Fine_tune_PubMedBert

This model is a fine-tuned version of microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4669
• Precision: 0.6359
• Recall: 0.7044
• F1: 0.6684
• Accuracy: 0.8802

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	11	0.8690	0.3548	0.0401	0.0721	0.7691
No log	2.0	22	0.6036	0.6005	0.4635	0.5232	0.8468
No log	3.0	33	0.4788	0.6160	0.5912	0.6034	0.8678
No log	4.0	44	0.4621	0.5331	0.6898	0.6014	0.8611
No log	5.0	55	0.4319	0.5795	0.6916	0.6306	0.8681
No log	6.0	66	0.4444	0.5754	0.7099	0.6356	0.8694
No log	7.0	77	0.4472	0.6069	0.7099	0.6543	0.8756
No log	8.0	88	0.4556	0.6227	0.6898	0.6545	0.8786
No log	9.0	99	0.4613	0.6118	0.7190	0.6611	0.8767
No log	10.0	110	0.4669	0.6359	0.7044	0.6684	0.8802

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0