| { | |
| "best_global_step": 145, | |
| "best_metric": 0.12023450434207916, | |
| "best_model_checkpoint": "saves_stability/prompt-tuning/llama-3-8b-instruct/train_cb_1757340265/checkpoint-145", | |
| "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": 29.0, | |
| "learning_rate": 3.5087719298245615e-06, | |
| "loss": 0.9423, | |
| "num_input_tokens_seen": 3648, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 14.25, | |
| "learning_rate": 7.894736842105263e-06, | |
| "loss": 1.0971, | |
| "num_input_tokens_seen": 6784, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 14.6875, | |
| "learning_rate": 1.2280701754385964e-05, | |
| "loss": 0.9456, | |
| "num_input_tokens_seen": 10496, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.0042, | |
| "num_input_tokens_seen": 13728, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 2.105263157894737e-05, | |
| "loss": 0.7269, | |
| "num_input_tokens_seen": 17632, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5087719298245614, | |
| "eval_loss": 0.7702681422233582, | |
| "eval_runtime": 0.8572, | |
| "eval_samples_per_second": 29.164, | |
| "eval_steps_per_second": 8.166, | |
| "num_input_tokens_seen": 19872, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 27.75, | |
| "learning_rate": 2.5438596491228074e-05, | |
| "loss": 0.7362, | |
| "num_input_tokens_seen": 20576, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6140350877192983, | |
| "grad_norm": 32.25, | |
| "learning_rate": 2.9824561403508772e-05, | |
| "loss": 0.8117, | |
| "num_input_tokens_seen": 23232, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 3.421052631578947e-05, | |
| "loss": 0.5184, | |
| "num_input_tokens_seen": 26016, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 3.859649122807018e-05, | |
| "loss": 0.331, | |
| "num_input_tokens_seen": 29344, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 47.0, | |
| "learning_rate": 4.298245614035088e-05, | |
| "loss": 0.2409, | |
| "num_input_tokens_seen": 32512, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9649122807017544, | |
| "grad_norm": 16.0, | |
| "learning_rate": 4.736842105263158e-05, | |
| "loss": 0.2065, | |
| "num_input_tokens_seen": 35168, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.0175438596491229, | |
| "eval_loss": 0.18226881325244904, | |
| "eval_runtime": 0.8467, | |
| "eval_samples_per_second": 29.527, | |
| "eval_steps_per_second": 8.268, | |
| "num_input_tokens_seen": 36432, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 85.5, | |
| "learning_rate": 4.999812487773597e-05, | |
| "loss": 0.3836, | |
| "num_input_tokens_seen": 37968, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 4.997703298253406e-05, | |
| "loss": 0.1872, | |
| "num_input_tokens_seen": 41168, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.2280701754385965, | |
| "grad_norm": 4.375, | |
| "learning_rate": 4.993252512887069e-05, | |
| "loss": 0.122, | |
| "num_input_tokens_seen": 44240, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 7.25, | |
| "learning_rate": 4.986464304284091e-05, | |
| "loss": 0.2443, | |
| "num_input_tokens_seen": 47120, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 11.0, | |
| "learning_rate": 4.977345036387331e-05, | |
| "loss": 0.2924, | |
| "num_input_tokens_seen": 49968, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 15.3125, | |
| "learning_rate": 4.965903258506806e-05, | |
| "loss": 0.1101, | |
| "num_input_tokens_seen": 52496, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "eval_loss": 0.1472374051809311, | |
| "eval_runtime": 0.8613, | |
| "eval_samples_per_second": 29.027, | |
| "eval_steps_per_second": 8.127, | |
| "num_input_tokens_seen": 53680, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 17.0, | |
| "learning_rate": 4.952149697304716e-05, | |
| "loss": 0.1742, | |
| "num_input_tokens_seen": 55504, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 19.0, | |
| "learning_rate": 4.9360972467392056e-05, | |
| "loss": 0.1738, | |
| "num_input_tokens_seen": 59504, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 16.375, | |
| "learning_rate": 4.917760955976277e-05, | |
| "loss": 0.1032, | |
| "num_input_tokens_seen": 62096, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 22.0, | |
| "learning_rate": 4.897158015281209e-05, | |
| "loss": 0.1261, | |
| "num_input_tokens_seen": 65808, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.9298245614035088, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 4.874307739902689e-05, | |
| "loss": 0.1384, | |
| "num_input_tokens_seen": 68848, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.017543859649123, | |
| "grad_norm": 24.5, | |
| "learning_rate": 4.849231551964771e-05, | |
| "loss": 0.1284, | |
| "num_input_tokens_seen": 71488, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "eval_loss": 0.12338735163211823, | |
| "eval_runtime": 0.8529, | |
| "eval_samples_per_second": 29.312, | |
| "eval_steps_per_second": 8.207, | |
| "num_input_tokens_seen": 72160, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 4.821952960383649e-05, | |
| "loss": 0.102, | |
| "num_input_tokens_seen": 74528, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.192982456140351, | |
| "grad_norm": 55.0, | |
| "learning_rate": 4.7924975388280524e-05, | |
| "loss": 0.293, | |
| "num_input_tokens_seen": 78112, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.280701754385965, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.760892901743944e-05, | |
| "loss": 0.0619, | |
| "num_input_tokens_seen": 81280, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 4.727168678465988e-05, | |
| "loss": 0.0129, | |
| "num_input_tokens_seen": 84384, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 46.75, | |
| "learning_rate": 4.6913564854400595e-05, | |
| "loss": 0.1818, | |
| "num_input_tokens_seen": 88448, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.543859649122807, | |
| "grad_norm": 16.5, | |
| "learning_rate": 4.6534898965828405e-05, | |
| "loss": 0.0319, | |
| "num_input_tokens_seen": 91904, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.543859649122807, | |
| "eval_loss": 0.12023450434207916, | |
| "eval_runtime": 0.8521, | |
| "eval_samples_per_second": 29.338, | |
| "eval_steps_per_second": 8.215, | |
| "num_input_tokens_seen": 91904, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 10.75, | |
| "learning_rate": 4.613604411806285e-05, | |
| "loss": 0.0305, | |
| "num_input_tokens_seen": 94944, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.719298245614035, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 4.5717374237364665e-05, | |
| "loss": 0.1112, | |
| "num_input_tokens_seen": 98368, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 66.0, | |
| "learning_rate": 4.5279281826580056e-05, | |
| "loss": 0.3174, | |
| "num_input_tokens_seen": 100992, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 39.25, | |
| "learning_rate": 4.482217759716946e-05, | |
| "loss": 0.2112, | |
| "num_input_tokens_seen": 103936, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.982456140350877, | |
| "grad_norm": 229.0, | |
| "learning_rate": 4.434649008416565e-05, | |
| "loss": 0.4179, | |
| "num_input_tokens_seen": 107296, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "eval_loss": 0.13938403129577637, | |
| "eval_runtime": 0.8421, | |
| "eval_samples_per_second": 29.689, | |
| "eval_steps_per_second": 8.313, | |
| "num_input_tokens_seen": 108856, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.0701754385964914, | |
| "grad_norm": 54.0, | |
| "learning_rate": 4.385266524442241e-05, | |
| "loss": 0.0899, | |
| "num_input_tokens_seen": 109720, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 4.334116603853007e-05, | |
| "loss": 0.0672, | |
| "num_input_tokens_seen": 112984, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.245614035087719, | |
| "grad_norm": 4.125, | |
| "learning_rate": 4.2812471996790206e-05, | |
| "loss": 0.0058, | |
| "num_input_tokens_seen": 116120, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 4.226707876965611e-05, | |
| "loss": 0.0599, | |
| "num_input_tokens_seen": 119192, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 24.75, | |
| "learning_rate": 4.1705497663060767e-05, | |
| "loss": 0.1797, | |
| "num_input_tokens_seen": 122840, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.5087719298245617, | |
| "grad_norm": 16.125, | |
| "learning_rate": 4.1128255159067665e-05, | |
| "loss": 0.21, | |
| "num_input_tokens_seen": 126328, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.56140350877193, | |
| "eval_loss": 0.13644804060459137, | |
| "eval_runtime": 0.8542, | |
| "eval_samples_per_second": 29.269, | |
| "eval_steps_per_second": 8.195, | |
| "num_input_tokens_seen": 128056, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.5964912280701755, | |
| "grad_norm": 0.375, | |
| "learning_rate": 4.053589242229412e-05, | |
| "loss": 0.1961, | |
| "num_input_tokens_seen": 129272, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 3.9928964792569655e-05, | |
| "loss": 0.0988, | |
| "num_input_tokens_seen": 131992, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.7719298245614032, | |
| "grad_norm": 42.75, | |
| "learning_rate": 3.930804126430513e-05, | |
| "loss": 0.1364, | |
| "num_input_tokens_seen": 135960, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.8596491228070176, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 3.867370395306068e-05, | |
| "loss": 0.2628, | |
| "num_input_tokens_seen": 139032, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 27.125, | |
| "learning_rate": 3.8026547549812665e-05, | |
| "loss": 0.1859, | |
| "num_input_tokens_seen": 142328, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.035087719298246, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 3.736717876343106e-05, | |
| "loss": 0.0244, | |
| "num_input_tokens_seen": 145096, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.0701754385964914, | |
| "eval_loss": 0.19157199561595917, | |
| "eval_runtime": 0.8508, | |
| "eval_samples_per_second": 29.384, | |
| "eval_steps_per_second": 8.228, | |
| "num_input_tokens_seen": 146952, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 4.12280701754386, | |
| "grad_norm": 111.0, | |
| "learning_rate": 3.66962157518902e-05, | |
| "loss": 0.1272, | |
| "num_input_tokens_seen": 148616, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 27.75, | |
| "learning_rate": 3.601428754274584e-05, | |
| "loss": 0.0763, | |
| "num_input_tokens_seen": 152008, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.298245614035087, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 3.532203344342212e-05, | |
| "loss": 0.0874, | |
| "num_input_tokens_seen": 155272, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 4.385964912280702, | |
| "grad_norm": 223.0, | |
| "learning_rate": 3.4620102441861143e-05, | |
| "loss": 0.1563, | |
| "num_input_tokens_seen": 158440, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 61.0, | |
| "learning_rate": 3.390915259809696e-05, | |
| "loss": 0.2761, | |
| "num_input_tokens_seen": 161192, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.56140350877193, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 3.318985042732461e-05, | |
| "loss": 0.1028, | |
| "num_input_tokens_seen": 164520, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "eval_loss": 0.18526895344257355, | |
| "eval_runtime": 0.8458, | |
| "eval_samples_per_second": 29.558, | |
| "eval_steps_per_second": 8.276, | |
| "num_input_tokens_seen": 165128, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 4.649122807017544, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 3.246287027504237e-05, | |
| "loss": 0.1062, | |
| "num_input_tokens_seen": 167336, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 19.375, | |
| "learning_rate": 3.172889368485311e-05, | |
| "loss": 0.055, | |
| "num_input_tokens_seen": 170216, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.824561403508772, | |
| "grad_norm": 4.625, | |
| "learning_rate": 3.0988608759517475e-05, | |
| "loss": 0.1101, | |
| "num_input_tokens_seen": 173192, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 4.912280701754386, | |
| "grad_norm": 0.041259765625, | |
| "learning_rate": 3.0242709515857758e-05, | |
| "loss": 0.1033, | |
| "num_input_tokens_seen": 177320, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 2.949189523411747e-05, | |
| "loss": 0.1186, | |
| "num_input_tokens_seen": 180120, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 5.087719298245614, | |
| "grad_norm": 50.25, | |
| "learning_rate": 2.8736869802386364e-05, | |
| "loss": 0.0302, | |
| "num_input_tokens_seen": 183224, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.087719298245614, | |
| "eval_loss": 0.21041154861450195, | |
| "eval_runtime": 0.8497, | |
| "eval_samples_per_second": 29.423, | |
| "eval_steps_per_second": 8.238, | |
| "num_input_tokens_seen": 183224, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.175438596491228, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 2.797834105670559e-05, | |
| "loss": 0.1013, | |
| "num_input_tokens_seen": 186264, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 5.2631578947368425, | |
| "grad_norm": 33.5, | |
| "learning_rate": 2.7217020117471793e-05, | |
| "loss": 0.1449, | |
| "num_input_tokens_seen": 189912, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.350877192982456, | |
| "grad_norm": 18.0, | |
| "learning_rate": 2.6453620722761896e-05, | |
| "loss": 0.0409, | |
| "num_input_tokens_seen": 193048, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 5.43859649122807, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 2.5688858559204053e-05, | |
| "loss": 0.2865, | |
| "num_input_tokens_seen": 196248, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.526315789473684, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 2.492345059102164e-05, | |
| "loss": 0.0024, | |
| "num_input_tokens_seen": 200152, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 5.5964912280701755, | |
| "eval_loss": 0.18615391850471497, | |
| "eval_runtime": 0.8593, | |
| "eval_samples_per_second": 29.093, | |
| "eval_steps_per_second": 8.146, | |
| "num_input_tokens_seen": 202424, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 5.614035087719298, | |
| "grad_norm": 37.5, | |
| "learning_rate": 2.4158114387879616e-05, | |
| "loss": 0.1476, | |
| "num_input_tokens_seen": 203096, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.701754385964913, | |
| "grad_norm": 49.5, | |
| "learning_rate": 2.3393567452163252e-05, | |
| "loss": 0.0759, | |
| "num_input_tokens_seen": 206328, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 5.7894736842105265, | |
| "grad_norm": 19.0, | |
| "learning_rate": 2.2630526546319914e-05, | |
| "loss": 0.015, | |
| "num_input_tokens_seen": 209080, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.87719298245614, | |
| "grad_norm": 75.0, | |
| "learning_rate": 2.186970702089457e-05, | |
| "loss": 0.0149, | |
| "num_input_tokens_seen": 212760, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 5.964912280701754, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 2.111182214388893e-05, | |
| "loss": 0.1325, | |
| "num_input_tokens_seen": 215160, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 6.052631578947368, | |
| "grad_norm": 32.25, | |
| "learning_rate": 2.0357582432072957e-05, | |
| "loss": 0.0953, | |
| "num_input_tokens_seen": 218336, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 6.105263157894737, | |
| "eval_loss": 0.21767759323120117, | |
| "eval_runtime": 0.8465, | |
| "eval_samples_per_second": 29.532, | |
| "eval_steps_per_second": 8.269, | |
| "num_input_tokens_seen": 220000, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 6.140350877192983, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 1.9607694984875754e-05, | |
| "loss": 0.0032, | |
| "num_input_tokens_seen": 221280, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.228070175438597, | |
| "grad_norm": 48.75, | |
| "learning_rate": 1.8862862821480025e-05, | |
| "loss": 0.268, | |
| "num_input_tokens_seen": 224256, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 6.315789473684211, | |
| "grad_norm": 77.5, | |
| "learning_rate": 1.8123784221741964e-05, | |
| "loss": 0.1344, | |
| "num_input_tokens_seen": 227392, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 6.4035087719298245, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 1.73911520715541e-05, | |
| "loss": 0.0021, | |
| "num_input_tokens_seen": 230432, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 6.491228070175438, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.666565321326512e-05, | |
| "loss": 0.1068, | |
| "num_input_tokens_seen": 233568, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 6.578947368421053, | |
| "grad_norm": 103.5, | |
| "learning_rate": 1.5947967801765345e-05, | |
| "loss": 0.0689, | |
| "num_input_tokens_seen": 236736, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 6.614035087719298, | |
| "eval_loss": 0.21969754993915558, | |
| "eval_runtime": 0.8481, | |
| "eval_samples_per_second": 29.479, | |
| "eval_steps_per_second": 8.254, | |
| "num_input_tokens_seen": 238272, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 1.5238768666841907e-05, | |
| "loss": 0.0025, | |
| "num_input_tokens_seen": 240288, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.754385964912281, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 1.4538720682400969e-05, | |
| "loss": 0.0008, | |
| "num_input_tokens_seen": 243392, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 6.842105263157895, | |
| "grad_norm": 113.0, | |
| "learning_rate": 1.3848480143148839e-05, | |
| "loss": 0.2647, | |
| "num_input_tokens_seen": 246496, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.9298245614035086, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 1.3168694149315796e-05, | |
| "loss": 0.0148, | |
| "num_input_tokens_seen": 250080, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 7.017543859649122, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 1.2500000000000006e-05, | |
| "loss": 0.0329, | |
| "num_input_tokens_seen": 252592, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.105263157894737, | |
| "grad_norm": 171.0, | |
| "learning_rate": 1.1843024595699805e-05, | |
| "loss": 0.2106, | |
| "num_input_tokens_seen": 255568, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 7.12280701754386, | |
| "eval_loss": 0.2411443293094635, | |
| "eval_runtime": 0.8552, | |
| "eval_samples_per_second": 29.234, | |
| "eval_steps_per_second": 8.186, | |
| "num_input_tokens_seen": 255984, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 7.192982456140351, | |
| "grad_norm": 0.12158203125, | |
| "learning_rate": 1.1198383850594758e-05, | |
| "loss": 0.0713, | |
| "num_input_tokens_seen": 258608, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 7.280701754385965, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.0566682115126344e-05, | |
| "loss": 0.0005, | |
| "num_input_tokens_seen": 262256, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 7.368421052631579, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 9.948511609419675e-06, | |
| "loss": 0.0826, | |
| "num_input_tokens_seen": 265488, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 7.456140350877193, | |
| "grad_norm": 23.5, | |
| "learning_rate": 9.344451868077353e-06, | |
| "loss": 0.1469, | |
| "num_input_tokens_seen": 268848, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 7.543859649122807, | |
| "grad_norm": 49.25, | |
| "learning_rate": 8.755069196866014e-06, | |
| "loss": 0.0301, | |
| "num_input_tokens_seen": 272016, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 7.631578947368421, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 8.180916141804906e-06, | |
| "loss": 0.006, | |
| "num_input_tokens_seen": 275536, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 7.631578947368421, | |
| "eval_loss": 0.2398126870393753, | |
| "eval_runtime": 0.8586, | |
| "eval_samples_per_second": 29.116, | |
| "eval_steps_per_second": 8.152, | |
| "num_input_tokens_seen": 275536, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 7.719298245614035, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 7.622530971154199e-06, | |
| "loss": 0.0075, | |
| "num_input_tokens_seen": 279024, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.807017543859649, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 7.080437170788723e-06, | |
| "loss": 0.0407, | |
| "num_input_tokens_seen": 281776, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 7.894736842105263, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 6.555142953430158e-06, | |
| "loss": 0.274, | |
| "num_input_tokens_seen": 284880, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 7.982456140350877, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 6.0471407821978135e-06, | |
| "loss": 0.0152, | |
| "num_input_tokens_seen": 288272, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 8.070175438596491, | |
| "grad_norm": 49.25, | |
| "learning_rate": 5.556906908924655e-06, | |
| "loss": 0.0119, | |
| "num_input_tokens_seen": 290800, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 8.140350877192983, | |
| "eval_loss": 0.24465905129909515, | |
| "eval_runtime": 0.8482, | |
| "eval_samples_per_second": 29.473, | |
| "eval_steps_per_second": 8.252, | |
| "num_input_tokens_seen": 293296, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 8.157894736842104, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 5.084900927671393e-06, | |
| "loss": 0.2666, | |
| "num_input_tokens_seen": 293712, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 8.24561403508772, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 4.631565343857239e-06, | |
| "loss": 0.003, | |
| "num_input_tokens_seen": 297680, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 4.19732515941125e-06, | |
| "loss": 0.1355, | |
| "num_input_tokens_seen": 300880, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 8.421052631578947, | |
| "grad_norm": 75.5, | |
| "learning_rate": 3.7825874743331907e-06, | |
| "loss": 0.0463, | |
| "num_input_tokens_seen": 304144, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 8.508771929824562, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 3.3877411050374424e-06, | |
| "loss": 0.0036, | |
| "num_input_tokens_seen": 307408, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 8.596491228070175, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 3.013156219837776e-06, | |
| "loss": 0.0882, | |
| "num_input_tokens_seen": 310384, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 8.649122807017545, | |
| "eval_loss": 0.23818281292915344, | |
| "eval_runtime": 0.8479, | |
| "eval_samples_per_second": 29.486, | |
| "eval_steps_per_second": 8.256, | |
| "num_input_tokens_seen": 312304, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 8.68421052631579, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 2.659183991914696e-06, | |
| "loss": 0.0367, | |
| "num_input_tokens_seen": 313360, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 8.771929824561404, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 2.326156270090735e-06, | |
| "loss": 0.0891, | |
| "num_input_tokens_seen": 316368, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.859649122807017, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 2.0143852677223075e-06, | |
| "loss": 0.0961, | |
| "num_input_tokens_seen": 319536, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 8.947368421052632, | |
| "grad_norm": 0.031005859375, | |
| "learning_rate": 1.7241632699998123e-06, | |
| "loss": 0.0286, | |
| "num_input_tokens_seen": 322928, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 9.035087719298245, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 1.4557623599303903e-06, | |
| "loss": 0.009, | |
| "num_input_tokens_seen": 325280, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 9.12280701754386, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 1.2094341632602064e-06, | |
| "loss": 0.0044, | |
| "num_input_tokens_seen": 327808, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 9.157894736842104, | |
| "eval_loss": 0.24589800834655762, | |
| "eval_runtime": 0.8628, | |
| "eval_samples_per_second": 28.974, | |
| "eval_steps_per_second": 8.113, | |
| "num_input_tokens_seen": 329216, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 9.210526315789474, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 9.85409612575411e-07, | |
| "loss": 0.2238, | |
| "num_input_tokens_seen": 331040, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 9.298245614035087, | |
| "grad_norm": 111.0, | |
| "learning_rate": 7.838987308029427e-07, | |
| "loss": 0.0863, | |
| "num_input_tokens_seen": 333824, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 9.385964912280702, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 6.050904343141095e-07, | |
| "loss": 0.1187, | |
| "num_input_tokens_seen": 336192, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 9.473684210526315, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 4.491523558155714e-07, | |
| "loss": 0.0226, | |
| "num_input_tokens_seen": 340224, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 9.56140350877193, | |
| "grad_norm": 0.06982421875, | |
| "learning_rate": 3.162306871937387e-07, | |
| "loss": 0.0135, | |
| "num_input_tokens_seen": 343008, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 9.649122807017545, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 2.064500424599436e-07, | |
| "loss": 0.0294, | |
| "num_input_tokens_seen": 346240, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 9.666666666666666, | |
| "eval_loss": 0.24775297939777374, | |
| "eval_runtime": 0.8551, | |
| "eval_samples_per_second": 29.236, | |
| "eval_steps_per_second": 8.186, | |
| "num_input_tokens_seen": 346944, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 9.736842105263158, | |
| "grad_norm": 160.0, | |
| "learning_rate": 1.1991334092484318e-07, | |
| "loss": 0.2084, | |
| "num_input_tokens_seen": 349472, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 9.824561403508772, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 5.6701710711626334e-08, | |
| "loss": 0.0398, | |
| "num_input_tokens_seen": 353088, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.912280701754385, | |
| "grad_norm": 38.25, | |
| "learning_rate": 1.6874412698408836e-08, | |
| "loss": 0.0542, | |
| "num_input_tokens_seen": 357088, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.041015625, | |
| "learning_rate": 4.687849611939576e-10, | |
| "loss": 0.0213, | |
| "num_input_tokens_seen": 359824, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "num_input_tokens_seen": 359824, | |
| "step": 570, | |
| "total_flos": 1.6202712091066368e+16, | |
| "train_loss": 0.1654530040347916, | |
| "train_runtime": 162.5762, | |
| "train_samples_per_second": 13.84, | |
| "train_steps_per_second": 3.506 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 570, | |
| "num_input_tokens_seen": 359824, | |
| "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": 1.6202712091066368e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |