| { | |
| "best_metric": 0.022033799439668655, | |
| "best_model_checkpoint": "./results/checkpoint-6000", | |
| "epoch": 0.687915615684476, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.9617824657953072e-05, | |
| "loss": 0.058, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_accuracy": 0.9880467809436451, | |
| "eval_loss": 0.04749465361237526, | |
| "eval_runtime": 369.9155, | |
| "eval_samples_per_second": 94.308, | |
| "eval_steps_per_second": 5.896, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 1.923564931590614e-05, | |
| "loss": 0.0513, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_accuracy": 0.9892220374935504, | |
| "eval_loss": 0.044388506561517715, | |
| "eval_runtime": 313.5583, | |
| "eval_samples_per_second": 111.258, | |
| "eval_steps_per_second": 6.956, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 1.885347397385921e-05, | |
| "loss": 0.0413, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_accuracy": 0.9901393108983546, | |
| "eval_loss": 0.044446878135204315, | |
| "eval_runtime": 322.6458, | |
| "eval_samples_per_second": 108.125, | |
| "eval_steps_per_second": 6.76, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 1.847129863181228e-05, | |
| "loss": 0.0364, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "eval_accuracy": 0.9911712434787594, | |
| "eval_loss": 0.037910301238298416, | |
| "eval_runtime": 258.562, | |
| "eval_samples_per_second": 134.923, | |
| "eval_steps_per_second": 8.435, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.8089123289765345e-05, | |
| "loss": 0.0389, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_accuracy": 0.9921171816774638, | |
| "eval_loss": 0.03007333353161812, | |
| "eval_runtime": 258.8834, | |
| "eval_samples_per_second": 134.756, | |
| "eval_steps_per_second": 8.425, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 1.7706947947718412e-05, | |
| "loss": 0.0343, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "eval_accuracy": 0.9923465000286648, | |
| "eval_loss": 0.03364783525466919, | |
| "eval_runtime": 258.3829, | |
| "eval_samples_per_second": 135.017, | |
| "eval_steps_per_second": 8.441, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 1.7324772605671482e-05, | |
| "loss": 0.0351, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_accuracy": 0.9918591985323626, | |
| "eval_loss": 0.030066516250371933, | |
| "eval_runtime": 258.5585, | |
| "eval_samples_per_second": 134.925, | |
| "eval_steps_per_second": 8.435, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 1.6942597263624552e-05, | |
| "loss": 0.0343, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_accuracy": 0.9925471535859657, | |
| "eval_loss": 0.02801605314016342, | |
| "eval_runtime": 308.2791, | |
| "eval_samples_per_second": 113.164, | |
| "eval_steps_per_second": 7.075, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 1.6560421921577622e-05, | |
| "loss": 0.0328, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_accuracy": 0.9937224101358711, | |
| "eval_loss": 0.022993654012680054, | |
| "eval_runtime": 304.7606, | |
| "eval_samples_per_second": 114.47, | |
| "eval_steps_per_second": 7.156, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 1.617824657953069e-05, | |
| "loss": 0.0322, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_accuracy": 0.9930631198761681, | |
| "eval_loss": 0.022875914350152016, | |
| "eval_runtime": 509.8625, | |
| "eval_samples_per_second": 68.422, | |
| "eval_steps_per_second": 4.278, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.579607123748376e-05, | |
| "loss": 0.0275, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_accuracy": 0.994353035601674, | |
| "eval_loss": 0.024034755304455757, | |
| "eval_runtime": 296.8109, | |
| "eval_samples_per_second": 117.536, | |
| "eval_steps_per_second": 7.348, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.541389589543683e-05, | |
| "loss": 0.0288, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_accuracy": 0.9946396835406753, | |
| "eval_loss": 0.022033799439668655, | |
| "eval_runtime": 293.7762, | |
| "eval_samples_per_second": 118.75, | |
| "eval_steps_per_second": 7.424, | |
| "step": 6000 | |
| } | |
| ], | |
| "max_steps": 26166, | |
| "num_train_epochs": 3, | |
| "total_flos": 1.2716870270976e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |