add self.subsets = ['med_qa_en_source'] in MedQA class in evaluation/benchmarks.py to fix the error datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset when evaluation on medqa
add wandb, scikit-learn packages in requirements.txt and remove unused import statement import openai in evaluation/inference.py
xxrjun
commented
3 months ago
PR to fix #38 , solution provided by @JulienVig
self.subsets = ['med_qa_en_source']inMedQAclass inevaluation/benchmarks.pyto fix the errordatasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetwhen evaluation onmedqawandb,scikit-learnpackages inrequirements.txtand remove unused import statementimport openaiinevaluation/inference.py