Repository navigation
chore(deploy): 新增自建识别服务,一条命令在 GPU 服务器上部署 Qwen3-ASR - #184
Merged
Merged
Conversation
deploy/selfhost/ 部署 vLLM 与 Qwen3-ASR,前面是检查访问令牌、按地址限流的 nginx 网关和 Caddy(有域名时 HTTPS);网站新增「自建识别服务」页面,附实测的用时、显存和调优说明。
Codecov Report✅ All modified and coverable lines are covered by tests. @@ Coverage Diff @@
## main #184 +/- ##
==========================================
- Coverage 93.79% 93.79% -0.01%
==========================================
Files 388 388
Lines 55271 55271
Branches 3743 3743
==========================================
- Hits 51842 51840 -2
- Misses 2933 2935 +2
Partials 496 496
Flags with carried forward coverage won't be shown. Click here to find out more. 🚀 New features to boost your workflow:
|
This was referenced Oct 24, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
概要
deploy/selfhost/:一条命令在装有 NVIDIA 显卡的 Linux 服务器上部署 vLLM 和 Qwen3-ASR,前面是 nginx 网关和 Caddy(有域名时自动 HTTPS)。Voltip 以「自定义接口」使用它。tokens文件中的令牌(日志只记行号),只开放POST /v1/audio/transcriptions和GET /v1/models,按客户端地址限流(只信任本机 Caddy 发来的X-Forwarded-For),上传 32 MB 以内且只在内存中处理,识别超过 120 秒返回 504。install.sh:检查显卡、驱动和 CUDA 版本,选择模型(8 GB 显存起默认 1.7B)、精度(计算能力低于 8.0 用 float16)和 vLLM 镜像(CUDA 13 或 12.9);--install-deps时安装 nvidia-utils、Docker 和 NVIDIA Container Toolkit;重跑即升级,令牌和设置保留。docker run+docker commit加上vllm[audio]:用docker build时 BuildKit 缓存会多留一份约 23 GB 的基础镜像。verify-all的 web 组新增selfhost-gateway(需要 Docker)。测试
deploy/selfhost/test_gateway.py:9 个测试,用真实 nginx 镜像加 vLLM 桩。去掉令牌检查和真实地址信任后有 4 个失败。curl | sh全新安装 12 分 20 秒,按页面的方法卸载后没有残留。未覆盖
config/zh.ts、config/en.ts中手写,内容同步后另开 firlab PR 加上本页。