Commits · 04ec6ba2ac7a4e4beee8be9dc15bc1922544ca82 · zhaoyu6 / sglang

"git@developer.sourcefind.cn:gaoqiong/yaml-cpp.git" did not exist on "7e129c9b6495a987019738fe32997f8184a1e15e"

25 Jul, 2024 2 commits
- Fix dockerfile and triton cache manager (#720) · 04ec6ba2
  Liangsheng Yin authored Jul 25, 2024
  
  04ec6ba2
- fix: fp8 config (#723) · d63f13c1
  Ying Sheng authored Jul 25, 2024
  
  d63f13c1
24 Jul, 2024 8 commits
- Bump version to 0.1.24 (#718) · 459abad2
  Ying Sheng authored Jul 24, 2024
  
  459abad2
- Improve benchmark scripts (#717) · 30d8e130
  Ying Sheng authored Jul 24, 2024
  
  30d8e130
- fix: llama 3.1 405b fp8 (#714) · e17deb27
  Yineng Zhang authored Jul 25, 2024
  
  e17deb27
- format · 83d2b30d
  Ying Sheng authored Jul 24, 2024
  
  83d2b30d
- Fix prefill size (#711) · 4367f4bb
  Ying Sheng authored Jul 24, 2024
  
  4367f4bb
- Update schedule_heuristic.py · 00e4baa7
  Lianmin Zheng authored Jul 24, 2024
  
  00e4baa7
- Auto adjust new ratio (#708) · 4cd64b8e
  Liangsheng Yin authored Jul 23, 2024
  
  4cd64b8e
- Fix multi-node deadlock (#709) · 01d66ae2
  Lianmin Zheng authored Jul 23, 2024
  
  01d66ae2
23 Jul, 2024 7 commits
- Reduce hardcoded logic of kernel usage (#707) · a523a3c1
  Mingyi authored Jul 23, 2024
  
  a523a3c1
- bump version to 0.1.23 (#706) · 9f94728f
  Ying Sheng authored Jul 23, 2024
  
  9f94728f
- Update vllm version to support llama3.1 (#705) · 444a0244
  Ying Sheng authored Jul 23, 2024
  
  444a0244
- feat: add e2e latency (#704) · fa7ccb33
  zhyncs authored Jul 24, 2024
  
  fa7ccb33
- Use min new token ratio at start (#701) · 26868443
  Liangsheng Yin authored Jul 23, 2024
  
  26868443
- Fix hf config loading (#702) · 824a77d0
  Ke Bao authored Jul 24, 2024
  
  824a77d0
- Fix flashinfer (#700) · cf99eab7
  Ying Sheng authored Jul 23, 2024
  
  cf99eab7
22 Jul, 2024 5 commits
- misc: fix typo (#698) · 9fdea29d
  zhyncs authored Jul 23, 2024
  
  9fdea29d
- Fix trt benchmark (#697) · df7c4c19
  Ying Sheng authored Jul 22, 2024
  
  df7c4c19
- Tune params (#696) · c3f1aac8
  Ying Sheng authored Jul 22, 2024
  
  c3f1aac8
- misc: update output token logic (#695) · d198791f
  zhyncs authored Jul 22, 2024
  
  d198791f
- fix: update bench serving (#694) · c07526e4
  zhyncs authored Jul 22, 2024
  
  c07526e4
21 Jul, 2024 8 commits
- Support Mistral-Nemo (#691) · 5303c1ed
  Ke Bao authored Jul 22, 2024
  
  5303c1ed
- misc: recommend to use chat model for benchmark (#690) · 65bd1338
  zhyncs authored Jul 22, 2024
  
  65bd1338
- Support Deepseek MoE Model (#689) · eedc12e1
  Liangsheng Yin authored Jul 21, 2024
  
  eedc12e1
- misc: update output file logic (#686) · 6a846bb1
  zhyncs authored Jul 21, 2024
  
  6a846bb1
- feat: update bench serving (#685) · 0fdb3127
  zhyncs authored Jul 21, 2024
  
  0fdb3127
- Fix StreamExecutor.fork() losing the current role start index. (#684) · 5ad033a0
  Max Shawabkeh authored Jul 20, 2024
  
  5ad033a0
- support non-streaming benchmark (#682) · 77e592e8
  Lianmin Zheng authored Jul 20, 2024
  
  77e592e8
- Support gpt-bigcode model class (#681) · caaad53b
  Liangsheng Yin authored Jul 20, 2024
  
  caaad53b
20 Jul, 2024 10 commits
- Decouple kv (#679) · 69d19188
  Liangsheng Yin authored Jul 20, 2024
  
  69d19188
- feat: support TRT LLM benchmark and multiple benchmarks (#670) · 4b4a67f8
  zhyncs authored Jul 21, 2024
  
  4b4a67f8
- Fallback when sampling failed (#678) · 0ac94c36
  Ke Bao authored Jul 21, 2024
  
  0ac94c36
- Update version to 0.1.22 (#677) · 2b4c6462
  Ying Sheng authored Jul 20, 2024
  
  2b4c6462
- Fix illegal tokens during sampling (#676) · f424e76d
  Liangsheng Yin authored Jul 20, 2024
  
  f424e76d
- Fix cuda graph with flashinfer (#675) · 490a1f39
  Lianmin Zheng authored Jul 20, 2024
  
  490a1f39
- refactor model loader: initial refactor (#664) · 06487f12
  Ying Sheng authored Jul 20, 2024
  
  06487f12
- Revert "Temporary fix invalid sample results" (#673) · 39c57317
  Liangsheng Yin authored Jul 20, 2024
  
  39c57317
- Fix random dataset (#671) · 9592a1f3
  Lianmin Zheng authored Jul 20, 2024
  
  9592a1f3
- Support random dataset in bench_serving.py (#669) · 35759efa
  Lianmin Zheng authored Jul 20, 2024
  
  35759efa