| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 22, | |
| "global_step": 87, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 19.59183260405101, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.3985, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "eval_loss": 1.5676738023757935, | |
| "eval_runtime": 3.9227, | |
| "eval_samples_per_second": 4.844, | |
| "eval_steps_per_second": 1.275, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 25.857640141956836, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.6073, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 19.53747244629775, | |
| "learning_rate": 6e-06, | |
| "loss": 1.5784, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 32.1611390841105, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4924, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 25.68491589990429, | |
| "learning_rate": 1e-05, | |
| "loss": 1.5215, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 21.800722806654328, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.5508, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 17.19812526688047, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.205, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 33.24510948526866, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.6376, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 14.986129451187, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.454, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 17.517194747344856, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6559, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 18.22606604955052, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.5816, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 26.577328122239592, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.7884, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 17.563463503652706, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.4405, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 19.98739837372538, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.7512, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 16.25364292099236, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7037, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 22.328347783615648, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.5476, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 20.71265405702363, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.8108, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 22.425069186679085, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.6751, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 18.646553829771403, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.7696, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 16.74801305116847, | |
| "learning_rate": 4e-05, | |
| "loss": 1.7699, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 30.99364381661585, | |
| "learning_rate": 4.2e-05, | |
| "loss": 2.1827, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 19.761245037191404, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.4776, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 1.856759786605835, | |
| "eval_runtime": 1.4309, | |
| "eval_samples_per_second": 13.279, | |
| "eval_steps_per_second": 3.494, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 27.821861354666428, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.8677, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 22.80354998875929, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.7424, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 19.920321630615994, | |
| "learning_rate": 5e-05, | |
| "loss": 1.769, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 23.82621534355921, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.8615, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 27.63826215875828, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.7164, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 30.810269474286084, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.8922, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 23.279443617002265, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.8507, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 18.091756285883704, | |
| "learning_rate": 6e-05, | |
| "loss": 1.6537, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 23.292568819411724, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.9125, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 52.21913805211888, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.8456, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 30.771918199992527, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.9398, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 40.19880102087157, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 2.3763, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 28.33779955799431, | |
| "learning_rate": 7e-05, | |
| "loss": 2.0815, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 54.379783823327905, | |
| "learning_rate": 7.2e-05, | |
| "loss": 2.1032, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 662.6041427579516, | |
| "learning_rate": 7.4e-05, | |
| "loss": 3.5917, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4011.501090614335, | |
| "learning_rate": 7.6e-05, | |
| "loss": 7.3933, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 753.5816146908044, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 15.8371, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 291.86609997137384, | |
| "learning_rate": 8e-05, | |
| "loss": 9.0249, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 922.0326051718562, | |
| "learning_rate": 8.2e-05, | |
| "loss": 9.8922, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 115.68135873083658, | |
| "learning_rate": 8.4e-05, | |
| "loss": 8.2719, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 693.1702757980898, | |
| "learning_rate": 8.6e-05, | |
| "loss": 24.1699, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 219.78828591396768, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 10.1246, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_loss": 8.758976936340332, | |
| "eval_runtime": 1.3819, | |
| "eval_samples_per_second": 13.749, | |
| "eval_steps_per_second": 3.618, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 101.32256273283853, | |
| "learning_rate": 9e-05, | |
| "loss": 9.1756, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 11.958039992364316, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 7.7273, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3232.8188937829473, | |
| "learning_rate": 9.4e-05, | |
| "loss": 8.7486, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 34.97640980952823, | |
| "learning_rate": 9.6e-05, | |
| "loss": 8.0394, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 28.785233378323184, | |
| "learning_rate": 9.8e-05, | |
| "loss": 7.6241, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 734.1733446536996, | |
| "learning_rate": 0.0001, | |
| "loss": 7.7813, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 59.66290191764372, | |
| "learning_rate": 0.00010200000000000001, | |
| "loss": 7.6734, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 18.23606863385719, | |
| "learning_rate": 0.00010400000000000001, | |
| "loss": 7.8783, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 15.152500341579357, | |
| "learning_rate": 0.00010600000000000002, | |
| "loss": 7.6016, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 25.1062172290986, | |
| "learning_rate": 0.00010800000000000001, | |
| "loss": 7.7206, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 30.75760910048317, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 7.7526, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 18.354297835205724, | |
| "learning_rate": 0.00011200000000000001, | |
| "loss": 7.5551, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 32.48061320811231, | |
| "learning_rate": 0.00011399999999999999, | |
| "loss": 7.6922, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 26.624723764338736, | |
| "learning_rate": 0.000116, | |
| "loss": 7.5791, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 13.92226298912218, | |
| "learning_rate": 0.000118, | |
| "loss": 7.3692, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 15.428105634893914, | |
| "learning_rate": 0.00012, | |
| "loss": 7.564, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 23.02235056392551, | |
| "learning_rate": 0.000122, | |
| "loss": 7.3948, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 16.470374126805776, | |
| "learning_rate": 0.000124, | |
| "loss": 7.4646, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 19.574553557429, | |
| "learning_rate": 0.000126, | |
| "loss": 7.3902, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 25.549182808604662, | |
| "learning_rate": 0.00012800000000000002, | |
| "loss": 7.5461, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 618.1147749407063, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 8.157, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 66.0610660913446, | |
| "learning_rate": 0.000132, | |
| "loss": 8.1284, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 8.004860877990723, | |
| "eval_runtime": 1.3748, | |
| "eval_samples_per_second": 13.82, | |
| "eval_steps_per_second": 3.637, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 41.635960690924975, | |
| "learning_rate": 0.000134, | |
| "loss": 8.0602, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 16.679869546723175, | |
| "learning_rate": 0.00013600000000000003, | |
| "loss": 7.8268, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 33.70142265056291, | |
| "learning_rate": 0.000138, | |
| "loss": 8.1206, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 43.29315748881759, | |
| "learning_rate": 0.00014, | |
| "loss": 7.7882, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 28.3799834727588, | |
| "learning_rate": 0.000142, | |
| "loss": 7.5454, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 30.911781055251613, | |
| "learning_rate": 0.000144, | |
| "loss": 7.7616, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 56.741836296758365, | |
| "learning_rate": 0.000146, | |
| "loss": 8.1145, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 13.967249574234195, | |
| "learning_rate": 0.000148, | |
| "loss": 7.6758, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 42.28071588262047, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 7.5507, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 27.40577229308579, | |
| "learning_rate": 0.000152, | |
| "loss": 7.6852, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 21.422615778584905, | |
| "learning_rate": 0.000154, | |
| "loss": 7.6071, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 28.22569647336005, | |
| "learning_rate": 0.00015600000000000002, | |
| "loss": 7.5556, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 20.075502929468385, | |
| "learning_rate": 0.00015800000000000002, | |
| "loss": 7.529, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 14.896626626967638, | |
| "learning_rate": 0.00016, | |
| "loss": 7.5352, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 16.98565517398293, | |
| "learning_rate": 0.000162, | |
| "loss": 7.5734, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 17.76803489316823, | |
| "learning_rate": 0.000164, | |
| "loss": 7.591, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 15.601674972992182, | |
| "learning_rate": 0.000166, | |
| "loss": 7.4251, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 21.36580777081032, | |
| "learning_rate": 0.000168, | |
| "loss": 7.728, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 19.73380900316605, | |
| "learning_rate": 0.00017, | |
| "loss": 7.3548, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 16.40313028374367, | |
| "learning_rate": 0.000172, | |
| "loss": 7.3652, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 11.846265862818466, | |
| "learning_rate": 0.000174, | |
| "loss": 7.4363, | |
| "step": 87 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 261, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 87, | |
| "total_flos": 6.081355071671501e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |