learning_rate: 2e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 20
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.93	1.0	526	1.0170	0.6341
0.533	2.0	1052	1.0406	0.6589
0.2317	3.0	1578	1.3728	0.6429
0.1081	4.0	2104	1.8683	0.6341
0.0647	5.0	2630	2.2965	0.6122
0.053	6.0	3156	2.3454	0.6385
0.0041	7.0	3682	2.5653	0.6399
0.0199	8.0	4208	2.7244	0.6414
0.0091	9.0	4734	2.7655	0.6312
0.0007	10.0	5260	2.8327	0.6370
0.0141	11.0	5786	2.9765	0.6283
0.0008	12.0	6312	2.9757	0.6283
0.0063	13.0	6838	2.9824	0.6370
0.0027	14.0	7364	2.8190	0.6341
0.0071	15.0	7890	3.0544	0.6254
0.0005	16.0	8416	2.9467	0.6385
0.006	17.0	8942	2.9506	0.6356
0.0035	18.0	9468	2.9577	0.6341
0.0097	19.0	9994	2.9790	0.6283
0.0004	19.9629	10500	2.9577	0.6327

levancuong17/visobert_model_UIT

visobert_model_UIT