| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 47.61904761904762, | |
| "eval_steps": 500, | |
| "global_step": 250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.32970941066741943, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.7065, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.34144940972328186, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.5961, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.3820774257183075, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.5497, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.359970360994339, | |
| "learning_rate": 0.00012800000000000002, | |
| "loss": 1.4212, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 0.3855084478855133, | |
| "learning_rate": 0.00016, | |
| "loss": 1.3231, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 0.28801724314689636, | |
| "learning_rate": 0.000192, | |
| "loss": 1.2461, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 0.34446007013320923, | |
| "learning_rate": 0.00019733333333333335, | |
| "loss": 1.1191, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.3313669264316559, | |
| "learning_rate": 0.0001937777777777778, | |
| "loss": 1.119, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "grad_norm": 0.2997862696647644, | |
| "learning_rate": 0.00019022222222222224, | |
| "loss": 1.0179, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "grad_norm": 0.2927868962287903, | |
| "learning_rate": 0.0001866666666666667, | |
| "loss": 0.9294, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 8.38, | |
| "grad_norm": 0.303634911775589, | |
| "learning_rate": 0.00018311111111111113, | |
| "loss": 0.8856, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 9.14, | |
| "grad_norm": 0.4033955931663513, | |
| "learning_rate": 0.00017955555555555558, | |
| "loss": 0.86, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.42172765731811523, | |
| "learning_rate": 0.00017600000000000002, | |
| "loss": 0.7868, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "grad_norm": 0.3995470702648163, | |
| "learning_rate": 0.00017244444444444444, | |
| "loss": 0.7428, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 11.43, | |
| "grad_norm": 0.4886798560619354, | |
| "learning_rate": 0.00016888888888888889, | |
| "loss": 0.6497, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 12.19, | |
| "grad_norm": 0.8596562743186951, | |
| "learning_rate": 0.00016533333333333333, | |
| "loss": 0.6389, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.95, | |
| "grad_norm": 5.950181007385254, | |
| "learning_rate": 0.00016177777777777778, | |
| "loss": 0.525, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 13.71, | |
| "grad_norm": 2.282550096511841, | |
| "learning_rate": 0.00015822222222222222, | |
| "loss": 0.447, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 14.48, | |
| "grad_norm": 0.877649188041687, | |
| "learning_rate": 0.00015466666666666667, | |
| "loss": 0.4354, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 15.24, | |
| "grad_norm": 0.941230833530426, | |
| "learning_rate": 0.0001511111111111111, | |
| "loss": 0.3089, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.8986572027206421, | |
| "learning_rate": 0.00014755555555555556, | |
| "loss": 0.3118, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 16.76, | |
| "grad_norm": 0.9937779307365417, | |
| "learning_rate": 0.000144, | |
| "loss": 0.2487, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 17.52, | |
| "grad_norm": 0.9299382567405701, | |
| "learning_rate": 0.00014044444444444445, | |
| "loss": 0.2026, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 18.29, | |
| "grad_norm": 1.0898679494857788, | |
| "learning_rate": 0.0001368888888888889, | |
| "loss": 0.1691, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 19.05, | |
| "grad_norm": 1.1475147008895874, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.1497, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 19.81, | |
| "grad_norm": 1.1220810413360596, | |
| "learning_rate": 0.00012977777777777779, | |
| "loss": 0.0984, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 20.57, | |
| "grad_norm": 1.19789457321167, | |
| "learning_rate": 0.00012622222222222223, | |
| "loss": 0.1001, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 21.33, | |
| "grad_norm": 0.8793210983276367, | |
| "learning_rate": 0.00012266666666666668, | |
| "loss": 0.072, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 22.1, | |
| "grad_norm": 0.9120911955833435, | |
| "learning_rate": 0.00011911111111111111, | |
| "loss": 0.0509, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 22.86, | |
| "grad_norm": 0.7733311653137207, | |
| "learning_rate": 0.00011555555555555555, | |
| "loss": 0.0429, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 23.62, | |
| "grad_norm": 0.6906972527503967, | |
| "learning_rate": 0.00011200000000000001, | |
| "loss": 0.0377, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 24.38, | |
| "grad_norm": 0.3450298607349396, | |
| "learning_rate": 0.00010844444444444446, | |
| "loss": 0.0271, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 25.14, | |
| "grad_norm": 0.40086209774017334, | |
| "learning_rate": 0.0001048888888888889, | |
| "loss": 0.0264, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 25.9, | |
| "grad_norm": 0.7334154844284058, | |
| "learning_rate": 0.00010133333333333335, | |
| "loss": 0.0212, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "grad_norm": 0.2674214243888855, | |
| "learning_rate": 9.777777777777778e-05, | |
| "loss": 0.0169, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 27.43, | |
| "grad_norm": 0.2615182399749756, | |
| "learning_rate": 9.422222222222223e-05, | |
| "loss": 0.0173, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 28.19, | |
| "grad_norm": 0.12926605343818665, | |
| "learning_rate": 9.066666666666667e-05, | |
| "loss": 0.0149, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 28.95, | |
| "grad_norm": 0.1451052874326706, | |
| "learning_rate": 8.711111111111112e-05, | |
| "loss": 0.0142, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 29.71, | |
| "grad_norm": 0.12337276339530945, | |
| "learning_rate": 8.355555555555556e-05, | |
| "loss": 0.0111, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 30.48, | |
| "grad_norm": 0.10490886121988297, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0126, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 31.24, | |
| "grad_norm": 0.12196756899356842, | |
| "learning_rate": 7.644444444444445e-05, | |
| "loss": 0.01, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.06647361814975739, | |
| "learning_rate": 7.28888888888889e-05, | |
| "loss": 0.0106, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 32.76, | |
| "grad_norm": 0.09191016852855682, | |
| "learning_rate": 6.933333333333334e-05, | |
| "loss": 0.0084, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 33.52, | |
| "grad_norm": 0.08465249091386795, | |
| "learning_rate": 6.577777777777779e-05, | |
| "loss": 0.0093, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 34.29, | |
| "grad_norm": 0.13242019712924957, | |
| "learning_rate": 6.222222222222222e-05, | |
| "loss": 0.0095, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 35.05, | |
| "grad_norm": 0.07912217080593109, | |
| "learning_rate": 5.866666666666667e-05, | |
| "loss": 0.0079, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 35.81, | |
| "grad_norm": 0.08500321954488754, | |
| "learning_rate": 5.511111111111111e-05, | |
| "loss": 0.0069, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 36.57, | |
| "grad_norm": 0.12945592403411865, | |
| "learning_rate": 5.1555555555555556e-05, | |
| "loss": 0.0076, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 37.33, | |
| "grad_norm": 0.05908092483878136, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.0079, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 38.1, | |
| "grad_norm": 0.05590814724564552, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.0071, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 38.86, | |
| "grad_norm": 0.07217204570770264, | |
| "learning_rate": 4.088888888888889e-05, | |
| "loss": 0.0071, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 39.62, | |
| "grad_norm": 0.08633767813444138, | |
| "learning_rate": 3.733333333333334e-05, | |
| "loss": 0.0061, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 40.38, | |
| "grad_norm": 0.08432795852422714, | |
| "learning_rate": 3.377777777777778e-05, | |
| "loss": 0.007, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 41.14, | |
| "grad_norm": 0.06549498438835144, | |
| "learning_rate": 3.0222222222222225e-05, | |
| "loss": 0.0062, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 41.9, | |
| "grad_norm": 0.07968047261238098, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.0056, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 42.67, | |
| "grad_norm": 0.054360050708055496, | |
| "learning_rate": 2.3111111111111112e-05, | |
| "loss": 0.0064, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 43.43, | |
| "grad_norm": 0.08684072643518448, | |
| "learning_rate": 1.9555555555555557e-05, | |
| "loss": 0.0052, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 44.19, | |
| "grad_norm": 0.08560307323932648, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.0066, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 44.95, | |
| "grad_norm": 0.08709636330604553, | |
| "learning_rate": 1.2444444444444445e-05, | |
| "loss": 0.0055, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 45.71, | |
| "grad_norm": 0.03685537353157997, | |
| "learning_rate": 8.88888888888889e-06, | |
| "loss": 0.0054, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 46.48, | |
| "grad_norm": 0.10162707418203354, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.0059, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 47.24, | |
| "grad_norm": 0.061479195952415466, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 0.0051, | |
| "step": 248 | |
| } | |
| ], | |
| "logging_steps": 4, | |
| "max_steps": 250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "total_flos": 4.0647058784256e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |