What does YOUR_INPUT_FILE look like? Can you provide an example? Thanks very much!

[XiaozhuLove]

cd retrieval_lm python passage_retrieval.py \ --model_name_or_path facebook/contriever-msmarco --passages psgs_w100.tsv \ --passages_embeddings "wikipedia_embeddings/*" \ --data YOUR_INPUT_FILE \ --output_dir YOUR_OUTPUT_FILE \ --n_docs 20

[innovation64]

cd retrieval_lm python passage_retrieval.py --model_name_or_path facebook/contriever-msmarco --passages psgs_w100.tsv --passages_embeddings "wikipedia_embeddings/*" --data YOUR_INPUT_FILE --output_dir YOUR_OUTPUT_FILE --n_docs 20

its script has some problems, it doesn't have --data argument, only has --dataset or --query, you need change it ,and i try to run it just by simple jsonl or json file with questions @XiaozhuLove

[WiFi-Lost]

@innovation64 Can I just use the json file in eval_data or Could you please share your json file?

cd retrieval_lm python passage_retrieval.py --model_name_or_path facebook/contriever-msmarco --passages psgs_w100.tsv --passages_embeddings "wikipedia_embeddings/*" --data YOUR_INPUT_FILE --output_dir YOUR_OUTPUT_FILE --n_docs 20

its script has some problems, it doesn't have --data argument, only has --dataset or --query, you need change it ,and i try to run it just by simple jsonl or json file with questions @XiaozhuLove

[chanmin0723]

They may have mistakenly written --data instead of --query. If you use --query YOUR_INPUT_FILE, you can see that it runs correctly, as shown in the image below.