bert-base-uncased-airlines-news-multi-label

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: 0.2807
• F1: 0.7124
• Roc Auc: 0.8100
• Accuracy: 0.6766

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: 7e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 150
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
No log	1.0	118	0.2992	0.2412	0.5680	0.5234
No log	2.0	236	0.2628	0.5603	0.7177	0.6255
No log	3.0	354	0.2785	0.5691	0.7044	0.6426
No log	4.0	472	0.2674	0.6309	0.7619	0.6340
0.2379	5.0	590	0.2640	0.6535	0.7768	0.6340
0.2379	6.0	708	0.2929	0.6596	0.7683	0.6596
0.2379	7.0	826	0.2778	0.7059	0.8189	0.6681
0.2379	8.0	944	0.2807	0.7124	0.8100	0.6766
0.0507	9.0	1062	0.3381	0.6688	0.7921	0.6511
0.0507	10.0	1180	0.3160	0.6919	0.8259	0.6468
0.0507	11.0	1298	0.3206	0.7063	0.8045	0.6936
0.0507	12.0	1416	0.3273	0.6943	0.8060	0.6766
0.0115	13.0	1534	0.3408	0.6794	0.7986	0.6638
0.0115	14.0	1652	0.3488	0.6817	0.7971	0.6681
0.0115	15.0	1770	0.3469	0.6962	0.8085	0.6766
0.0115	16.0	1888	0.3517	0.6795	0.7966	0.6596
0.0045	17.0	2006	0.3537	0.6814	0.8011	0.6596
0.0045	18.0	2124	0.3566	0.6857	0.8021	0.6638
0.0045	19.0	2242	0.3587	0.6795	0.7966	0.6596
0.0045	20.0	2360	0.3596	0.6795	0.7966	0.6596

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1