| { | |
| "best_metric": 0.6506040692329407, | |
| "best_model_checkpoint": "tiny_bert_rand_50_v1_qnli/checkpoint-410", | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 2460, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.4577155113220215, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.6648, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6146805784367564, | |
| "eval_loss": 0.6506040692329407, | |
| "eval_runtime": 1.956, | |
| "eval_samples_per_second": 2792.966, | |
| "eval_steps_per_second": 11.248, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.5135442018508911, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.6346, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6143144792238696, | |
| "eval_loss": 0.6588263511657715, | |
| "eval_runtime": 1.924, | |
| "eval_samples_per_second": 2839.373, | |
| "eval_steps_per_second": 11.434, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.518192768096924, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.5929, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6144975288303131, | |
| "eval_loss": 0.6646158695220947, | |
| "eval_runtime": 1.9265, | |
| "eval_samples_per_second": 2835.689, | |
| "eval_steps_per_second": 11.42, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.417586088180542, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.5346, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6057111477210324, | |
| "eval_loss": 0.7260009050369263, | |
| "eval_runtime": 1.9205, | |
| "eval_samples_per_second": 2844.601, | |
| "eval_steps_per_second": 11.455, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 5.0499773025512695, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4633, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6020501555921655, | |
| "eval_loss": 0.7664932608604431, | |
| "eval_runtime": 1.9267, | |
| "eval_samples_per_second": 2835.443, | |
| "eval_steps_per_second": 11.419, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 5.864147186279297, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.3949, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5938129233022149, | |
| "eval_loss": 0.975488007068634, | |
| "eval_runtime": 1.9229, | |
| "eval_samples_per_second": 2840.982, | |
| "eval_steps_per_second": 11.441, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 2460, | |
| "total_flos": 1.6480417453111296e+16, | |
| "train_loss": 0.5475072597100483, | |
| "train_runtime": 429.8187, | |
| "train_samples_per_second": 12184.555, | |
| "train_steps_per_second": 47.695 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "total_flos": 1.6480417453111296e+16, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |