Commits · 2216a4e5c5e95be00376562237c2324ea1ff5272 · OpenDAS / vllm_cscc · GitLab

23 Oct, 2024 2 commits
- [Model] Support E5-V (#9576) · 831540cf
  Cyrus Leung authored Oct 23, 2024
  
  831540cf
- [Misc] Add an env var VLLM_LOGGING_PREFIX, if set, it will be prepend to all... · 29061ed9
  Flex Wang authored Oct 22, 2024
```
[Misc] Add an env var VLLM_LOGGING_PREFIX, if set, it will be prepend to all logging messages (#9590)
```
  29061ed9
22 Oct, 2024 17 commits
- [BugFix] Fix metrics error for --num-scheduler-steps > 1 (#8234) · b17046e2
  yulei authored Oct 23, 2024
  
  b17046e2
- [Bugfix] fix detokenizer shallow copy (#5919) · 23b899a8
  Aurick Qiao authored Oct 22, 2024
  
  23b899a8
- [torch.compile] auto infer dynamic_arg_dims from type annotation (#9589) · 17c79f3c
  youkaichao authored Oct 22, 2024
  
  17c79f3c
- [BugFix] Prevent exporting duplicate OpenTelemetry spans (#9017) · cd5601ac
  Ronen Schaffer authored Oct 22, 2024
  
  cd5601ac
- [Frontend] Support custom request_id from request (#9550) · 434984e6
  Yuhong Guo authored Oct 23, 2024
```
Co-authored-by: Yuhong Guo <yuhong.gyh@antgroup.com>
```
  434984e6
- [Bugfix] Eagle: change config name for fc bias (#9580) · 08075c34
  gopalsarda authored Oct 22, 2024
  
  08075c34
- [Model][VLM] Initialize support for Mono-InternVL model (#9528) · bb392ea2
  Isotr0py authored Oct 23, 2024
  
  bb392ea2
- [Neuron] [Bugfix] Fix neuron startup (#9374) · 9dbcce84
  xendo authored Oct 22, 2024
```
Co-authored-by: Jerzy Zagorski <jzagorsk@amazon.com>
```
  9dbcce84
- [V1] Implement vLLM V1 [1/N] (#9289) · 6c5af09b
  Woosuk Kwon authored Oct 22, 2024
  
  6c5af09b
- [Hardware][CPU] using current_platform.is_cpu (#9536) · 3ddbe255
  wangshuai09 authored Oct 22, 2024
  
  3ddbe255
- support TP in qwen2 bnb (#9574) · 0d02747f
  chenqianfzh authored Oct 22, 2024
  
  0d02747f
- [Core] Remove evictor_v1 (#9572) · ca30c3c8
  Kuntai Du authored Oct 21, 2024
  
  ca30c3c8
- [Bugfix]: phi.py get rope_theta from config file (#9503) · 74692421
  Falko1 authored Oct 22, 2024
```
Co-authored-by: Isotr0py <2037008807@qq.com>
```
  74692421
- [Bugfix]: serialize config by value for --trust-remote-code (#6751) · b7299011
  Travis Johnson authored Oct 21, 2024
```
Signed-off-by: Travis Johnson <tsjohnso@us.ibm.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
```
  b7299011
- 增加w8a8的triton环境变量控制 · 2009d4a1
  zhuwenwen authored Oct 22, 2024
  
  2009d4a1
- remove notes · f7512877
  zhuwenwen authored Oct 22, 2024
  
  f7512877
- [core] move parallel sampling out from vllm core (#9302) · 76a5e132
  youkaichao authored Oct 21, 2024
  
  76a5e132
21 Oct, 2024 7 commits
- [Frontend] Don't log duplicate error stacktrace for every request in the batch (#9023) · 711f3a78
  Wallas Henrique authored Oct 21, 2024
```
Signed-off-by: Wallas Santos <wallashss@ibm.com>
```
  711f3a78
- [BugFix] Update draft model TP size check to allow matching target TP size (#9394) · 15713e3b
  Nick Hill authored Oct 21, 2024
```
Co-authored-by: Baoyuan Qi <qibaoyuan@126.com>
```
  15713e3b
- [Frontend] Reduce frequency of client cancellation checking (#7959) · 9d9186be
  Nick Hill authored Oct 21, 2024
  
  9d9186be
- [Model][Bugfix] Fix batching with multi-image in PixtralHF (#9518) · 5241aa14
  Michael Goin authored Oct 21, 2024
  
  5241aa14
- [Bugfix][Misc]: fix graph capture for decoder (#9549) · 8ca89548
  yudian0504 authored Oct 22, 2024
  
  8ca89548
- [Model] FalconMamba Support (#9325) · f6b97293
  Dhia Eddine Rhaiem authored Oct 21, 2024
  
  f6b97293
- [Doc] Consistent naming of attention backends (#9498) · 496e991d
  Thomas Parnell authored Oct 21, 2024
```
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
```
  496e991d
20 Oct, 2024 3 commits
- [Kernel] Support sliding window in flash attention backend (#9403) · 4fa3e333
  Chen Zhang authored Oct 20, 2024
  
  4fa3e333
- [Model][Pixtral] Use memory_efficient_attention for PixtralHFVision (#9520) · 962d2c63
  Michael Goin authored Oct 20, 2024
  
  962d2c63
- [Bugfix] Pass json-schema to GuidedDecodingParams and make test stronger (#9530) · 5b59fe0f
  Chen Zhang authored Oct 19, 2024
  
  5b59fe0f
19 Oct, 2024 10 commits
- [Model][Pixtral] Optimizations for input_processor_for_pixtral_hf (#9514) · 8e3e7f27
  Michael Goin authored Oct 19, 2024
  
  8e3e7f27
- update _custom_ops.py · 2c7f740a
  zhuwenwen authored Oct 19, 2024
  
  2c7f740a
- update _custom_ops.py · 5289f7c4
  zhuwenwen authored Oct 19, 2024
  
  5289f7c4
- [Bugfix] Fix missing task for speculative decoding (#9524) · 263d8ee1
  Cyrus Leung authored Oct 19, 2024
  
  263d8ee1
- [Frontend] Support simpler image input format (#9478) · c5eea3c8
  Yue Zhang authored Oct 18, 2024
  
  c5eea3c8
- [Doc] update gpu-memory-utilization flag docs (#9507) · 82c25151
  Joe Runde authored Oct 18, 2024
```
Signed-off-by: Joe Runde <Joseph.Runde@ibm.com>
```
  82c25151
- [Frontend] Avoid creating guided decoding LogitsProcessor unnecessarily (#9521) · 1325872e
  Nick Hill authored Oct 19, 2024
  
  1325872e
- 🐛 fix torch memory profiling (#9516) · 380e1863
  Joe Runde authored Oct 18, 2024
```
Signed-off-by: Joe Runde <Joseph.Runde@ibm.com>
```
  380e1863
- [Bugfix] Fix offline mode when using `mistral_common` (#9457) · 337ed766
  sasha0552 authored Oct 19, 2024
  
  337ed766
- [Kernel] Add env variable to force flashinfer backend to enable tensor cores (#9497) · 0c9a5258
  Thomas Parnell authored Oct 19, 2024
```
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
Co-authored-by: Chih-Chieh Yang <chih.chieh.yang@ibm.com>
Co-authored-by: Cody Yu <hao.yu.cody@gmail.com>
```
  0c9a5258
18 Oct, 2024 1 commit
- [MISC] Consolidate cleanup() and refactor offline_inference_with_prefix.py (#9510) · d11bf435
  Cody Yu authored Oct 18, 2024
  
  d11bf435