rom1504

Easily turn large sets of image urls to an image dataset. Can download, resize and package 100M urls in 20h on one machine.

Easily compute clip embeddings and build a clip retrieval system with them

Easily convert common crawl to a dataset of caption and document. Image/text Audio/text Video/text, ...

Get hundred of million of image+url from the crawling at home dataset and preprocess them

Using efficientnet to provide embeddings for retrieval

Semantic search with embeddings: index anything

Minecraft web based chat client

Efficiently read embedding in streaming from any filesystem

bot made with mineflayer which can do task

gpu tester detects broken and slow gpus in a cluster

Dalle service