IOST-ASCOL/nepali-datasets ? reverse-engineered prompt

Build me a simple, clean website for the Nepali datasets collection, so people can browse all the resources in one place without digging through a huge README.

I want it to organize the datasets into clear sections like text and NLP, audio and speech, images, geospatial, time series, financial, and special datasets. Each entry should show the dataset name, a short description if available, and a link out to the source. It should also feel easy to scan, with a search box, category filters, and a nice layout for both desktop and mobile.

Please keep the tone polished and helpful, since this is meant to be a public resource for researchers and students. If it makes sense, turn the existing README content into the site content so the repo stays easy to maintain. Use current docs online if you need to.