cva-quant-weighted-classifier

This model is a fine-tuned version of alex-miller/ODABert on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1539
• Accuracy: 0.9643
• F1: 0.9630
• Precision: 0.9286
• Recall: 1.0

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: 6e-06
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.6832	1.0	4	0.6577	0.6786	0.7097	0.6111	0.8462
0.6332	2.0	8	0.6121	0.8571	0.8462	0.8462	0.8462
0.587	3.0	12	0.5636	0.8571	0.8462	0.8462	0.8462
0.5308	4.0	16	0.5053	0.8571	0.8462	0.8462	0.8462
0.4738	5.0	20	0.4425	0.8929	0.8800	0.9167	0.8462
0.3972	6.0	24	0.3848	0.8929	0.8800	0.9167	0.8462
0.3347	7.0	28	0.3371	0.8929	0.8800	0.9167	0.8462
0.2769	8.0	32	0.2950	0.8929	0.8800	0.9167	0.8462
0.2321	9.0	36	0.2621	0.8929	0.8800	0.9167	0.8462
0.1847	10.0	40	0.2343	0.8929	0.8800	0.9167	0.8462
0.1524	11.0	44	0.2120	0.8929	0.8800	0.9167	0.8462
0.1374	12.0	48	0.1935	0.8929	0.8800	0.9167	0.8462
0.1112	13.0	52	0.1792	0.9286	0.9231	0.9231	0.9231
0.0881	14.0	56	0.1687	0.9643	0.9630	0.9286	1.0
0.0785	15.0	60	0.1623	0.9643	0.9630	0.9286	1.0
0.065	16.0	64	0.1585	0.9643	0.9630	0.9286	1.0
0.0625	17.0	68	0.1570	0.9643	0.9630	0.9286	1.0
0.0566	18.0	72	0.1554	0.9643	0.9630	0.9286	1.0
0.0587	19.0	76	0.1544	0.9643	0.9630	0.9286	1.0
0.0537	20.0	80	0.1539	0.9643	0.9630	0.9286	1.0

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1