CS221-bert-base-uncased-finetuned-semeval-NT-tir

This model is a fine-tuned version of google-bert/bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

Model description

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 2e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
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: 100
num_epochs: 20

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.339	1.0	369	0.2944	0.2091	0.5579	0.5190
0.2326	2.0	738	0.2562	0.4035	0.6321	0.5829
0.17	3.0	1107	0.2748	0.4852	0.6821	0.6005
0.1252	4.0	1476	0.3174	0.4954	0.7209	0.5285
0.0954	5.0	1845	0.3520	0.4669	0.6766	0.5774
0.0337	6.0	2214	0.3920	0.5091	0.7215	0.5516
0.0105	7.0	2583	0.4154	0.5054	0.7026	0.5883
0.0431	8.0	2952	0.4430	0.4925	0.7084	0.5516
0.0063	9.0	3321	0.4653	0.5164	0.7194	0.5870
0.0034	10.0	3690	0.4786	0.4960	0.7012	0.5761
0.0054	11.0	4059	0.4896	0.5211	0.7290	0.5611
0.0029	12.0	4428	0.5173	0.4954	0.7026	0.5734
0.0024	13.0	4797	0.5175	0.5058	0.7094	0.5720
0.0017	14.0	5166	0.5217	0.5026	0.7065	0.5734