Synthesize Cases

Phenotype set sampling

python sample/phenotype_sampling.py \
    --orpha ../data/Orphanet/disease2phenotype.json \
    --hpo-meta ../data/HPO/HPO_meta.json \
    --out sampled_phenotypes.json \
    --gen-num 3 \
    --max-key 3 \
    --seed 42

These sampled phenotype sets will be used as input to GPT-4o to generate free-text rare disease cases.

Model training

Installation

conda create -n radar python=3.11
conda activate radar

pip install -e ".[torch,metrics]" --no-build-isolation

pip install ninja
MAX_JOBS=4 pip install flash-attn --no-build-isolation --use-pep517

pip install liger-kernel
pip install deepspeed

Training

# SFT
llamafactory-cli train sft.yaml

# DPO
llamafactory-cli train dpo.yaml

Acknowledgement

This repo benefits from the wonderful work LLaMA-Factory.