learning_rate: 2e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 64
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 500
num_epochs: 3
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
1.4644	0.1263	500	0.6023
1.2133	0.2527	1000	0.5163
1.1196	0.3790	1500	0.4730
1.0349	0.5053	2000	0.4462
0.9930	0.6316	2500	0.4258
0.9242	0.7580	3000	0.4137
0.9228	0.8843	3500	0.3976
0.8742	1.0106	4000	0.3768
0.8703	1.1369	4500	0.3681
0.8529	1.2633	5000	0.3661
0.7940	1.3896	5500	0.3589
0.7693	1.5159	6000	0.3478
0.7683	1.6422	6500	0.3391
0.7758	1.7686	7000	0.3345
0.7811	1.8949	7500	0.3281
0.7714	2.0212	8000	0.3281
0.7521	2.1475	8500	0.3251
0.7587	2.2739	9000	0.3215
0.7364	2.4002	9500	0.3220
0.7300	2.5265	10000	0.3162
0.7471	2.6529	10500	0.3158
0.7388	2.7792	11000	0.3134
0.7209	2.9055	11500	0.3162

sithu015/MyanBERTa-legal-finetuned

MyanBERTa-legal-finetuned