learning_rate: 1e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 64
total_eval_batch_size: 64
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_steps: 400
num_epochs: 5

Training Loss	Epoch	Step	Validation Loss	F1	Threshold	F1 At 05
2.7309	0.3876	200	2.7389	0.5506	0.51	0.5506
2.5841	0.7752	400	2.5592	0.6117	0.51	0.6083
2.3249	1.1628	600	2.5262	0.6301	0.51	0.6276
2.2814	1.5504	800	2.4446	0.6494	0.4200	0.6252
2.2382	1.9380	1000	2.4209	0.6562	0.4600	0.6525
1.9427	2.3256	1200	2.6415	0.6581	0.3400	0.6421
1.9336	2.7132	1400	2.6539	0.6606	0.58	0.6478
1.6468	3.1008	1600	2.7483	0.6536	0.4600	0.6516
1.5632	3.4884	1800	2.8935	0.6539	0.52	0.6471
1.6254	3.8760	2000	2.8170	0.6533	0.4100	0.6500
1.3267	4.2636	2200	3.0258	0.6546	0.47	0.6489
1.2919	4.6512	2400	3.1296	0.6560	0.38	0.6489

DrinkIcedT/roberta-large_MBTI_P_MBTI_agg_balanced_100

roberta-large_MBTI_P_MBTI_agg_balanced_100