learning_rate: 3e-05
train_batch_size: 192
eval_batch_size: 384
seed: 42
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_ratio: 0.1
num_epochs: 10
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Exact Match Accuracy	F1 Micro	F1 Macro
No log	1.0	181	0.3253	0.3249	0.4663	0.1127
0.3712	2.0	362	0.2824	0.4945	0.6017	0.4137
0.3712	3.0	543	0.2745	0.5170	0.6203	0.4506
0.2642	4.0	724	0.2748	0.5065	0.6140	0.4561
0.2642	5.0	905	0.2783	0.5227	0.6209	0.4594
0.2374	6.0	1086	0.2832	0.5155	0.6191	0.4729
0.2374	7.0	1267	0.2903	0.5159	0.6192	0.4824
0.2182	8.0	1448	0.2951	0.5181	0.6198	0.4841
0.2182	9.0	1629	0.2982	0.5112	0.6129	0.4838
0.2046	10.0	1810	0.3035	0.5185	0.6169	0.4825

lindtsey/xlm_roberta_emotion_detection

xlm_roberta_emotion_detection