learning_rate: 3e-05
train_batch_size: 96
eval_batch_size: 192
seed: 42
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: linear
num_epochs: 10
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss	Exact Match Accuracy	F1 Micro	F1 Macro
0.1926	1.0	393	0.1482	0.0339	0.0843	0.0336
0.136	2.0	786	0.1269	0.1808	0.3256	0.2574
0.1202	3.0	1179	0.1205	0.2185	0.3621	0.3371
0.1119	4.0	1572	0.1189	0.2601	0.4065	0.3926
0.1055	5.0	1965	0.1193	0.2801	0.4218	0.4206
0.1003	6.0	2358	0.1200	0.2811	0.4179	0.4180
0.0956	7.0	2751	0.1221	0.2829	0.4206	0.4269
0.0918	8.0	3144	0.1232	0.2929	0.4269	0.4353
0.0886	9.0	3537	0.1246	0.2974	0.4279	0.4397
0.086	10.0	3930	0.1252	0.2962	0.4263	0.4377

lindtsey/roberta_emotion_detection

roberta_emotion_detection