Commits · 1edd4e07d6ad52f4f63e7f6beaa5987c1e1cf621 · change / sglang · GitLab

30 Jul, 2024 10 commits
- chore: bump v0.2.7 (#830) · 1edd4e07
  Yineng Zhang authored Jul 30, 2024
  
  1edd4e07
- misc: update e2e test benchmark config (#825) · f52eda35
  Yineng Zhang authored Jul 30, 2024
  
  f52eda35
- Add awq_marlin (#826) · b579ecf0
  Ying Sheng authored Jul 30, 2024
  
  b579ecf0
- Adjust default mem fraction to avoid OOM (#823) · e7487b08
  Ying Sheng authored Jul 30, 2024
  
  e7487b08
- Support disable_ignore_eos in bench_serving.py (#824) · ae5c0fc4
  Ying Sheng authored Jul 30, 2024
  
  ae5c0fc4
- Fix streaming bug (#820) · daf593a3
  ObjectNotFound authored Jul 30, 2024
  
  daf593a3
- Code structure refactor (#807) · cdcbde5f
  Liangsheng Yin authored Jul 29, 2024
  
  cdcbde5f
- Fix LiteLLM kwargs (#817) · 21e22b9e
  Enrique Shockwave authored Jul 30, 2024
  
  21e22b9e
- Revert "Organize public APIs" (#815) · db6089e6
  Ying Sheng authored Jul 29, 2024
  
  db6089e6
- Remove inf value for chunked prefill size (#812) · 3520f75f
  Liangsheng Yin authored Jul 29, 2024
  
  3520f75f
29 Jul, 2024 14 commits
- Organize public APIs (#809) · c8e9fed8
  Liangsheng Yin authored Jul 29, 2024
  
  c8e9fed8
- Add support for OpenAI API : offline batch(file) processing (#699) · 084fa54d
  yichuan~ authored Jul 30, 2024
```
Co-authored-by: hnyls2002 <hnyls2002@gmail.com>
```
  084fa54d
- Revert "Revert "fix: update flashinfer to 0.1.2 to fix sampling for cu118"" (#806) · eba458bd
  Ying Sheng authored Jul 29, 2024
  
  eba458bd
- feat: add chat template for internlm2-chat (#802) · 3d1cb0af
  Yineng Zhang authored Jul 30, 2024
  
  3d1cb0af
- Revert "fix: update flashinfer to 0.1.2 to fix sampling for cu118" (#805) · 7d352b4f
  Ying Sheng authored Jul 29, 2024
  
  7d352b4f
- fix: update flashinfer to 0.1.2 to fix sampling for cu118 (#803) · 87064015
  Yineng Zhang authored Jul 30, 2024
  
  87064015
- Chunked prefill (#800) · 7cd4f244
  Liangsheng Yin authored Jul 29, 2024
  
  7cd4f244
- Revert "Chunked prefill support" (#799) · 98111fbe
  Ying Sheng authored Jul 29, 2024
  
  98111fbe
- Chunked prefill support (#797) · 2ec39ab7
  Liangsheng Yin authored Jul 29, 2024
  
  2ec39ab7
- Add role documentation, add system begin & end tokens (#793) · 8f6274c8
  ObjectNotFound authored Jul 29, 2024
  
  8f6274c8
- Fix logging (#796) · 325a06c2
  Ying Sheng authored Jul 28, 2024
  
  325a06c2
- Fix lazy import location (#795) · 79f81629
  Ying Sheng authored Jul 28, 2024
  
  79f81629
- Lazy-import third-party backends (#794) · b688fd85
  Eric Yoon authored Jul 29, 2024
  
  b688fd85
- Fix echo + lobprob for OpenAI API when the prompt is a list (#791) · 8d908a93
  Ying Sheng authored Jul 28, 2024
  
  8d908a93
28 Jul, 2024 11 commits
- chore: add copyright for srt (#790) · dd7e8b94
  Yineng Zhang authored Jul 28, 2024
  
  dd7e8b94
- Vectorize logprobs computation (#787) · c71880f8
  Ying Sheng authored Jul 28, 2024
  
  c71880f8
- docs: init readthedocs support (#783) · 94862579
  Yineng Zhang authored Jul 28, 2024
  
  94862579
- fix: replace pillow with PIL in PACKAGE_LIST (#781) · 68e52626
  Yineng Zhang authored Jul 28, 2024
  
  68e52626
- Bump version to 0.2.6 (#779) · bc1154c3
  Lianmin Zheng authored Jul 27, 2024
  
  bc1154c3
- Allow disabling flashinfer sampling kernel (#778) · 752e6430
  Lianmin Zheng authored Jul 27, 2024
  
  752e6430
- Rename prefill_token_logprobs -> input_token_logprobs; decode_token_logprobs... · 30db99b3
  Lianmin Zheng authored Jul 27, 2024
```
Rename prefill_token_logprobs -> input_token_logprobs; decode_token_logprobs -> output_token_logprobs (#776)
```
  30db99b3
- Fix return_log_probs with cuda graph (#775) · 0a409bd4
  Lianmin Zheng authored Jul 27, 2024
  
  0a409bd4
- minor refactor: move check server args to server_args.py (#774) · e4db4e5b
  Mingyi authored Jul 27, 2024
  
  e4db4e5b
- Move sampling logits to float32 (#773) · bbc07c41
  Lianmin Zheng authored Jul 27, 2024
  
  bbc07c41
- Fix max new tokens (#772) · a036d419
  Lianmin Zheng authored Jul 27, 2024
  
  a036d419
27 Jul, 2024 5 commits
- Fix max_tokens for OpenAI chat completion API (#766) · f95e6617
  Lianmin Zheng authored Jul 27, 2024
  
  f95e6617
- [Minor] Improve the code style in TokenizerManager (#767) · 0736b270
  Lianmin Zheng authored Jul 27, 2024
  
  0736b270
- Fix TransformerTokenizer init for chatglm2 & 3 (#761) · 3fdab919
  Ke Bao authored Jul 27, 2024
  
  3fdab919
- Fix context length (#757) · d9fccfef
  Liangsheng Yin authored Jul 26, 2024
  
  d9fccfef
- Deepseek v2 support (#693) · 679ebcbb
  Liangsheng Yin authored Jul 26, 2024
  
  679ebcbb