learning_rate: 1e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Use adamw_torch 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: 1500
num_epochs: 10
label_smoothing_factor: 0.1

Training Loss	Epoch	Step	Validation Loss	Accuracy	Micro Precision	Micro Recall	Micro F1	Macro Precision	Macro Recall	Macro F1
1.0896	1.0000	11305	1.0603	0.8892	0.8892	0.8892	0.8892	0.8688	0.8362	0.8450
1.01	2.0	22611	1.0248	0.8998	0.8998	0.8998	0.8998	0.8749	0.8490	0.8548
0.9338	3.0000	33916	1.0257	0.9012	0.9012	0.9012	0.9012	0.8821	0.8564	0.8645
0.8806	4.0	45222	1.0314	0.9079	0.9079	0.9079	0.9079	0.8786	0.8690	0.8703
0.8391	5.0000	56527	1.0573	0.9054	0.9054	0.9054	0.9054	0.8729	0.8684	0.8675
0.8133	6.0	67833	1.0680	0.9088	0.9088	0.9088	0.9088	0.8780	0.8755	0.8741
0.7886	7.0000	79138	1.0885	0.9110	0.9110	0.9110	0.9110	0.8859	0.8720	0.8758
0.7756	8.0	90444	1.1009	0.9116	0.9116	0.9116	0.9116	0.8810	0.8771	0.8762
0.7656	9.0000	101749	1.1112	0.9122	0.9122	0.9122	0.9122	0.8811	0.8775	0.8763
0.7608	9.9996	113050	1.1111	0.9119	0.9119	0.9119	0.9119	0.8810	0.8759	0.8754

stulcrad/XLM-RoBERTa-CERED2

XLM-RoBERTa-CERED2