| { | |
| "best_metric": 0.00030570320086553693, | |
| "best_model_checkpoint": "/media/aiteam/Data/LuminSoft/projects/ocr/results/checkpoint-212814", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 364824, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.07173027098178864, | |
| "learning_rate": 9.375e-05, | |
| "loss": 0.0796, | |
| "step": 30402 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 0.9691155403795684, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.0006197646143846214, | |
| "eval_runtime": 2732.1107, | |
| "eval_samples_per_second": 11.117, | |
| "eval_steps_per_second": 1.39, | |
| "step": 30402 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.011926465667784214, | |
| "learning_rate": 8.75e-05, | |
| "loss": 0.0008, | |
| "step": 60804 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 0.976564760945452, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.0004327517526689917, | |
| "eval_runtime": 2738.2395, | |
| "eval_samples_per_second": 11.092, | |
| "eval_steps_per_second": 1.387, | |
| "step": 60804 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.006019719876348972, | |
| "learning_rate": 8.125000000000001e-05, | |
| "loss": 0.0005, | |
| "step": 91206 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 0.9809933294323522, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.0003870356595143676, | |
| "eval_runtime": 2734.2135, | |
| "eval_samples_per_second": 11.108, | |
| "eval_steps_per_second": 1.389, | |
| "step": 91206 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.0002722485805861652, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0004, | |
| "step": 121608 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 0.981663601147577, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.0003594958398025483, | |
| "eval_runtime": 2730.1307, | |
| "eval_samples_per_second": 11.125, | |
| "eval_steps_per_second": 1.391, | |
| "step": 121608 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.00043152127182111144, | |
| "learning_rate": 6.875e-05, | |
| "loss": 0.0003, | |
| "step": 152010 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 0.9837872090873163, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.000314301869366318, | |
| "eval_runtime": 2695.597, | |
| "eval_samples_per_second": 11.268, | |
| "eval_steps_per_second": 1.409, | |
| "step": 152010 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.0002681140904314816, | |
| "learning_rate": 6.25e-05, | |
| "loss": 0.0002, | |
| "step": 182412 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 0.9853297104414079, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.00031128423870541155, | |
| "eval_runtime": 2742.3832, | |
| "eval_samples_per_second": 11.075, | |
| "eval_steps_per_second": 1.385, | |
| "step": 182412 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.00042421137914061546, | |
| "learning_rate": 5.6250000000000005e-05, | |
| "loss": 0.0002, | |
| "step": 212814 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_bleu": 0.9852415464502277, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.00030570320086553693, | |
| "eval_runtime": 2763.5679, | |
| "eval_samples_per_second": 10.991, | |
| "eval_steps_per_second": 1.374, | |
| "step": 212814 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.013042913749814034, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0002, | |
| "step": 243216 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_bleu": 0.9855596017686066, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.0003345392469782382, | |
| "eval_runtime": 2754.3836, | |
| "eval_samples_per_second": 11.027, | |
| "eval_steps_per_second": 1.379, | |
| "step": 243216 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.009026958607137203, | |
| "learning_rate": 4.375e-05, | |
| "loss": 0.0001, | |
| "step": 273618 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_bleu": 0.9856968436402628, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.0003159026673529297, | |
| "eval_runtime": 2750.1172, | |
| "eval_samples_per_second": 11.044, | |
| "eval_steps_per_second": 1.381, | |
| "step": 273618 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.4567546713806223e-05, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.0001, | |
| "step": 304020 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_bleu": 0.9858578320415788, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.00036075865500606596, | |
| "eval_runtime": 2744.9657, | |
| "eval_samples_per_second": 11.065, | |
| "eval_steps_per_second": 1.383, | |
| "step": 304020 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.00015938151045702398, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.0001, | |
| "step": 334422 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_bleu": 0.9860261333017161, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.00039157370338216424, | |
| "eval_runtime": 2750.6108, | |
| "eval_samples_per_second": 11.042, | |
| "eval_steps_per_second": 1.38, | |
| "step": 334422 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.00013136252528056502, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0001, | |
| "step": 364824 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_bleu": 0.9866485075861869, | |
| "eval_cer": 0.0, | |
| "eval_loss": 0.00037165774847380817, | |
| "eval_runtime": 2749.3584, | |
| "eval_samples_per_second": 11.047, | |
| "eval_steps_per_second": 1.381, | |
| "step": 364824 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 486432, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 16, | |
| "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": 2.681400041091367e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |