Commits

Branch selector

 main

User selector

Yard1

Datepicker

All time

Commit History

Commits on Jul 1, 2024

• [Bugfix] Add explicit end_forward calls to flashinfer (vllm-project#6044)

  Yard1

  authored
  c4059ea

  Copy full SHA for c4059ea

• [Bugfix] Use RayActorError for older versions of Ray in RayTokenizerGroupPool (vllm-project#6039)

  Yard1

  authored
  dec6fc6

  Copy full SHA for dec6fc6

Commits on Jun 29, 2024

• [Core] Optimize SequenceStatus.is_finished by switching to IntEnum (vllm-project#5974)

  Yard1

  authored
  7c01f70

  Copy full SHA for 7c01f70

Commits on Jun 25, 2024

• [Core] Add fault tolerance for RayTokenizerGroupPool (vllm-project#5748)

  Yard1

  authored
  67882db

  Copy full SHA for 67882db

Commits on Jun 16, 2024

• Add basic correctness 2 GPU tests to 4 GPU pipeline (vllm-project#5518)

  Yard1

  authored
  f31c1f9

  Copy full SHA for f31c1f9

Commits on Jun 13, 2024

• Add cuda_device_count_stateless (vllm-project#5473)

  Yard1

  authored
  50eed24

  Copy full SHA for 50eed24

• Revert "[Core] Remove unnecessary copies in flash attn backend" (vllm-project#5478)

  Yard1

  authored
  6b0511a

  Copy full SHA for 6b0511a

• Seperate dev requirements into lint and test (vllm-project#5474)

  Yard1

  authored
  a8fda4f

  Copy full SHA for a8fda4f

Commits on Jun 7, 2024

• Remove Ray health check (vllm-project#4693)

  Yard1

  authored
  18a277b

  Copy full SHA for 18a277b

• [Core] Change LoRA embedding sharding to support loading methods (vllm-project#5038)

  Yard1

  authored
  ccdc490

  Copy full SHA for ccdc490

• [Core] Avoid copying prompt/output tokens if no penalties are used (vllm-project#5289)

  Yard1

  authored
  a31cab7

  Copy full SHA for a31cab7

Commits on Jun 3, 2024

• [Core] Remove unnecessary copies in flash attn backend (vllm-project#5138)

  Yard1

  authored
  0ab278c

  Copy full SHA for 0ab278c

Commits on May 21, 2024

• [Core] Fix scheduler considering "no LoRA" as "LoRA" (vllm-project#4897)

  Yard1

  authored
  65ae8c2

  Copy full SHA for 65ae8c2

Commits on May 17, 2024

• [Doc] Update Ray Data distributed offline inference example (vllm-project#4871)

  Yard1

  authored
  c5711ef

  Copy full SHA for c5711ef

Commits on May 8, 2024

• [Core] Faster startup for LoRA enabled models (vllm-project#4634)

  Yard1

  authored
  ad932a2

  Copy full SHA for ad932a2

Commits on Apr 16, 2024

• [Core] Refactor model loading code (vllm-project#4097)

  Yard1

  authored
  69e1d2f

  Copy full SHA for 69e1d2f

Commits on Apr 11, 2024

• Add extra punica sizes to support bigger vocabs (vllm-project#4015)

  Yard1

  authored
  1e96c33

  Copy full SHA for 1e96c33

• [Core] Set linear_weights directly on the layer (vllm-project#3977)

  Yard1

  authored
  a10d305

  Copy full SHA for a10d305

Commits on Mar 25, 2024

• Optimize _get_ranks in Sampler (vllm-project#3623)

  Yard1

  authored
  3a24309

  Copy full SHA for 3a24309

Commits on Mar 22, 2024

• [Core] Improve detokenization performance for prefill (vllm-project#3469)

  Show description for bfdb1ba
  

  Yard1

  and

  MeloYang05

  authored
  bfdb1ba

  Copy full SHA for bfdb1ba

Commits on Mar 20, 2024

• [1/n] Triton sampling kernel (vllm-project#3186)

  Show description for 426ec4e
  

  Yard1

  and

  ywang96

  authored
  426ec4e

  Copy full SHA for 426ec4e

Commits on Mar 19, 2024

• [Core] Cache some utils (vllm-project#3474)

  Yard1

  authored
  b37cdce

  Copy full SHA for b37cdce

Commits on Mar 18, 2024

• [Core] Zero-copy asdict for InputMetadata (vllm-project#3475)

  Yard1

  authored
  49eedea

  Copy full SHA for 49eedea

Commits on Mar 15, 2024

• Asynchronous tokenization (vllm-project#2879)

  Yard1

  authored
  fb96c1e

  Copy full SHA for fb96c1e

Commits on Mar 13, 2024

• Fix lint (vllm-project#3388)

  Yard1

  authored
  c33afd8

  Copy full SHA for c33afd8

Commits on Mar 4, 2024

• Add health check, make async Engine more robust (vllm-project#3015)

  Show description for ff578ca
  

  Yard1

  and

  zhuohan123

  authored
  ff578ca

  Copy full SHA for ff578ca

• Push logprob generation to LLMEngine (vllm-project#3065)

  Show description for 22de452
  

  Yard1

  and

  avnishn

  authored
  22de452

  Copy full SHA for 22de452

Commits on Feb 21, 2024

• Add metrics to RequestOutput (vllm-project#2876)

  Yard1

  authored
  017d9f1

  Copy full SHA for 017d9f1

Commits on Feb 20, 2024

• Make vLLM logging formatting optional (vllm-project#2877)

  Yard1

  authored
  181b27d

  Copy full SHA for 181b27d

Commits on Jan 23, 2024

• [Experimental] Add multi-LoRA support (vllm-project#1804)

  Show description for 9b945da
  

  4 peopleauthored
  9b945da

  Copy full SHA for 9b945da

Commits on Dec 20, 2023

• Remove Sampler copy stream (vllm-project#2209)

  Yard1

  authored
  bd29cf3

  Copy full SHA for bd29cf3

Commits on Dec 17, 2023

• Make sampler less blocking (vllm-project#1889)

  Yard1

  authored
  a7347d9

  Copy full SHA for a7347d9

Commits on Dec 14, 2023

• Optimize Mixtral with expert parallelism (vllm-project#2090)

  Yard1

  authored
  21d93c1

  Copy full SHA for 21d93c1

Commits on Dec 6, 2023

• Save pytorch profiler output for latency benchmark (vllm-project#1871)

  Show description for 05ff90b
  Yard1

  authored
  05ff90b

  Copy full SHA for 05ff90b

Commits on Nov 3, 2023

• Support YaRN models (vllm-project#1264)

  Show description for 9f669a9
  

  3 peopleauthored
  9f669a9

  Copy full SHA for 9f669a9

Pagination

Previous

Next