{
  "best_global_step": 493,
  "best_metric": 0.14917685091495514,
  "best_model_checkpoint": "saves_stability/ia3/llama-3-8b-instruct/train_cb_1757340195/checkpoint-493",
  "epoch": 10.0,
  "eval_steps": 29,
  "global_step": 570,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08771929824561403,
      "grad_norm": 3.249423027038574,
      "learning_rate": 3.5087719298245615e-06,
      "loss": 1.1728,
      "num_input_tokens_seen": 3552,
      "step": 5
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 2.8306596279144287,
      "learning_rate": 7.894736842105263e-06,
      "loss": 1.2997,
      "num_input_tokens_seen": 7264,
      "step": 10
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 2.9023263454437256,
      "learning_rate": 1.2280701754385964e-05,
      "loss": 1.1782,
      "num_input_tokens_seen": 10528,
      "step": 15
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 2.783803939819336,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.084,
      "num_input_tokens_seen": 14720,
      "step": 20
    },
    {
      "epoch": 0.43859649122807015,
      "grad_norm": 3.0386464595794678,
      "learning_rate": 2.105263157894737e-05,
      "loss": 1.2233,
      "num_input_tokens_seen": 18016,
      "step": 25
    },
    {
      "epoch": 0.5087719298245614,
      "eval_loss": 1.1734217405319214,
      "eval_runtime": 0.6212,
      "eval_samples_per_second": 40.245,
      "eval_steps_per_second": 11.269,
      "num_input_tokens_seen": 20064,
      "step": 29
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 2.617579698562622,
      "learning_rate": 2.5438596491228074e-05,
      "loss": 1.3131,
      "num_input_tokens_seen": 20640,
      "step": 30
    },
    {
      "epoch": 0.6140350877192983,
      "grad_norm": 3.382931709289551,
      "learning_rate": 2.9824561403508772e-05,
      "loss": 1.1836,
      "num_input_tokens_seen": 24800,
      "step": 35
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 3.0703821182250977,
      "learning_rate": 3.421052631578947e-05,
      "loss": 1.1926,
      "num_input_tokens_seen": 28064,
      "step": 40
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 2.6126792430877686,
      "learning_rate": 3.859649122807018e-05,
      "loss": 1.3449,
      "num_input_tokens_seen": 30944,
      "step": 45
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 3.3939754962921143,
      "learning_rate": 4.298245614035088e-05,
      "loss": 1.105,
      "num_input_tokens_seen": 33664,
      "step": 50
    },
    {
      "epoch": 0.9649122807017544,
      "grad_norm": 2.854220151901245,
      "learning_rate": 4.736842105263158e-05,
      "loss": 1.1998,
      "num_input_tokens_seen": 36320,
      "step": 55
    },
    {
      "epoch": 1.0175438596491229,
      "eval_loss": 1.1734217405319214,
      "eval_runtime": 0.6146,
      "eval_samples_per_second": 40.677,
      "eval_steps_per_second": 11.39,
      "num_input_tokens_seen": 37832,
      "step": 58
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 3.078892707824707,
      "learning_rate": 4.999812487773597e-05,
      "loss": 0.969,
      "num_input_tokens_seen": 39080,
      "step": 60
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 3.7424535751342773,
      "learning_rate": 4.997703298253406e-05,
      "loss": 1.0843,
      "num_input_tokens_seen": 42536,
      "step": 65
    },
    {
      "epoch": 1.2280701754385965,
      "grad_norm": 2.660959243774414,
      "learning_rate": 4.993252512887069e-05,
      "loss": 1.1454,
      "num_input_tokens_seen": 45608,
      "step": 70
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 2.7965874671936035,
      "learning_rate": 4.986464304284091e-05,
      "loss": 1.0633,
      "num_input_tokens_seen": 49352,
      "step": 75
    },
    {
      "epoch": 1.4035087719298245,
      "grad_norm": 3.675288200378418,
      "learning_rate": 4.977345036387331e-05,
      "loss": 1.0067,
      "num_input_tokens_seen": 52328,
      "step": 80
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 3.109158992767334,
      "learning_rate": 4.965903258506806e-05,
      "loss": 0.6419,
      "num_input_tokens_seen": 56328,
      "step": 85
    },
    {
      "epoch": 1.526315789473684,
      "eval_loss": 0.5481208562850952,
      "eval_runtime": 0.6151,
      "eval_samples_per_second": 40.642,
      "eval_steps_per_second": 11.38,
      "num_input_tokens_seen": 57288,
      "step": 87
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 2.5073697566986084,
      "learning_rate": 4.952149697304716e-05,
      "loss": 0.695,
      "num_input_tokens_seen": 59048,
      "step": 90
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 2.0390727519989014,
      "learning_rate": 4.9360972467392056e-05,
      "loss": 0.3522,
      "num_input_tokens_seen": 62504,
      "step": 95
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 1.2837077379226685,
      "learning_rate": 4.917760955976277e-05,
      "loss": 0.3394,
      "num_input_tokens_seen": 65832,
      "step": 100
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 1.2724180221557617,
      "learning_rate": 4.897158015281209e-05,
      "loss": 0.3744,
      "num_input_tokens_seen": 68808,
      "step": 105
    },
    {
      "epoch": 1.9298245614035088,
      "grad_norm": 0.29305028915405273,
      "learning_rate": 4.874307739902689e-05,
      "loss": 0.197,
      "num_input_tokens_seen": 71848,
      "step": 110
    },
    {
      "epoch": 2.017543859649123,
      "grad_norm": 1.2843000888824463,
      "learning_rate": 4.849231551964771e-05,
      "loss": 0.2289,
      "num_input_tokens_seen": 74040,
      "step": 115
    },
    {
      "epoch": 2.0350877192982457,
      "eval_loss": 0.19221718609333038,
      "eval_runtime": 0.622,
      "eval_samples_per_second": 40.192,
      "eval_steps_per_second": 11.254,
      "num_input_tokens_seen": 74520,
      "step": 116
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 1.219214916229248,
      "learning_rate": 4.821952960383649e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 77272,
      "step": 120
    },
    {
      "epoch": 2.192982456140351,
      "grad_norm": 0.49150121212005615,
      "learning_rate": 4.7924975388280524e-05,
      "loss": 0.1044,
      "num_input_tokens_seen": 80280,
      "step": 125
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 0.23392945528030396,
      "learning_rate": 4.760892901743944e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 83480,
      "step": 130
    },
    {
      "epoch": 2.3684210526315788,
      "grad_norm": 0.5713793635368347,
      "learning_rate": 4.727168678465988e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 86232,
      "step": 135
    },
    {
      "epoch": 2.456140350877193,
      "grad_norm": 1.2394146919250488,
      "learning_rate": 4.6913564854400595e-05,
      "loss": 0.1514,
      "num_input_tokens_seen": 89656,
      "step": 140
    },
    {
      "epoch": 2.543859649122807,
      "grad_norm": 0.1592731475830078,
      "learning_rate": 4.6534898965828405e-05,
      "loss": 0.571,
      "num_input_tokens_seen": 93080,
      "step": 145
    },
    {
      "epoch": 2.543859649122807,
      "eval_loss": 0.17445145547389984,
      "eval_runtime": 0.6155,
      "eval_samples_per_second": 40.62,
      "eval_steps_per_second": 11.374,
      "num_input_tokens_seen": 93080,
      "step": 145
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 1.722758412361145,
      "learning_rate": 4.613604411806285e-05,
      "loss": 0.1959,
      "num_input_tokens_seen": 96440,
      "step": 150
    },
    {
      "epoch": 2.719298245614035,
      "grad_norm": 2.1143929958343506,
      "learning_rate": 4.5717374237364665e-05,
      "loss": 0.1432,
      "num_input_tokens_seen": 100280,
      "step": 155
    },
    {
      "epoch": 2.807017543859649,
      "grad_norm": 0.3370112180709839,
      "learning_rate": 4.5279281826580056e-05,
      "loss": 0.17,
      "num_input_tokens_seen": 103512,
      "step": 160
    },
    {
      "epoch": 2.8947368421052633,
      "grad_norm": 0.8938125967979431,
      "learning_rate": 4.482217759716946e-05,
      "loss": 0.1916,
      "num_input_tokens_seen": 106168,
      "step": 165
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 0.7071975469589233,
      "learning_rate": 4.434649008416565e-05,
      "loss": 0.1435,
      "num_input_tokens_seen": 109624,
      "step": 170
    },
    {
      "epoch": 3.0526315789473686,
      "eval_loss": 0.16302163898944855,
      "eval_runtime": 0.6157,
      "eval_samples_per_second": 40.603,
      "eval_steps_per_second": 11.369,
      "num_input_tokens_seen": 111928,
      "step": 174
    },
    {
      "epoch": 3.0701754385964914,
      "grad_norm": 1.815433144569397,
      "learning_rate": 4.385266524442241e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 112472,
      "step": 175
    },
    {
      "epoch": 3.1578947368421053,
      "grad_norm": 0.9170152544975281,
      "learning_rate": 4.334116603853007e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 115576,
      "step": 180
    },
    {
      "epoch": 3.245614035087719,
      "grad_norm": 0.5121091604232788,
      "learning_rate": 4.2812471996790206e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 119000,
      "step": 185
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 1.585691213607788,
      "learning_rate": 4.226707876965611e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 122360,
      "step": 190
    },
    {
      "epoch": 3.4210526315789473,
      "grad_norm": 0.11304627358913422,
      "learning_rate": 4.1705497663060767e-05,
      "loss": 0.293,
      "num_input_tokens_seen": 125496,
      "step": 195
    },
    {
      "epoch": 3.5087719298245617,
      "grad_norm": 1.571065902709961,
      "learning_rate": 4.1128255159067665e-05,
      "loss": 0.3009,
      "num_input_tokens_seen": 128760,
      "step": 200
    },
    {
      "epoch": 3.56140350877193,
      "eval_loss": 0.1575746238231659,
      "eval_runtime": 0.6184,
      "eval_samples_per_second": 40.426,
      "eval_steps_per_second": 11.319,
      "num_input_tokens_seen": 131160,
      "step": 203
    },
    {
      "epoch": 3.5964912280701755,
      "grad_norm": 0.6615663170814514,
      "learning_rate": 4.053589242229412e-05,
      "loss": 0.1217,
      "num_input_tokens_seen": 132248,
      "step": 205
    },
    {
      "epoch": 3.6842105263157894,
      "grad_norm": 1.515011191368103,
      "learning_rate": 3.9928964792569655e-05,
      "loss": 0.0966,
      "num_input_tokens_seen": 135160,
      "step": 210
    },
    {
      "epoch": 3.7719298245614032,
      "grad_norm": 0.22244925796985626,
      "learning_rate": 3.930804126430513e-05,
      "loss": 0.309,
      "num_input_tokens_seen": 138200,
      "step": 215
    },
    {
      "epoch": 3.8596491228070176,
      "grad_norm": 0.9162463545799255,
      "learning_rate": 3.867370395306068e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 141752,
      "step": 220
    },
    {
      "epoch": 3.9473684210526314,
      "grad_norm": 1.127389907836914,
      "learning_rate": 3.8026547549812665e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 146008,
      "step": 225
    },
    {
      "epoch": 4.035087719298246,
      "grad_norm": 0.13011278212070465,
      "learning_rate": 3.736717876343106e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 148648,
      "step": 230
    },
    {
      "epoch": 4.0701754385964914,
      "eval_loss": 0.15967491269111633,
      "eval_runtime": 0.621,
      "eval_samples_per_second": 40.255,
      "eval_steps_per_second": 11.271,
      "num_input_tokens_seen": 150056,
      "step": 232
    },
    {
      "epoch": 4.12280701754386,
      "grad_norm": 1.293678879737854,
      "learning_rate": 3.66962157518902e-05,
      "loss": 0.188,
      "num_input_tokens_seen": 151656,
      "step": 235
    },
    {
      "epoch": 4.2105263157894735,
      "grad_norm": 2.862987518310547,
      "learning_rate": 3.601428754274584e-05,
      "loss": 0.3198,
      "num_input_tokens_seen": 154280,
      "step": 240
    },
    {
      "epoch": 4.298245614035087,
      "grad_norm": 0.9852196574211121,
      "learning_rate": 3.532203344342212e-05,
      "loss": 0.2077,
      "num_input_tokens_seen": 157160,
      "step": 245
    },
    {
      "epoch": 4.385964912280702,
      "grad_norm": 1.4380601644515991,
      "learning_rate": 3.4620102441861143e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 160072,
      "step": 250
    },
    {
      "epoch": 4.473684210526316,
      "grad_norm": 0.6360757946968079,
      "learning_rate": 3.390915259809696e-05,
      "loss": 0.0836,
      "num_input_tokens_seen": 163752,
      "step": 255
    },
    {
      "epoch": 4.56140350877193,
      "grad_norm": 0.5739740133285522,
      "learning_rate": 3.318985042732461e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 166600,
      "step": 260
    },
    {
      "epoch": 4.578947368421053,
      "eval_loss": 0.15982376039028168,
      "eval_runtime": 0.6188,
      "eval_samples_per_second": 40.399,
      "eval_steps_per_second": 11.312,
      "num_input_tokens_seen": 167208,
      "step": 261
    },
    {
      "epoch": 4.649122807017544,
      "grad_norm": 0.23819059133529663,
      "learning_rate": 3.246287027504237e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 169704,
      "step": 265
    },
    {
      "epoch": 4.7368421052631575,
      "grad_norm": 2.7764995098114014,
      "learning_rate": 3.172889368485311e-05,
      "loss": 0.1414,
      "num_input_tokens_seen": 173000,
      "step": 270
    },
    {
      "epoch": 4.824561403508772,
      "grad_norm": 0.6186763644218445,
      "learning_rate": 3.0988608759517475e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 177128,
      "step": 275
    },
    {
      "epoch": 4.912280701754386,
      "grad_norm": 0.8575865626335144,
      "learning_rate": 3.0242709515857758e-05,
      "loss": 0.0839,
      "num_input_tokens_seen": 180680,
      "step": 280
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.012724075466394424,
      "learning_rate": 2.949189523411747e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 183280,
      "step": 285
    },
    {
      "epoch": 5.087719298245614,
      "grad_norm": 0.9721237421035767,
      "learning_rate": 2.8736869802386364e-05,
      "loss": 0.0956,
      "num_input_tokens_seen": 186160,
      "step": 290
    },
    {
      "epoch": 5.087719298245614,
      "eval_loss": 0.15423022210597992,
      "eval_runtime": 0.6256,
      "eval_samples_per_second": 39.964,
      "eval_steps_per_second": 11.19,
      "num_input_tokens_seen": 186160,
      "step": 290
    },
    {
      "epoch": 5.175438596491228,
      "grad_norm": 1.3833911418914795,
      "learning_rate": 2.797834105670559e-05,
      "loss": 0.106,
      "num_input_tokens_seen": 189584,
      "step": 295
    },
    {
      "epoch": 5.2631578947368425,
      "grad_norm": 1.4466298818588257,
      "learning_rate": 2.7217020117471793e-05,
      "loss": 0.1565,
      "num_input_tokens_seen": 193296,
      "step": 300
    },
    {
      "epoch": 5.350877192982456,
      "grad_norm": 1.2966651916503906,
      "learning_rate": 2.6453620722761896e-05,
      "loss": 0.1551,
      "num_input_tokens_seen": 197328,
      "step": 305
    },
    {
      "epoch": 5.43859649122807,
      "grad_norm": 0.6950919032096863,
      "learning_rate": 2.5688858559204053e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 200560,
      "step": 310
    },
    {
      "epoch": 5.526315789473684,
      "grad_norm": 0.7273321747779846,
      "learning_rate": 2.492345059102164e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 203632,
      "step": 315
    },
    {
      "epoch": 5.5964912280701755,
      "eval_loss": 0.1532665342092514,
      "eval_runtime": 0.6187,
      "eval_samples_per_second": 40.408,
      "eval_steps_per_second": 11.314,
      "num_input_tokens_seen": 206000,
      "step": 319
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 1.7336548566818237,
      "learning_rate": 2.4158114387879616e-05,
      "loss": 0.2487,
      "num_input_tokens_seen": 206480,
      "step": 320
    },
    {
      "epoch": 5.701754385964913,
      "grad_norm": 1.8356577157974243,
      "learning_rate": 2.3393567452163252e-05,
      "loss": 0.3011,
      "num_input_tokens_seen": 209232,
      "step": 325
    },
    {
      "epoch": 5.7894736842105265,
      "grad_norm": 1.1948175430297852,
      "learning_rate": 2.2630526546319914e-05,
      "loss": 0.1642,
      "num_input_tokens_seen": 213168,
      "step": 330
    },
    {
      "epoch": 5.87719298245614,
      "grad_norm": 1.3118547201156616,
      "learning_rate": 2.186970702089457e-05,
      "loss": 0.1599,
      "num_input_tokens_seen": 216752,
      "step": 335
    },
    {
      "epoch": 5.964912280701754,
      "grad_norm": 0.24697716534137726,
      "learning_rate": 2.111182214388893e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 219536,
      "step": 340
    },
    {
      "epoch": 6.052631578947368,
      "grad_norm": 0.1937997192144394,
      "learning_rate": 2.0357582432072957e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 222272,
      "step": 345
    },
    {
      "epoch": 6.105263157894737,
      "eval_loss": 0.15324875712394714,
      "eval_runtime": 0.6137,
      "eval_samples_per_second": 40.735,
      "eval_steps_per_second": 11.406,
      "num_input_tokens_seen": 224064,
      "step": 348
    },
    {
      "epoch": 6.140350877192983,
      "grad_norm": 0.5285282731056213,
      "learning_rate": 1.9607694984875754e-05,
      "loss": 0.1867,
      "num_input_tokens_seen": 225664,
      "step": 350
    },
    {
      "epoch": 6.228070175438597,
      "grad_norm": 1.2666800022125244,
      "learning_rate": 1.8862862821480025e-05,
      "loss": 0.1796,
      "num_input_tokens_seen": 229408,
      "step": 355
    },
    {
      "epoch": 6.315789473684211,
      "grad_norm": 0.9785047769546509,
      "learning_rate": 1.8123784221741964e-05,
      "loss": 0.278,
      "num_input_tokens_seen": 232832,
      "step": 360
    },
    {
      "epoch": 6.4035087719298245,
      "grad_norm": 0.6506590843200684,
      "learning_rate": 1.73911520715541e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 236288,
      "step": 365
    },
    {
      "epoch": 6.491228070175438,
      "grad_norm": 0.482577383518219,
      "learning_rate": 1.666565321326512e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 239296,
      "step": 370
    },
    {
      "epoch": 6.578947368421053,
      "grad_norm": 1.2154335975646973,
      "learning_rate": 1.5947967801765345e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 242848,
      "step": 375
    },
    {
      "epoch": 6.614035087719298,
      "eval_loss": 0.15488915145397186,
      "eval_runtime": 0.6182,
      "eval_samples_per_second": 40.438,
      "eval_steps_per_second": 11.323,
      "num_input_tokens_seen": 243840,
      "step": 377
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3786420524120331,
      "learning_rate": 1.5238768666841907e-05,
      "loss": 0.1958,
      "num_input_tokens_seen": 245344,
      "step": 380
    },
    {
      "epoch": 6.754385964912281,
      "grad_norm": 0.35596537590026855,
      "learning_rate": 1.4538720682400969e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 248256,
      "step": 385
    },
    {
      "epoch": 6.842105263157895,
      "grad_norm": 0.5632941126823425,
      "learning_rate": 1.3848480143148839e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 250976,
      "step": 390
    },
    {
      "epoch": 6.9298245614035086,
      "grad_norm": 0.5877909660339355,
      "learning_rate": 1.3168694149315796e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 254944,
      "step": 395
    },
    {
      "epoch": 7.017543859649122,
      "grad_norm": 0.5971133708953857,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 257888,
      "step": 400
    },
    {
      "epoch": 7.105263157894737,
      "grad_norm": 0.12039807438850403,
      "learning_rate": 1.1843024595699805e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 261088,
      "step": 405
    },
    {
      "epoch": 7.12280701754386,
      "eval_loss": 0.15452837944030762,
      "eval_runtime": 0.6206,
      "eval_samples_per_second": 40.28,
      "eval_steps_per_second": 11.279,
      "num_input_tokens_seen": 261504,
      "step": 406
    },
    {
      "epoch": 7.192982456140351,
      "grad_norm": 0.4003351926803589,
      "learning_rate": 1.1198383850594758e-05,
      "loss": 0.0568,
      "num_input_tokens_seen": 263840,
      "step": 410
    },
    {
      "epoch": 7.280701754385965,
      "grad_norm": 2.261807680130005,
      "learning_rate": 1.0566682115126344e-05,
      "loss": 0.2437,
      "num_input_tokens_seen": 267424,
      "step": 415
    },
    {
      "epoch": 7.368421052631579,
      "grad_norm": 1.0385570526123047,
      "learning_rate": 9.948511609419675e-06,
      "loss": 0.2337,
      "num_input_tokens_seen": 270720,
      "step": 420
    },
    {
      "epoch": 7.456140350877193,
      "grad_norm": 1.3235441446304321,
      "learning_rate": 9.344451868077353e-06,
      "loss": 0.2239,
      "num_input_tokens_seen": 273504,
      "step": 425
    },
    {
      "epoch": 7.543859649122807,
      "grad_norm": 0.8284580111503601,
      "learning_rate": 8.755069196866014e-06,
      "loss": 0.1115,
      "num_input_tokens_seen": 276608,
      "step": 430
    },
    {
      "epoch": 7.631578947368421,
      "grad_norm": 0.41157782077789307,
      "learning_rate": 8.180916141804906e-06,
      "loss": 0.1625,
      "num_input_tokens_seen": 280352,
      "step": 435
    },
    {
      "epoch": 7.631578947368421,
      "eval_loss": 0.1503370851278305,
      "eval_runtime": 0.62,
      "eval_samples_per_second": 40.325,
      "eval_steps_per_second": 11.291,
      "num_input_tokens_seen": 280352,
      "step": 435
    },
    {
      "epoch": 7.719298245614035,
      "grad_norm": 0.722707986831665,
      "learning_rate": 7.622530971154199e-06,
      "loss": 0.0874,
      "num_input_tokens_seen": 283808,
      "step": 440
    },
    {
      "epoch": 7.807017543859649,
      "grad_norm": 1.2332518100738525,
      "learning_rate": 7.080437170788723e-06,
      "loss": 0.1117,
      "num_input_tokens_seen": 286688,
      "step": 445
    },
    {
      "epoch": 7.894736842105263,
      "grad_norm": 0.7050608992576599,
      "learning_rate": 6.555142953430158e-06,
      "loss": 0.1637,
      "num_input_tokens_seen": 290240,
      "step": 450
    },
    {
      "epoch": 7.982456140350877,
      "grad_norm": 0.6452037692070007,
      "learning_rate": 6.0471407821978135e-06,
      "loss": 0.1281,
      "num_input_tokens_seen": 293568,
      "step": 455
    },
    {
      "epoch": 8.070175438596491,
      "grad_norm": 0.11075533181428909,
      "learning_rate": 5.556906908924655e-06,
      "loss": 0.0503,
      "num_input_tokens_seen": 296720,
      "step": 460
    },
    {
      "epoch": 8.140350877192983,
      "eval_loss": 0.1512749046087265,
      "eval_runtime": 0.6186,
      "eval_samples_per_second": 40.414,
      "eval_steps_per_second": 11.316,
      "num_input_tokens_seen": 299344,
      "step": 464
    },
    {
      "epoch": 8.157894736842104,
      "grad_norm": 0.3659641444683075,
      "learning_rate": 5.084900927671393e-06,
      "loss": 0.1795,
      "num_input_tokens_seen": 300112,
      "step": 465
    },
    {
      "epoch": 8.24561403508772,
      "grad_norm": 2.306575298309326,
      "learning_rate": 4.631565343857239e-06,
      "loss": 0.1017,
      "num_input_tokens_seen": 303664,
      "step": 470
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.44194552302360535,
      "learning_rate": 4.19732515941125e-06,
      "loss": 0.1015,
      "num_input_tokens_seen": 306480,
      "step": 475
    },
    {
      "epoch": 8.421052631578947,
      "grad_norm": 1.1002084016799927,
      "learning_rate": 3.7825874743331907e-06,
      "loss": 0.1745,
      "num_input_tokens_seen": 309744,
      "step": 480
    },
    {
      "epoch": 8.508771929824562,
      "grad_norm": 1.1627758741378784,
      "learning_rate": 3.3877411050374424e-06,
      "loss": 0.1327,
      "num_input_tokens_seen": 313360,
      "step": 485
    },
    {
      "epoch": 8.596491228070175,
      "grad_norm": 1.3900173902511597,
      "learning_rate": 3.013156219837776e-06,
      "loss": 0.3406,
      "num_input_tokens_seen": 316784,
      "step": 490
    },
    {
      "epoch": 8.649122807017545,
      "eval_loss": 0.14917685091495514,
      "eval_runtime": 0.6239,
      "eval_samples_per_second": 40.072,
      "eval_steps_per_second": 11.22,
      "num_input_tokens_seen": 318672,
      "step": 493
    },
    {
      "epoch": 8.68421052631579,
      "grad_norm": 0.21432092785835266,
      "learning_rate": 2.659183991914696e-06,
      "loss": 0.0523,
      "num_input_tokens_seen": 319760,
      "step": 495
    },
    {
      "epoch": 8.771929824561404,
      "grad_norm": 0.13650639355182648,
      "learning_rate": 2.326156270090735e-06,
      "loss": 0.0745,
      "num_input_tokens_seen": 322512,
      "step": 500
    },
    {
      "epoch": 8.859649122807017,
      "grad_norm": 0.21135732531547546,
      "learning_rate": 2.0143852677223075e-06,
      "loss": 0.0987,
      "num_input_tokens_seen": 325936,
      "step": 505
    },
    {
      "epoch": 8.947368421052632,
      "grad_norm": 0.8339365124702454,
      "learning_rate": 1.7241632699998123e-06,
      "loss": 0.1183,
      "num_input_tokens_seen": 329360,
      "step": 510
    },
    {
      "epoch": 9.035087719298245,
      "grad_norm": 0.7296983003616333,
      "learning_rate": 1.4557623599303903e-06,
      "loss": 0.1854,
      "num_input_tokens_seen": 332168,
      "step": 515
    },
    {
      "epoch": 9.12280701754386,
      "grad_norm": 0.28494396805763245,
      "learning_rate": 1.2094341632602064e-06,
      "loss": 0.0851,
      "num_input_tokens_seen": 336296,
      "step": 520
    },
    {
      "epoch": 9.157894736842104,
      "eval_loss": 0.15301954746246338,
      "eval_runtime": 0.6247,
      "eval_samples_per_second": 40.017,
      "eval_steps_per_second": 11.205,
      "num_input_tokens_seen": 337480,
      "step": 522
    },
    {
      "epoch": 9.210526315789474,
      "grad_norm": 0.34183087944984436,
      "learning_rate": 9.85409612575411e-07,
      "loss": 0.2518,
      "num_input_tokens_seen": 339080,
      "step": 525
    },
    {
      "epoch": 9.298245614035087,
      "grad_norm": 0.09366191923618317,
      "learning_rate": 7.838987308029427e-07,
      "loss": 0.076,
      "num_input_tokens_seen": 342568,
      "step": 530
    },
    {
      "epoch": 9.385964912280702,
      "grad_norm": 0.7235726118087769,
      "learning_rate": 6.050904343141095e-07,
      "loss": 0.1136,
      "num_input_tokens_seen": 345576,
      "step": 535
    },
    {
      "epoch": 9.473684210526315,
      "grad_norm": 0.38851621747016907,
      "learning_rate": 4.491523558155714e-07,
      "loss": 0.0561,
      "num_input_tokens_seen": 348392,
      "step": 540
    },
    {
      "epoch": 9.56140350877193,
      "grad_norm": 0.5667129755020142,
      "learning_rate": 3.162306871937387e-07,
      "loss": 0.0778,
      "num_input_tokens_seen": 351912,
      "step": 545
    },
    {
      "epoch": 9.649122807017545,
      "grad_norm": 0.6705904006958008,
      "learning_rate": 2.064500424599436e-07,
      "loss": 0.1179,
      "num_input_tokens_seen": 355432,
      "step": 550
    },
    {
      "epoch": 9.666666666666666,
      "eval_loss": 0.15663054585456848,
      "eval_runtime": 0.6217,
      "eval_samples_per_second": 40.211,
      "eval_steps_per_second": 11.259,
      "num_input_tokens_seen": 356456,
      "step": 551
    },
    {
      "epoch": 9.736842105263158,
      "grad_norm": 1.3486438989639282,
      "learning_rate": 1.1991334092484318e-07,
      "loss": 0.1534,
      "num_input_tokens_seen": 358760,
      "step": 555
    },
    {
      "epoch": 9.824561403508772,
      "grad_norm": 0.07935890555381775,
      "learning_rate": 5.6701710711626334e-08,
      "loss": 0.227,
      "num_input_tokens_seen": 362088,
      "step": 560
    },
    {
      "epoch": 9.912280701754385,
      "grad_norm": 0.39842548966407776,
      "learning_rate": 1.6874412698408836e-08,
      "loss": 0.2402,
      "num_input_tokens_seen": 365544,
      "step": 565
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.19832293689250946,
      "learning_rate": 4.687849611939576e-10,
      "loss": 0.1567,
      "num_input_tokens_seen": 367864,
      "step": 570
    },
    {
      "epoch": 10.0,
      "num_input_tokens_seen": 367864,
      "step": 570,
      "total_flos": 1.656518360727552e+16,
      "train_loss": 0.3069157079123614,
      "train_runtime": 158.2968,
      "train_samples_per_second": 14.214,
      "train_steps_per_second": 3.601
    }
  ],
  "logging_steps": 5,
  "max_steps": 570,
  "num_input_tokens_seen": 367864,
  "num_train_epochs": 10,
  "save_steps": 29,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.656518360727552e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}