This is the model mistralai/Mistral-7B-Instruct-v0.2 which I cut all the intermediate(feed_forward_length) size with 14336 down to 3072, resulting in a ~2.81B model.
It's necessary to pre-train this model, cause at the moment is generating just gibberish.
model-00001-of-00003.safetensors 2.0 GBmodel-00002-of-00003.safetensors 1.9 GBmodel-00003-of-00003.safetensors 1.8 GBHave the files? Submit a torrent.
Submit a torrentHave the complete files and permission to share them? Replace the example folder path. The script checks the files, makes a torrent and submits its magnet and checksums. Keep your torrent client seeding. Get your community key from your settings, and read the script before running it.
curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo Aryanne/Mistral-3B-Instruct-v0.2-init ./model-folder