biomedical-ner-all-finetuned-ner-combined-v1

This model is a fine-tuned version of d4data/biomedical-ner-all on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2720
• Precision: 0.7536
• Recall: 0.7415
• F1: 0.7475
• Accuracy: 0.9264

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	240	0.2967	0.6682	0.6905	0.6791	0.9113
No log	2.0	480	0.2534	0.7135	0.7351	0.7241	0.9210
0.3367	3.0	720	0.2390	0.7477	0.7281	0.7378	0.9236
0.3367	4.0	960	0.2386	0.7335	0.7462	0.7398	0.9235
0.1877	5.0	1200	0.2473	0.7302	0.7597	0.7446	0.9243
0.1877	6.0	1440	0.2435	0.7449	0.7477	0.7463	0.9261
0.1508	7.0	1680	0.2535	0.7251	0.7712	0.7474	0.9246
0.1508	8.0	1920	0.2564	0.7331	0.7604	0.7465	0.9254
0.1206	9.0	2160	0.2642	0.7435	0.7481	0.7458	0.9256
0.1206	10.0	2400	0.2720	0.7536	0.7415	0.7475	0.9264
0.0984	11.0	2640	0.2866	0.7462	0.7431	0.7446	0.9254
0.0984	12.0	2880	0.2982	0.7332	0.7561	0.7445	0.9250
0.08	13.0	3120	0.3123	0.7289	0.7584	0.7434	0.9252
0.08	14.0	3360	0.3167	0.7239	0.7633	0.7430	0.9237
0.0658	15.0	3600	0.3238	0.7408	0.7499	0.7453	0.9252

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 3.6.0
• Tokenizers 0.22.1