atdev-company/atdev-serve
在 GitHub 打开 ↗Up to 14.2x faster Qwen3.8-27B inference on a single A100 - a drop-in plugin for vLLM 0.31.0. Models: huggingface.co/atdev-company
Star24
24h 新增+1
7 天新增—
Fork2
Issue0
语言Python
项目简介
Up to 14.2x faster Qwen3.8-27B inference on a single A100 - a drop-in plugin for vLLM 0.31.0. Models: huggingface.co/atdev-company
Star 趋势
话题标签
a100cudallm-inferencequantizationqwenspeculative-decodingvllm
仓库信息
| License | Apache-2.0 |
| 创建时间 | 2026-10-30 |
| 最后推送 | 21 小时前 |
| 热度分 | 8.0 |