learning_rate: 2e-05
train_batch_size: 12
eval_batch_size: 12
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 24
optimizer: Use 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: 200
num_epochs: 10

Training Loss	Epoch	Step	Validation Loss	Accuracy	Micro Precision	Micro Recall	Micro F1	Macro Precision	Macro Recall	Macro F1
No log	1.0	434	0.7391	0.7789	0.7789	0.7789	0.7789	0.4774	0.4592	0.4480
1.7944	2.0	868	0.4150	0.8895	0.8895	0.8895	0.8895	0.6905	0.7064	0.6950
0.5614	3.0	1302	0.3435	0.8842	0.8842	0.8842	0.8842	0.7358	0.7143	0.7162
0.3415	4.0	1736	0.2973	0.9053	0.9053	0.9053	0.9053	0.7661	0.7584	0.7589
0.2116	5.0	2170	0.3134	0.9211	0.9211	0.9211	0.9211	0.8494	0.8258	0.8326
0.127	6.0	2604	0.4035	0.8947	0.8947	0.8947	0.8947	0.7205	0.7026	0.7077
0.0631	7.0	3038	0.3703	0.9263	0.9263	0.9263	0.9263	0.8011	0.8041	0.7988
0.0631	8.0	3472	0.3783	0.9211	0.9211	0.9211	0.9211	0.8011	0.8005	0.7966
0.0366	9.0	3906	0.3840	0.9263	0.9263	0.9263	0.9263	0.8053	0.8088	0.8053
0.0163	10.0	4340	0.4008	0.9211	0.9211	0.9211	0.9211	0.7905	0.7945	0.7860

stulcrad/XLM-RoBERTa-CERED4

XLM-RoBERTa-CERED4