| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 260, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.038461538461538464, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 3.3093, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 3.317, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 3.1254, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 2.6178, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 2.3151, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 2.0459, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.863295793533325, | |
| "eval_runtime": 0.4843, | |
| "eval_samples_per_second": 43.364, | |
| "eval_steps_per_second": 2.065, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00019985583705641418, | |
| "loss": 1.6915, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.3461538461538463, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001992708874098054, | |
| "loss": 1.4892, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00019823877374156647, | |
| "loss": 1.3906, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.7307692307692308, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.00019676414519013781, | |
| "loss": 1.3301, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00019485364419471454, | |
| "loss": 1.2724, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.706272602081299, | |
| "eval_runtime": 0.4877, | |
| "eval_samples_per_second": 43.056, | |
| "eval_steps_per_second": 2.05, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.1153846153846154, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00019251587657449236, | |
| "loss": 1.2358, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.0001897613727639014, | |
| "loss": 1.2118, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 1.1899, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.00018305360832480117, | |
| "loss": 1.1662, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8846153846153846, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0001791305627069662, | |
| "loss": 1.1654, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.669499397277832, | |
| "eval_runtime": 0.492, | |
| "eval_samples_per_second": 42.682, | |
| "eval_steps_per_second": 2.032, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00017485107481711012, | |
| "loss": 1.1435, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.269230769230769, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00017023442153554777, | |
| "loss": 1.1188, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0001653013984983585, | |
| "loss": 1.124, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.6538461538461537, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.0001600742264237979, | |
| "loss": 1.108, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.00015457645101945046, | |
| "loss": 1.1056, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.654463291168213, | |
| "eval_runtime": 0.4845, | |
| "eval_samples_per_second": 43.339, | |
| "eval_steps_per_second": 2.064, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.038461538461538, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00014883283692099112, | |
| "loss": 1.0918, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00014286925614030542, | |
| "loss": 1.0814, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.423076923076923, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.00013671257152545277, | |
| "loss": 1.0778, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.0001303905157574247, | |
| "loss": 1.0717, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.8076923076923075, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.0001239315664287558, | |
| "loss": 1.0643, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 1.0615, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.6669278144836426, | |
| "eval_runtime": 0.4883, | |
| "eval_samples_per_second": 43.009, | |
| "eval_steps_per_second": 2.048, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.1923076923076925, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00011071984957874479, | |
| "loss": 1.0503, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00010402659401094152, | |
| "loss": 1.05, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.576923076923077, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.73152007189939e-05, | |
| "loss": 1.0449, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 9.061590105968208e-05, | |
| "loss": 1.0451, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.961538461538462, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 8.395887191422397e-05, | |
| "loss": 1.0347, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.668196439743042, | |
| "eval_runtime": 0.4907, | |
| "eval_samples_per_second": 42.799, | |
| "eval_steps_per_second": 2.038, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 7.73740997570278e-05, | |
| "loss": 1.0339, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 6.346153846153846, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 7.089124558212871e-05, | |
| "loss": 1.0281, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 6.538461538461538, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 6.453951129574644e-05, | |
| "loss": 1.0306, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 6.730769230769231, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 5.834750817679606e-05, | |
| "loss": 1.0268, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 5.234312799786921e-05, | |
| "loss": 1.0216, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.674699544906616, | |
| "eval_runtime": 0.4901, | |
| "eval_samples_per_second": 42.844, | |
| "eval_steps_per_second": 2.04, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 7.115384615384615, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 4.6553417387219886e-05, | |
| "loss": 1.0217, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 7.3076923076923075, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 4.100445599768774e-05, | |
| "loss": 1.0184, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 1.0163, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 3.072756464904006e-05, | |
| "loss": 1.0133, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.884615384615385, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 2.6045926771976303e-05, | |
| "loss": 1.0204, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.679577112197876, | |
| "eval_runtime": 0.4883, | |
| "eval_samples_per_second": 43.003, | |
| "eval_steps_per_second": 2.048, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 8.076923076923077, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 2.1697413758237784e-05, | |
| "loss": 1.0097, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.26923076923077, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 1.7701613410634365e-05, | |
| "loss": 1.0168, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 8.461538461538462, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 1.4076524743778319e-05, | |
| "loss": 1.0141, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 8.653846153846153, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 1.083847690782972e-05, | |
| "loss": 1.0125, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 8.846153846153847, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 8.002055634117578e-06, | |
| "loss": 1.0071, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.6812119483947754, | |
| "eval_runtime": 0.4891, | |
| "eval_samples_per_second": 42.934, | |
| "eval_steps_per_second": 2.044, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 9.038461538461538, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 5.580037533961546e-06, | |
| "loss": 1.0137, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 3.5833325466437694e-06, | |
| "loss": 1.0116, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 9.423076923076923, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 2.0209347957732328e-06, | |
| "loss": 1.0093, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 9.615384615384615, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 8.998820754091531e-07, | |
| "loss": 1.0136, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 9.807692307692308, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 2.2522414843748618e-07, | |
| "loss": 1.008, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.0, | |
| "loss": 1.0088, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.681701183319092, | |
| "eval_runtime": 0.4897, | |
| "eval_samples_per_second": 42.88, | |
| "eval_steps_per_second": 2.042, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 260, | |
| "total_flos": 4.092505436769485e+17, | |
| "train_loss": 1.2536941931797907, | |
| "train_runtime": 1396.0468, | |
| "train_samples_per_second": 47.677, | |
| "train_steps_per_second": 0.186 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.092505436769485e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |