910448,
      "grad_norm": 3.165227174758911,
      "learning_rate": 1.4199030917512329e-05,
      "loss": 0.4776,
      "num_input_tokens_seen": 1034792,
      "step": 3635
    },
    {
      "epoch": 6.791044776119403,
      "grad_norm": 1.9558213949203491,
      "learning_rate": 1.412567254423589e-05,
      "loss": 0.6419,
      "num_input_tokens_seen": 1036200,
      "step": 3640
    },
    {
      "epoch": 6.800373134328359,
      "grad_norm": 1.3405424356460571,
      "learning_rate": 1.4052429470394352e-05,
      "loss": 0.4179,
      "num_input_tokens_seen": 1037608,
      "step": 3645
    },
    {
      "epoch": 6.809701492537314,
      "grad_norm": 2.206340789794922,
      "learning_rate": 1.3979302472576868e-05,
      "loss": 0.7917,
      "num_input_tokens_seen": 1038952,
      "step": 3650
    },
    {
      "epoch": 6.8190298507462686,
      "grad_norm": 1.9380404949188232,
      "learning_rate": 1.3906292326141904e-05,
      "loss": 0.6035,
      "num_input_tokens_seen": 1040584,
      "step": 3655
    },
    {
      "epoch": 6.8283582089552235,
      "grad_norm": 2.089273452758789,
      "learning_rate": 1.3833399805208904e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 1041960,
      "step": 3660
    },
    {
      "epoch": 6.837686567164179,
      "grad_norm": 2.4019381999969482,
      "learning_rate": 1.376062568265018e-05,
      "loss": 0.309,
      "num_input_tokens_seen": 1043336,
      "step": 3665
    },
    {
      "epoch": 6.847014925373134,
      "grad_norm": 1.7107216119766235,
      "learning_rate": 1.3687970730082678e-05,
      "loss": 0.557,
      "num_input_tokens_seen": 1044712,
      "step": 3670
    },
    {
      "epoch": 6.856343283582089,
      "grad_norm": 2.393645763397217,
      "learning_rate": 1.361543571785976e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 1046120,
      "step": 3675
    },
    {
      "epoch": 6.865671641791045,
      "grad_norm": 2.115851879119873,
      "learning_rate": 1.3543021415063117e-05,
      "loss": 0.5313,
      "num_input_tokens_seen": 1047592,
      "step": 3680
    },
    {
      "epoch": 6.875,
      "grad_norm": 1.8831361532211304,
      "learning_rate": 1.347072858949453e-05,
      "loss": 0.6291,
      "num_input_tokens_seen": 1048936,
      "step": 3685
    },
    {
      "epoch": 6.884328358208955,
      "grad_norm": 3.5699737071990967,
      "learning_rate": 1.3398558007667806e-05,
      "loss": 0.664,
      "num_input_tokens_seen": 1050376,
      "step": 3690
    },
    {
      "epoch": 6.893656716417911,
      "grad_norm": 2.01493239402771,
      "learning_rate": 1.3326510434800566e-05,
      "loss": 0.6376,
      "num_input_tokens_seen": 1051816,
      "step": 3695
    },
    {
      "epoch": 6.902985074626866,
      "grad_norm": 1.5419564247131348,
      "learning_rate": 1.3254586634806249e-05,
      "loss": 0.6527,
      "num_input_tokens_seen": 1053128,
      "step": 3700
    },
    {
      "epoch": 6.912313432835821,
      "grad_norm": 1.8879196643829346,
      "learning_rate": 1.3182787370285865e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 1054440,
      "step": 3705
    },
    {
      "epoch": 6.9216417910447765,
      "grad_norm": 2.5752100944519043,
      "learning_rate": 1.3111113402520053e-05,
      "loss": 0.6328,
      "num_input_tokens_seen": 1055816,
      "step": 3710
    },
    {
      "epoch": 6.9309701492537314,
      "grad_norm": 1.0279862880706787,
      "learning_rate": 1.3039565491460882e-05,
      "loss": 0.361,
      "num_input_tokens_seen": 1057288,
      "step": 3715
    },
    {
      "epoch": 6.940298507462686,
      "grad_norm": 1.3198506832122803,
      "learning_rate": 1.2968144395723914e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 1058888,
      "step": 3720
    },
    {
      "epoch": 6.949626865671641,
      "grad_norm": 1.5974754095077515,
      "learning_rate": 1.289685087258004e-05,
      "loss": 0.5993,
      "num_input_tokens_seen": 1060296,
      "step": 3725
    },
    {
      "epoch": 6.958955223880597,
      "grad_norm": 1.6485826969146729,
      "learning_rate": 1.2825685677947563e-05,
      "loss": 0.3289,
      "num_input_tokens_seen": 1061640,
      "step": 3730
    },
    {
      "epoch": 6.968283582089552,
      "grad_norm": 1.8757879734039307,
      "learning_rate": 1.2754649566384114e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 1062888,
      "step": 3735
    },
    {
      "epoch": 6.977611940298507,
      "grad_norm": 1.1854349374771118,
      "learning_rate": 1.2683743291078649e-05,
      "loss": 0.4168,
      "num_input_tokens_seen": 1064328,
      "step": 3740
    },
    {
      "epoch": 6.986940298507463,
      "grad_norm": 3.1413753032684326,
      "learning_rate": 1.2612967603843512e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 1065928,
      "step": 3745
    },
    {
      "epoch": 6.996268656716418,
      "grad_norm": 1.9983711242675781,
      "learning_rate": 1.2542323255106408e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 1067176,
      "step": 3750
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.6158590912818909,
      "eval_runtime": 4.5706,
      "eval_samples_per_second": 52.072,
      "eval_steps_per_second": 13.127,
      "num_input_tokens_seen": 1067520,
      "step": 3752
    },
    {
      "epoch": 7.005597014925373,
      "grad_norm": 2.0818207263946533,
      "learning_rate": 1.24718109939025e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 1068192,
      "step": 3755
    },
    {
      "epoch": 7.014925373134329,
      "grad_norm": 1.568673014640808,
      "learning_rate": 1.2401431567866425e-05,
      "loss": 0.6157,
      "num_input_tokens_seen": 1069600,
      "step": 3760
    },
    {
      "epoch": 7.024253731343284,
      "grad_norm": 3.2669010162353516,
      "learning_rate": 1.233118572322437e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 1070880,
      "step": 3765
    },
    {
      "epoch": 7.0335820895522385,
      "grad_norm": 1.107926845550537,
      "learning_rate": 1.22610742047862e-05,
      "loss": 0.343,
      "num_input_tokens_seen": 1072416,
      "step": 3770
    },
    {
      "epoch": 7.042910447761194,
      "grad_norm": 1.9720746278762817,
      "learning_rate": 1.21910977559375e-05,
      "loss": 0.6301,
      "num_input_tokens_seen": 1073824,
      "step": 3775
    },
    {
      "epoch": 7.052238805970149,
      "grad_norm": 2.9087188243865967,
      "learning_rate": 1.212125711863176e-05,
      "loss": 0.3029,
      "num_input_tokens_seen": 1074976,
      "step": 3780
    },
    {
      "epoch": 7.061567164179104,
      "grad_norm": 1.509771704673767,
      "learning_rate": 1.2051553033382425e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 1076416,
      "step": 3785
    },
    {
      "epoch": 7.07089552238806,
      "grad_norm": 1.4959207773208618,
      "learning_rate": 1.1981986239255164e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 1077920,
      "step": 3790
    },
    {
      "epoch": 7.080223880597015,
      "grad_norm": 1.7031633853912354,
      "learning_rate": 1.1912557473859895e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 1079776,
      "step": 3795
    },
    {
      "epoch": 7.08955223880597,
      "grad_norm": 1.9472756385803223,
      "learning_rate": 1.1843267473343048e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 1081088,
      "step": 3800
    },
    {
      "epoch": 7.098880597014926,
      "grad_norm": 2.3733177185058594,
      "learning_rate": 1.177411697237977e-05,
      "loss": 0.4094,
      "num_input_tokens_seen": 1082624,
      "step": 3805
    },
    {
      "epoch": 7.108208955223881,
      "grad_norm": 1.843748927116394,
      "learning_rate": 1.1705106704166069e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 1084000,
      "step": 3810
    },
    {
      "epoch": 7.117537313432836,
      "grad_norm": 2.0206353664398193,
      "learning_rate": 1.1636237400411107e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 1085504,
      "step": 3815
    },
    {
      "epoch": 7.126865671641791,
      "grad_norm": 1.6327645778656006,
      "learning_rate": 1.1567509791329401e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 1086848,
      "step": 3820
    },
    {
      "epoch": 7.1361940298507465,
      "grad_norm": 2.0612289905548096,
      "learning_rate": 1.1498924605633111e-05,
      "loss": 0.6318,
      "num_input_tokens_seen": 1088160,
      "step": 3825
    },
    {
      "epoch": 7.145522388059701,
      "grad_norm": 1.6673834323883057,
      "learning_rate": 1.1430482570524253e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 1089568,
      "step": 3830
    },
    {
      "epoch": 7.154850746268656,
      "grad_norm": 2.0685958862304688,
      "learning_rate": 1.136218441168709e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 1090752,
      "step": 3835
    },
    {
      "epoch": 7.164179104477612,
      "grad_norm": 3.803478240966797,
      "learning_rate": 1.1294030853280321e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 1092128,
      "step": 3840
    },
    {
      "epoch": 7.173507462686567,
      "grad_norm": 2.5569252967834473,
      "learning_rate": 1.12260226179295e-05,
      "loss": 0.5357,
      "num_input_tokens_seen": 1093376,
      "step": 3845
    },
    {
      "epoch": 7.182835820895522,
      "grad_norm": 1.3899413347244263,
      "learning_rate": 1.115816042671933e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 1094848,
      "step": 3850
    },
    {
      "epoch": 7.192164179104478,
      "grad_norm": 2.2758994102478027,
      "learning_rate": 1.1090444999186e-05,
      "loss": 0.52,
      "num_input_tokens_seen": 1096352,
      "step": 3855
    },
    {
      "epoch": 7.201492537313433,
      "grad_norm": 2.409092903137207,
      "learning_rate": 1.102287705330961e-05,
      "loss": 0.5499,
      "num_input_tokens_seen": 1097824,
      "step": 3860
    },
    {
      "epoch": 7.210820895522388,
      "grad_norm": 2.772160291671753,
      "learning_rate": 1.095545730550649e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 1099136,
      "step": 3865
    },
    {
      "epoch": 7.220149253731344,
      "grad_norm": 2.6501264572143555,
      "learning_rate": 1.0888186470621675e-05,
      "loss": 0.3578,
      "num_input_tokens_seen": 1100512,
      "step": 3870
    },
    {
      "epoch": 7.229477611940299,
      "grad_norm": 2.0545010566711426,
      "learning_rate": 1.0821065261921262e-05,
      "loss": 0.5803,
      "num_input_tokens_seen": 1101856,
      "step": 3875
    },
    {
      "epoch": 7.2388059701492535,
      "grad_norm": 1.6310173273086548,
      "learning_rate": 1.0754094391084896e-05,
      "loss": 0.4801,
      "num_input_tokens_seen": 1103264,
      "step": 3880
    },
    {
      "epoch": 7.248134328358209,
      "grad_norm": 3.859694242477417,
      "learning_rate": 1.0687274568198208e-05,
      "loss": 0.6809,
      "num_input_tokens_seen": 1104736,
      "step": 3885
    },
    {
      "epoch": 7.257462686567164,
      "grad_norm": 1.5826741456985474,
      "learning_rate": 1.0620606501745251e-05,
      "loss": 0.4409,
      "num_input_tokens_seen": 1106336,
      "step": 3890
    },
    {
      "epoch": 7.266791044776119,
      "grad_norm": 2.3241894245147705,
      "learning_rate": 1.0554090898601062e-05,
      "loss": 0.6239,
      "num_input_tokens_seen": 1107808,
      "step": 3895
    },
    {
      "epoch": 7.276119402985074,
      "grad_norm": 2.595885992050171,
      "learning_rate": 1.0487728464024086e-05,
      "loss": 0.5718,
      "num_input_tokens_seen": 1109184,
      "step": 3900
    },
    {
      "epoch": 7.28544776119403,
      "grad_norm": 1.1373722553253174,
      "learning_rate": 1.0421519901648758e-05,
      "loss": 0.4891,
      "num_input_tokens_seen": 1110720,
      "step": 3905
    },
    {
      "epoch": 7.294776119402985,
      "grad_norm": 2.01339054107666,
      "learning_rate": 1.0355465913478019e-05,
      "loss": 0.525,
      "num_input_tokens_seen": 1112096,
      "step": 3910
    },
    {
      "epoch": 7.30410447761194,
      "grad_norm": 2.7450828552246094,
      "learning_rate": 1.0289567199875878e-05,
      "loss": 0.6742,
      "num_input_tokens_seen": 1113408,
      "step": 3915
    },
    {
      "epoch": 7.313432835820896,
      "grad_norm": 1.9322885274887085,
      "learning_rate": 1.0223824459559971e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 1114784,
      "step": 3920
    },
    {
      "epoch": 7.322761194029851,
      "grad_norm": 1.186610460281372,
      "learning_rate": 1.0158238389594164e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 1116416,
      "step": 3925
    },
    {
      "epoch": 7.332089552238806,
      "grad_norm": 2.4434452056884766,
      "learning_rate": 1.009280968538118e-05,
      "loss": 0.4438,
      "num_input_tokens_seen": 1117824,
      "step": 3930
    },
    {
      "epoch": 7.3414179104477615,
      "grad_norm": 2.344008207321167,
      "learning_rate": 1.0027539040655173e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 1119136,
      "step": 3935
    },
    {
      "epoch": 7.350746268656716,
      "grad_norm": 1.8992395401000977,
      "learning_rate": 9.962427147474474e-06,
      "loss": 0.4128,
      "num_input_tokens_seen": 1120480,
      "step": 3940
    },
    {
      "epoch": 7.360074626865671,
      "grad_norm": 1.5992515087127686,
      "learning_rate": 9.89747469621411e-06,
      "loss": 0.3397,
      "num_input_tokens_seen": 1121856,
      "step": 3945
    },
    {
      "epoch": 7.369402985074627,
      "grad_norm": 2.3671560287475586,
      "learning_rate": 9.832682375558614e-06,
      "loss": 0.4918,
      "num_input_tokens_seen": 1123136,
      "step": 3950
    },
    {
      "epoch": 7.378731343283582,
      "grad_norm": 2.5741052627563477,
      "learning_rate": 9.768050872494638e-06,
      "loss": 0.5549,
      "num_input_tokens_seen": 1124480,
      "step": 3955
    },
    {
      "epoch": 7.388059701492537,
      "grad_norm": 2.1324446201324463,
      "learning_rate": 9.70358087230373e-06,
      "loss": 0.6926,
      "num_input_tokens_seen": 1125856,
      "step": 3960
    },
    {
      "epoch": 7.397388059701493,
      "grad_norm": 1.248674750328064,
      "learning_rate": 9.639273058555004e-06,
      "loss": 0.4454,
      "num_input_tokens_seen": 1127264,
      "step": 3965
    },
    {
      "epoch": 7.406716417910448,
      "grad_norm": 1.7027866840362549,
      "learning_rate": 9.575128113097952e-06,
      "loss": 0.5325,
      "num_input_tokens_seen": 1128640,
      "step": 3970
    },
    {
      "epoch": 7.416044776119403,
      "grad_norm": 3.6465561389923096,
      "learning_rate": 9.5111467160552e-06,
      "loss": 0.7194,
      "num_input_tokens_seen": 1129984,
      "step": 3975
    },
    {
      "epoch": 7.425373134328359,
      "grad_norm": 2.1068220138549805,
      "learning_rate": 9.447329545815246e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 1131552,
      "step": 3980
    },
    {
      "epoch": 7.434701492537314,
      "grad_norm": 2.559316635131836,
      "learning_rate": 9.383677279025347e-06,
      "loss": 0.5531,
      "num_input_tokens_seen": 1132928,
      "step": 3985
    },
    {
      "epoch": 7.4440298507462686,
      "grad_norm": 1.828503966331482,
      "learning_rate": 9.320190590584273e-06,
      "loss": 0.3553,
      "num_input_tokens_seen": 1134336,
      "step": 3990
    },
    {
      "epoch": 7.4533582089552235,
      "grad_norm": 1.927769660949707,
      "learning_rate": 9.256870153635208e-06,
      "loss": 0.3777,
      "num_input_tokens_seen": 1135840,
      "step": 3995
    },
    {
      "epoch": 7.462686567164179,
      "grad_norm": 3.1055734157562256,
      "learning_rate": 9.193716639558587e-06,
      "loss": 0.4107,
      "num_input_tokens_seen": 1137216,
      "step": 4000
    },
    {
      "epoch": 7.472014925373134,
      "grad_norm": 2.3692989349365234,
      "learning_rate": 9.130730717964948e-06,
      "loss": 0.5373,
      "num_input_tokens_seen": 1138528,
      "step": 4005
    },
    {
      "epoch": 7.481343283582089,
      "grad_norm": 2.7855849266052246,
      "learning_rate": 9.067913056687899e-06,
      "loss": 0.5778,
      "num_input_tokens_seen": 1139872,
      "step": 4010
    },
    {
      "epoch": 7.490671641791045,
      "grad_norm": 2.222172737121582,
      "learning_rate": 9.005264321776974e-06,
      "loss": 0.3903,
      "num_input_tokens_seen": 1141408,
      "step": 4015
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.6290847063064575,
      "learning_rate": 8.942785177490617e-06,
      "loss": 0.5722,
      "num_input_tokens_seen": 1142912,
      "step": 4020
    },
    {
      "epoch": 7.5,
      "eval_loss": 0.6261455416679382,
      "eval_runtime": 4.5999,
      "eval_samples_per_second": 51.741,
      "eval_steps_per_second": 13.044,
      "num_input_tokens_seen": 1142912,
      "step": 4020
    },
    {
      "epoch": 7.509328358208955,
      "grad_norm": 1.0721356868743896,
      "learning_rate": 8.880476286289091e-06,
      "loss": 0.296,
      "num_input_tokens_seen": 1144256,
      "step": 4025
    },
    {
      "epoch": 7.518656716417911,
      "grad_norm": 2.5110669136047363,
      "learning_rate": 8.818338308827529e-06,
      "loss": 0.602,
      "num_input_tokens_seen": 1145568,
      "step": 4030
    },
    {
      "epoch": 7.527985074626866,
      "grad_norm": 2.085482358932495,
      "learning_rate": 8.75637190394884e-06,
      "loss": 0.4672,
      "num_input_tokens_seen": 1147008,
      "step": 4035
    },
    {
      "epoch": 7.537313432835821,
      "grad_norm": 2.5544304847717285,
      "learning_rate": 8.69457772867677e-06,
      "loss": 0.3043,
      "num_input_tokens_seen": 1148416,
      "step": 4040
    },
    {
      "epoch": 7.5466417910447765,
      "grad_norm": 2.796832799911499,
      "learning_rate": 8.632956438208961e-06,
      "loss": 0.2587,
      "num_input_tokens_seen": 1149888,
      "step": 4045
    },
    {
      "epoch": 7.5559701492537314,
      "grad_norm": 3.5357625484466553,
      "learning_rate": 8.571508685909931e-06,
      "loss": 0.3835,
      "num_input_tokens_seen": 1151360,
      "step": 4050
    },
    {
      "epoch": 7.565298507462686,
      "grad_norm": 2.2749745845794678,
      "learning_rate": 8.510235123304227e-06,
      "loss": 0.5345,
      "num_input_tokens_seen": 1152800,
      "step": 4055
    },
    {
      "epoch": 7.574626865671641,
      "grad_norm": 1.88466215133667,
      "learning_rate": 8.449136400069464e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 1154272,
      "step": 4060
    },
    {
      "epoch": 7.583955223880597,
      "grad_norm": 1.90096116065979,
      "learning_rate": 8.38821316402946e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 1155584,
      "step": 4065
    },
    {
      "epoch": 7.593283582089552,
      "grad_norm": 2.2478063106536865,
      "learning_rate": 8.327466061147337e-06,
      "loss": 0.3993,
      "num_input_tokens_seen": 1157152,
      "step": 4070
    },
    {
      "epoch": 7.602611940298507,
      "grad_norm": 3.0016376972198486,
      "learning_rate": 8.266895735518729e-06,
      "loss": 0.6667,
      "num_input_tokens_seen": 1158432,
      "step": 4075
    },
    {
      "epoch": 7.611940298507463,
      "grad_norm": 1.9572343826293945,
      "learning_rate": 8.20650282936488e-06,
      "loss": 0.3697,
      "num_input_tokens_seen": 1160032,
      "step": 4080
    },
    {
      "epoch": 7.621268656716418,
      "grad_norm": 3.4903042316436768,
      "learning_rate": 8.146287983025903e-06,
      "loss": 0.5299,
      "num_input_tokens_seen": 1161536,
      "step": 4085
    },
    {
      "epoch": 7.630597014925373,
      "grad_norm": 2.2559237480163574,
      "learning_rate": 8.086251834953953e-06,
      "loss": 0.5299,
      "num_input_tokens_seen": 1163072,
      "step": 4090
    },
    {
      "epoch": 7.639925373134329,
      "grad_norm": 0.959821879863739,
      "learning_rate": 8.02639502170645e-06,
      "loss": 0.3647,
      "num_input_tokens_seen": 1164384,
      "step": 4095
    },
    {
      "epoch": 7.649253731343284,
      "grad_norm": 4.66791296005249,
      "learning_rate": 7.966718177939366e-06,
      "loss": 0.5227,
      "num_input_tokens_seen": 1165728,
      "step": 4100
    },
    {
      "epoch": 7.6585820895522385,
      "grad_norm": 1.5075697898864746,
      "learning_rate": 7.907221936400453e-06,
      "loss": 0.4548,
      "num_input_tokens_seen": 1167264,
      "step": 4105
    },
    {
      "epoch": 7.667910447761194,
      "grad_norm": 1.6090877056121826,
      "learning_rate": 7.847906927922574e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 1168800,
      "step": 4110
    },
    {
      "epoch": 7.677238805970149,
      "grad_norm": 1.5544253587722778,
      "learning_rate": 7.788773781416974e-06,
      "loss": 0.3454,
      "num_input_tokens_seen": 1170240,
      "step": 4115
    },
    {
      "epoch": 7.686567164179104,
      "grad_norm": 1.9958131313323975,
      "learning_rate": 7.729823123866656e-06,
      "loss": 0.3855,
      "num_input_tokens_seen": 1171936,
      "step": 4120
    },
    {
      "epoch": 7.69589552238806,
      "grad_norm": 2.5896942615509033,
      "learning_rate": 7.671055580319706e-06,
      "loss": 0.4985,
      "num_input_tokens_seen": 1173216,
      "step": 4125
    },
    {
      "epoch": 7.705223880597015,
      "grad_norm": 2.0339889526367188,
      "learning_rate": 7.612471773882651e-06,
      "loss": 0.5138,
      "num_input_tokens_seen": 1174496,
      "step": 4130
    },
    {
      "epoch": 7.71455223880597,
      "grad_norm": 2.518007755279541,
      "learning_rate": 7.554072325713896e-06,
      "loss": 0.4515,
      "num_input_tokens_seen": 1176032,
      "step": 4135
    },
    {
      "epoch": 7.723880597014926,
      "grad_norm": 1.5726410150527954,
      "learning_rate": 7.495857855017091e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 1177472,
      "step": 4140
    },
    {
      "epoch": 7.733208955223881,
      "grad_norm": 2.1392624378204346,
      "learning_rate": 7.437828979034606e-06,
      "loss": 0.5393,
      "num_input_tokens_seen": 1178976,
      "step": 4145
    },
    {
      "epoch": 7.742537313432836,
      "grad_norm": 2.03029727935791,
      "learning_rate": 7.379986313040959e-06,
      "loss": 0.5328,
      "num_input_tokens_seen": 1180448,
      "step": 4150
    },
    {
      "epoch": 7.7518656716417915,
      "grad_norm": 3.305720806121826,
      "learning_rate": 7.3223304703363135e-06,
      "loss": 0.501,
      "num_input_tokens_seen": 1181888,
      "step": 4155
    },
    {
      "epoch": 7.7611940298507465,
      "grad_norm": 2.7143356800079346,
      "learning_rate": 7.264862062239947e-06,
      "loss": 0.3812,
      "num_input_tokens_seen": 1183456,
      "step": 4160
    },
    {
      "epoch": 7.770522388059701,
      "grad_norm": 2.367859125137329,
      "learning_rate": 7.2075816980837814e-06,
      "loss": 0.6018,
      "num_input_tokens_seen": 1184800,
      "step": 4165
    },
    {
      "epoch": 7.779850746268656,
      "grad_norm": 2.3008596897125244,
      "learning_rate": 7.150489985205952e-06,
      "loss": 0.5585,
      "num_input_tokens_seen": 1186304,
      "step": 4170
    },
    {
      "epoch": 7.789179104477612,
      "grad_norm": 2.7711384296417236,
      "learning_rate": 7.093587528944326e-06,
      "loss": 0.4951,
      "num_input_tokens_seen": 1187744,
      "step": 4175
    },
    {
      "epoch": 7.798507462686567,
      "grad_norm": 2.4321558475494385,
      "learning_rate": 7.036874932630117e-06,
      "loss": 0.4268,
      "num_input_tokens_seen": 1189376,
      "step": 4180
    },
    {
      "epoch": 7.807835820895522,
      "grad_norm": 2.5192575454711914,
      "learning_rate": 6.980352797581438e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 1190912,
      "step": 4185
    },
    {
      "epoch": 7.817164179104478,
      "grad_norm": 2.684894561767578,
      "learning_rate": 6.924021723097004e-06,
      "loss": 0.5439,
      "num_input_tokens_seen": 1192416,
      "step": 4190
    },
    {
      "epoch": 7.826492537313433,
      "grad_norm": 2.2324812412261963,
      "learning_rate": 6.867882306449694e-06,
      "loss": 0.5855,
      "num_input_tokens_seen": 1193792,
      "step": 4195
    },
    {
      "epoch": 7.835820895522388,
      "grad_norm": 1.5660499334335327,
      "learning_rate": 6.8119351428802796e-06,
      "loss": 0.2858,
      "num_input_tokens_seen": 1195200,
      "step": 4200
    },
    {
      "epoch": 7.845149253731344,
      "grad_norm": 2.165149211883545,
      "learning_rate": 6.756180825591099e-06,
      "loss": 0.4692,
      "num_input_tokens_seen": 1196672,
      "step": 4205
    },
    {
      "epoch": 7.854477611940299,
      "grad_norm": 1.8318179845809937,
      "learning_rate": 6.700619945739728e-06,
      "loss": 0.6301,
      "num_input_tokens_seen": 1198336,
      "step": 4210
    },
    {
      "epoch": 7.8638059701492535,
      "grad_norm": 2.358863115310669,
      "learning_rate": 6.645253092432785e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 1199712,
      "step": 4215
    },
    {
      "epoch": 7.8731343283582085,
      "grad_norm": 1.0567878484725952,
      "learning_rate": 6.590080852719602e-06,
      "loss": 0.2913,
      "num_input_tokens_seen": 1201088,
      "step": 4220
    },
    {
      "epoch": 7.882462686567164,
      "grad_norm": 2.786796808242798,
      "learning_rate": 6.535103811586085e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 1202432,
      "step": 4225
    },
    {
      "epoch": 7.891791044776119,
      "grad_norm": 2.3607053756713867,
      "learning_rate": 6.480322551948426e-06,
      "loss": 0.5051,
      "num_input_tokens_seen": 1203872,
      "step": 4230
    },
    {
      "epoch": 7.901119402985074,
      "grad_norm": 1.5261811017990112,
      "learning_rate": 6.425737654646993e-06,
      "loss": 0.4074,
      "num_input_tokens_seen": 1205312,
      "step": 4235
    },
    {
      "epoch": 7.91044776119403,
      "grad_norm": 2.907742738723755,
      "learning_rate": 6.371349698440149e-06,
      "loss": 0.5316,
      "num_input_tokens_seen": 1206688,
      "step": 4240
    },
    {
      "epoch": 7.919776119402985,
      "grad_norm": 1.6161439418792725,
      "learning_rate": 6.317159259998073e-06,
      "loss": 0.5485,
      "num_input_tokens_seen": 1208064,
      "step": 4245
    },
    {
      "epoch": 7.92910447761194,
      "grad_norm": 2.0374107360839844,
      "learning_rate": 6.2631669138967285e-06,
      "loss": 0.3606,
      "num_input_tokens_seen": 1209600,
      "step": 4250
    },
    {
      "epoch": 7.938432835820896,
      "grad_norm": 2.0685784816741943,
      "learning_rate": 6.209373232611682e-06,
      "loss": 0.3313,
      "num_input_tokens_seen": 1210976,
      "step": 4255
    },
    {
      "epoch": 7.947761194029851,
      "grad_norm": 3.264538049697876,
      "learning_rate": 6.155778786512109e-06,
      "loss": 0.5725,
      "num_input_tokens_seen": 1212416,
      "step": 4260
    },
    {
      "epoch": 7.957089552238806,
      "grad_norm": 3.492642402648926,
      "learning_rate": 6.102384143854698e-06,
      "loss": 0.4159,
      "num_input_tokens_seen": 1213920,
      "step": 4265
    },
    {
      "epoch": 7.9664179104477615,
      "grad_norm": 2.456319808959961,
      "learning_rate": 6.049189870777652e-06,
      "loss": 0.4899,
      "num_input_tokens_seen": 1215360,
      "step": 4270
    },
    {
      "epoch": 7.975746268656716,
      "grad_norm": 2.4835474491119385,
      "learning_rate": 5.996196531294657e-06,
      "loss": 0.3994,
      "num_input_tokens_seen": 1216736,
      "step": 4275
    },
    {
      "epoch": 7.985074626865671,
      "grad_norm": 1.0539278984069824,
      "learning_rate": 5.943404687288939e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 1218240,
      "step": 4280
    },
    {
      "epoch": 7.994402985074627,
      "grad_norm": 2.199493885040283,
      "learning_rate": 5.890814898507277e-06,
      "loss": 0.4968,
      "num_input_tokens_seen": 1219488,
      "step": 4285
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.619342565536499,
      "eval_runtime": 4.5732,
      "eval_samples_per_second": 52.043,
      "eval_steps_per_second": 13.12,
      "num_input_tokens_seen": 1220200,
      "step": 4288
    },
    {
      "epoch": 8.003731343283581,
      "grad_norm": 1.655181646347046,
      "learning_rate": 5.838427722554057e-06,
      "loss": 0.4535,
      "num_input_tokens_seen": 1220808,
      "step": 4290
    },
    {
      "epoch": 8.013059701492537,
      "grad_norm": 2.2056469917297363,
      "learning_rate": 5.786243714885442e-06,
      "loss": 0.4311,
      "num_input_tokens_seen": 1222280,
      "step": 4295
    },
    {
      "epoch": 8.022388059701493,
      "grad_norm": 2.0312516689300537,
      "learning_rate": 5.734263428803352e-06,
      "loss": 0.3933,
      "num_input_tokens_seen": 1223688,
      "step": 4300
    },
    {
      "epoch": 8.031716417910447,
      "grad_norm": 3.28053879737854,
      "learning_rate": 5.6824874154497194e-06,
      "loss": 0.6709,
      "num_input_tokens_seen": 1225352,
      "step": 4305
    },
    {
      "epoch": 8.041044776119403,
      "grad_norm": 2.0707650184631348,
      "learning_rate": 5.63091622380055e-06,
      "loss": 0.5449,
      "num_input_tokens_seen": 1226824,
      "step": 4310
    },
    {
      "epoch": 8.050373134328359,
      "grad_norm": 2.0669567584991455,
      "learning_rate": 5.5795504006601855e-06,
      "loss": 0.3813,
      "num_input_tokens_seen": 1228232,
      "step": 4315
    },
    {
      "epoch": 8.059701492537313,
      "grad_norm": 2.769364595413208,
      "learning_rate": 5.528390490655428e-06,
      "loss": 0.5027,
      "num_input_tokens_seen": 1229672,
      "step": 4320
    },
    {
      "epoch": 8.069029850746269,
      "grad_norm": 2.5455830097198486,
      "learning_rate": 5.477437036229833e-06,
      "loss": 0.5468,
      "num_input_tokens_seen": 1230920,
      "step": 4325
    },
    {
      "epoch": 8.078358208955224,
      "grad_norm": 2.167426824569702,
      "learning_rate": 5.426690577637913e-06,
      "loss": 0.3645,
      "num_input_tokens_seen": 1232232,
      "step": 4330
    },
    {
      "epoch": 8.087686567164178,
      "grad_norm": 1.3536076545715332,
      "learning_rate": 5.37615165293942e-06,
      "loss": 0.45,
      "num_input_tokens_seen": 1233640,
      "step": 4335
    },
    {
      "epoch": 8.097014925373134,
      "grad_norm": 2.5364761352539062,
      "learning_rate": 5.325820797993661e-06,
      "loss": 0.4156,
      "num_input_tokens_seen": 1235048,
      "step": 4340
    },
    {
      "epoch": 8.10634328358209,
      "grad_norm": 2.0572941303253174,
      "learning_rate": 5.275698546453775e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 1236552,
      "step": 4345
    },
    {
      "epoch": 8.115671641791044,
      "grad_norm": 1.9713358879089355,
      "learning_rate": 5.225785429761124e-06,
      "loss": 0.5024,
      "num_input_tokens_seen": 1237896,
      "step": 4350
    },
    {
      "epoch": 8.125,
      "grad_norm": 2.744194746017456,
      "learning_rate": 5.176081977139621e-06,
      "loss": 0.5425,
      "num_input_tokens_seen": 1239112,
      "step": 4355
    },
    {
      "epoch": 8.134328358208956,
      "grad_norm": 2.1876399517059326,
      "learning_rate": 5.12658871559013e-06,
      "loss": 0.4017,
      "num_input_tokens_seen": 1240744,
      "step": 4360
    },
    {
      "epoch": 8.14365671641791,
      "grad_norm": 3.0552144050598145,
      "learning_rate": 5.077306169884888e-06,
      "loss": 0.5013,
      "num_input_tokens_seen": 1242152,
      "step": 4365
    },
    {
      "epoch": 8.152985074626866,
      "grad_norm": 2.8527369499206543,
      "learning_rate": 5.0282348625619175e-06,
      "loss": 0.469,
      "num_input_tokens_seen": 1243496,
      "step": 4370
    },
    {
      "epoch": 8.162313432835822,
      "grad_norm": 3.371128797531128,
      "learning_rate": 4.979375313919526e-06,
      "loss": 0.3981,
      "num_input_tokens_seen": 1245064,
      "step": 4375
    },
    {
      "epoch": 8.171641791044776,
      "grad_norm": 2.2645492553710938,
      "learning_rate": 4.930728042010724e-06,
      "loss": 0.3872,
      "num_input_tokens_seen": 1246408,
      "step": 4380
    },
    {
      "epoch": 8.180970149253731,
      "grad_norm": 3.0741164684295654,
      "learning_rate": 4.882293562637827e-06,
      "loss": 0.5113,
      "num_input_tokens_seen": 1247784,
      "step": 4385
    },
    {
      "epoch": 8.190298507462687,
      "grad_norm": 1.9678038358688354,
      "learning_rate": 4.834072389346883e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 1249192,
      "step": 4390
    },
    {
      "epoch": 8.199626865671641,
      "grad_norm": 2.471418857574463,
      "learning_rate": 4.78606503342231e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 1250664,
      "step": 4395
    },
    {
      "epoch": 8.208955223880597,
      "grad_norm": 2.0927133560180664,
      "learning_rate": 4.738272003881417e-06,
      "loss": 0.3944,
      "num_input_tokens_seen": 1252072,
      "step": 4400
    },
    {
      "epoch": 8.218283582089553,
      "grad_norm": 3.9715166091918945,
      "learning_rate": 4.690693807469035e-06,
      "loss": 0.5444,
      "num_input_tokens_seen": 1253416,
      "step": 4405
    },
    {
      "epoch": 8.227611940298507,
      "grad_norm": 1.7606277465820312,
      "learning_rate": 4.643330948652155e-06,
      "loss": 0.6846,
      "num_input_tokens_seen": 1254728,
      "step": 4410
    },
    {
      "epoch": 8.236940298507463,
      "grad_norm": 2.299985885620117,
      "learning_rate": 4.596183929614559e-06,
      "loss": 0.3973,
      "num_input_tokens_seen": 1255976,
      "step": 4415
    },
    {
      "epoch": 8.246268656716419,
      "grad_norm": 1.714682698249817,
      "learning_rate": 4.549253250251498e-06,
      "loss": 0.3542,
      "num_input_tokens_seen": 1257288,
      "step": 4420
    },
    {
      "epoch": 8.255597014925373,
      "grad_norm": 2.002528190612793,
      "learning_rate": 4.502539408164386e-06,
      "loss": 0.333,
      "num_input_tokens_seen": 1258664,
      "step": 4425
    },
    {
      "epoch": 8.264925373134329,
      "grad_norm": 1.7194960117340088,
      "learning_rate": 4.4560428986555516e-06,
      "loss": 0.3111,
      "num_input_tokens_seen": 1260104,
      "step": 4430
    },
    {
      "epoch": 8.274253731343283,
      "grad_norm": 2.598156213760376,
      "learning_rate": 4.409764214722945e-06,
      "loss": 0.4924,
      "num_input_tokens_seen": 1261448,
      "step": 4435
    },
    {
      "epoch": 8.283582089552239,
      "grad_norm": 1.8464361429214478,
      "learning_rate": 4.363703847054948e-06,
      "loss": 0.4135,
      "num_input_tokens_seen": 1263048,
      "step": 4440
    },
    {
      "epoch": 8.292910447761194,
      "grad_norm": 2.283355951309204,
      "learning_rate": 4.317862284025165e-06,
      "loss": 0.4873,
      "num_input_tokens_seen": 1264360,
      "step": 4445
    },
    {
      "epoch": 8.302238805970148,
      "grad_norm": 2.9945428371429443,
      "learning_rate": 4.272240011687206e-06,
      "loss": 0.3323,
      "num_input_tokens_seen": 1265640,
      "step": 4450
    },
    {
      "epoch": 8.311567164179104,
      "grad_norm": 1.82478928565979,
      "learning_rate": 4.226837513769596e-06,
      "loss": 0.3319,
      "num_input_tokens_seen": 1267016,
      "step": 4455
    },
    {
      "epoch": 8.32089552238806,
      "grad_norm": 2.510647773742676,
      "learning_rate": 4.181655271670587e-06,
      "loss": 0.5204,
      "num_input_tokens_seen": 1268744,
      "step": 4460
    },
    {
      "epoch": 8.330223880597014,
      "grad_norm": 1.6120847463607788,
      "learning_rate": 4.1366937644531e-06,
      "loss": 0.3081,
      "num_input_tokens_seen": 1270120,
      "step": 4465
    },
    {
      "epoch": 8.33955223880597,
      "grad_norm": 3.1077470779418945,
      "learning_rate": 4.091953468839607e-06,
      "loss": 0.6197,
      "num_input_tokens_seen": 1271656,
      "step": 4470
    },
    {
      "epoch": 8.348880597014926,
      "grad_norm": 1.3565236330032349,
      "learning_rate": 4.047434859207114e-06,
      "loss": 0.5291,
      "num_input_tokens_seen": 1273128,
      "step": 4475
    },
    {
      "epoch": 8.35820895522388,
      "grad_norm": 2.6521074771881104,
      "learning_rate": 4.003138407582102e-06,
      "loss": 0.4993,
      "num_input_tokens_seen": 1274632,
      "step": 4480
    },
    {
      "epoch": 8.367537313432836,
      "grad_norm": 3.736938714981079,
      "learning_rate": 3.959064583635527e-06,
      "loss": 0.5247,
      "num_input_tokens_seen": 1275848,
      "step": 4485
    },
    {
      "epoch": 8.376865671641792,
      "grad_norm": 2.025364875793457,
      "learning_rate": 3.9152138546778625e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 1277224,
      "step": 4490
    },
    {
      "epoch": 8.386194029850746,
      "grad_norm": 3.4433794021606445,
      "learning_rate": 3.871586685654102e-06,
      "loss": 0.6758,
      "num_input_tokens_seen": 1278504,
      "step": 4495
    },
    {
      "epoch": 8.395522388059701,
      "grad_norm": 2.872206449508667,
      "learning_rate": 3.82818353913888e-06,
      "loss": 0.5574,
      "num_input_tokens_seen": 1279784,
      "step": 4500
    },
    {
      "epoch": 8.404850746268657,
      "grad_norm": 2.34690523147583,
      "learning_rate": 3.785004875331527e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 1281256,
      "step": 4505
    },
    {
      "epoch": 8.414179104477611,
      "grad_norm": 1.1326650381088257,
      "learning_rate": 3.742051152051221e-06,
      "loss": 0.522,
      "num_input_tokens_seen": 1282792,
      "step": 4510
    },
    {
      "epoch": 8.423507462686567,
      "grad_norm": 2.4032540321350098,
      "learning_rate": 3.6993228247320877e-06,
      "loss": 0.3458,
      "num_input_tokens_seen": 1284232,
      "step": 4515
    },
    {
      "epoch": 8.432835820895523,
      "grad_norm": 2.8545684814453125,
      "learning_rate": 3.6568203464184383e-06,
      "loss": 0.35,
      "num_input_tokens_seen": 1285448,
      "step": 4520
    },
    {
      "epoch": 8.442164179104477,
      "grad_norm": 3.48283314704895,
      "learning_rate": 3.614544167759901e-06,
      "loss": 0.7026,
      "num_input_tokens_seen": 1286824,
      "step": 4525
    },
    {
      "epoch": 8.451492537313433,
      "grad_norm": 3.3355185985565186,
      "learning_rate": 3.5724947370066708e-06,
      "loss": 0.5298,
      "num_input_tokens_seen": 1288136,
      "step": 4530
    },
    {
      "epoch": 8.460820895522389,
      "grad_norm": 0.8718127608299255,
      "learning_rate": 3.5306725000047918e-06,
      "loss": 0.3352,
      "num_input_tokens_seen": 1289512,
      "step": 4535
    },
    {
      "epoch": 8.470149253731343,
      "grad_norm": 1.9637565612792969,
      "learning_rate": 3.4890779001913452e-06,
      "loss": 0.2454,
      "num_input_tokens_seen": 1291144,
      "step": 4540
    },
    {
      "epoch": 8.479477611940299,
      "grad_norm": 2.4974539279937744,
      "learning_rate": 3.4477113785898407e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 1292680,
      "step": 4545
    },
    {
      "epoch": 8.488805970149254,
      "grad_norm": 1.2249170541763306,
      "learning_rate": 3.4065733738054606e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 1294056,
      "step": 4550
    },
    {
      "epoch": 8.498134328358208,
      "grad_norm": 2.1300222873687744,
      "learning_rate": 3.3656643220204785e-06,
      "loss": 0.6181,
      "num_input_tokens_seen": 1295400,
      "step": 4555
    },
    {
      "epoch": 8.5,
      "eval_loss": 0.6357511878013611,
      "eval_runtime": 4.5832,
      "eval_samples_per_second": 51.928,
      "eval_steps_per_second": 13.091,
      "num_input_tokens_seen": 1295720,
      "step": 4556
    },
    {
      "epoch": 8.507462686567164,
      "grad_norm": 3.0027620792388916,
      "learning_rate": 3.3249846569895744e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 1296872,
      "step": 4560
    },
    {
      "epoch": 8.51679104477612,
      "grad_norm": 2.129547595977783,
      "learning_rate": 3.284534810035278e-06,
      "loss": 0.436,
      "num_input_tokens_seen": 1298408,
      "step": 4565
    },
    {
      "epoch": 8.526119402985074,
      "grad_norm": 1.4341539144515991,
      "learning_rate": 3.2443152100433832e-06,
      "loss": 0.4667,
      "num_input_tokens_seen": 1299848,
      "step": 4570
    },
    {
      "epoch": 8.53544776119403,
      "grad_norm": 2.716299057006836,
      "learning_rate": 3.204326283458381e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 1301224,
      "step": 4575
    },
    {
      "epoch": 8.544776119402986,
      "grad_norm": 2.760127305984497,
      "learning_rate": 3.1645684542789765e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 1302856,
      "step": 4580
    },
    {
      "epoch": 8.55410447761194,
      "grad_norm": 1.7544218301773071,
      "learning_rate": 3.12504214405355e-06,
      "loss": 0.4942,
      "num_input_tokens_seen": 1304168,
      "step": 4585
    },
    {
      "epoch": 8.563432835820896,
      "grad_norm": 3.9774012565612793,
      "learning_rate": 3.0857477718757187e-06,
      "loss": 0.4554,
      "num_input_tokens_seen": 1305704,
      "step": 4590
    },
    {
      "epoch": 8.572761194029852,
      "grad_norm": 2.934168577194214,
      "learning_rate": 3.0466857543798966e-06,
      "loss": 0.4791,
      "num_input_tokens_seen": 1307080,
      "step": 4595
    },
    {
      "epoch": 8.582089552238806,
      "grad_norm": 4.7950825691223145,
      "learning_rate": 3.007856505736836e-06,
      "loss": 0.4386,
      "num_input_tokens_seen": 1308296,
      "step": 4600
    },
    {
      "epoch": 8.591417910447761,
      "grad_norm": 2.212034225463867,
      "learning_rate": 2.969260437649293e-06,
      "loss": 0.2698,
      "num_input_tokens_seen": 1309832,
      "step": 4605
    },
    {
      "epoch": 8.600746268656717,
      "grad_norm": 1.4387774467468262,
      "learning_rate": 2.9308979593476004e-06,
      "loss": 0.4645,
      "num_input_tokens_seen": 1311368,
      "step": 4610
    },
    {
      "epoch": 8.610074626865671,
      "grad_norm": 1.2265300750732422,
      "learning_rate": 2.892769477585397e-06,
      "loss": 0.3086,
      "num_input_tokens_seen": 1312680,
      "step": 4615
    },
    {
      "epoch": 8.619402985074627,
      "grad_norm": 1.7413617372512817,
      "learning_rate": 2.8548753966352426e-06,
      "loss": 0.3109,
      "num_input_tokens_seen": 1314152,
      "step": 4620
    },
    {
      "epoch": 8.628731343283581,
      "grad_norm": 0.7740257978439331,
      "learning_rate": 2.8172161182844075e-06,
      "loss": 0.4478,
      "num_input_tokens_seen": 1315496,
      "step": 4625
    },
    {
      "epoch": 8.638059701492537,
      "grad_norm": 1.5663965940475464,
      "learning_rate": 2.779792041830537e-06,
      "loss": 0.4937,
      "num_input_tokens_seen": 1317192,
      "step": 4630
    },
    {
      "epoch": 8.647388059701493,
      "grad_norm": 2.562016487121582,
      "learning_rate": 2.742603564077478e-06,
      "loss": 0.3145,
      "num_input_tokens_seen": 1318632,
      "step": 4635
    },
    {
      "epoch": 8.656716417910447,
      "grad_norm": 2.4282116889953613,
      "learning_rate": 2.7056510793310314e-06,
      "loss": 0.4102,
      "num_input_tokens_seen": 1320232,
      "step": 4640
    },
    {
      "epoch": 8.666044776119403,
      "grad_norm": 2.2467234134674072,
      "learning_rate": 2.6689349793947994e-06,
      "loss": 0.271,
      "num_input_tokens_seen": 1321768,
      "step": 4645
    },
    {
      "epoch": 8.675373134328359,
      "grad_norm": 2.7613186836242676,
      "learning_rate": 2.6324556535660045e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 1323272,
      "step": 4650
    },
    {
      "epoch": 8.684701492537313,
      "grad_norm": 3.0170986652374268,
      "learning_rate": 2.5962134886313936e-06,
      "loss": 0.4227,
      "num_input_tokens_seen": 1324680,
      "step": 4655
    },
    {
      "epoch": 8.694029850746269,
      "grad_norm": 2.4644272327423096,
      "learning_rate": 2.5602088688631148e-06,
      "loss": 0.5604,
      "num_input_tokens_seen": 1325992,
      "step": 4660
    },
    {
      "epoch": 8.703358208955224,
      "grad_norm": 1.1744383573532104,
      "learning_rate": 2.5244421760146355e-06,
      "loss": 0.5224,
      "num_input_tokens_seen": 1327400,
      "step": 4665
    },
    {
      "epoch": 8.712686567164178,
      "grad_norm": 2.171969413757324,
      "learning_rate": 2.488913789316724e-06,
      "loss": 0.6707,
      "num_input_tokens_seen": 1328744,
      "step": 4670
    },
    {
      "epoch": 8.722014925373134,
      "grad_norm": 2.1137423515319824,
      "learning_rate": 2.4536240854733967e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 1330280,
      "step": 4675
    },
    {
      "epoch": 8.73134328358209,
      "grad_norm": 2.141444444656372,
      "learning_rate": 2.4185734386579506e-06,
      "loss": 0.446,
      "num_input_tokens_seen": 1331784,
      "step": 4680
    },
    {
      "epoch": 8.740671641791044,
      "grad_norm": 2.5716073513031006,
      "learning_rate": 2.383762220508984e-06,
      "loss": 0.4745,
      "num_input_tokens_seen": 1333128,
      "step": 4685
    },
    {
      "epoch": 8.75,
      "grad_norm": 2.491004705429077,
      "learning_rate": 2.349190800126444e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 1334696,
      "step": 4690
    },
    {
      "epoch": 8.759328358208956,
      "grad_norm": 2.808584451675415,
      "learning_rate": 2.3148595440677405e-06,
      "loss": 0.4539,
      "num_input_tokens_seen": 1336264,
      "step": 4695
    },
    {
      "epoch": 8.76865671641791,
      "grad_norm": 2.693746566772461,
      "learning_rate": 2.280768816343834e-06,
      "loss": 0.4404,
      "num_input_tokens_seen": 1337608,
      "step": 4700
    },
    {
      "epoch": 8.777985074626866,
      "grad_norm": 1.5714143514633179,
      "learning_rate": 2.246918978415394e-06,
      "loss": 0.5246,
      "num_input_tokens_seen": 1339144,
      "step": 4705
    },
    {
      "epoch": 8.787313432835822,
      "grad_norm": 1.6074087619781494,
      "learning_rate": 2.2133103891889438e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 1340488,
      "step": 4710
    },
    {
      "epoch": 8.796641791044776,
      "grad_norm": 2.204249382019043,
      "learning_rate": 2.1799434050131018e-06,
      "loss": 0.5227,
      "num_input_tokens_seen": 1341960,
      "step": 4715
    },
    {
      "epoch": 8.805970149253731,
      "grad_norm": 1.2797794342041016,
      "learning_rate": 2.1468183796747365e-06,
      "loss": 0.4973,
      "num_input_tokens_seen": 1343624,
      "step": 4720
    },
    {
      "epoch": 8.815298507462687,
      "grad_norm": 1.6312012672424316,
      "learning_rate": 2.1139356643952666e-06,
      "loss": 0.3485,
      "num_input_tokens_seen": 1344936,
      "step": 4725
    },
    {
      "epoch": 8.824626865671641,
      "grad_norm": 2.1299076080322266,
      "learning_rate": 2.0812956078269275e-06,
      "loss": 0.5275,
      "num_input_tokens_seen": 1346408,
      "step": 4730
    },
    {
      "epoch": 8.833955223880597,
      "grad_norm": 2.5503058433532715,
      "learning_rate": 2.0488985560490477e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 1347784,
      "step": 4735
    },
    {
      "epoch": 8.843283582089553,
      "grad_norm": 1.6357694864273071,
      "learning_rate": 2.01674485256442e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 1349128,
      "step": 4740
    },
    {
      "epoch": 8.852611940298507,
      "grad_norm": 3.4971728324890137,
      "learning_rate": 1.9848348382956293e-06,
      "loss": 0.5343,
      "num_input_tokens_seen": 1350312,
      "step": 4745
    },
    {
      "epoch": 8.861940298507463,
      "grad_norm": 2.7725703716278076,
      "learning_rate": 1.953168851581452e-06,
      "loss": 0.3899,
      "num_input_tokens_seen": 1351784,
      "step": 4750
    },
    {
      "epoch": 8.871268656716419,
      "grad_norm": 2.3643572330474854,
      "learning_rate": 1.921747228173254e-06,
      "loss": 0.4928,
      "num_input_tokens_seen": 1353416,
      "step": 4755
    },
    {
      "epoch": 8.880597014925373,
      "grad_norm": 2.367875337600708,
      "learning_rate": 1.8905703012314563e-06,
      "loss": 0.6247,
      "num_input_tokens_seen": 1354760,
      "step": 4760
    },
    {
      "epoch": 8.889925373134329,
      "grad_norm": 5.695141315460205,
      "learning_rate": 1.8596384013219725e-06,
      "loss": 0.5189,
      "num_input_tokens_seen": 1355976,
      "step": 4765
    },
    {
      "epoch": 8.899253731343283,
      "grad_norm": 2.473560094833374,
      "learning_rate": 1.8289518564127223e-06,
      "loss": 0.373,
      "num_input_tokens_seen": 1357416,
      "step": 4770
    },
    {
      "epoch": 8.908582089552239,
      "grad_norm": 2.315114736557007,
      "learning_rate": 1.7985109918701643e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 1359080,
      "step": 4775
    },
    {
      "epoch": 8.917910447761194,
      "grad_norm": 1.7605212926864624,
      "learning_rate": 1.7683161304558076e-06,
      "loss": 0.4283,
      "num_input_tokens_seen": 1360552,
      "step": 4780
    },
    {
      "epoch": 8.927238805970148,
      "grad_norm": 2.084561586380005,
      "learning_rate": 1.738367592322837e-06,
      "loss": 0.3866,
      "num_input_tokens_seen": 1361960,
      "step": 4785
    },
    {
      "epoch": 8.936567164179104,
      "grad_norm": 1.5463670492172241,
      "learning_rate": 1.708665695012676e-06,
      "loss": 0.6771,
      "num_input_tokens_seen": 1363368,
      "step": 4790
    },
    {
      "epoch": 8.94589552238806,
      "grad_norm": 3.5588579177856445,
      "learning_rate": 1.6792107534516571e-06,
      "loss": 0.5205,
      "num_input_tokens_seen": 1364840,
      "step": 4795
    },
    {
      "epoch": 8.955223880597014,
      "grad_norm": 2.2765579223632812,
      "learning_rate": 1.6500030799476513e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 1366152,
      "step": 4800
    },
    {
      "epoch": 8.96455223880597,
      "grad_norm": 1.4313689470291138,
      "learning_rate": 1.621042984186777e-06,
      "loss": 0.4464,
      "num_input_tokens_seen": 1367528,
      "step": 4805
    },
    {
      "epoch": 8.973880597014926,
      "grad_norm": 1.6111520528793335,
      "learning_rate": 1.5923307732301136e-06,
      "loss": 0.3448,
      "num_input_tokens_seen": 1368968,
      "step": 4810
    },
    {
      "epoch": 8.98320895522388,
      "grad_norm": 3.212109088897705,
      "learning_rate": 1.5638667515104288e-06,
      "loss": 0.527,
      "num_input_tokens_seen": 1370248,
      "step": 4815
    },
    {
      "epoch": 8.992537313432836,
      "grad_norm": 2.8777408599853516,
      "learning_rate": 1.5356512208289846e-06,
      "loss": 0.6561,
      "num_input_tokens_seen": 1371496,
      "step": 4820
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.6277225017547607,
      "eval_runtime": 4.5768,
      "eval_samples_per_second": 52.001,
      "eval_steps_per_second": 13.109,
      "num_input_tokens_seen": 1372560,
      "step": 4824
    },
    {
      "epoch": 9.001865671641792,
      "grad_norm": 3.2055232524871826,
      "learning_rate": 1.5076844803522922e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 1372784,
      "step": 4825
    },
    {
      "epoch": 9.011194029850746,
      "grad_norm": 2.6281862258911133,
      "learning_rate": 1.4799668266089834e-06,
      "loss": 0.4729,
      "num_input_tokens_seen": 1374160,
      "step": 4830
    },
    {
      "epoch": 9.020522388059701,
      "grad_norm": 2.359022378921509,
      "learning_rate": 1.452498553486642e-06,
      "loss": 0.3241,
      "num_input_tokens_seen": 1375472,
      "step": 4835
    },
    {
      "epoch": 9.029850746268657,
      "grad_norm": 1.9456086158752441,
      "learning_rate": 1.4252799522286892e-06,
      "loss": 0.5157,
      "num_input_tokens_seen": 1376976,
      "step": 4840
    },
    {
      "epoch": 9.039179104477611,
      "grad_norm": 4.123142242431641,
      "learning_rate": 1.398311311431308e-06,
      "loss": 0.4337,
      "num_input_tokens_seen": 1378256,
      "step": 4845
    },
    {
      "epoch": 9.048507462686567,
      "grad_norm": 2.243795394897461,
      "learning_rate": 1.3715929170403623e-06,
      "loss": 0.4303,
      "num_input_tokens_seen": 1379856,
      "step": 4850
    },
    {
      "epoch": 9.057835820895523,
      "grad_norm": 1.5125812292099,
      "learning_rate": 1.3451250523483976e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 1381584,
      "step": 4855
    },
    {
      "epoch": 9.067164179104477,
      "grad_norm": 1.867136836051941,
      "learning_rate": 1.3189079979915864e-06,
      "loss": 0.4755,
      "num_input_tokens_seen": 1382896,
      "step": 4860
    },
    {
      "epoch": 9.076492537313433,
      "grad_norm": 2.8726887702941895,
      "learning_rate": 1.2929420319468255e-06,
      "loss": 0.4512,
      "num_input_tokens_seen": 1384240,
      "step": 4865
    },
    {
      "epoch": 9.085820895522389,
      "grad_norm": 2.245140790939331,
      "learning_rate": 1.2672274295287057e-06,
      "loss": 0.386,
      "num_input_tokens_seen": 1385808,
      "step": 4870
    },
    {
      "epoch": 9.095149253731343,
      "grad_norm": 2.384204626083374,
      "learning_rate": 1.2417644633866632e-06,
      "loss": 0.1665,
      "num_input_tokens_seen": 1387312,
      "step": 4875
    },
    {
      "epoch": 9.104477611940299,
      "grad_norm": 1.9132916927337646,
      "learning_rate": 1.2165534035020409e-06,
      "loss": 0.5733,
      "num_input_tokens_seen": 1388688,
      "step": 4880
    },
    {
      "epoch": 9.113805970149254,
      "grad_norm": 1.9066709280014038,
      "learning_rate": 1.1915945171852572e-06,
      "loss": 0.546,
      "num_input_tokens_seen": 1390000,
      "step": 4885
    },
    {
      "epoch": 9.123134328358208,
      "grad_norm": 2.021293878555298,
      "learning_rate": 1.1668880690729467e-06,
      "loss": 0.5948,
      "num_input_tokens_seen": 1391408,
      "step": 4890
    },
    {
      "epoch": 9.132462686567164,
      "grad_norm": 2.532153606414795,
      "learning_rate": 1.142434321125177e-06,
      "loss": 0.5248,
      "num_input_tokens_seen": 1392848,
      "step": 4895
    },
    {
      "epoch": 9.14179104477612,
      "grad_norm": 1.7421011924743652,
      "learning_rate": 1.1182335326226533e-06,
      "loss": 0.4546,
      "num_input_tokens_seen": 1394320,
      "step": 4900
    },
    {
      "epoch": 9.151119402985074,
      "grad_norm": 3.4395647048950195,
      "learning_rate": 1.0942859601639794e-06,
      "loss": 0.5438,
      "num_input_tokens_seen": 1395504,
      "step": 4905
    },
    {
      "epoch": 9.16044776119403,
      "grad_norm": 1.494896650314331,
      "learning_rate": 1.0705918576629364e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 1396880,
      "step": 4910
    },
    {
      "epoch": 9.169776119402986,
      "grad_norm": 1.6920926570892334,
      "learning_rate": 1.0471514763457814e-06,
      "loss": 0.5815,
      "num_input_tokens_seen": 1398064,
      "step": 4915
    },
    {
      "epoch": 9.17910447761194,
      "grad_norm": 2.563417673110962,
      "learning_rate": 1.023965064748597e-06,
      "loss": 0.5522,
      "num_input_tokens_seen": 1399504,
      "step": 4920
    },
    {
      "epoch": 9.188432835820896,
      "grad_norm": 2.9966654777526855,
      "learning_rate": 1.0010328687146464e-06,
      "loss": 0.468,
      "num_input_tokens_seen": 1400848,
      "step": 4925
    },
    {
      "epoch": 9.197761194029852,
      "grad_norm": 2.2512927055358887,
      "learning_rate": 9.783551313917699e-07,
      "loss": 0.3237,
      "num_input_tokens_seen": 1402128,
      "step": 4930
    },
    {
      "epoch": 9.207089552238806,
      "grad_norm": 2.1150429248809814,
      "learning_rate": 9.559320932298111e-07,
      "loss": 0.4635,
      "num_input_tokens_seen": 1403632,
      "step": 4935
    },
    {
      "epoch": 9.216417910447761,
      "grad_norm": 2.3147387504577637,
      "learning_rate": 9.337639919780539e-07,
      "loss": 0.4044,
      "num_input_tokens_seen": 1405072,
      "step": 4940
    },
    {
      "epoch": 9.225746268656717,
      "grad_norm": 2.0629079341888428,
      "learning_rate": 9.118510626827198e-07,
      "loss": 0.513,
      "num_input_tokens_seen": 1406480,
      "step": 4945
    },
    {
      "epoch": 9.235074626865671,
      "grad_norm": 3.321600914001465,
      "learning_rate": 8.901935376844611e-07,
      "loss": 0.4136,
      "num_input_tokens_seen": 1407792,
      "step": 4950
    },
    {
      "epoch": 9.244402985074627,
      "grad_norm": 1.9444159269332886,
      "learning_rate": 8.687916466159157e-07,
      "loss": 0.5911,
      "num_input_tokens_seen": 1409072,
      "step": 4955
    },
    {
      "epoch": 9.253731343283581,
      "grad_norm": 1.3653244972229004,
      "learning_rate": 8.476456163992397e-07,
      "loss": 0.3762,
      "num_input_tokens_seen": 1410576,
      "step": 4960
    },
    {
      "epoch": 9.263059701492537,
      "grad_norm": 2.5764822959899902,
      "learning_rate": 8.267556712437341e-07,
      "loss": 0.3984,
      "num_input_tokens_seen": 1411856,
      "step": 4965
    },
    {
      "epoch": 9.272388059701493,
      "grad_norm": 2.708169460296631,
      "learning_rate": 8.061220326434582e-07,
      "loss": 0.4947,
      "num_input_tokens_seen": 1413200,
      "step": 4970
    },
    {
      "epoch": 9.281716417910447,
      "grad_norm": 3.179684638977051,
      "learning_rate": 7.857449193748645e-07,
      "loss": 0.3574,
      "num_input_tokens_seen": 1414576,
      "step": 4975
    },
    {
      "epoch": 9.291044776119403,
      "grad_norm": 3.3870224952697754,
      "learning_rate": 7.656245474945034e-07,
      "loss": 0.4467,
      "num_input_tokens_seen": 1415952,
      "step": 4980
    },
    {
      "epoch": 9.300373134328359,
      "grad_norm": 2.9979724884033203,
      "learning_rate": 7.457611303367196e-07,
      "loss": 0.4911,
      "num_input_tokens_seen": 1417136,
      "step": 4985
    },
    {
      "epoch": 9.309701492537313,
      "grad_norm": 1.9281370639801025,
      "learning_rate": 7.261548785113925e-07,
      "loss": 0.29,
      "num_input_tokens_seen": 1418736,
      "step": 4990
    },
    {
      "epoch": 9.319029850746269,
      "grad_norm": 2.808270215988159,
      "learning_rate": 7.068059999016969e-07,
      "loss": 0.3442,
      "num_input_tokens_seen": 1420144,
      "step": 4995
    },
    {
      "epoch": 9.328358208955224,
      "grad_norm": 2.7481400966644287,
      "learning_rate": 6.877146996619122e-07,
      "loss": 0.4768,
      "num_input_tokens_seen": 1421552,
      "step": 5000
    },
    {
      "epoch": 9.337686567164178,
      "grad_norm": 2.0273537635803223,
      "learning_rate": 6.688811802152279e-07,
      "loss": 0.3508,
      "num_input_tokens_seen": 1423024,
      "step": 5005
    },
    {
      "epoch": 9.347014925373134,
      "grad_norm": 3.794126510620117,
      "learning_rate": 6.503056412516223e-07,
      "loss": 0.4901,
      "num_input_tokens_seen": 1424208,
      "step": 5010
    },
    {
      "epoch": 9.35634328358209,
      "grad_norm": 3.785670757293701,
      "learning_rate": 6.31988279725726e-07,
      "loss": 0.3562,
      "num_input_tokens_seen": 1425648,
      "step": 5015
    },
    {
      "epoch": 9.365671641791044,
      "grad_norm": 2.2090444564819336,
      "learning_rate": 6.139292898547366e-07,
      "loss": 0.3572,
      "num_input_tokens_seen": 1426896,
      "step": 5020
    },
    {
      "epoch": 9.375,
      "grad_norm": 3.258925199508667,
      "learning_rate": 5.961288631163687e-07,
      "loss": 0.4678,
      "num_input_tokens_seen": 1428336,
      "step": 5025
    },
    {
      "epoch": 9.384328358208956,
      "grad_norm": 1.9619524478912354,
      "learning_rate": 5.785871882468069e-07,
      "loss": 0.5158,
      "num_input_tokens_seen": 1429712,
      "step": 5030
    },
    {
      "epoch": 9.39365671641791,
      "grad_norm": 3.102452278137207,
      "learning_rate": 5.613044512387283e-07,
      "loss": 0.4429,
      "num_input_tokens_seen": 1430928,
      "step": 5035
    },
    {
      "epoch": 9.402985074626866,
      "grad_norm": 2.9860949516296387,
      "learning_rate": 5.442808353393059e-07,
      "loss": 0.3878,
      "num_input_tokens_seen": 1432272,
      "step": 5040
    },
    {
      "epoch": 9.412313432835822,
      "grad_norm": 2.3907999992370605,
      "learning_rate": 5.275165210482824e-07,
      "loss": 0.4624,
      "num_input_tokens_seen": 1433616,
      "step": 5045
    },
    {
      "epoch": 9.421641791044776,
      "grad_norm": 2.4058892726898193,
      "learning_rate": 5.110116861160502e-07,
      "loss": 0.4543,
      "num_input_tokens_seen": 1435216,
      "step": 5050
    },
    {
      "epoch": 9.430970149253731,
      "grad_norm": 1.7519402503967285,
      "learning_rate": 4.947665055417605e-07,
      "loss": 0.3967,
      "num_input_tokens_seen": 1436432,
      "step": 5055
    },
    {
      "epoch": 9.440298507462687,
      "grad_norm": 2.9587061405181885,
      "learning_rate": 4.787811515714919e-07,
      "loss": 0.5496,
      "num_input_tokens_seen": 1437808,
      "step": 5060
    },
    {
      "epoch": 9.449626865671641,
      "grad_norm": 2.776102066040039,
      "learning_rate": 4.6305579369638475e-07,
      "loss": 0.4914,
      "num_input_tokens_seen": 1439152,
      "step": 5065
    },
    {
      "epoch": 9.458955223880597,
      "grad_norm": 2.4112508296966553,
      "learning_rate": 4.4759059865088494e-07,
      "loss": 0.4194,
      "num_input_tokens_seen": 1440752,
      "step": 5070
    },
    {
      "epoch": 9.468283582089553,
      "grad_norm": 2.1456336975097656,
      "learning_rate": 4.323857304109419e-07,
      "loss": 0.2387,
      "num_input_tokens_seen": 1442128,
      "step": 5075
    },
    {
      "epoch": 9.477611940298507,
      "grad_norm": 3.1962382793426514,
      "learning_rate": 4.1744135019230193e-07,
      "loss": 0.4713,
      "num_input_tokens_seen": 1443728,
      "step": 5080
    },
    {
      "epoch": 9.486940298507463,
      "grad_norm": 2.96341609954834,
      "learning_rate": 4.0275761644876787e-07,
      "loss": 0.4764,
      "num_input_tokens_seen": 1445136,
      "step": 5085
    },
    {
      "epoch": 9.496268656716419,
      "grad_norm": 1.7467764616012573,
      "learning_rate": 3.883346848705338e-07,
      "loss": 0.4826,
      "num_input_tokens_seen": 1446736,
      "step": 5090
    },
    {
      "epoch": 9.5,
      "eval_loss": 0.6321566104888916,
      "eval_runtime": 4.5829,
      "eval_samples_per_second": 51.932,
      "eval_steps_per_second": 13.092,
      "num_input_tokens_seen": 1447376,
      "step": 5092
    },
    {
      "epoch": 9.505597014925373,
      "grad_norm": 2.2064566612243652,
      "learning_rate": 3.741727083825475e-07,
      "loss": 0.4404,
      "num_input_tokens_seen": 1448112,
      "step": 5095
    },
    {
      "epoch": 9.514925373134329,
      "grad_norm": 4.11940336227417,
      "learning_rate": 3.6027183714285595e-07,
      "loss": 0.3827,
      "num_input_tokens_seen": 1449744,
      "step": 5100
    },
    {
      "epoch": 9.524253731343283,
      "grad_norm": 2.5974295139312744,
      "learning_rate": 3.4663221854105423e-07,
      "loss": 0.8389,
      "num_input_tokens_seen": 1451024,
      "step": 5105
    },
    {
      "epoch": 9.533582089552239,
      "grad_norm": 2.6577701568603516,
      "learning_rate": 3.332539971966836e-07,
      "loss": 0.4868,
      "num_input_tokens_seen": 1452304,
      "step": 5110
    },
    {
      "epoch": 9.542910447761194,
      "grad_norm": 2.184831142425537,
      "learning_rate": 3.201373149577247e-07,
      "loss": 0.4929,
      "num_input_tokens_seen": 1453968,
      "step": 5115
    },
    {
      "epoch": 9.552238805970148,
      "grad_norm": 1.9858027696609497,
      "learning_rate": 3.0728231089907634e-07,
      "loss": 0.4344,
      "num_input_tokens_seen": 1455376,
      "step": 5120
    },
    {
      "epoch": 9.561567164179104,
      "grad_norm": 2.8102385997772217,
      "learning_rate": 2.946891213211012e-07,
      "loss": 0.3555,
      "num_input_tokens_seen": 1456816,
      "step": 5125
    },
    {
      "epoch": 9.57089552238806,
      "grad_norm": 1.5339605808258057,
      "learning_rate": 2.823578797481574e-07,
      "loss": 0.4273,
      "num_input_tokens_seen": 1458320,
      "step": 5130
    },
    {
      "epoch": 9.580223880597014,
      "grad_norm": 3.5844898223876953,
      "learning_rate": 2.7028871692720003e-07,
      "loss": 0.5191,
      "num_input_tokens_seen": 1459632,
      "step": 5135
    },
    {
      "epoch": 9.58955223880597,
      "grad_norm": 2.262385129928589,
      "learning_rate": 2.5848176082639007e-07,
      "loss": 0.4946,
      "num_input_tokens_seen": 1461008,
      "step": 5140
    },
    {
      "epoch": 9.598880597014926,
      "grad_norm": 1.10933518409729,
      "learning_rate": 2.4693713663372644e-07,
      "loss": 0.2777,
      "num_input_tokens_seen": 1462224,
      "step": 5145
    },
    {
      "epoch": 9.60820895522388,
      "grad_norm": 2.5455191135406494,
      "learning_rate": 2.3565496675574118e-07,
      "loss": 0.5296,
      "num_input_tokens_seen": 1463664,
      "step": 5150
    },
    {
      "epoch": 9.617537313432836,
      "grad_norm": 1.3662569522857666,
      "learning_rate": 2.246353708161758e-07,
      "loss": 0.2898,
      "num_input_tokens_seen": 1465168,
      "step": 5155
    },
    {
      "epoch": 9.626865671641792,
      "grad_norm": 3.4847514629364014,
      "learning_rate": 2.1387846565474045e-07,
      "loss": 0.4143,
      "num_input_tokens_seen": 1466544,
      "step": 5160
    },
    {
      "epoch": 9.636194029850746,
      "grad_norm": 4.003303527832031,
      "learning_rate": 2.0338436532584827e-07,
      "loss": 0.6039,
      "num_input_tokens_seen": 1468176,
      "step": 5165
    },
    {
      "epoch": 9.645522388059701,
      "grad_norm": 2.9795074462890625,
      "learning_rate": 1.9315318109742465e-07,
      "loss": 0.7007,
      "num_input_tokens_seen": 1469680,
      "step": 5170
    },
    {
      "epoch": 9.654850746268657,
      "grad_norm": 3.2243094444274902,
      "learning_rate": 1.831850214497194e-07,
      "loss": 0.5124,
      "num_input_tokens_seen": 1471024,
      "step": 5175
    },
    {
      "epoch": 9.664179104477611,
      "grad_norm": 2.32784366607666,
      "learning_rate": 1.7347999207415478e-07,
      "loss": 0.2848,
      "num_input_tokens_seen": 1472464,
      "step": 5180
    },
    {
      "epoch": 9.673507462686567,
      "grad_norm": 2.816188097000122,
      "learning_rate": 1.6403819587221814e-07,
      "loss": 0.4261,
      "num_input_tokens_seen": 1473904,
      "step": 5185
    },
    {
      "epoch": 9.682835820895523,
      "grad_norm": 1.9632453918457031,
      "learning_rate": 1.5485973295434885e-07,
      "loss": 0.575,
      "num_input_tokens_seen": 1475408,
      "step": 5190
    },
    {
      "epoch": 9.692164179104477,
      "grad_norm": 1.970475673675537,
      "learning_rate": 1.4594470063890308e-07,
      "loss": 0.3295,
      "num_input_tokens_seen": 1476784,
      "step": 5195
    },
    {
      "epoch": 9.701492537313433,
      "grad_norm": 2.192255973815918,
      "learning_rate": 1.3729319345109348e-07,
      "loss": 0.5133,
      "num_input_tokens_seen": 1478160,
      "step": 5200
    },
    {
      "epoch": 9.710820895522389,
      "grad_norm": 1.5208903551101685,
      "learning_rate": 1.2890530312200945e-07,
      "loss": 0.3318,
      "num_input_tokens_seen": 1479600,
      "step": 5205
    },
    {
      "epoch": 9.720149253731343,
      "grad_norm": 2.261181592941284,
      "learning_rate": 1.207811185876373e-07,
      "loss": 0.2828,
      "num_input_tokens_seen": 1480944,
      "step": 5210
    },
    {
      "epoch": 9.729477611940299,
      "grad_norm": 2.315556526184082,
      "learning_rate": 1.1292072598791114e-07,
      "loss": 0.3973,
      "num_input_tokens_seen": 1482448,
      "step": 5215
    },
    {
      "epoch": 9.738805970149254,
      "grad_norm": 1.6816129684448242,
      "learning_rate": 1.0532420866581072e-07,
      "loss": 0.4369,
      "num_input_tokens_seen": 1483920,
      "step": 5220
    },
    {
      "epoch": 9.748134328358208,
      "grad_norm": 2.8451807498931885,
      "learning_rate": 9.799164716646769e-08,
      "loss": 0.3919,
      "num_input_tokens_seen": 1485232,
      "step": 5225
    },
    {
      "epoch": 9.757462686567164,
      "grad_norm": 1.6358333826065063,
      "learning_rate": 9.092311923632191e-08,
      "loss": 0.364,
      "num_input_tokens_seen": 1486640,
      "step": 5230
    },
    {
      "epoch": 9.76679104477612,
      "grad_norm": 3.8773183822631836,
      "learning_rate": 8.411869982228038e-08,
      "loss": 0.6139,
      "num_input_tokens_seen": 1488080,
      "step": 5235
    },
    {
      "epoch": 9.776119402985074,
      "grad_norm": 2.255976438522339,
      "learning_rate": 7.757846107094291e-08,
      "loss": 0.3338,
      "num_input_tokens_seen": 1489456,
      "step": 5240
    },
    {
      "epoch": 9.78544776119403,
      "grad_norm": 3.1358773708343506,
      "learning_rate": 7.130247232782216e-08,
      "loss": 0.3959,
      "num_input_tokens_seen": 1491024,
      "step": 5245
    },
    {
      "epoch": 9.794776119402986,
      "grad_norm": 2.1085126399993896,
      "learning_rate": 6.529080013661648e-08,
      "loss": 0.3863,
      "num_input_tokens_seen": 1492432,
      "step": 5250
    },
    {
      "epoch": 9.80410447761194,
      "grad_norm": 2.5028364658355713,
      "learning_rate": 5.954350823850208e-08,
      "loss": 0.3354,
      "num_input_tokens_seen": 1494064,
      "step": 5255
    },
    {
      "epoch": 9.813432835820896,
      "grad_norm": 3.430619478225708,
      "learning_rate": 5.4060657571453064e-08,
      "loss": 0.3973,
      "num_input_tokens_seen": 1495536,
      "step": 5260
    },
    {
      "epoch": 9.822761194029852,
      "grad_norm": 3.1927783489227295,
      "learning_rate": 4.884230626960307e-08,
      "loss": 0.4862,
      "num_input_tokens_seen": 1496848,
      "step": 5265
    },
    {
      "epoch": 9.832089552238806,
      "grad_norm": 2.27006459236145,
      "learning_rate": 4.388850966261793e-08,
      "loss": 0.3312,
      "num_input_tokens_seen": 1498352,
      "step": 5270
    },
    {
      "epoch": 9.841417910447761,
      "grad_norm": 2.7164342403411865,
      "learning_rate": 3.919932027512674e-08,
      "loss": 0.477,
      "num_input_tokens_seen": 1499824,
      "step": 5275
    },
    {
      "epoch": 9.850746268656717,
      "grad_norm": 3.0617809295654297,
      "learning_rate": 3.477478782614452e-08,
      "loss": 0.5223,
      "num_input_tokens_seen": 1501328,
      "step": 5280
    },
    {
      "epoch": 9.860074626865671,
      "grad_norm": 1.9799847602844238,
      "learning_rate": 3.061495922855873e-08,
      "loss": 0.4075,
      "num_input_tokens_seen": 1502640,
      "step": 5285
    },
    {
      "epoch": 9.869402985074627,
      "grad_norm": 2.8700857162475586,
      "learning_rate": 2.67198785886269e-08,
      "loss": 0.4224,
      "num_input_tokens_seen": 1504144,
      "step": 5290
    },
    {
      "epoch": 9.878731343283581,
      "grad_norm": 2.700885772705078,
      "learning_rate": 2.3089587205507578e-08,
      "loss": 0.531,
      "num_input_tokens_seen": 1505584,
      "step": 5295
    },
    {
      "epoch": 9.888059701492537,
      "grad_norm": 1.563413381576538,
      "learning_rate": 1.972412357083009e-08,
      "loss": 0.4155,
      "num_input_tokens_seen": 1507184,
      "step": 5300
    },
    {
      "epoch": 9.897388059701493,
      "grad_norm": 3.931457281112671,
      "learning_rate": 1.662352336827544e-08,
      "loss": 0.4789,
      "num_input_tokens_seen": 1508656,
      "step": 5305
    },
    {
      "epoch": 9.906716417910447,
      "grad_norm": 2.2047131061553955,
      "learning_rate": 1.3787819473207176e-08,
      "loss": 0.5689,
      "num_input_tokens_seen": 1510192,
      "step": 5310
    },
    {
      "epoch": 9.916044776119403,
      "grad_norm": 2.1585283279418945,
      "learning_rate": 1.1217041952313323e-08,
      "loss": 0.2944,
      "num_input_tokens_seen": 1511632,
      "step": 5315
    },
    {
      "epoch": 9.925373134328359,
      "grad_norm": 1.4426147937774658,
      "learning_rate": 8.91121806330386e-09,
      "loss": 0.2992,
      "num_input_tokens_seen": 1513008,
      "step": 5320
    },
    {
      "epoch": 9.934701492537313,
      "grad_norm": 1.604980230331421,
      "learning_rate": 6.870372254602631e-09,
      "loss": 0.3149,
      "num_input_tokens_seen": 1514416,
      "step": 5325
    },
    {
      "epoch": 9.944029850746269,
      "grad_norm": 1.2609069347381592,
      "learning_rate": 5.09452616509476e-09,
      "loss": 0.6028,
      "num_input_tokens_seen": 1515856,
      "step": 5330
    },
    {
      "epoch": 9.953358208955224,
      "grad_norm": 2.265486240386963,
      "learning_rate": 3.58369862391017e-09,
      "loss": 0.4039,
      "num_input_tokens_seen": 1517328,
      "step": 5335
    },
    {
      "epoch": 9.962686567164178,
      "grad_norm": 1.3175299167633057,
      "learning_rate": 2.3379056502015327e-09,
      "loss": 0.3266,
      "num_input_tokens_seen": 1518800,
      "step": 5340
    },
    {
      "epoch": 9.972014925373134,
      "grad_norm": 1.9460862874984741,
      "learning_rate": 1.357160452988837e-09,
      "loss": 0.4888,
      "num_input_tokens_seen": 1520208,
      "step": 5345
    },
    {
      "epoch": 9.98134328358209,
      "grad_norm": 2.076957941055298,
      "learning_rate": 6.414734310233872e-10,
      "loss": 0.319,
      "num_input_tokens_seen": 1521616,
      "step": 5350
    },
    {
      "epoch": 9.990671641791044,
      "grad_norm": 2.4709126949310303,
      "learning_rate": 1.9085217266290312e-10,
      "loss": 0.4874,
      "num_input_tokens_seen": 1522992,
      "step": 5355
    },
    {
      "epoch": 10.0,
      "grad_norm": 7.597536563873291,
      "learning_rate": 5.3014557993558144e-12,
      "loss": 0.2695,
      "num_input_tokens_seen": 1524216,
      "step": 5360
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.6314803957939148,
      "eval_runtime": 4.5761,
      "eval_samples_per_second": 52.009,
      "eval_steps_per_second": 13.112,
      "num_input_tokens_seen": 1524216,
      "step": 5360
    },
    {
      "epoch": 10.0,
      "num_input_tokens_seen": 1524216,
      "step": 5360,
      "total_flos": 6.863475758314291e+16,
      "train_loss": 0.5810689434520344,
      "train_runtime": 1028.1731,
      "train_samples_per_second": 20.823,
      "train_steps_per_second": 5.213
    }
  ],
  "logging_steps": 5,
  "max_steps": 5360,
  "num_input_tokens_seen": 1524216,
  "num_train_epochs": 10,
  "save_steps": 268,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.863475758314291e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}