ade_biobert_output

This model is a fine-tuned version of jay0911/fine-tuned-aemodel on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3619
• Precision: 0.9353
• Recall: 0.9358
• F1: 0.9355
• Recall Positive: 0.8686
• Recall Negative: 0.9613

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: 5e-05
• 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
• lr_scheduler_warmup_steps: 500
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Recall Positive	Recall Negative
0.1921	0.2126	500	0.2565	0.9347	0.9332	0.9337	0.9147	0.9412
0.1893	0.4252	1000	0.2461	0.9409	0.9392	0.9397	0.9289	0.9436
0.2207	0.6378	1500	0.2583	0.9421	0.9418	0.9419	0.9104	0.9551
0.1706	0.8503	2000	0.3926	0.9216	0.9205	0.9183	0.7866	0.9776
0.1219	1.0629	2500	0.3413	0.9373	0.9354	0.9359	0.9246	0.9400
0.1097	1.2755	3000	0.3073	0.9453	0.9456	0.9453	0.8919	0.9685
0.1645	1.4881	3500	0.2700	0.9433	0.9430	0.9431	0.9118	0.9563
0.2348	1.7007	4000	0.2449	0.9452	0.9456	0.9452	0.8876	0.9703
0.2718	1.9133	4500	0.2304	0.9425	0.9426	0.9425	0.8990	0.9612

Framework versions

• Transformers 4.55.0
• Pytorch 2.6.0+cu124
• Datasets 4.0.0
• Tokenizers 0.21.4