| { | |
| "best_metric": 1.6572628021240234, | |
| "best_model_checkpoint": "./modele-socratique-sft\\checkpoint-171", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 171, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11695906432748537, | |
| "grad_norm": 4.373384475708008, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 5.2688, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.23391812865497075, | |
| "grad_norm": 7.579962253570557, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 4.2978, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 3.2114369869232178, | |
| "learning_rate": 0.0001, | |
| "loss": 2.864, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4678362573099415, | |
| "grad_norm": 2.231703758239746, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 2.2745, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5847953216374269, | |
| "grad_norm": 1.8353674411773682, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 2.0056, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 3.654597282409668, | |
| "learning_rate": 0.0002, | |
| "loss": 1.9087, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8187134502923976, | |
| "grad_norm": 1.4905436038970947, | |
| "learning_rate": 0.00019626168224299065, | |
| "loss": 1.751, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.935672514619883, | |
| "grad_norm": 1.4564584493637085, | |
| "learning_rate": 0.00019252336448598133, | |
| "loss": 1.7465, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9941520467836257, | |
| "eval_loss": 1.743960976600647, | |
| "eval_runtime": 1.1773, | |
| "eval_samples_per_second": 64.554, | |
| "eval_steps_per_second": 8.494, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 1.3438069820404053, | |
| "learning_rate": 0.00018878504672897197, | |
| "loss": 1.6509, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1695906432748537, | |
| "grad_norm": 1.4461097717285156, | |
| "learning_rate": 0.00018504672897196262, | |
| "loss": 1.5477, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.286549707602339, | |
| "grad_norm": 1.7722139358520508, | |
| "learning_rate": 0.0001813084112149533, | |
| "loss": 1.6173, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 1.3352079391479492, | |
| "learning_rate": 0.00017757009345794393, | |
| "loss": 1.6119, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.52046783625731, | |
| "grad_norm": 1.5055865049362183, | |
| "learning_rate": 0.00017383177570093458, | |
| "loss": 1.5867, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6374269005847952, | |
| "grad_norm": 1.6256341934204102, | |
| "learning_rate": 0.00017009345794392523, | |
| "loss": 1.5639, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 1.4370089769363403, | |
| "learning_rate": 0.0001663551401869159, | |
| "loss": 1.526, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.871345029239766, | |
| "grad_norm": 1.7688274383544922, | |
| "learning_rate": 0.00016261682242990654, | |
| "loss": 1.5438, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.9883040935672516, | |
| "grad_norm": 1.3985792398452759, | |
| "learning_rate": 0.0001588785046728972, | |
| "loss": 1.5231, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.6572628021240234, | |
| "eval_runtime": 1.1323, | |
| "eval_samples_per_second": 67.121, | |
| "eval_steps_per_second": 8.832, | |
| "step": 171 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 595, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "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": 935836976234496.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |