learning_rate: 2e-05
train_batch_size: 2
eval_batch_size: 2
seed: 3407
gradient_accumulation_steps: 4
total_train_batch_size: 8
optimizer: Use OptimizerNames.ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 10

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.4193	1.0	161	0.3617	0.2254	0.0814	0.1196	0.9191
0.277	2.0	322	0.2614	0.3917	0.2375	0.2957	0.9273
0.2348	3.0	483	0.2318	0.3973	0.3461	0.3700	0.9278
0.1894	4.0	644	0.2259	0.4120	0.3656	0.3874	0.9301
0.1643	5.0	805	0.2334	0.3633	0.4267	0.3925	0.9238
0.161	6.0	966	0.2254	0.3885	0.4118	0.3998	0.9273
0.1289	7.0	1127	0.2446	0.3706	0.4750	0.4164	0.9211
0.121	8.0	1288	0.2437	0.3739	0.4416	0.4049	0.9239
0.1075	9.0	1449	0.2428	0.4001	0.4250	0.4122	0.9284

yihong-gongzi/outputs

outputs