{
  "best_global_step": 551,
  "best_metric": 3.8482749462127686,
  "best_model_checkpoint": "saves/prefix-tuning/llama-3-8b-instruct/train_cb_1754652159/checkpoint-551",
  "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": 1.9907561540603638,
      "learning_rate": 3.5087719298245615e-06,
      "loss": 12.757,
      "num_input_tokens_seen": 3552,
      "step": 5
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 2.235771894454956,
      "learning_rate": 7.894736842105263e-06,
      "loss": 12.4922,
      "num_input_tokens_seen": 7264,
      "step": 10
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 2.3024075031280518,
      "learning_rate": 1.2280701754385964e-05,
      "loss": 12.5261,
      "num_input_tokens_seen": 10528,
      "step": 15
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 1.8383631706237793,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 12.3027,
      "num_input_tokens_seen": 14720,
      "step": 20
    },
    {
      "epoch": 0.43859649122807015,
      "grad_norm": 2.205447196960449,
      "learning_rate": 2.105263157894737e-05,
      "loss": 12.6939,
      "num_input_tokens_seen": 18016,
      "step": 25
    },
    {
      "epoch": 0.5087719298245614,
      "eval_loss": 12.153342247009277,
      "eval_runtime": 0.6127,
      "eval_samples_per_second": 40.802,
      "eval_steps_per_second": 11.424,
      "num_input_tokens_seen": 20064,
      "step": 29
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 2.085423707962036,
      "learning_rate": 2.5438596491228074e-05,
      "loss": 12.5997,
      "num_input_tokens_seen": 20640,
      "step": 30
    },
    {
      "epoch": 0.6140350877192983,
      "grad_norm": 2.133491277694702,
      "learning_rate": 2.9824561403508772e-05,
      "loss": 12.6701,
      "num_input_tokens_seen": 24800,
      "step": 35
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 1.9534471035003662,
      "learning_rate": 3.421052631578947e-05,
      "loss": 11.6756,
      "num_input_tokens_seen": 28064,
      "step": 40
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 2.0729124546051025,
      "learning_rate": 3.859649122807018e-05,
      "loss": 11.3441,
      "num_input_tokens_seen": 30944,
      "step": 45
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 2.010507106781006,
      "learning_rate": 4.298245614035088e-05,
      "loss": 11.7411,
      "num_input_tokens_seen": 33664,
      "step": 50
    },
    {
      "epoch": 0.9649122807017544,
      "grad_norm": 2.055694580078125,
      "learning_rate": 4.736842105263158e-05,
      "loss": 11.0368,
      "num_input_tokens_seen": 36320,
      "step": 55
    },
    {
      "epoch": 1.0175438596491229,
      "eval_loss": 10.890871047973633,
      "eval_runtime": 0.6065,
      "eval_samples_per_second": 41.218,
      "eval_steps_per_second": 11.541,
      "num_input_tokens_seen": 37832,
      "step": 58
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 2.116218090057373,
      "learning_rate": 4.999812487773597e-05,
      "loss": 11.3165,
      "num_input_tokens_seen": 39080,
      "step": 60
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 2.034322738647461,
      "learning_rate": 4.997703298253406e-05,
      "loss": 11.1092,
      "num_input_tokens_seen": 42536,
      "step": 65
    },
    {
      "epoch": 1.2280701754385965,
      "grad_norm": 1.8908439874649048,
      "learning_rate": 4.993252512887069e-05,
      "loss": 10.783,
      "num_input_tokens_seen": 45608,
      "step": 70
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 2.179896116256714,
      "learning_rate": 4.986464304284091e-05,
      "loss": 9.9158,
      "num_input_tokens_seen": 49352,
      "step": 75
    },
    {
      "epoch": 1.4035087719298245,
      "grad_norm": 1.8387644290924072,
      "learning_rate": 4.977345036387331e-05,
      "loss": 10.1963,
      "num_input_tokens_seen": 52328,
      "step": 80
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 2.1669516563415527,
      "learning_rate": 4.965903258506806e-05,
      "loss": 10.4486,
      "num_input_tokens_seen": 56328,
      "step": 85
    },
    {
      "epoch": 1.526315789473684,
      "eval_loss": 9.39639949798584,
      "eval_runtime": 0.6062,
      "eval_samples_per_second": 41.24,
      "eval_steps_per_second": 11.547,
      "num_input_tokens_seen": 57288,
      "step": 87
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 2.1507346630096436,
      "learning_rate": 4.952149697304716e-05,
      "loss": 9.7679,
      "num_input_tokens_seen": 59048,
      "step": 90
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.8538447618484497,
      "learning_rate": 4.9360972467392056e-05,
      "loss": 8.7756,
      "num_input_tokens_seen": 62504,
      "step": 95
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 2.315293550491333,
      "learning_rate": 4.917760955976277e-05,
      "loss": 8.4432,
      "num_input_tokens_seen": 65832,
      "step": 100
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 1.8978357315063477,
      "learning_rate": 4.897158015281209e-05,
      "loss": 8.6132,
      "num_input_tokens_seen": 68808,
      "step": 105
    },
    {
      "epoch": 1.9298245614035088,
      "grad_norm": 1.8035889863967896,
      "learning_rate": 4.874307739902689e-05,
      "loss": 8.4403,
      "num_input_tokens_seen": 71848,
      "step": 110
    },
    {
      "epoch": 2.017543859649123,
      "grad_norm": 1.7996368408203125,
      "learning_rate": 4.849231551964771e-05,
      "loss": 7.9533,
      "num_input_tokens_seen": 74040,
      "step": 115
    },
    {
      "epoch": 2.0350877192982457,
      "eval_loss": 7.922963619232178,
      "eval_runtime": 0.6109,
      "eval_samples_per_second": 40.923,
      "eval_steps_per_second": 11.458,
      "num_input_tokens_seen": 74520,
      "step": 116
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 1.788129448890686,
      "learning_rate": 4.821952960383649e-05,
      "loss": 8.1754,
      "num_input_tokens_seen": 77272,
      "step": 120
    },
    {
      "epoch": 2.192982456140351,
      "grad_norm": 1.6834213733673096,
      "learning_rate": 4.7924975388280524e-05,
      "loss": 8.2941,
      "num_input_tokens_seen": 80280,
      "step": 125
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 1.727159857749939,
      "learning_rate": 4.760892901743944e-05,
      "loss": 8.0539,
      "num_input_tokens_seen": 83480,
      "step": 130
    },
    {
      "epoch": 2.3684210526315788,
      "grad_norm": 1.7604347467422485,
      "learning_rate": 4.727168678465988e-05,
      "loss": 7.7214,
      "num_input_tokens_seen": 86232,
      "step": 135
    },
    {
      "epoch": 2.456140350877193,
      "grad_norm": 1.5606811046600342,
      "learning_rate": 4.6913564854400595e-05,
      "loss": 7.7251,
      "num_input_tokens_seen": 89656,
      "step": 140
    },
    {
      "epoch": 2.543859649122807,
      "grad_norm": 1.685321569442749,
      "learning_rate": 4.6534898965828405e-05,
      "loss": 7.4909,
      "num_input_tokens_seen": 93080,
      "step": 145
    },
    {
      "epoch": 2.543859649122807,
      "eval_loss": 6.920214653015137,
      "eval_runtime": 0.6094,
      "eval_samples_per_second": 41.022,
      "eval_steps_per_second": 11.486,
      "num_input_tokens_seen": 93080,
      "step": 145
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 1.6069295406341553,
      "learning_rate": 4.613604411806285e-05,
      "loss": 6.7691,
      "num_input_tokens_seen": 96440,
      "step": 150
    },
    {
      "epoch": 2.719298245614035,
      "grad_norm": 1.6741470098495483,
      "learning_rate": 4.5717374237364665e-05,
      "loss": 7.0387,
      "num_input_tokens_seen": 100280,
      "step": 155
    },
    {
      "epoch": 2.807017543859649,
      "grad_norm": 1.6800447702407837,
      "learning_rate": 4.5279281826580056e-05,
      "loss": 6.33,
      "num_input_tokens_seen": 103512,
      "step": 160
    },
    {
      "epoch": 2.8947368421052633,
      "grad_norm": 1.4865342378616333,
      "learning_rate": 4.482217759716946e-05,
      "loss": 6.4127,
      "num_input_tokens_seen": 106168,
      "step": 165
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 1.5618782043457031,
      "learning_rate": 4.434649008416565e-05,
      "loss": 6.0778,
      "num_input_tokens_seen": 109624,
      "step": 170
    },
    {
      "epoch": 3.0526315789473686,
      "eval_loss": 6.102657794952393,
      "eval_runtime": 0.6108,
      "eval_samples_per_second": 40.931,
      "eval_steps_per_second": 11.461,
      "num_input_tokens_seen": 111928,
      "step": 174
    },
    {
      "epoch": 3.0701754385964914,
      "grad_norm": 1.8464562892913818,
      "learning_rate": 4.385266524442241e-05,
      "loss": 6.0082,
      "num_input_tokens_seen": 112472,
      "step": 175
    },
    {
      "epoch": 3.1578947368421053,
      "grad_norm": 1.7316503524780273,
      "learning_rate": 4.334116603853007e-05,
      "loss": 6.3538,
      "num_input_tokens_seen": 115576,
      "step": 180
    },
    {
      "epoch": 3.245614035087719,
      "grad_norm": 2.099435329437256,
      "learning_rate": 4.2812471996790206e-05,
      "loss": 6.1122,
      "num_input_tokens_seen": 119000,
      "step": 185
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 1.4580914974212646,
      "learning_rate": 4.226707876965611e-05,
      "loss": 6.0652,
      "num_input_tokens_seen": 122360,
      "step": 190
    },
    {
      "epoch": 3.4210526315789473,
      "grad_norm": 1.372891902923584,
      "learning_rate": 4.1705497663060767e-05,
      "loss": 5.6347,
      "num_input_tokens_seen": 125496,
      "step": 195
    },
    {
      "epoch": 3.5087719298245617,
      "grad_norm": 1.5532852411270142,
      "learning_rate": 4.1128255159067665e-05,
      "loss": 6.4246,
      "num_input_tokens_seen": 128760,
      "step": 200
    },
    {
      "epoch": 3.56140350877193,
      "eval_loss": 5.588784217834473,
      "eval_runtime": 0.6092,
      "eval_samples_per_second": 41.039,
      "eval_steps_per_second": 11.491,
      "num_input_tokens_seen": 131160,
      "step": 203
    },
    {
      "epoch": 3.5964912280701755,
      "grad_norm": 1.3230892419815063,
      "learning_rate": 4.053589242229412e-05,
      "loss": 5.8833,
      "num_input_tokens_seen": 132248,
      "step": 205
    },
    {
      "epoch": 3.6842105263157894,
      "grad_norm": 1.3423362970352173,
      "learning_rate": 3.9928964792569655e-05,
      "loss": 5.4105,
      "num_input_tokens_seen": 135160,
      "step": 210
    },
    {
      "epoch": 3.7719298245614032,
      "grad_norm": 1.3119643926620483,
      "learning_rate": 3.930804126430513e-05,
      "loss": 5.8935,
      "num_input_tokens_seen": 138200,
      "step": 215
    },
    {
      "epoch": 3.8596491228070176,
      "grad_norm": 2.161855697631836,
      "learning_rate": 3.867370395306068e-05,
      "loss": 5.5327,
      "num_input_tokens_seen": 141752,
      "step": 220
    },
    {
      "epoch": 3.9473684210526314,
      "grad_norm": 1.5290467739105225,
      "learning_rate": 3.8026547549812665e-05,
      "loss": 6.2326,
      "num_input_tokens_seen": 146008,
      "step": 225
    },
    {
      "epoch": 4.035087719298246,
      "grad_norm": 1.4180949926376343,
      "learning_rate": 3.736717876343106e-05,
      "loss": 5.7896,
      "num_input_tokens_seen": 148648,
      "step": 230
    },
    {
      "epoch": 4.0701754385964914,
      "eval_loss": 5.226391792297363,
      "eval_runtime": 0.6095,
      "eval_samples_per_second": 41.015,
      "eval_steps_per_second": 11.484,
      "num_input_tokens_seen": 150056,
      "step": 232
    },
    {
      "epoch": 4.12280701754386,
      "grad_norm": 1.4968349933624268,
      "learning_rate": 3.66962157518902e-05,
      "loss": 5.3999,
      "num_input_tokens_seen": 151656,
      "step": 235
    },
    {
      "epoch": 4.2105263157894735,
      "grad_norm": 1.1194733381271362,
      "learning_rate": 3.601428754274584e-05,
      "loss": 5.3437,
      "num_input_tokens_seen": 154280,
      "step": 240
    },
    {
      "epoch": 4.298245614035087,
      "grad_norm": 1.4553688764572144,
      "learning_rate": 3.532203344342212e-05,
      "loss": 5.4678,
      "num_input_tokens_seen": 157160,
      "step": 245
    },
    {
      "epoch": 4.385964912280702,
      "grad_norm": 1.4682071208953857,
      "learning_rate": 3.4620102441861143e-05,
      "loss": 5.4829,
      "num_input_tokens_seen": 160072,
      "step": 250
    },
    {
      "epoch": 4.473684210526316,
      "grad_norm": 1.5301012992858887,
      "learning_rate": 3.390915259809696e-05,
      "loss": 5.1392,
      "num_input_tokens_seen": 163752,
      "step": 255
    },
    {
      "epoch": 4.56140350877193,
      "grad_norm": 1.37416410446167,
      "learning_rate": 3.318985042732461e-05,
      "loss": 4.6737,
      "num_input_tokens_seen": 166600,
      "step": 260
    },
    {
      "epoch": 4.578947368421053,
      "eval_loss": 4.920877933502197,
      "eval_runtime": 0.6082,
      "eval_samples_per_second": 41.106,
      "eval_steps_per_second": 11.51,
      "num_input_tokens_seen": 167208,
      "step": 261
    },
    {
      "epoch": 4.649122807017544,
      "grad_norm": 1.3864611387252808,
      "learning_rate": 3.246287027504237e-05,
      "loss": 5.0123,
      "num_input_tokens_seen": 169704,
      "step": 265
    },
    {
      "epoch": 4.7368421052631575,
      "grad_norm": 1.261285424232483,
      "learning_rate": 3.172889368485311e-05,
      "loss": 4.5466,
      "num_input_tokens_seen": 173000,
      "step": 270
    },
    {
      "epoch": 4.824561403508772,
      "grad_norm": 2.1658103466033936,
      "learning_rate": 3.0988608759517475e-05,
      "loss": 5.6943,
      "num_input_tokens_seen": 177128,
      "step": 275
    },
    {
      "epoch": 4.912280701754386,
      "grad_norm": 1.3422341346740723,
      "learning_rate": 3.0242709515857758e-05,
      "loss": 4.4906,
      "num_input_tokens_seen": 180680,
      "step": 280
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.7265498638153076,
      "learning_rate": 2.949189523411747e-05,
      "loss": 4.466,
      "num_input_tokens_seen": 183280,
      "step": 285
    },
    {
      "epoch": 5.087719298245614,
      "grad_norm": 1.6141693592071533,
      "learning_rate": 2.8736869802386364e-05,
      "loss": 4.9442,
      "num_input_tokens_seen": 186160,
      "step": 290
    },
    {
      "epoch": 5.087719298245614,
      "eval_loss": 4.636929512023926,
      "eval_runtime": 0.6133,
      "eval_samples_per_second": 40.763,
      "eval_steps_per_second": 11.414,
      "num_input_tokens_seen": 186160,
      "step": 290
    },
    {
      "epoch": 5.175438596491228,
      "grad_norm": 1.650472640991211,
      "learning_rate": 2.797834105670559e-05,
      "loss": 4.6986,
      "num_input_tokens_seen": 189584,
      "step": 295
    },
    {
      "epoch": 5.2631578947368425,
      "grad_norm": 1.516791582107544,
      "learning_rate": 2.7217020117471793e-05,
      "loss": 4.6943,
      "num_input_tokens_seen": 193296,
      "step": 300
    },
    {
      "epoch": 5.350877192982456,
      "grad_norm": 1.355210542678833,
      "learning_rate": 2.6453620722761896e-05,
      "loss": 4.8374,
      "num_input_tokens_seen": 197328,
      "step": 305
    },
    {
      "epoch": 5.43859649122807,
      "grad_norm": 1.489431619644165,
      "learning_rate": 2.5688858559204053e-05,
      "loss": 4.4153,
      "num_input_tokens_seen": 200560,
      "step": 310
    },
    {
      "epoch": 5.526315789473684,
      "grad_norm": 1.0992668867111206,
      "learning_rate": 2.492345059102164e-05,
      "loss": 4.2959,
      "num_input_tokens_seen": 203632,
      "step": 315
    },
    {
      "epoch": 5.5964912280701755,
      "eval_loss": 4.427004337310791,
      "eval_runtime": 0.6091,
      "eval_samples_per_second": 41.044,
      "eval_steps_per_second": 11.492,
      "num_input_tokens_seen": 206000,
      "step": 319
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 1.5276824235916138,
      "learning_rate": 2.4158114387879616e-05,
      "loss": 4.7125,
      "num_input_tokens_seen": 206480,
      "step": 320
    },
    {
      "epoch": 5.701754385964913,
      "grad_norm": 1.5830577611923218,
      "learning_rate": 2.3393567452163252e-05,
      "loss": 4.4515,
      "num_input_tokens_seen": 209232,
      "step": 325
    },
    {
      "epoch": 5.7894736842105265,
      "grad_norm": 1.2052812576293945,
      "learning_rate": 2.2630526546319914e-05,
      "loss": 4.7611,
      "num_input_tokens_seen": 213168,
      "step": 330
    },
    {
      "epoch": 5.87719298245614,
      "grad_norm": 1.3054100275039673,
      "learning_rate": 2.186970702089457e-05,
      "loss": 4.8574,
      "num_input_tokens_seen": 216752,
      "step": 335
    },
    {
      "epoch": 5.964912280701754,
      "grad_norm": 1.9110170602798462,
      "learning_rate": 2.111182214388893e-05,
      "loss": 4.0179,
      "num_input_tokens_seen": 219536,
      "step": 340
    },
    {
      "epoch": 6.052631578947368,
      "grad_norm": 1.5975571870803833,
      "learning_rate": 2.0357582432072957e-05,
      "loss": 4.0552,
      "num_input_tokens_seen": 222272,
      "step": 345
    },
    {
      "epoch": 6.105263157894737,
      "eval_loss": 4.272974014282227,
      "eval_runtime": 0.6081,
      "eval_samples_per_second": 41.114,
      "eval_steps_per_second": 11.512,
      "num_input_tokens_seen": 224064,
      "step": 348
    },
    {
      "epoch": 6.140350877192983,
      "grad_norm": 1.9832843542099,
      "learning_rate": 1.9607694984875754e-05,
      "loss": 4.9111,
      "num_input_tokens_seen": 225664,
      "step": 350
    },
    {
      "epoch": 6.228070175438597,
      "grad_norm": 1.5513900518417358,
      "learning_rate": 1.8862862821480025e-05,
      "loss": 4.4043,
      "num_input_tokens_seen": 229408,
      "step": 355
    },
    {
      "epoch": 6.315789473684211,
      "grad_norm": 1.2998956441879272,
      "learning_rate": 1.8123784221741964e-05,
      "loss": 4.8502,
      "num_input_tokens_seen": 232832,
      "step": 360
    },
    {
      "epoch": 6.4035087719298245,
      "grad_norm": 1.7501558065414429,
      "learning_rate": 1.73911520715541e-05,
      "loss": 4.1639,
      "num_input_tokens_seen": 236288,
      "step": 365
    },
    {
      "epoch": 6.491228070175438,
      "grad_norm": 1.292130708694458,
      "learning_rate": 1.666565321326512e-05,
      "loss": 3.9357,
      "num_input_tokens_seen": 239296,
      "step": 370
    },
    {
      "epoch": 6.578947368421053,
      "grad_norm": 1.1817779541015625,
      "learning_rate": 1.5947967801765345e-05,
      "loss": 4.3442,
      "num_input_tokens_seen": 242848,
      "step": 375
    },
    {
      "epoch": 6.614035087719298,
      "eval_loss": 4.141702175140381,
      "eval_runtime": 0.6064,
      "eval_samples_per_second": 41.225,
      "eval_steps_per_second": 11.543,
      "num_input_tokens_seen": 243840,
      "step": 377
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 1.2892791032791138,
      "learning_rate": 1.5238768666841907e-05,
      "loss": 4.0103,
      "num_input_tokens_seen": 245344,
      "step": 380
    },
    {
      "epoch": 6.754385964912281,
      "grad_norm": 1.3167837858200073,
      "learning_rate": 1.4538720682400969e-05,
      "loss": 4.0767,
      "num_input_tokens_seen": 248256,
      "step": 385
    },
    {
      "epoch": 6.842105263157895,
      "grad_norm": 1.1590781211853027,
      "learning_rate": 1.3848480143148839e-05,
      "loss": 4.2847,
      "num_input_tokens_seen": 250976,
      "step": 390
    },
    {
      "epoch": 6.9298245614035086,
      "grad_norm": 1.3808413743972778,
      "learning_rate": 1.3168694149315796e-05,
      "loss": 4.0075,
      "num_input_tokens_seen": 254944,
      "step": 395
    },
    {
      "epoch": 7.017543859649122,
      "grad_norm": 1.1876074075698853,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 4.0259,
      "num_input_tokens_seen": 257888,
      "step": 400
    },
    {
      "epoch": 7.105263157894737,
      "grad_norm": 1.1949429512023926,
      "learning_rate": 1.1843024595699805e-05,
      "loss": 3.8654,
      "num_input_tokens_seen": 261088,
      "step": 405
    },
    {
      "epoch": 7.12280701754386,
      "eval_loss": 4.0376129150390625,
      "eval_runtime": 0.6104,
      "eval_samples_per_second": 40.957,
      "eval_steps_per_second": 11.468,
      "num_input_tokens_seen": 261504,
      "step": 406
    },
    {
      "epoch": 7.192982456140351,
      "grad_norm": 1.077356219291687,
      "learning_rate": 1.1198383850594758e-05,
      "loss": 3.8608,
      "num_input_tokens_seen": 263840,
      "step": 410
    },
    {
      "epoch": 7.280701754385965,
      "grad_norm": 1.5111844539642334,
      "learning_rate": 1.0566682115126344e-05,
      "loss": 4.3957,
      "num_input_tokens_seen": 267424,
      "step": 415
    },
    {
      "epoch": 7.368421052631579,
      "grad_norm": 1.4843523502349854,
      "learning_rate": 9.948511609419675e-06,
      "loss": 4.5217,
      "num_input_tokens_seen": 270720,
      "step": 420
    },
    {
      "epoch": 7.456140350877193,
      "grad_norm": 1.4406427145004272,
      "learning_rate": 9.344451868077353e-06,
      "loss": 3.904,
      "num_input_tokens_seen": 273504,
      "step": 425
    },
    {
      "epoch": 7.543859649122807,
      "grad_norm": 1.4402191638946533,
      "learning_rate": 8.755069196866014e-06,
      "loss": 4.3844,
      "num_input_tokens_seen": 276608,
      "step": 430
    },
    {
      "epoch": 7.631578947368421,
      "grad_norm": 1.8447635173797607,
      "learning_rate": 8.180916141804906e-06,
      "loss": 4.2193,
      "num_input_tokens_seen": 280352,
      "step": 435
    },
    {
      "epoch": 7.631578947368421,
      "eval_loss": 3.954362154006958,
      "eval_runtime": 0.6092,
      "eval_samples_per_second": 41.035,
      "eval_steps_per_second": 11.49,
      "num_input_tokens_seen": 280352,
      "step": 435
    },
    {
      "epoch": 7.719298245614035,
      "grad_norm": 1.0336785316467285,
      "learning_rate": 7.622530971154199e-06,
      "loss": 4.0324,
      "num_input_tokens_seen": 283808,
      "step": 440
    },
    {
      "epoch": 7.807017543859649,
      "grad_norm": 1.221795678138733,
      "learning_rate": 7.080437170788723e-06,
      "loss": 3.9307,
      "num_input_tokens_seen": 286688,
      "step": 445
    },
    {
      "epoch": 7.894736842105263,
      "grad_norm": 1.4437804222106934,
      "learning_rate": 6.555142953430158e-06,
      "loss": 3.933,
      "num_input_tokens_seen": 290240,
      "step": 450
    },
    {
      "epoch": 7.982456140350877,
      "grad_norm": 1.7613487243652344,
      "learning_rate": 6.0471407821978135e-06,
      "loss": 3.8559,
      "num_input_tokens_seen": 293568,
      "step": 455
    },
    {
      "epoch": 8.070175438596491,
      "grad_norm": 1.53519606590271,
      "learning_rate": 5.556906908924655e-06,
      "loss": 3.6867,
      "num_input_tokens_seen": 296720,
      "step": 460
    },
    {
      "epoch": 8.140350877192983,
      "eval_loss": 3.900949239730835,
      "eval_runtime": 0.6112,
      "eval_samples_per_second": 40.901,
      "eval_steps_per_second": 11.452,
      "num_input_tokens_seen": 299344,
      "step": 464
    },
    {
      "epoch": 8.157894736842104,
      "grad_norm": 1.4721977710723877,
      "learning_rate": 5.084900927671393e-06,
      "loss": 4.4772,
      "num_input_tokens_seen": 300112,
      "step": 465
    },
    {
      "epoch": 8.24561403508772,
      "grad_norm": 1.5723037719726562,
      "learning_rate": 4.631565343857239e-06,
      "loss": 4.0608,
      "num_input_tokens_seen": 303664,
      "step": 470
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 1.189079761505127,
      "learning_rate": 4.19732515941125e-06,
      "loss": 3.7438,
      "num_input_tokens_seen": 306480,
      "step": 475
    },
    {
      "epoch": 8.421052631578947,
      "grad_norm": 1.3771969079971313,
      "learning_rate": 3.7825874743331907e-06,
      "loss": 4.0063,
      "num_input_tokens_seen": 309744,
      "step": 480
    },
    {
      "epoch": 8.508771929824562,
      "grad_norm": 1.3309853076934814,
      "learning_rate": 3.3877411050374424e-06,
      "loss": 3.9324,
      "num_input_tokens_seen": 313360,
      "step": 485
    },
    {
      "epoch": 8.596491228070175,
      "grad_norm": 1.665574550628662,
      "learning_rate": 3.013156219837776e-06,
      "loss": 3.9532,
      "num_input_tokens_seen": 316784,
      "step": 490
    },
    {
      "epoch": 8.649122807017545,
      "eval_loss": 3.8729708194732666,
      "eval_runtime": 0.6067,
      "eval_samples_per_second": 41.205,
      "eval_steps_per_second": 11.537,
      "num_input_tokens_seen": 318672,
      "step": 493
    },
    {
      "epoch": 8.68421052631579,
      "grad_norm": 1.3207027912139893,
      "learning_rate": 2.659183991914696e-06,
      "loss": 3.6781,
      "num_input_tokens_seen": 319760,
      "step": 495
    },
    {
      "epoch": 8.771929824561404,
      "grad_norm": 0.9984751343727112,
      "learning_rate": 2.326156270090735e-06,
      "loss": 3.7426,
      "num_input_tokens_seen": 322512,
      "step": 500
    },
    {
      "epoch": 8.859649122807017,
      "grad_norm": 1.2388505935668945,
      "learning_rate": 2.0143852677223075e-06,
      "loss": 3.9789,
      "num_input_tokens_seen": 325936,
      "step": 505
    },
    {
      "epoch": 8.947368421052632,
      "grad_norm": 1.271156668663025,
      "learning_rate": 1.7241632699998123e-06,
      "loss": 4.0003,
      "num_input_tokens_seen": 329360,
      "step": 510
    },
    {
      "epoch": 9.035087719298245,
      "grad_norm": 1.4883043766021729,
      "learning_rate": 1.4557623599303903e-06,
      "loss": 4.2719,
      "num_input_tokens_seen": 332168,
      "step": 515
    },
    {
      "epoch": 9.12280701754386,
      "grad_norm": 1.2969716787338257,
      "learning_rate": 1.2094341632602064e-06,
      "loss": 4.1488,
      "num_input_tokens_seen": 336296,
      "step": 520
    },
    {
      "epoch": 9.157894736842104,
      "eval_loss": 3.859020948410034,
      "eval_runtime": 0.6085,
      "eval_samples_per_second": 41.082,
      "eval_steps_per_second": 11.503,
      "num_input_tokens_seen": 337480,
      "step": 522
    },
    {
      "epoch": 9.210526315789474,
      "grad_norm": 1.2461988925933838,
      "learning_rate": 9.85409612575411e-07,
      "loss": 4.0027,
      "num_input_tokens_seen": 339080,
      "step": 525
    },
    {
      "epoch": 9.298245614035087,
      "grad_norm": 1.2477507591247559,
      "learning_rate": 7.838987308029427e-07,
      "loss": 4.0184,
      "num_input_tokens_seen": 342568,
      "step": 530
    },
    {
      "epoch": 9.385964912280702,
      "grad_norm": 1.2659870386123657,
      "learning_rate": 6.050904343141095e-07,
      "loss": 3.9123,
      "num_input_tokens_seen": 345576,
      "step": 535
    },
    {
      "epoch": 9.473684210526315,
      "grad_norm": 1.4222140312194824,
      "learning_rate": 4.491523558155714e-07,
      "loss": 3.7029,
      "num_input_tokens_seen": 348392,
      "step": 540
    },
    {
      "epoch": 9.56140350877193,
      "grad_norm": 1.7531884908676147,
      "learning_rate": 3.162306871937387e-07,
      "loss": 4.0105,
      "num_input_tokens_seen": 351912,
      "step": 545
    },
    {
      "epoch": 9.649122807017545,
      "grad_norm": 1.3468551635742188,
      "learning_rate": 2.064500424599436e-07,
      "loss": 4.0229,
      "num_input_tokens_seen": 355432,
      "step": 550
    },
    {
      "epoch": 9.666666666666666,
      "eval_loss": 3.8482749462127686,
      "eval_runtime": 0.6125,
      "eval_samples_per_second": 40.816,
      "eval_steps_per_second": 11.429,
      "num_input_tokens_seen": 356456,
      "step": 551
    },
    {
      "epoch": 9.736842105263158,
      "grad_norm": 1.8819421529769897,
      "learning_rate": 1.1991334092484318e-07,
      "loss": 3.8035,
      "num_input_tokens_seen": 358760,
      "step": 555
    },
    {
      "epoch": 9.824561403508772,
      "grad_norm": 1.1540199518203735,
      "learning_rate": 5.6701710711626334e-08,
      "loss": 4.1963,
      "num_input_tokens_seen": 362088,
      "step": 560
    },
    {
      "epoch": 9.912280701754385,
      "grad_norm": 1.252097249031067,
      "learning_rate": 1.6874412698408836e-08,
      "loss": 3.7447,
      "num_input_tokens_seen": 365544,
      "step": 565
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.58130145072937,
      "learning_rate": 4.687849611939576e-10,
      "loss": 3.8419,
      "num_input_tokens_seen": 367864,
      "step": 570
    },
    {
      "epoch": 10.0,
      "num_input_tokens_seen": 367864,
      "step": 570,
      "total_flos": 1.6564749657243648e+16,
      "train_loss": 6.090754040500574,
      "train_runtime": 121.9228,
      "train_samples_per_second": 18.454,
      "train_steps_per_second": 4.675
    }
  ],
  "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": 1.6564749657243648e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}