ClinicalTrialBioBERT

This model is a fine-tuned version of dmis-lab/biobert-v1.1 on Clinical Trial Texts Dataset.

Model description

A Clinical Trial Language Model.

Intended uses & limitations

Use when you need domain knowledge from the clinical trial domain.

Training and evaluation data

Trained on 500k steps of Clinical Trial Texts Dataset

Perplexity of BioBERT: Perplexity of ClinicalTrialBioBERT:

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 128
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 500000
• mixed_precision_training: Native AMP

Training results

10k step training loss: 0.92 500k step training loss: 0.50

Framework versions

• Transformers 4.20.1
• Pytorch 1.11.0
• Datasets 2.7.1
• Tokenizers 0.12.1