bert-relevance-classifier

This model is a fine-tuned version of google-bert/bert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2526
• Accuracy: 0.898
• Auc: N/A

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: 8
• eval_batch_size: 8
• seed: 42
• 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
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Auc
0.367	1.0	627	0.3122	0.851	N/A
0.326	2.0	1254	0.2637	0.889	N/A
0.314	3.0	1881	0.2688	0.888	N/A
0.3106	4.0	2508	0.2499	0.902	N/A
0.2994	5.0	3135	0.2667	0.899	N/A
0.2984	6.0	3762	0.2515	0.9	N/A
0.29	7.0	4389	0.2524	0.903	N/A
0.2838	8.0	5016	0.2594	0.898	N/A
0.2793	9.0	5643	0.2501	0.9	N/A
0.278	10.0	6270	0.2526	0.898	N/A

Framework versions

• Transformers 4.51.3
• Pytorch 2.4.1.post100
• Datasets 3.5.0
• Tokenizers 0.21.0