Benchmarking ASR models for medical use

Organization

• scripts_transcription contains instructions and scripts to perform inference with ASR models

• scripts_simsamu contains instructions and scripts to preprocess to benchmark models with the simsamu dataset

• scripts_evaluation contains instructions and scripts to compute evaluation metrics.

Each subdirectory has its own README.md with instructions.

General requirements

• Python >3.9

Tested with Python 3.12.13

Required packaged can be installed with:

pip install requirements.txt

Other files

• figuring_out.ipynb contains information about the models, requirements etc.

• requirements.txt contains required packages and versions.