ner-biomedical-maccrobat2018

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: 0.7250
• Accuracy: 0.7665

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.6575	1.0	10	1.7587	0.2575
1.4576	2.0	20	1.1109	0.5523
0.9363	3.0	30	0.8227	0.6823
0.6519	4.0	40	0.7157	0.7200
0.4871	5.0	50	0.6950	0.7298
0.3808	6.0	60	0.6697	0.7455
0.3106	7.0	70	0.6761	0.7481
0.2558	8.0	80	0.6697	0.7568
0.2151	9.0	90	0.6689	0.7546
0.1856	10.0	100	0.6743	0.7587
0.1617	11.0	110	0.7075	0.7569
0.1426	12.0	120	0.7087	0.7561
0.1282	13.0	130	0.6996	0.7624
0.1163	14.0	140	0.7200	0.7646
0.1068	15.0	150	0.7044	0.7673
0.0968	16.0	160	0.7150	0.7641
0.0934	17.0	170	0.7239	0.7673
0.0871	18.0	180	0.7264	0.7652
0.085	19.0	190	0.7269	0.7655
0.083	20.0	200	0.7250	0.7665

Framework versions

• Transformers 4.50.3
• Pytorch 2.6.0+cu124
• Tokenizers 0.21.1