[!NOTE] This is a copy of ZixiQi/DeepSeek-V3-4layers-MTP-FP8 at commit
6553d61, made on 2026-10-30. vLLM CI uses this checkpoint and the original is on a personal account, so we keep a copy here to reduce supply-chain risk.
DeepSeek-V3 architecture with 4 layers + 8 experts per MoE + MTP module + FP8 weights from original model without further tuning
To be used in CI testing
Have the files? Submit a torrent.
Submit a torrent →Have 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 vllm-project/DeepSeek-V3-4layers-MTP-FP8 ./model-folder