lna-lab

lna-lab/gemma4-12b-vllm-sm120

Reproducible recipe: serve abliterated Gemma-4-12B (gemma4_unified) at 50-118 tok/s on no-NVLink Blackwell (SM120) via vLLM nightly + ModelOpt FP8/NVFP4 + MTP spec-decode.

⭐ 14 ⑂ 0 Python Apache-2.0 · 2026-06-07推送
14
Watchers
0
贡献者
0
Commits
0
Releases
0
Open Issues
2026-06-07
最近推送
原文 中文
暂无 README