mpasila/Discord-short-sharegpt download history

mpasila/Discord-short-sharegpt is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 44 times (8 in the last 7 days), and 868 times in total. It ranks #294,341 among datasets by monthly downloads.

Single channel's conversations turned into ShareGPT format from Discord-Data. It has also been optimized for Llama 3.1 tokenizer with each conversation being max 8192 tokens. Since I'm using Unsloth I had to add another adjustment, for some reason it adds a 28 token length system prompt to each conv

Models trained on Discord-short-sharegpt

4 models list it as training data.

Open mpasila/Discord-short-sharegpt on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.