bert-finetuned-char-classification-e15

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

• Loss: 0.1233
• F1: 0.5572
• Roc Auc: 0.7345
• Accuracy: 0.3761

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 512
• eval_batch_size: 512
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.1881	1.2	500	0.1694	0.0104	0.5024	0.0007
0.1563	2.39	1000	0.1415	0.3037	0.5948	0.1221
0.1374	3.59	1500	0.1290	0.3997	0.6353	0.1958
0.127	4.78	2000	0.1231	0.4467	0.6579	0.2376
0.1194	5.98	2500	0.1188	0.4801	0.6748	0.2665
0.1114	7.18	3000	0.1183	0.5071	0.6944	0.3016
0.1048	8.37	3500	0.1177	0.5202	0.7032	0.3216
0.0983	9.57	4000	0.1172	0.5344	0.7123	0.3371
0.0917	10.77	4500	0.1164	0.5414	0.7175	0.3458
0.0852	11.96	5000	0.1175	0.5495	0.7240	0.3580
0.0787	13.16	5500	0.1226	0.5545	0.7328	0.3742
0.0738	14.35	6000	0.1233	0.5572	0.7345	0.3761

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3