| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 0.0003, | |
| "loss": 1.9296, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 0.0003, | |
| "loss": 1.7554, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6093164556962025, | |
| "eval_loss": 1.7939746379852295, | |
| "eval_runtime": 4.7481, | |
| "eval_samples_per_second": 105.304, | |
| "eval_steps_per_second": 13.268, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 0.0003, | |
| "loss": 1.6394, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 0.0003, | |
| "loss": 1.5315, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.0003, | |
| "loss": 1.5248, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6085063291139241, | |
| "eval_loss": 1.8273799419403076, | |
| "eval_runtime": 4.904, | |
| "eval_samples_per_second": 101.957, | |
| "eval_steps_per_second": 12.847, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 0.0003, | |
| "loss": 1.1692, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 0.0003, | |
| "loss": 1.2054, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6027088607594937, | |
| "eval_loss": 1.9717934131622314, | |
| "eval_runtime": 4.6675, | |
| "eval_samples_per_second": 107.124, | |
| "eval_steps_per_second": 13.498, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.0003, | |
| "loss": 1.0252, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.8608, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.8989, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5986835443037974, | |
| "eval_loss": 2.151914596557617, | |
| "eval_runtime": 5.4432, | |
| "eval_samples_per_second": 91.857, | |
| "eval_steps_per_second": 11.574, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5842, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.6306, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5960506329113924, | |
| "eval_loss": 2.329284191131592, | |
| "eval_runtime": 4.4146, | |
| "eval_samples_per_second": 113.261, | |
| "eval_steps_per_second": 14.271, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5368, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4495, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4712, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5935696202531645, | |
| "eval_loss": 2.5598793029785156, | |
| "eval_runtime": 4.5669, | |
| "eval_samples_per_second": 109.484, | |
| "eval_steps_per_second": 13.795, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3625, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3797, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5935696202531645, | |
| "eval_loss": 2.732860803604126, | |
| "eval_runtime": 4.5141, | |
| "eval_samples_per_second": 110.763, | |
| "eval_steps_per_second": 13.956, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3582, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3399, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3527, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5912911392405064, | |
| "eval_loss": 2.8185083866119385, | |
| "eval_runtime": 4.7158, | |
| "eval_samples_per_second": 106.026, | |
| "eval_steps_per_second": 13.359, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3202, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3314, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.592, | |
| "eval_loss": 2.824962854385376, | |
| "eval_runtime": 4.8288, | |
| "eval_samples_per_second": 103.546, | |
| "eval_steps_per_second": 13.047, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3174, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3157, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3265, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5911392405063292, | |
| "eval_loss": 2.9242382049560547, | |
| "eval_runtime": 4.7138, | |
| "eval_samples_per_second": 106.072, | |
| "eval_steps_per_second": 13.365, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2989, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3148, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.591240506329114, | |
| "eval_loss": 3.0012593269348145, | |
| "eval_runtime": 5.1318, | |
| "eval_samples_per_second": 97.431, | |
| "eval_steps_per_second": 12.276, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3028, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3047, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3184, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.590632911392405, | |
| "eval_loss": 2.931525707244873, | |
| "eval_runtime": 4.7893, | |
| "eval_samples_per_second": 104.399, | |
| "eval_steps_per_second": 13.154, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3101, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.5896962025316456, | |
| "eval_loss": 2.9116382598876953, | |
| "eval_runtime": 4.5748, | |
| "eval_samples_per_second": 109.295, | |
| "eval_steps_per_second": 13.771, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3063, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3041, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3164, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5902025316455696, | |
| "eval_loss": 2.920793056488037, | |
| "eval_runtime": 4.718, | |
| "eval_samples_per_second": 105.977, | |
| "eval_steps_per_second": 13.353, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2957, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3074, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.5908607594936709, | |
| "eval_loss": 2.9385440349578857, | |
| "eval_runtime": 5.108, | |
| "eval_samples_per_second": 97.887, | |
| "eval_steps_per_second": 12.334, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3013, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3002, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3107, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.5891645569620253, | |
| "eval_loss": 2.9519243240356445, | |
| "eval_runtime": 4.8208, | |
| "eval_samples_per_second": 103.717, | |
| "eval_steps_per_second": 13.068, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2892, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3054, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.5898227848101266, | |
| "eval_loss": 3.010847568511963, | |
| "eval_runtime": 5.1511, | |
| "eval_samples_per_second": 97.066, | |
| "eval_steps_per_second": 12.23, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2959, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.297, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.309, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.5903544303797469, | |
| "eval_loss": 3.003683567047119, | |
| "eval_runtime": 4.8904, | |
| "eval_samples_per_second": 102.241, | |
| "eval_steps_per_second": 12.882, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2883, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3005, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.5898481012658228, | |
| "eval_loss": 3.0279438495635986, | |
| "eval_runtime": 5.1333, | |
| "eval_samples_per_second": 97.403, | |
| "eval_steps_per_second": 12.273, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2959, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2911, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.0003, | |
| "loss": 0.3127, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.5883291139240506, | |
| "eval_loss": 2.9650285243988037, | |
| "eval_runtime": 4.8904, | |
| "eval_samples_per_second": 102.242, | |
| "eval_steps_per_second": 12.882, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 5000, | |
| "total_flos": 3.1967425075347456e+17, | |
| "train_loss": 0.5438678237915039, | |
| "train_runtime": 3497.065, | |
| "train_samples_per_second": 45.753, | |
| "train_steps_per_second": 1.43 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 3.1967425075347456e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |