nlp_te_ner_pubmedbert

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

• Loss: 0.0181
• Precision: 0.9817
• Recall: 0.9840
• F1: 0.9829
• Accuracy: 0.9959

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	21	0.0206	0.9783	0.9832	0.9808	0.9950
No log	2.0	42	0.0209	0.9766	0.9820	0.9793	0.9948
No log	3.0	63	0.0206	0.9788	0.9827	0.9807	0.9949
No log	4.0	84	0.0181	0.9817	0.9840	0.9829	0.9959
No log	5.0	105	0.0200	0.9785	0.9827	0.9806	0.9950
No log	6.0	126	0.0231	0.9705	0.9799	0.9752	0.9933
No log	7.0	147	0.0227	0.9738	0.9789	0.9764	0.9937
No log	8.0	168	0.0234	0.9726	0.9777	0.9751	0.9936
No log	9.0	189	0.0213	0.9768	0.9794	0.9781	0.9945

Framework versions

• Transformers 4.41.2
• Pytorch 2.2.1
• Datasets 2.19.2
• Tokenizers 0.19.1