让语音合成更快速、更高效。核心价值在于通过vLLM加速IndexTTS的推理过程，显著提升语音合成的速度和并发能力。 单个请求RTF从0.3降至0.1；GPT模型decode速度提升至280 token/s；支持多角色音频混合，为语音合成带来更多创意可能
Python · active 2025-05-03 → 2026-06-16 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 410 days from 2025-05-03 to 2026-06-16. Pushes: 32 total, peak 3 in a day. Pull requests: 10 total, peak 2 in a day. Issues: 109 total, peak 5 in a day. Comments: 181 total, peak 6 in a day. Stars: 524 total, peak 14 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| Ksuriuri | 67 | 30 | 5 | 30 |
| e1732a364fed | 30 | 0 | 2 | 18 |
| hjj-lmx | 9 | 0 | 0 | 1 |
| lhhssd | 7 | 0 | 0 | 6 |
| Bluedyson | 5 | 2 | 0 | 2 |
| AiBoBoMaker | 5 | 0 | 0 | 4 |
| ZVladN | 4 | 0 | 0 | 2 |
| ljtnine | 4 | 0 | 0 | 2 |
| yizhi2632438 | 4 | 0 | 0 | 2 |
| qkwu | 3 | 0 | 0 | 1 |
| vonchenplus | 3 | 0 | 0 | 3 |
| Jandown | 3 | 0 | 0 | 2 |
| robin12jbj | 3 | 0 | 0 | 3 |
| fkmaan-commits | 3 | 0 | 0 | 3 |
| jujunchen | 3 | 0 | 0 | 3 |
| supy0001 | 3 | 0 | 0 | 2 |
| shao-ssq | 3 | 0 | 0 | 3 |
| ghostplant | 3 | 0 | 0 | 2 |
| a779159990 | 3 | 0 | 0 | 2 |
| armynew | 3 | 0 | 0 | 2 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#208unkindled12026-04-16 01:27生成速度非常快，但是和原版对比，我生成的语音有很大的噪音
- Issue comment#197Bluedyson2026-04-07 09:18请问作者，对于indextts2的加速支持预计什么时候能上线。感谢
- Issue comment#204EilinH2026-03-26 05:575090lap上跑并发与串行总耗时一样
- Issue#204cacard2026-03-24 08:585090lap上跑并发与串行总耗时一样
- Issue comment#201EilinH2026-03-09 04:16高并发卡死
- Issue#200Bluedyson2026-03-08 14:17Small factual mistake in README.md
- Issue comment#200Bluedyson2026-03-08 14:17Small factual mistake in README.md
- Issue#200Idiotabtcodes2026-03-07 13:23Small factual mistake in README.md
- Issue comment#1971208054812026-03-07 02:17请问作者，对于indextts2的加速支持预计什么时候能上线。感谢
- Issue comment#198Ksuriuri2026-03-03 15:06Feat: Update vLLM version to 0.16.0 and adjust BigVGAN compilation method to improve inference speed
- Issue comment#196creed24152026-03-02 07:32环境错误
- Issue#196creed24152026-03-02 07:28环境错误
- Issue#195JYHang2026-01-26 09:00V2比V1慢很多，加速好像没有效果
- Issue#194resautu62026-01-08 11:25希望能加入流式推理
- Issue comment#185q5906240-bot2026-01-08 06:41是不是不支持Volta和Turing 架构，要Ampere以上？
- Issue#193q5906240-bot2026-01-08 05:45为什么我docker部署之后都是utf-8的编码 返回的音频数据应该是乱码了
- Issue comment#191liutianye2025-12-30 02:26为什么5090 使用 v2模型 显存占用与参数约定不一致
- Issue comment#180Crestina20012025-12-29 03:44有tts2的各个环境rtf数据吗？一个快速版本（rtf=0.07 on A100）：
- Issue comment#180luluff2025-12-26 23:20有tts2的各个环境rtf数据吗？一个快速版本（rtf=0.07 on A100）：
- Issue comment#189momomoi2025-12-26 07:32为什么1.5版本的比2.0的推理快这么多啊
- Issue#191Degree-212025-12-26 04:02为什么5090 使用 v2模型 显存占用与参数约定不一致
- Issue#190junwei08122025-12-25 07:33编译过程报错
- Issue#190junwei08122025-12-25 03:36编译过程报错
- Issue comment#145DDiabolous2025-12-24 10:43成功部署到A100 上，但是 速度还是不够快。
- Issue#176Busboy31292025-12-21 08:09upsteram index tts support to vllm
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 524 stars here means stars gained during the window, not the repo's star count.