| { | |
| "best_global_step": 464, | |
| "best_metric": 0.14825789630413055, | |
| "best_model_checkpoint": "saves/ia3/llama-3-8b-instruct/train_cb_1756128976/checkpoint-464", | |
| "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": 3.2857978343963623, | |
| "learning_rate": 3.5087719298245615e-06, | |
| "loss": 1.1877, | |
| "num_input_tokens_seen": 3552, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 2.826841354370117, | |
| "learning_rate": 7.894736842105263e-06, | |
| "loss": 1.309, | |
| "num_input_tokens_seen": 7264, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 2.9366278648376465, | |
| "learning_rate": 1.2280701754385964e-05, | |
| "loss": 1.1793, | |
| "num_input_tokens_seen": 10528, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 2.8773298263549805, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.0865, | |
| "num_input_tokens_seen": 14720, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 3.032203435897827, | |
| "learning_rate": 2.105263157894737e-05, | |
| "loss": 1.2393, | |
| "num_input_tokens_seen": 18016, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5087719298245614, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 1.1732473373413086, | |
| "eval_runtime": 0.604, | |
| "eval_samples_per_second": 41.394, | |
| "eval_steps_per_second": 11.59, | |
| "num_input_tokens_seen": 20064, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 2.570289373397827, | |
| "learning_rate": 2.5438596491228074e-05, | |
| "loss": 1.3136, | |
| "num_input_tokens_seen": 20640, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6140350877192983, | |
| "grad_norm": 3.3799400329589844, | |
| "learning_rate": 2.9824561403508772e-05, | |
| "loss": 1.1864, | |
| "num_input_tokens_seen": 24800, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 3.0833399295806885, | |
| "learning_rate": 3.421052631578947e-05, | |
| "loss": 1.1892, | |
| "num_input_tokens_seen": 28064, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 2.591188430786133, | |
| "learning_rate": 3.859649122807018e-05, | |
| "loss": 1.3389, | |
| "num_input_tokens_seen": 30944, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 3.4618539810180664, | |
| "learning_rate": 4.298245614035088e-05, | |
| "loss": 1.1066, | |
| "num_input_tokens_seen": 33664, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9649122807017544, | |
| "grad_norm": 2.843844175338745, | |
| "learning_rate": 4.736842105263158e-05, | |
| "loss": 1.196, | |
| "num_input_tokens_seen": 36320, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.0175438596491229, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 1.1732473373413086, | |
| "eval_runtime": 0.5781, | |
| "eval_samples_per_second": 43.244, | |
| "eval_steps_per_second": 12.108, | |
| "num_input_tokens_seen": 37832, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 3.0758798122406006, | |
| "learning_rate": 4.999812487773597e-05, | |
| "loss": 0.9704, | |
| "num_input_tokens_seen": 39080, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 3.7676706314086914, | |
| "learning_rate": 4.997703298253406e-05, | |
| "loss": 1.0932, | |
| "num_input_tokens_seen": 42536, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.2280701754385965, | |
| "grad_norm": 2.6574699878692627, | |
| "learning_rate": 4.993252512887069e-05, | |
| "loss": 1.1462, | |
| "num_input_tokens_seen": 45608, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 2.809377431869507, | |
| "learning_rate": 4.986464304284091e-05, | |
| "loss": 1.0703, | |
| "num_input_tokens_seen": 49352, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 3.8227696418762207, | |
| "learning_rate": 4.977345036387331e-05, | |
| "loss": 1.0132, | |
| "num_input_tokens_seen": 52328, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 3.1258037090301514, | |
| "learning_rate": 4.965903258506806e-05, | |
| "loss": 0.6503, | |
| "num_input_tokens_seen": 56328, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "eval_accuracy": 0.81, | |
| "eval_loss": 0.559042751789093, | |
| "eval_runtime": 0.5844, | |
| "eval_samples_per_second": 42.776, | |
| "eval_steps_per_second": 11.977, | |
| "num_input_tokens_seen": 57288, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 2.510925769805908, | |
| "learning_rate": 4.952149697304716e-05, | |
| "loss": 0.6947, | |
| "num_input_tokens_seen": 59048, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 2.0533506870269775, | |
| "learning_rate": 4.9360972467392056e-05, | |
| "loss": 0.3543, | |
| "num_input_tokens_seen": 62504, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 1.303076982498169, | |
| "learning_rate": 4.917760955976277e-05, | |
| "loss": 0.3442, | |
| "num_input_tokens_seen": 65832, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 1.217045545578003, | |
| "learning_rate": 4.897158015281209e-05, | |
| "loss": 0.3672, | |
| "num_input_tokens_seen": 68808, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.9298245614035088, | |
| "grad_norm": 0.27959978580474854, | |
| "learning_rate": 4.874307739902689e-05, | |
| "loss": 0.194, | |
| "num_input_tokens_seen": 71848, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.017543859649123, | |
| "grad_norm": 1.2869893312454224, | |
| "learning_rate": 4.849231551964771e-05, | |
| "loss": 0.225, | |
| "num_input_tokens_seen": 74040, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "eval_accuracy": 0.91, | |
| "eval_loss": 0.18880803883075714, | |
| "eval_runtime": 0.5923, | |
| "eval_samples_per_second": 42.212, | |
| "eval_steps_per_second": 11.819, | |
| "num_input_tokens_seen": 74520, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 1.2986481189727783, | |
| "learning_rate": 4.821952960383649e-05, | |
| "loss": 0.0831, | |
| "num_input_tokens_seen": 77272, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.192982456140351, | |
| "grad_norm": 0.49514040350914, | |
| "learning_rate": 4.7924975388280524e-05, | |
| "loss": 0.1083, | |
| "num_input_tokens_seen": 80280, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.280701754385965, | |
| "grad_norm": 0.2440158575773239, | |
| "learning_rate": 4.760892901743944e-05, | |
| "loss": 0.0922, | |
| "num_input_tokens_seen": 83480, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.6071472764015198, | |
| "learning_rate": 4.727168678465988e-05, | |
| "loss": 0.1993, | |
| "num_input_tokens_seen": 86232, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 1.3904006481170654, | |
| "learning_rate": 4.6913564854400595e-05, | |
| "loss": 0.153, | |
| "num_input_tokens_seen": 89656, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.543859649122807, | |
| "grad_norm": 0.17373153567314148, | |
| "learning_rate": 4.6534898965828405e-05, | |
| "loss": 0.568, | |
| "num_input_tokens_seen": 93080, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.543859649122807, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.1757977306842804, | |
| "eval_runtime": 0.5644, | |
| "eval_samples_per_second": 44.294, | |
| "eval_steps_per_second": 12.402, | |
| "num_input_tokens_seen": 93080, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 1.6921156644821167, | |
| "learning_rate": 4.613604411806285e-05, | |
| "loss": 0.1949, | |
| "num_input_tokens_seen": 96440, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.719298245614035, | |
| "grad_norm": 2.1323416233062744, | |
| "learning_rate": 4.5717374237364665e-05, | |
| "loss": 0.1408, | |
| "num_input_tokens_seen": 100280, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 0.32603445649147034, | |
| "learning_rate": 4.5279281826580056e-05, | |
| "loss": 0.1741, | |
| "num_input_tokens_seen": 103512, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.8819119334220886, | |
| "learning_rate": 4.482217759716946e-05, | |
| "loss": 0.1908, | |
| "num_input_tokens_seen": 106168, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.982456140350877, | |
| "grad_norm": 0.6695258021354675, | |
| "learning_rate": 4.434649008416565e-05, | |
| "loss": 0.1509, | |
| "num_input_tokens_seen": 109624, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "eval_accuracy": 0.93, | |
| "eval_loss": 0.16153018176555634, | |
| "eval_runtime": 0.5197, | |
| "eval_samples_per_second": 48.107, | |
| "eval_steps_per_second": 13.47, | |
| "num_input_tokens_seen": 111928, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.0701754385964914, | |
| "grad_norm": 1.8589503765106201, | |
| "learning_rate": 4.385266524442241e-05, | |
| "loss": 0.1831, | |
| "num_input_tokens_seen": 112472, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.9382370710372925, | |
| "learning_rate": 4.334116603853007e-05, | |
| "loss": 0.0704, | |
| "num_input_tokens_seen": 115576, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.245614035087719, | |
| "grad_norm": 0.520358145236969, | |
| "learning_rate": 4.2812471996790206e-05, | |
| "loss": 0.0847, | |
| "num_input_tokens_seen": 119000, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 1.6390992403030396, | |
| "learning_rate": 4.226707876965611e-05, | |
| "loss": 0.0842, | |
| "num_input_tokens_seen": 122360, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.11116215586662292, | |
| "learning_rate": 4.1705497663060767e-05, | |
| "loss": 0.2927, | |
| "num_input_tokens_seen": 125496, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.5087719298245617, | |
| "grad_norm": 1.5918360948562622, | |
| "learning_rate": 4.1128255159067665e-05, | |
| "loss": 0.3084, | |
| "num_input_tokens_seen": 128760, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.56140350877193, | |
| "eval_accuracy": 0.93, | |
| "eval_loss": 0.1591552346944809, | |
| "eval_runtime": 0.5662, | |
| "eval_samples_per_second": 44.158, | |
| "eval_steps_per_second": 12.364, | |
| "num_input_tokens_seen": 131160, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.5964912280701755, | |
| "grad_norm": 0.6156125664710999, | |
| "learning_rate": 4.053589242229412e-05, | |
| "loss": 0.117, | |
| "num_input_tokens_seen": 132248, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 1.5829564332962036, | |
| "learning_rate": 3.9928964792569655e-05, | |
| "loss": 0.0982, | |
| "num_input_tokens_seen": 135160, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.7719298245614032, | |
| "grad_norm": 0.23042283952236176, | |
| "learning_rate": 3.930804126430513e-05, | |
| "loss": 0.3142, | |
| "num_input_tokens_seen": 138200, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.8596491228070176, | |
| "grad_norm": 0.9264437556266785, | |
| "learning_rate": 3.867370395306068e-05, | |
| "loss": 0.0683, | |
| "num_input_tokens_seen": 141752, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 1.1045851707458496, | |
| "learning_rate": 3.8026547549812665e-05, | |
| "loss": 0.1347, | |
| "num_input_tokens_seen": 146008, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.035087719298246, | |
| "grad_norm": 0.13252288103103638, | |
| "learning_rate": 3.736717876343106e-05, | |
| "loss": 0.1682, | |
| "num_input_tokens_seen": 148648, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.0701754385964914, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.16104494035243988, | |
| "eval_runtime": 0.5441, | |
| "eval_samples_per_second": 45.949, | |
| "eval_steps_per_second": 12.866, | |
| "num_input_tokens_seen": 150056, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 4.12280701754386, | |
| "grad_norm": 1.2706420421600342, | |
| "learning_rate": 3.66962157518902e-05, | |
| "loss": 0.1885, | |
| "num_input_tokens_seen": 151656, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 2.8882739543914795, | |
| "learning_rate": 3.601428754274584e-05, | |
| "loss": 0.3189, | |
| "num_input_tokens_seen": 154280, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.298245614035087, | |
| "grad_norm": 0.9904075264930725, | |
| "learning_rate": 3.532203344342212e-05, | |
| "loss": 0.2102, | |
| "num_input_tokens_seen": 157160, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 4.385964912280702, | |
| "grad_norm": 1.374855875968933, | |
| "learning_rate": 3.4620102441861143e-05, | |
| "loss": 0.1598, | |
| "num_input_tokens_seen": 160072, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.5782589912414551, | |
| "learning_rate": 3.390915259809696e-05, | |
| "loss": 0.0827, | |
| "num_input_tokens_seen": 163752, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.56140350877193, | |
| "grad_norm": 0.5617624521255493, | |
| "learning_rate": 3.318985042732461e-05, | |
| "loss": 0.0562, | |
| "num_input_tokens_seen": 166600, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.16242551803588867, | |
| "eval_runtime": 0.5591, | |
| "eval_samples_per_second": 44.718, | |
| "eval_steps_per_second": 12.521, | |
| "num_input_tokens_seen": 167208, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 4.649122807017544, | |
| "grad_norm": 0.22714340686798096, | |
| "learning_rate": 3.246287027504237e-05, | |
| "loss": 0.1684, | |
| "num_input_tokens_seen": 169704, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 2.7295970916748047, | |
| "learning_rate": 3.172889368485311e-05, | |
| "loss": 0.1376, | |
| "num_input_tokens_seen": 173000, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.824561403508772, | |
| "grad_norm": 0.5690297484397888, | |
| "learning_rate": 3.0988608759517475e-05, | |
| "loss": 0.1687, | |
| "num_input_tokens_seen": 177128, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 4.912280701754386, | |
| "grad_norm": 0.8604775071144104, | |
| "learning_rate": 3.0242709515857758e-05, | |
| "loss": 0.0827, | |
| "num_input_tokens_seen": 180680, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.012729275971651077, | |
| "learning_rate": 2.949189523411747e-05, | |
| "loss": 0.0615, | |
| "num_input_tokens_seen": 183280, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 5.087719298245614, | |
| "grad_norm": 0.9690691232681274, | |
| "learning_rate": 2.8736869802386364e-05, | |
| "loss": 0.096, | |
| "num_input_tokens_seen": 186160, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.087719298245614, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.15590623021125793, | |
| "eval_runtime": 0.585, | |
| "eval_samples_per_second": 42.735, | |
| "eval_steps_per_second": 11.966, | |
| "num_input_tokens_seen": 186160, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.175438596491228, | |
| "grad_norm": 1.4275689125061035, | |
| "learning_rate": 2.797834105670559e-05, | |
| "loss": 0.1093, | |
| "num_input_tokens_seen": 189584, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 5.2631578947368425, | |
| "grad_norm": 1.4390819072723389, | |
| "learning_rate": 2.7217020117471793e-05, | |
| "loss": 0.157, | |
| "num_input_tokens_seen": 193296, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.350877192982456, | |
| "grad_norm": 1.3548004627227783, | |
| "learning_rate": 2.6453620722761896e-05, | |
| "loss": 0.16, | |
| "num_input_tokens_seen": 197328, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 5.43859649122807, | |
| "grad_norm": 0.6790382266044617, | |
| "learning_rate": 2.5688858559204053e-05, | |
| "loss": 0.0681, | |
| "num_input_tokens_seen": 200560, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.526315789473684, | |
| "grad_norm": 0.7437301278114319, | |
| "learning_rate": 2.492345059102164e-05, | |
| "loss": 0.0737, | |
| "num_input_tokens_seen": 203632, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 5.5964912280701755, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.15260252356529236, | |
| "eval_runtime": 0.6032, | |
| "eval_samples_per_second": 41.449, | |
| "eval_steps_per_second": 11.606, | |
| "num_input_tokens_seen": 206000, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 5.614035087719298, | |
| "grad_norm": 1.7895355224609375, | |
| "learning_rate": 2.4158114387879616e-05, | |
| "loss": 0.2497, | |
| "num_input_tokens_seen": 206480, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.701754385964913, | |
| "grad_norm": 1.9388424158096313, | |
| "learning_rate": 2.3393567452163252e-05, | |
| "loss": 0.2984, | |
| "num_input_tokens_seen": 209232, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 5.7894736842105265, | |
| "grad_norm": 1.203923225402832, | |
| "learning_rate": 2.2630526546319914e-05, | |
| "loss": 0.1651, | |
| "num_input_tokens_seen": 213168, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.87719298245614, | |
| "grad_norm": 1.3108657598495483, | |
| "learning_rate": 2.186970702089457e-05, | |
| "loss": 0.1575, | |
| "num_input_tokens_seen": 216752, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 5.964912280701754, | |
| "grad_norm": 0.2699100375175476, | |
| "learning_rate": 2.111182214388893e-05, | |
| "loss": 0.0937, | |
| "num_input_tokens_seen": 219536, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 6.052631578947368, | |
| "grad_norm": 0.17759937047958374, | |
| "learning_rate": 2.0357582432072957e-05, | |
| "loss": 0.1728, | |
| "num_input_tokens_seen": 222272, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 6.105263157894737, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.15527071058750153, | |
| "eval_runtime": 0.5492, | |
| "eval_samples_per_second": 45.524, | |
| "eval_steps_per_second": 12.747, | |
| "num_input_tokens_seen": 224064, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 6.140350877192983, | |
| "grad_norm": 0.5311676859855652, | |
| "learning_rate": 1.9607694984875754e-05, | |
| "loss": 0.1903, | |
| "num_input_tokens_seen": 225664, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.228070175438597, | |
| "grad_norm": 1.297541856765747, | |
| "learning_rate": 1.8862862821480025e-05, | |
| "loss": 0.1822, | |
| "num_input_tokens_seen": 229408, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 6.315789473684211, | |
| "grad_norm": 0.9657095074653625, | |
| "learning_rate": 1.8123784221741964e-05, | |
| "loss": 0.2763, | |
| "num_input_tokens_seen": 232832, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 6.4035087719298245, | |
| "grad_norm": 0.640754759311676, | |
| "learning_rate": 1.73911520715541e-05, | |
| "loss": 0.1021, | |
| "num_input_tokens_seen": 236288, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 6.491228070175438, | |
| "grad_norm": 0.4857073724269867, | |
| "learning_rate": 1.666565321326512e-05, | |
| "loss": 0.0561, | |
| "num_input_tokens_seen": 239296, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 6.578947368421053, | |
| "grad_norm": 1.2188024520874023, | |
| "learning_rate": 1.5947967801765345e-05, | |
| "loss": 0.1545, | |
| "num_input_tokens_seen": 242848, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 6.614035087719298, | |
| "eval_accuracy": 0.93, | |
| "eval_loss": 0.15140964090824127, | |
| "eval_runtime": 0.5211, | |
| "eval_samples_per_second": 47.974, | |
| "eval_steps_per_second": 13.433, | |
| "num_input_tokens_seen": 243840, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.3681733012199402, | |
| "learning_rate": 1.5238768666841907e-05, | |
| "loss": 0.1936, | |
| "num_input_tokens_seen": 245344, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.754385964912281, | |
| "grad_norm": 0.35686439275741577, | |
| "learning_rate": 1.4538720682400969e-05, | |
| "loss": 0.0681, | |
| "num_input_tokens_seen": 248256, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 6.842105263157895, | |
| "grad_norm": 0.5755576491355896, | |
| "learning_rate": 1.3848480143148839e-05, | |
| "loss": 0.0965, | |
| "num_input_tokens_seen": 250976, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.9298245614035086, | |
| "grad_norm": 0.5823020935058594, | |
| "learning_rate": 1.3168694149315796e-05, | |
| "loss": 0.0593, | |
| "num_input_tokens_seen": 254944, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 7.017543859649122, | |
| "grad_norm": 0.6067392230033875, | |
| "learning_rate": 1.2500000000000006e-05, | |
| "loss": 0.0929, | |
| "num_input_tokens_seen": 257888, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.105263157894737, | |
| "grad_norm": 0.11550065875053406, | |
| "learning_rate": 1.1843024595699805e-05, | |
| "loss": 0.0507, | |
| "num_input_tokens_seen": 261088, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 7.12280701754386, | |
| "eval_accuracy": 0.93, | |
| "eval_loss": 0.1483900249004364, | |
| "eval_runtime": 0.5416, | |
| "eval_samples_per_second": 46.156, | |
| "eval_steps_per_second": 12.924, | |
| "num_input_tokens_seen": 261504, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 7.192982456140351, | |
| "grad_norm": 0.35420212149620056, | |
| "learning_rate": 1.1198383850594758e-05, | |
| "loss": 0.0565, | |
| "num_input_tokens_seen": 263840, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 7.280701754385965, | |
| "grad_norm": 2.2538681030273438, | |
| "learning_rate": 1.0566682115126344e-05, | |
| "loss": 0.2486, | |
| "num_input_tokens_seen": 267424, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 7.368421052631579, | |
| "grad_norm": 1.0232347249984741, | |
| "learning_rate": 9.948511609419675e-06, | |
| "loss": 0.2373, | |
| "num_input_tokens_seen": 270720, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 7.456140350877193, | |
| "grad_norm": 1.3066163063049316, | |
| "learning_rate": 9.344451868077353e-06, | |
| "loss": 0.2271, | |
| "num_input_tokens_seen": 273504, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 7.543859649122807, | |
| "grad_norm": 0.8171947002410889, | |
| "learning_rate": 8.755069196866014e-06, | |
| "loss": 0.1129, | |
| "num_input_tokens_seen": 276608, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 7.631578947368421, | |
| "grad_norm": 0.3773968815803528, | |
| "learning_rate": 8.180916141804906e-06, | |
| "loss": 0.166, | |
| "num_input_tokens_seen": 280352, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 7.631578947368421, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.1568061262369156, | |
| "eval_runtime": 0.5702, | |
| "eval_samples_per_second": 43.846, | |
| "eval_steps_per_second": 12.277, | |
| "num_input_tokens_seen": 280352, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 7.719298245614035, | |
| "grad_norm": 0.7733727693557739, | |
| "learning_rate": 7.622530971154199e-06, | |
| "loss": 0.0905, | |
| "num_input_tokens_seen": 283808, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.807017543859649, | |
| "grad_norm": 1.2004691362380981, | |
| "learning_rate": 7.080437170788723e-06, | |
| "loss": 0.1109, | |
| "num_input_tokens_seen": 286688, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 7.894736842105263, | |
| "grad_norm": 0.672938883304596, | |
| "learning_rate": 6.555142953430158e-06, | |
| "loss": 0.164, | |
| "num_input_tokens_seen": 290240, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 7.982456140350877, | |
| "grad_norm": 0.6395520567893982, | |
| "learning_rate": 6.0471407821978135e-06, | |
| "loss": 0.1285, | |
| "num_input_tokens_seen": 293568, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 8.070175438596491, | |
| "grad_norm": 0.10756539553403854, | |
| "learning_rate": 5.556906908924655e-06, | |
| "loss": 0.0494, | |
| "num_input_tokens_seen": 296720, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 8.140350877192983, | |
| "eval_accuracy": 0.92, | |
| "eval_loss": 0.14825789630413055, | |
| "eval_runtime": 0.5287, | |
| "eval_samples_per_second": 47.284, | |
| "eval_steps_per_second": 13.239, | |
| "num_input_tokens_seen": 299344, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 8.157894736842104, | |
| "grad_norm": 0.36664772033691406, | |
| "learning_rate": 5.084900927671393e-06, | |
| "loss": 0.1731, | |
| "num_input_tokens_seen": 300112, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 8.24561403508772, | |
| "grad_norm": 2.390585422515869, | |
| "learning_rate": 4.631565343857239e-06, | |
| "loss": 0.1065, | |
| "num_input_tokens_seen": 303664, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 0.45317980647087097, | |
| "learning_rate": 4.19732515941125e-06, | |
| "loss": 0.1003, | |
| "num_input_tokens_seen": 306480, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 8.421052631578947, | |
| "grad_norm": 1.0741758346557617, | |
| "learning_rate": 3.7825874743331907e-06, | |
| "loss": 0.1752, | |
| "num_input_tokens_seen": 309744, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 8.508771929824562, | |
| "grad_norm": 1.1840258836746216, | |
| "learning_rate": 3.3877411050374424e-06, | |
| "loss": 0.1333, | |
| "num_input_tokens_seen": 313360, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 8.596491228070175, | |
| "grad_norm": 1.450769305229187, | |
| "learning_rate": 3.013156219837776e-06, | |
| "loss": 0.3338, | |
| "num_input_tokens_seen": 316784, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 8.649122807017545, | |
| "eval_accuracy": 0.93, | |
| "eval_loss": 0.1508120447397232, | |
| "eval_runtime": 0.5683, | |
| "eval_samples_per_second": 43.988, | |
| "eval_steps_per_second": 12.317, | |
| "num_input_tokens_seen": 318672, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 8.68421052631579, | |
| "grad_norm": 0.23999197781085968, | |
| "learning_rate": 2.659183991914696e-06, | |
| "loss": 0.0512, | |
| "num_input_tokens_seen": 319760, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 8.771929824561404, | |
| "grad_norm": 0.13647691905498505, | |
| "learning_rate": 2.326156270090735e-06, | |
| "loss": 0.0747, | |
| "num_input_tokens_seen": 322512, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.859649122807017, | |
| "grad_norm": 0.21567077934741974, | |
| "learning_rate": 2.0143852677223075e-06, | |
| "loss": 0.098, | |
| "num_input_tokens_seen": 325936, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 8.947368421052632, | |
| "grad_norm": 0.8078049421310425, | |
| "learning_rate": 1.7241632699998123e-06, | |
| "loss": 0.1179, | |
| "num_input_tokens_seen": 329360, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 9.035087719298245, | |
| "grad_norm": 0.7871196269989014, | |
| "learning_rate": 1.4557623599303903e-06, | |
| "loss": 0.187, | |
| "num_input_tokens_seen": 332168, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 9.12280701754386, | |
| "grad_norm": 0.29638534784317017, | |
| "learning_rate": 1.2094341632602064e-06, | |
| "loss": 0.0833, | |
| "num_input_tokens_seen": 336296, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 9.157894736842104, | |
| "eval_accuracy": 0.93, | |
| "eval_loss": 0.1507260799407959, | |
| "eval_runtime": 0.5932, | |
| "eval_samples_per_second": 42.148, | |
| "eval_steps_per_second": 11.801, | |
| "num_input_tokens_seen": 337480, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 9.210526315789474, | |
| "grad_norm": 0.34370607137680054, | |
| "learning_rate": 9.85409612575411e-07, | |
| "loss": 0.2548, | |
| "num_input_tokens_seen": 339080, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 9.298245614035087, | |
| "grad_norm": 0.09956851601600647, | |
| "learning_rate": 7.838987308029427e-07, | |
| "loss": 0.0774, | |
| "num_input_tokens_seen": 342568, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 9.385964912280702, | |
| "grad_norm": 0.7313379645347595, | |
| "learning_rate": 6.050904343141095e-07, | |
| "loss": 0.1128, | |
| "num_input_tokens_seen": 345576, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 9.473684210526315, | |
| "grad_norm": 0.40931132435798645, | |
| "learning_rate": 4.491523558155714e-07, | |
| "loss": 0.0578, | |
| "num_input_tokens_seen": 348392, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 9.56140350877193, | |
| "grad_norm": 0.5584717988967896, | |
| "learning_rate": 3.162306871937387e-07, | |
| "loss": 0.0793, | |
| "num_input_tokens_seen": 351912, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 9.649122807017545, | |
| "grad_norm": 0.6730838418006897, | |
| "learning_rate": 2.064500424599436e-07, | |
| "loss": 0.1172, | |
| "num_input_tokens_seen": 355432, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 9.666666666666666, | |
| "eval_accuracy": 0.93, | |
| "eval_loss": 0.14835186302661896, | |
| "eval_runtime": 0.536, | |
| "eval_samples_per_second": 46.646, | |
| "eval_steps_per_second": 13.061, | |
| "num_input_tokens_seen": 356456, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 9.736842105263158, | |
| "grad_norm": 1.4842658042907715, | |
| "learning_rate": 1.1991334092484318e-07, | |
| "loss": 0.1617, | |
| "num_input_tokens_seen": 358760, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 9.824561403508772, | |
| "grad_norm": 0.08213107287883759, | |
| "learning_rate": 5.6701710711626334e-08, | |
| "loss": 0.23, | |
| "num_input_tokens_seen": 362088, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.912280701754385, | |
| "grad_norm": 0.4282042682170868, | |
| "learning_rate": 1.6874412698408836e-08, | |
| "loss": 0.2394, | |
| "num_input_tokens_seen": 365544, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.19724462926387787, | |
| "learning_rate": 4.687849611939576e-10, | |
| "loss": 0.1568, | |
| "num_input_tokens_seen": 367864, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "num_input_tokens_seen": 367864, | |
| "step": 570, | |
| "total_flos": 1.656518360727552e+16, | |
| "train_loss": 0.30795817226171496, | |
| "train_runtime": 112.7666, | |
| "train_samples_per_second": 19.953, | |
| "train_steps_per_second": 5.055 | |
| } | |
| ], | |
| "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 | |
| } | |