Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jan 22, 2024

• Add a 1-line docstring to explain why calling context_attention_fwd twice in test_prefix_prefill.py (vllm-project#2553)

  JasonZhu1313

  authored
  7a0b011

• Fix progress bar and allow HTTPS in benchmark_serving.py (vllm-project#2552)

  hmellor

  authored
  63e835c

• Add qwen2 (vllm-project#2495)

  JustinLin610

  authored
  94b5ede

• Fix vllm-project#2540 (vllm-project#2545)

  pcmoritz

  authored
  ab7e600

• [Speculative decoding 2/9] Multi-step worker for draft model (vllm-project#2424)

  cadedaniel

  authored
  18bfcdd

• migrate pydantic from v1 to v2 (vllm-project#2531)

  joennlae

  authored
  71d63ed

Commits on Jan 21, 2024

• [Fix] Keep scheduler.running as deque (vllm-project#2523)

  njhill

  authored
  d75c407

• Add group as an argument in broadcast ops (vllm-project#2522)

  GindaChen

  authored
  5b23c3f

Commits on Jan 20, 2024

• Add benchmark serving to CI (vllm-project#2505)

  simon-mo

  authored
  00efdc8

• [Bugfix] fix load local safetensors model (vllm-project#2512)

  esmeetu

  authored
  91a61da

Commits on Jan 19, 2024

• Simplify broadcast logic for control messages (vllm-project#2501)

  zhuohan123

  authored
  ef9b636

• Support OpenAI API server in benchmark_serving.py (vllm-project#2172)

  hmellor

  authored
  2709c00

• refactor complemention api for readability (vllm-project#2499)

  simon-mo

  authored
  dd7e8f5

Commits on Jan 18, 2024

• [BugFix] Fix abort_seq_group (vllm-project#2463)

  beginlner

  authored
  d2a6836

• Don't download both safetensor and bin files. (vllm-project#2480)

  NikolaBorisov

  authored
  7e10811

• [Neuron] Add an option to build with neuron (vllm-project#2065)

  liangfu

  authored
  18473cf

• fix: fix some args desc (vllm-project#2487)

  zspo

  authored
  4df417d

• Minor fix in prefill cache example (vllm-project#2494)

  JasonZhu1313

  authored
  5d80a91

• fix stablelm.py tensor-parallel-size bug (vllm-project#2482)

  YingchaoX

  authored
  8a25d3a

• [Experimental] Prefix Caching Support (vllm-project#1669)

  Show description for d10f8e1
  

  3 peopleauthored
  d10f8e1

Commits on Jan 17, 2024

• OpenAI Server refactoring (vllm-project#2360)

  FlorianJoncour

  authored
  14cc317

• Add StableLM3B model (vllm-project#2372)

  ita9naiwa

  authored
  e1957c6

Commits on Jan 16, 2024

• ci: retry on build failure as well (vllm-project#2457)

  simon-mo

  authored
  8cd5a99

• CI: make sure benchmark script exit on error (vllm-project#2449)

  simon-mo

  authored
  947f0b2

Commits on Jan 15, 2024

• fix weigit loading for GQA with TP (vllm-project#2379)

  zhangch9

  authored
  f780504

• Allow buildkite to retry build on agent lost (vllm-project#2446)

  simon-mo

  authored
  bfc072a

• Announce the second vLLM meetup (vllm-project#2444)

  WoosukKwon

  authored
  2a18da2

Commits on Jan 14, 2024

• [CI] Add Buildkite (vllm-project#2355)

  simon-mo

  authored
  6e01e8c

• [Minor] Optimize cuda graph memory usage (vllm-project#2437)

  esmeetu

  authored
  9f659bf

Commits on Jan 12, 2024

• [Minor] Fix err msg (vllm-project#2431)

  WoosukKwon

  authored
  35c4bc2

• Aligning top_p and top_k Sampling (vllm-project#1885)

  Show description for 218dc2c
  chenxu2048

  authored
  218dc2c

• Update quickstart.rst (vllm-project#2369)

  nautsimon

  authored
  827cbcd

• Suggest using dtype=half when OOM.

  chuanzhubin

  authored
  cb7a1c1

• Address Phi modeling update 2 (vllm-project#2428)

  huiwy

  authored
  7878958

• Allow setting fastapi root_path argument (vllm-project#2341)

  chiragjn

  authored
  ce03624

Pagination

PreviousNext