| { | |
| "best_metric": 0.16280439496040344, | |
| "best_model_checkpoint": "./results/checkpoint-500", | |
| "epoch": 2.4038461538461537, | |
| "eval_steps": 20, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09615384615384616, | |
| "grad_norm": 1.8630799055099487, | |
| "learning_rate": 1.9615384615384617e-05, | |
| "loss": 0.4117, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09615384615384616, | |
| "eval_accuracy": 0.8387096774193549, | |
| "eval_loss": 0.29503169655799866, | |
| "eval_runtime": 2.9188, | |
| "eval_samples_per_second": 31.863, | |
| "eval_steps_per_second": 8.223, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 9.186605453491211, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 0.3727, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "eval_accuracy": 0.8709677419354839, | |
| "eval_loss": 0.20473997294902802, | |
| "eval_runtime": 2.8638, | |
| "eval_samples_per_second": 32.475, | |
| "eval_steps_per_second": 8.381, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.28846153846153844, | |
| "grad_norm": 8.3997802734375, | |
| "learning_rate": 1.8846153846153846e-05, | |
| "loss": 0.2863, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.28846153846153844, | |
| "eval_accuracy": 0.956989247311828, | |
| "eval_loss": 0.2754175662994385, | |
| "eval_runtime": 2.7512, | |
| "eval_samples_per_second": 33.803, | |
| "eval_steps_per_second": 8.723, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.0750838965177536, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.2698, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "eval_accuracy": 0.946236559139785, | |
| "eval_loss": 0.16919000446796417, | |
| "eval_runtime": 2.6759, | |
| "eval_samples_per_second": 34.754, | |
| "eval_steps_per_second": 8.969, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4807692307692308, | |
| "grad_norm": 0.0878710225224495, | |
| "learning_rate": 1.807692307692308e-05, | |
| "loss": 0.1232, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4807692307692308, | |
| "eval_accuracy": 0.946236559139785, | |
| "eval_loss": 0.16618885099887848, | |
| "eval_runtime": 2.6422, | |
| "eval_samples_per_second": 35.197, | |
| "eval_steps_per_second": 9.083, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 0.038383062928915024, | |
| "learning_rate": 1.7692307692307694e-05, | |
| "loss": 0.0806, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "eval_accuracy": 0.946236559139785, | |
| "eval_loss": 0.15975366532802582, | |
| "eval_runtime": 2.6652, | |
| "eval_samples_per_second": 34.894, | |
| "eval_steps_per_second": 9.005, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6730769230769231, | |
| "grad_norm": 0.022622136399149895, | |
| "learning_rate": 1.730769230769231e-05, | |
| "loss": 0.2571, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6730769230769231, | |
| "eval_accuracy": 0.978494623655914, | |
| "eval_loss": 0.0726698786020279, | |
| "eval_runtime": 2.6685, | |
| "eval_samples_per_second": 34.85, | |
| "eval_steps_per_second": 8.994, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 9.369976043701172, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 0.1374, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "eval_accuracy": 0.978494623655914, | |
| "eval_loss": 0.0701598972082138, | |
| "eval_runtime": 2.6859, | |
| "eval_samples_per_second": 34.626, | |
| "eval_steps_per_second": 8.936, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8653846153846154, | |
| "grad_norm": 7.6264119148254395, | |
| "learning_rate": 1.653846153846154e-05, | |
| "loss": 0.165, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8653846153846154, | |
| "eval_accuracy": 0.956989247311828, | |
| "eval_loss": 0.08651256561279297, | |
| "eval_runtime": 2.7028, | |
| "eval_samples_per_second": 34.409, | |
| "eval_steps_per_second": 8.88, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 0.03415932506322861, | |
| "learning_rate": 1.6153846153846154e-05, | |
| "loss": 0.528, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "eval_accuracy": 0.967741935483871, | |
| "eval_loss": 0.12613914906978607, | |
| "eval_runtime": 2.7168, | |
| "eval_samples_per_second": 34.231, | |
| "eval_steps_per_second": 8.834, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0576923076923077, | |
| "grad_norm": 0.17008128762245178, | |
| "learning_rate": 1.576923076923077e-05, | |
| "loss": 0.1023, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.0576923076923077, | |
| "eval_accuracy": 0.9139784946236559, | |
| "eval_loss": 0.37878894805908203, | |
| "eval_runtime": 2.7414, | |
| "eval_samples_per_second": 33.924, | |
| "eval_steps_per_second": 8.755, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.010861682705581188, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0278, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "eval_accuracy": 0.946236559139785, | |
| "eval_loss": 0.1893940567970276, | |
| "eval_runtime": 2.7091, | |
| "eval_samples_per_second": 34.329, | |
| "eval_steps_per_second": 8.859, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.017718007788062096, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.1642, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_accuracy": 0.9139784946236559, | |
| "eval_loss": 0.3861010670661926, | |
| "eval_runtime": 2.6955, | |
| "eval_samples_per_second": 34.503, | |
| "eval_steps_per_second": 8.904, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.3461538461538463, | |
| "grad_norm": 0.021516259759664536, | |
| "learning_rate": 1.4615384615384615e-05, | |
| "loss": 0.1108, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.3461538461538463, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.20804111659526825, | |
| "eval_runtime": 2.9167, | |
| "eval_samples_per_second": 31.886, | |
| "eval_steps_per_second": 8.229, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.4423076923076923, | |
| "grad_norm": 0.013441790826618671, | |
| "learning_rate": 1.4230769230769232e-05, | |
| "loss": 0.1514, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.4423076923076923, | |
| "eval_accuracy": 0.967741935483871, | |
| "eval_loss": 0.15365849435329437, | |
| "eval_runtime": 2.8225, | |
| "eval_samples_per_second": 32.95, | |
| "eval_steps_per_second": 8.503, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.02140916883945465, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 0.0025, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.946236559139785, | |
| "eval_loss": 0.30473944544792175, | |
| "eval_runtime": 2.6926, | |
| "eval_samples_per_second": 34.539, | |
| "eval_steps_per_second": 8.913, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.6346153846153846, | |
| "grad_norm": 0.011240499094128609, | |
| "learning_rate": 1.3461538461538463e-05, | |
| "loss": 0.1124, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.6346153846153846, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.24811075627803802, | |
| "eval_runtime": 2.6909, | |
| "eval_samples_per_second": 34.561, | |
| "eval_steps_per_second": 8.919, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.7307692307692308, | |
| "grad_norm": 0.009658828377723694, | |
| "learning_rate": 1.3076923076923078e-05, | |
| "loss": 0.3081, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.7307692307692308, | |
| "eval_accuracy": 0.978494623655914, | |
| "eval_loss": 0.1271650344133377, | |
| "eval_runtime": 2.6952, | |
| "eval_samples_per_second": 34.505, | |
| "eval_steps_per_second": 8.905, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.8269230769230769, | |
| "grad_norm": 48.972293853759766, | |
| "learning_rate": 1.2692307692307693e-05, | |
| "loss": 0.0764, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.8269230769230769, | |
| "eval_accuracy": 0.989247311827957, | |
| "eval_loss": 0.0935346931219101, | |
| "eval_runtime": 2.7195, | |
| "eval_samples_per_second": 34.198, | |
| "eval_steps_per_second": 8.825, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.01087904442101717, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.0923, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "eval_accuracy": 0.9354838709677419, | |
| "eval_loss": 0.23266802728176117, | |
| "eval_runtime": 2.7195, | |
| "eval_samples_per_second": 34.197, | |
| "eval_steps_per_second": 8.825, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.019230769230769, | |
| "grad_norm": 19.043825149536133, | |
| "learning_rate": 1.1923076923076925e-05, | |
| "loss": 0.137, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.019230769230769, | |
| "eval_accuracy": 0.956989247311828, | |
| "eval_loss": 0.16817408800125122, | |
| "eval_runtime": 2.7341, | |
| "eval_samples_per_second": 34.015, | |
| "eval_steps_per_second": 8.778, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.1153846153846154, | |
| "grad_norm": 0.036368228495121, | |
| "learning_rate": 1.1538461538461538e-05, | |
| "loss": 0.0796, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.1153846153846154, | |
| "eval_accuracy": 0.956989247311828, | |
| "eval_loss": 0.16539430618286133, | |
| "eval_runtime": 2.7265, | |
| "eval_samples_per_second": 34.11, | |
| "eval_steps_per_second": 8.803, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.2115384615384617, | |
| "grad_norm": 0.03690864145755768, | |
| "learning_rate": 1.1153846153846154e-05, | |
| "loss": 0.0089, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.2115384615384617, | |
| "eval_accuracy": 0.956989247311828, | |
| "eval_loss": 0.2199721783399582, | |
| "eval_runtime": 2.7071, | |
| "eval_samples_per_second": 34.354, | |
| "eval_steps_per_second": 8.865, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.011880586855113506, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 0.1514, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "eval_accuracy": 0.956989247311828, | |
| "eval_loss": 0.20590856671333313, | |
| "eval_runtime": 2.697, | |
| "eval_samples_per_second": 34.482, | |
| "eval_steps_per_second": 8.899, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.4038461538461537, | |
| "grad_norm": 0.008282246068120003, | |
| "learning_rate": 1.0384615384615386e-05, | |
| "loss": 0.0044, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4038461538461537, | |
| "eval_accuracy": 0.956989247311828, | |
| "eval_loss": 0.16280439496040344, | |
| "eval_runtime": 2.7133, | |
| "eval_samples_per_second": 34.275, | |
| "eval_steps_per_second": 8.845, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 1040, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 480317453780160.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |