kbulutozler/distilbert-base-uncased-FT-ner-JNLPBA

Model Card for Model ID

Fine-tuned distilbert model. Trained on train set of JNLPBA dataset taken from BLURB.

Model Details

Model Sources [optional]

• Repository: https://github.com/kbulutozler/medical-llm-benchmark

Training Details

Training Data

Train set of JNLPBA dataset.

Training Procedure

Classical fine-tuning.

Training Hyperparameters

• Training regime: [More Information Needed]

learning_rate=5e-5 per_device_train_batch_size=16 per_device_eval_batch_size=16 num_train_epochs=3 weight_decay=0.01

Evaluation

Testing Data

Test set of JNLPBA dataset.

Results

Precision: 0.73 Recall: 0.83 Micro-F1: 0.78

Environmental Impact

• Hardware Type: 1xRTX A4000
• Hours used: 00:19:00