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: 0.6384
• Accuracy: 0.8007

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
2.7438	1.0	10	1.8410	0.2970
1.4729	2.0	20	1.0655	0.5852
0.9161	3.0	30	0.7775	0.7075
0.6509	4.0	40	0.6808	0.7477
0.4923	5.0	50	0.6315	0.7603
0.3818	6.0	60	0.6120	0.7756
0.3096	7.0	70	0.6025	0.7742
0.2546	8.0	80	0.5992	0.7861
0.2089	9.0	90	0.6075	0.7883
0.178	10.0	100	0.6149	0.7877
0.159	11.0	110	0.6219	0.8012
0.139	12.0	120	0.6282	0.7997
0.1239	13.0	130	0.6222	0.7970
0.1115	14.0	140	0.6311	0.7915
0.1015	15.0	150	0.6336	0.7976
0.0958	16.0	160	0.6321	0.7955
0.0898	17.0	170	0.6352	0.7990
0.0874	18.0	180	0.6464	0.7981
0.0841	19.0	190	0.6380	0.7992
0.0819	20.0	200	0.6384	0.8007

Framework versions

• Transformers 4.48.2
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0