Codes and Datasets of paper "Investigating and Mitigating Catastrophic Forgetting in Medical Knowledge Injection through Internal Knowledge Augmentation Learning"

Requirement

vllm==0.8.5
torch==2.6.0
transformers==4.51.3
pandas==2.2.3

bash generate_datasets.sh

bash run_RefInject.sh
bash run_InternAL.sh

python analysis.py

Name		Name	Last commit message	Last commit date
data		data
primekg		primekg
analysis.py		analysis.py
elicit_internal_knowledge.py		elicit_internal_knowledge.py
evaluate_llama_other_benchmarks.py		evaluate_llama_other_benchmarks.py
evaluate_llama_other_benchmarks_lora.py		evaluate_llama_other_benchmarks_lora.py
evaluate_llama_primekg_probe.py		evaluate_llama_primekg_probe.py
evaluate_llama_primekg_probe_lora.py		evaluate_llama_primekg_probe_lora.py
extract_internal_knowledge.py		extract_internal_knowledge.py
generate_datasets.sh		generate_datasets.sh
generate_internal.py		generate_internal.py
generate_refinject.py		generate_refinject.py
parsing_internal_knowledge.py		parsing_internal_knowledge.py
primekg_relations.xlsx		primekg_relations.xlsx
readme.md		readme.md
run_AlphaEdit.sh		run_AlphaEdit.sh
run_InternAL.sh		run_InternAL.sh
run_MEMIT.sh		run_MEMIT.sh
run_RefInject.sh		run_RefInject.sh
run_RefInject_lora.sh		run_RefInject_lora.sh
train_genft.py		train_genft.py
train_sft.py		train_sft.py
train_sft_lora.py		train_sft_lora.py