| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 68, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14705882352941177, | |
| "grad_norm": 0.9781005831627027, | |
| "learning_rate": 2.8571428571428574e-05, | |
| "loss": 1.1084, | |
| "mean_token_accuracy": 0.7578839182853698, | |
| "num_tokens": 640462.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 0.6955683337172678, | |
| "learning_rate": 3.8032786885245905e-05, | |
| "loss": 0.9348, | |
| "mean_token_accuracy": 0.7874236106872559, | |
| "num_tokens": 1259362.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4411764705882353, | |
| "grad_norm": 1.8154067226337278, | |
| "learning_rate": 3.475409836065574e-05, | |
| "loss": 0.9954, | |
| "mean_token_accuracy": 0.7773380041122436, | |
| "num_tokens": 1895339.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.6357695058074742, | |
| "learning_rate": 3.1475409836065576e-05, | |
| "loss": 0.8677, | |
| "mean_token_accuracy": 0.8019283294677735, | |
| "num_tokens": 2505918.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7352941176470589, | |
| "grad_norm": 0.8133712484258966, | |
| "learning_rate": 2.8196721311475412e-05, | |
| "loss": 0.8142, | |
| "mean_token_accuracy": 0.8124492406845093, | |
| "num_tokens": 3153639.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 0.5210471105697568, | |
| "learning_rate": 2.4918032786885248e-05, | |
| "loss": 0.8708, | |
| "mean_token_accuracy": 0.797481119632721, | |
| "num_tokens": 3789164.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0294117647058822, | |
| "grad_norm": 3.7021334939852113, | |
| "learning_rate": 2.1639344262295087e-05, | |
| "loss": 0.846, | |
| "mean_token_accuracy": 0.809739100933075, | |
| "num_tokens": 4419827.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.6754015523647303, | |
| "learning_rate": 1.836065573770492e-05, | |
| "loss": 0.7306, | |
| "mean_token_accuracy": 0.8303763628005981, | |
| "num_tokens": 5044302.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.3235294117647058, | |
| "grad_norm": 1.2809188857705796, | |
| "learning_rate": 1.5081967213114754e-05, | |
| "loss": 0.6667, | |
| "mean_token_accuracy": 0.8389277696609497, | |
| "num_tokens": 5677897.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 0.5380085475385603, | |
| "learning_rate": 1.1803278688524591e-05, | |
| "loss": 0.6416, | |
| "mean_token_accuracy": 0.8458507895469666, | |
| "num_tokens": 6307507.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.6176470588235294, | |
| "grad_norm": 0.46342793767814744, | |
| "learning_rate": 8.524590163934427e-06, | |
| "loss": 0.6526, | |
| "mean_token_accuracy": 0.8422300696372986, | |
| "num_tokens": 6947098.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 0.4489151107136719, | |
| "learning_rate": 5.245901639344263e-06, | |
| "loss": 0.5468, | |
| "mean_token_accuracy": 0.8634997725486755, | |
| "num_tokens": 7567565.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.9117647058823528, | |
| "grad_norm": 0.7612436131924979, | |
| "learning_rate": 1.9672131147540985e-06, | |
| "loss": 0.6499, | |
| "mean_token_accuracy": 0.8438345670700074, | |
| "num_tokens": 8199491.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "mean_token_accuracy": 0.8247674107551575, | |
| "num_tokens": 8590054.0, | |
| "step": 68, | |
| "total_flos": 12887022829568.0, | |
| "train_loss": 0.7892868659075569, | |
| "train_runtime": 182.7339, | |
| "train_samples_per_second": 5.899, | |
| "train_steps_per_second": 0.372 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 68, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 12887022829568.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |