Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 26, 2024

• I think this is the root cause of the bug that prevents BNB from playing nicely with tensor parallel stuff

  JoshuaDavid

  authored
  83b1e18

  Copy full SHA for 83b1e18

Commits on Jul 25, 2024

• [model][meta-llama-3.1] Optimistically remove warning about tensor parallelism + bnb quantization being unsupported

  JoshuaDavid

  authored
  c6d7b5e

  Copy full SHA for c6d7b5e

• [bugfix][model] Update index calculation in vllm.model_executor.layers.linear.weight_loader() for tensor parallelism

  Show description for ee70d56
  JoshuaDavid

  authored
  ee70d56

  Copy full SHA for ee70d56

• [Bugfix] Fix empty (nullptr) channelwise scales when loading wNa16 using compressed tensors (vllm-project#6798)

  LucasWilkinson

  authored
  cd7edc4

  Copy full SHA for cd7edc4

• [Doc] Add documentations for nightly benchmarks (vllm-project#6412)

  KuntaiDu

  authored
  6a1e25b

  Copy full SHA for 6a1e25b

• [Bugfix] Add synchronize to prevent possible data race (vllm-project#6788)

  Show description for 95db75d
  

  tlrmchlsmth

  and

  LucasWilkinson

  authored
  95db75d

  Copy full SHA for 95db75d

• [Bugfix] Fix kv_cache_dtype=fp8 without scales for FP8 checkpoints (vllm-project#6761)

  mgoin

  authored
  65b1f12

  Copy full SHA for 65b1f12

• [ Misc ] fp8-marlin channelwise via compressed-tensors (vllm-project#6524)

  Show description for 889da13
  

  robertgshaw2-redhat

  and

  mgoin

  authored
  889da13

  Copy full SHA for 889da13

• [Bugfix] Add image placeholder for OpenAI Compatible Server of MiniCPM-V (vllm-project#6787)

  Show description for b75e314
  

  3 peopleauthored
  b75e314

  Copy full SHA for b75e314

• [Bugfix] Fix encoding_format in examples/openai_embedding_client.py (vllm-project#6755)

  CatherineSue

  authored
  316a41a

  Copy full SHA for 316a41a

• [Bugfix] Fix awq_marlin and gptq_marlin flags (vllm-project#6745)

  alexm-redhat

  authored
  0310029

  Copy full SHA for 0310029

• [Bugfix] Fix decode tokens w. CUDA graph (vllm-project#6757)

  comaniac

  authored
  309aaef

  Copy full SHA for 309aaef

• [Model] Adding support for MiniCPM-V (vllm-project#4087)

  HwwwwwwwH

  authored
  9e169a4

  Copy full SHA for 9e169a4

• [Frontend] Represent tokens with identifiable strings (vllm-project#6626)

  ezliu

  authored
  5689e25

  Copy full SHA for 5689e25

• [core][distributed] fix zmq hang (vllm-project#6759)

  youkaichao

  authored
  740374d

  Copy full SHA for 740374d

Commits on Jul 24, 2024

• [Doc][AMD][ROCm]Added tips to refer to mi300x tuning guide for mi300x users (vllm-project#6754)

  hongxiayang

  authored
  d88c458

  Copy full SHA for d88c458

• [Bugfix] Bump transformers to 4.43.2 (vllm-project#6752)

  mgoin

  authored
  421e218

  Copy full SHA for 421e218

• [Core] Tweaks to model runner/input builder developer APIs (vllm-project#6712)

  Yard1

  authored
  5448f67

  Copy full SHA for 5448f67

• Add fp8 support to reshape_and_cache_flash (vllm-project#6667)

  Yard1

  authored
  0e63494

  Copy full SHA for 0e63494

• [Frontend] split run_server into build_server and run_server (vllm-project#6740)

  dtrifiro

  authored
  ee81258

  Copy full SHA for ee81258

• [Bugfix] Miscalculated latency lead to time_to_first_token_seconds inaccurate. (vllm-project#6686)

  AllenDou

  authored
  40468b1

  Copy full SHA for 40468b1

• [Bugfix] Fix speculative decode seeded test (vllm-project#6743)

  njhill

  authored
  2cf0df3

  Copy full SHA for 2cf0df3

• Adding f-string to validation error which is missing (vllm-project#6748)

  luizanao

  authored
  5451463

  Copy full SHA for 5451463

• [Bugfix]fix modelscope compatible issue (vllm-project#6730)

  liuyhwangyh

  authored
  f4f8a9d

  Copy full SHA for f4f8a9d

• [Build/CI] Update run-amd-test.sh. Enable Docker Hub login. (vllm-project#6711)

  Alexei-V-Ivanov-AMD

  authored
  b570811

  Copy full SHA for b570811

• [Docs][ROCm] Detailed instructions to build from source (vllm-project#6680)

  WoosukKwon

  authored
  ccc4a73

  Copy full SHA for ccc4a73

• [Bugfix] Fix token padding for chameleon (vllm-project#6724)

  ywang96

  authored
  0a740a1

  Copy full SHA for 0a740a1

• [SpecDecoding] Update MLPSpeculator CI tests to use smaller model (vllm-project#6714)

  njhill

  authored
  c882a7f

  Copy full SHA for c882a7f

• [Bugfix] fix flashinfer cudagraph capture for PP (vllm-project#6708)

  SolitaryThinker

  authored
  5e8ca97

  Copy full SHA for 5e8ca97

Commits on Jul 23, 2024

• [bitsandbytes]: support read bnb pre-quantized model (vllm-project#5753)

  Show description for 87525fa
  

  thesues

  and

  mgoin

  authored
  87525fa

  Copy full SHA for 87525fa

• [Bugfix] StatLoggers: cache spec decode metrics when they get collected. (vllm-project#6645)

  Show description for 2f808e6
  tdoublep

  authored
  2f808e6

  Copy full SHA for 2f808e6

• [CI] Add smoke test for non-uniform AutoFP8 quantization (vllm-project#6702)

  mgoin

  authored
  01c16ed

  Copy full SHA for 01c16ed

• [build] relax wheel size limit (vllm-project#6704)

  youkaichao

  authored
  72fc704

  Copy full SHA for 72fc704

• Bump transformers version for Llama 3.1 hotfix and patch Chameleon (vllm-project#6690)

  ywang96

  authored
  1bedf21

  Copy full SHA for 1bedf21

• [Model] Pipeline Parallel Support for DeepSeek v2 (vllm-project#6519)

  Show description for 507ef78
  tjohnson31415

  authored
  507ef78

  Copy full SHA for 507ef78

Pagination

Previous

Next