|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 4.0, |
|
"eval_steps": 500, |
|
"global_step": 2116, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.1890359168241966, |
|
"grad_norm": 0.5558584332466125, |
|
"learning_rate": 3e-05, |
|
"loss": 2.1103, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.3780718336483932, |
|
"grad_norm": 0.4733140468597412, |
|
"learning_rate": 3e-05, |
|
"loss": 1.931, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.5671077504725898, |
|
"grad_norm": 0.48580050468444824, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8948, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.7561436672967864, |
|
"grad_norm": 0.40655234456062317, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8988, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.945179584120983, |
|
"grad_norm": 0.6303955316543579, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8745, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.5701538461538461, |
|
"eval_loss": 1.6455246210098267, |
|
"eval_runtime": 5.0805, |
|
"eval_samples_per_second": 98.415, |
|
"eval_steps_per_second": 12.4, |
|
"step": 529 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_exact_match": 15.0, |
|
"eval_f1": 23.927619047619068, |
|
"step": 529 |
|
}, |
|
{ |
|
"epoch": 1.1342155009451795, |
|
"grad_norm": 0.4278818666934967, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8798, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 1.3232514177693762, |
|
"grad_norm": 0.449830025434494, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8613, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 1.5122873345935728, |
|
"grad_norm": 0.5130725502967834, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8709, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 1.7013232514177694, |
|
"grad_norm": 0.4889715611934662, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8448, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 1.8903591682419658, |
|
"grad_norm": 11.186380386352539, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8764, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.5723589743589743, |
|
"eval_loss": 1.6143561601638794, |
|
"eval_runtime": 5.0466, |
|
"eval_samples_per_second": 99.077, |
|
"eval_steps_per_second": 12.484, |
|
"step": 1058 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_exact_match": 15.6, |
|
"eval_f1": 25.825238095238113, |
|
"step": 1058 |
|
}, |
|
{ |
|
"epoch": 2.0793950850661624, |
|
"grad_norm": 0.5414931774139404, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8398, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 2.268431001890359, |
|
"grad_norm": 7.441005706787109, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8348, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 2.4574669187145557, |
|
"grad_norm": 24.312145233154297, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8421, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 2.6465028355387523, |
|
"grad_norm": 0.7407662868499756, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8292, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 2.835538752362949, |
|
"grad_norm": 0.9653642177581787, |
|
"learning_rate": 3e-05, |
|
"loss": 1.828, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.5734358974358974, |
|
"eval_loss": 1.617864727973938, |
|
"eval_runtime": 5.2153, |
|
"eval_samples_per_second": 95.872, |
|
"eval_steps_per_second": 12.08, |
|
"step": 1587 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_exact_match": 16.2, |
|
"eval_f1": 25.811428571428586, |
|
"step": 1587 |
|
}, |
|
{ |
|
"epoch": 3.0245746691871456, |
|
"grad_norm": 450.6055603027344, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8204, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 3.213610586011342, |
|
"grad_norm": 2.1127028465270996, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8361, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 3.402646502835539, |
|
"grad_norm": 564.9320068359375, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8338, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 3.5916824196597354, |
|
"grad_norm": 52.5078010559082, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8075, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 3.780718336483932, |
|
"grad_norm": 27.87557601928711, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8175, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 3.9697542533081287, |
|
"grad_norm": 4.810656547546387, |
|
"learning_rate": 3e-05, |
|
"loss": 1.8218, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.5728205128205128, |
|
"eval_loss": 1.6224406957626343, |
|
"eval_runtime": 5.0588, |
|
"eval_samples_per_second": 98.838, |
|
"eval_steps_per_second": 12.454, |
|
"step": 2116 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_exact_match": 17.4, |
|
"eval_f1": 27.194285714285726, |
|
"step": 2116 |
|
} |
|
], |
|
"logging_steps": 100, |
|
"max_steps": 26450, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 50, |
|
"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": 1.722440409583452e+17, |
|
"train_batch_size": 2, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|