.
├── indatasets # Input dataset (NQ, TQA, WEBQ)
├── outdatasets # Output dataset (NQ, TQA, WEBQ)
└── main.py # Main code file
Use the following command to retrieve generated documents using the Document Generator Retriever (DGR):
python main.py --dataset [nq, tqa, webq] --split [train, dev, test] --model_name [sentence-transformers/gtr-t5-large, sentence-transformers/all-MiniLM-L6-v2] --top_k [3, 5]
Feel free to customize the parameters according to your needs.