Reverse Engineering of Supervised Semantic Speech Tokenizer (S3Tokenizer) proposed in CosyVoice
active 2025-03-28 → 2026-05-12 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 411 days from 2025-03-28 to 2026-05-12. Pushes: 14 total, peak 5 in a day. Pull requests: 7 total, peak 3 in a day. Issues: 10 total, peak 2 in a day. Comments: 24 total, peak 8 in a day. Stars: 118 total, peak 6 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 |
|---|---|---|---|---|
| coderabbitai[bot] | 20 | 0 | 0 | 12 |
| xingchensong | 15 | 8 | 4 | 3 |
| Mddct | 10 | 6 | 0 | 3 |
| Li-XYi | 4 | 0 | 0 | 0 |
| zly-idleness | 3 | 0 | 1 | 2 |
| splinter21 | 2 | 0 | 0 | 1 |
| ArvinNotMe | 2 | 0 | 0 | 1 |
| ylzz1997 | 1 | 0 | 0 | 1 |
| dimbleby | 1 | 0 | 1 | 0 |
| lsy-92 | 1 | 0 | 0 | 0 |
| WangGewu | 1 | 0 | 0 | 0 |
| robin1001 | 1 | 0 | 1 | 0 |
| xinwang0304 | 1 | 0 | 0 | 1 |
| trimonster233 | 1 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue#51trimonster2332026-03-11 06:50想问下视频展示中命令行中同时看CPU和GPU还能写代码是怎么做到的
- Issue comment#50coderabbitai[bot]2026-03-07 14:10fix: remove dead cos/sin computation in AudioEncoderV2
- Issue#49Mddct2026-01-21 15:07speech_tokenizer_v3_25hz tokens produce very different reconstruction vs CosyVoice tokens (same shape/length, very different codes)
- Issue comment#48coderabbitai[bot]2025-12-22 11:51cosyvoice3更新了tokenizer
- Issue comment#45splinter212025-12-19 02:44cosyvoice开源v3的版本了，这个仓库会更新吗？
- ReleaseMddct2025-12-18 15:50🎉🎉🎉 S3Tokenizer now support cosyvoice3 🎉🎉🎉
- Issue comment#47ylzz19972025-12-16 07:06feat: support speech_tokenizer_v3_25hz
- Issue comment#47Mddct2025-12-15 16:18feat: support speech_tokenizer_v3_25hz
- Issue#45splinter212025-12-15 09:03cosyvoice开源v3的版本了，这个仓库会更新吗？
- Issue comment#43zly-idleness2025-09-29 14:37Update model_v2.py
- Issue comment#43zly-idleness2025-09-29 14:16Update model_v2.py
- Pull request#43zly-idleness2025-09-29 14:16
- Issue comment#43Mddct2025-09-29 13:56Update model_v2.py : Refine FSQ encode
- Issue comment#42xingchensong2025-09-29 13:28Problem reconstruct audio with s3tokenizer
- Issue#41ArvinNotMe2025-09-26 03:49性能提升问题
- Issue comment#31ArvinNotMe2025-09-26 03:45Model file
- Issue comment#40xinwang03042025-08-14 02:07为什么用这个库提取的结果，和onnx提取的结果有点差别呢？
- Issue comment#39xingchensong2025-08-07 13:00Limit the number of segements when processing long audio segments.
- Issue comment#38xingchensong2025-07-29 02:55V2是否可以将切片长度改为60s？
- Issue#38WangGewu2025-07-29 02:31V2是否可以将切片长度改为60s？
- Pull request#37xingchensong2025-07-10 02:07
- Issue comment#37coderabbitai[bot]2025-07-10 01:39missrate 0.2% -> 0.5%
- Pull request#37xingchensong2025-07-10 01:39
- Pull request#36xingchensong2025-07-10 01:11
- Pull request#36dimbleby2025-07-09 20:50
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 118 stars here means stars gained during the window, not the repo's star count.