Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 23, 2024

• Bump version to v0.5.5 (vllm-project#7823)

  simon-mo

  authored
  09c7792

• [Misc] Update marlin to use vLLMParameters (vllm-project#7803)

  dsikka

  authored
  f1df5db

• [github][misc] promote asking llm first (vllm-project#7809)

  youkaichao

  authored
  35ee2ad

• [BugFix] Fix server crash on empty prompt (vllm-project#7746)

  Show description for e25fee5
  maxdebayser

  authored
  e25fee5

• [misc] Add Torch profiler support for CPU-only devices (vllm-project#7806)

  DamonFool

  authored
  faeddb5

• [Hardware][Intel GPU] refactor xpu_model_runner for tp (vllm-project#7712)

  jikunshang

  authored
  fc5ebbd

• [Ray backend] Better error when pg topology is bad. (vllm-project#7584)

  Show description for c01a6cb
  

  rkooo567

  and

  youkaichao

  authored
  c01a6cb

Commits on Aug 22, 2024

• [Frontend] error suppression cleanup (vllm-project#7786)

  Show description for b903e1b
  joerunde

  authored
  b903e1b

• [Misc] fix typo in triton import warning (vllm-project#7794)

  lsy323

  authored
  a152246

• [ci] Cleanup & refactor Dockerfile to pass different Python versions and sccache bucket via build args (vllm-project#7705)

  Show description for 666ad0a
  khluu

  authored
  666ad0a

• [Bugfix] Use LoadFormat values for vllm serve --load-format (vllm-project#7784)

  mgoin

  authored
  15310b5

• [Doc] Fix incorrect docs from vllm-project#7615 (vllm-project#7788)

  petersalas

  authored
  57792ed

• [Misc] Enhance prefix-caching benchmark tool (vllm-project#6568)

  Jeffwan

  authored
  d3b5b98

• [Bugfix] spec decode handle None entries in topk args in create_sequence_group_output (vllm-project#7232)

  Show description for cc0eaf1
  tjohnson31415

  authored
  cc0eaf1

• [Misc] update fp8 to use vLLMParameter (vllm-project#7437)

  dsikka

  authored
  955b519

• [Bugfix] Don't build machete on cuda <12.0 (vllm-project#7757)

  LucasWilkinson

  authored
  55d63b1

• Fix ShardedStateLoader for vllm fp8 quantization (vllm-project#7708)

  sfc-gh-zhwang

  authored
  4f419c0

• [Speculative Decoding] EAGLE Implementation with Top-1 proposer (vllm-project#6830)

  abhigoyal1997

  authored
  a3fce56

• [Misc] Use torch.compile for GemmaRMSNorm (vllm-project#7642)

  WoosukKwon

  authored
  b3856be

• [ci] refine dependency for distributed tests (vllm-project#7776)

  youkaichao

  authored
  8c6f694

• [TPU] Avoid initializing TPU runtime in is_tpu (vllm-project#7763)

  WoosukKwon

  authored
  eeee1c3

• Revert "[Kernel] Expand MoE weight loading + Add Fused Marlin MoE Kernel (vllm-project#7527)" (vllm-project#7764)

  mgoin

  authored
  aae74ef

• [Bug][Frontend] Improve ZMQ client robustness (vllm-project#7443)

  Show description for cde9183
  joerunde

  authored
  cde9183

• [Model] Fix Phi-3.5-vision-instruct 'num_crops' issue (vllm-project#7710)

  zifeitong

  authored
  df1a211

• [Kernel] Replaced blockReduce[...] functions with cub::BlockReduce (vllm-project#7233)

  Show description for 7937009
  

  ProExpertProg

  and

  mgoin

  authored
  7937009

Commits on Aug 21, 2024

• [AMD][CI/Build] Disambiguation of the function call for ROCm 6.2 headers compatibility (vllm-project#7477)

  Show description for 9984605
  

  gshtras

  and

  charlifu

  authored
  9984605

• [distributed][misc] error on same VLLM_HOST_IP setting (vllm-project#7756)

  youkaichao

  authored
  7eebe8c

• [Kernel] Expand MoE weight loading + Add Fused Marlin MoE Kernel (vllm-project#7527)

  Show description for 8678a69
  

  dsikka

  and

  ElizaWszola

  authored
  8678a69

• [ci] [multi-step] narrow multi-step test dependency paths (vllm-project#7760)

  SolitaryThinker

  authored
  5844017

• [Model] Add UltravoxModel and UltravoxConfig (vllm-project#7615)

  petersalas

  authored
  1ca0d4f

• [misc] Add Torch profiler support (vllm-project#7451)

  Show description for dd53c4b
  

  SolitaryThinker

  and

  comaniac

  authored
  dd53c4b

• [Frontend] Improve Startup Failure UX (vllm-project#7716)

  robertgshaw2-neuralmagic

  authored
  970dfdc

• [multi-step] Raise error if not using async engine (vllm-project#7703)

  SolitaryThinker

  authored
  91f4522

• [Bugfix] Pass PYTHONPATH from setup.py to CMake (vllm-project#7730)

  sasha0552

  authored
  1b32e02

• [Bugfix][Frontend] Fix Issues Under High Load With zeromq Frontend (vllm-project#7394)

  Show description for f7e3b0c
  

  robertgshaw2-neuralmagic

  and

  njhill

  authored
  f7e3b0c

Pagination

PreviousNext