Commits

Branch selector

 main

User selector

njhill

Datepicker

All time

Commit History

Commits on Feb 14, 2025

• [V1] Simplify GPUModelRunner._update_states check (vllm-project#13265)

  njhill

  authored
  085b7b2

  Copy full SHA for 085b7b2

Commits on Feb 10, 2025

• [V1] Use msgpack for core request serialization (vllm-project#12918)

  Show description for 67c4637
  njhill

  authored
  67c4637

  Copy full SHA for 67c4637

Commits on Feb 5, 2025

• [V1][Misc] Shorten FinishReason enum and use constant strings (vllm-project#12760)

  njhill

  authored
  3d09e59

  Copy full SHA for 3d09e59

Commits on Jan 24, 2025

• [V1][Frontend] Coalesce bunched RequestOutputs (vllm-project#12298)

  Show description for 24b0205
  

  njhill

  and

  robertgshaw2-redhat

  authored
  24b0205

  Copy full SHA for 24b0205

Commits on Jan 22, 2025

• [Frontend][V1] Online serving performance improvements (vllm-project#12287)

  njhill

  authored
  aea9436

  Copy full SHA for aea9436

• [Benchmark] More accurate TPOT calc in benchmark_serving.py (vllm-project#12288)

  Show description for 222a9dc
  njhill

  authored
  222a9dc

  Copy full SHA for 222a9dc

Commits on Nov 7, 2024

• [V1][BugFix] Fix Generator construction in greedy + seed case (vllm-project#10097)

  Show description for 1fa020c
  njhill

  authored
  1fa020c

  Copy full SHA for 1fa020c

• [Frontend] Adjust try/except blocks in API impl (vllm-project#10056)

  Show description for 29862b8
  njhill

  authored
  29862b8

  Copy full SHA for 29862b8

Commits on Nov 3, 2024

• [V1] Support per-request seed (vllm-project#9945)

  Show description for 1f1b6d6
  njhill

  authored
  1f1b6d6

  Copy full SHA for 1f1b6d6

Commits on Nov 2, 2024

• [Docs] Update Granite 3.0 models in supported models table (vllm-project#9930)

  Show description for eed92f1
  

  njhill

  and

  DarkLight1337

  authored
  eed92f1

  Copy full SHA for eed92f1

Commits on Oct 21, 2024

• [BugFix] Update draft model TP size check to allow matching target TP size (vllm-project#9394)

  Show description for 15713e3
  

  njhill

  and

  qibaoyuan

  authored
  15713e3

  Copy full SHA for 15713e3

• [Frontend] Reduce frequency of client cancellation checking (vllm-project#7959)

  njhill

  authored
  9d9186b

  Copy full SHA for 9d9186b

Commits on Oct 19, 2024

• [Frontend] Avoid creating guided decoding LogitsProcessor unnecessarily (vllm-project#9521)

  njhill

  authored
  1325872

  Copy full SHA for 1325872

Commits on Oct 18, 2024

• [BugFix] Fix and simplify completion API usage streaming (vllm-project#9475)

  njhill

  authored
  25aeb7d

  Copy full SHA for 25aeb7d

• [BugFix] Typing fixes to RequestOutput.prompt and beam search (vllm-project#9473)

  njhill

  authored
  1ffc8a7

  Copy full SHA for 1ffc8a7

Commits on Oct 15, 2024

• [BugFix] Fix chat API continuous usage stats (vllm-project#9357)

  njhill

  authored
  e9d517f

  Copy full SHA for e9d517f

Commits on Oct 3, 2024

• [Doc] Update Granite model docs (vllm-project#9025)

  njhill

  authored
  18c2e30

  Copy full SHA for 18c2e30

Commits on Sep 29, 2024

• [BugFix] Fix seeded random sampling with encoder-decoder models (vllm-project#8870)

  Show description for 31f46a0
  

  njhill

  and

  ywang96

  authored
  31f46a0

  Copy full SHA for 31f46a0

Commits on Sep 26, 2024

• [BugFix] Fix test breakages from transformers 4.45 upgrade (vllm-project#8829)

  njhill

  authored
  4b377d6

  Copy full SHA for 4b377d6

Commits on Sep 19, 2024

• [Frontend] Use MQLLMEngine for embeddings models too (vllm-project#8584)

  njhill

  authored
  76515f3

  Copy full SHA for 76515f3

Commits on Sep 18, 2024

• [BugFix] Nonzero exit code if MQLLMEngine startup fails (vllm-project#8572)

  njhill

  authored
  d9cd78e

  Copy full SHA for d9cd78e

Commits on Sep 17, 2024

• [Misc] Don't dump contents of kvcache tensors on errors (vllm-project#8527)

  njhill

  authored
  56c3de0

  Copy full SHA for 56c3de0

Commits on Sep 16, 2024

• [BugFix] Fix clean shutdown issues (vllm-project#8492)

  njhill

  authored
  acd5511

  Copy full SHA for acd5511

Commits on Sep 13, 2024

• [HotFix] Fix final output truncation with stop string + streaming (vllm-project#8468)

  njhill

  authored
  18e9e1f

  Copy full SHA for 18e9e1f

Commits on Sep 12, 2024

• [Core] Add engine option to return only deltas or final output (vllm-project#7381)

  njhill

  authored
  551ce01

  Copy full SHA for 551ce01

Commits on Sep 6, 2024

• [BugFix] Fix Granite model configuration (vllm-project#8216)

  njhill

  authored
  baa5467

  Copy full SHA for baa5467

Commits on Sep 4, 2024

• [Benchmark] Add --async-engine option to benchmark_throughput.py (vllm-project#7964)

  njhill

  authored
  d4db9f5

  Copy full SHA for d4db9f5

Commits on Aug 29, 2024

• [Frontend] Minor optimizations to zmq decoupled front-end (vllm-project#7957)

  Show description for 4289cad
  

  njhill

  and

  Robert Shaw

  authored
  4289cad

  Copy full SHA for 4289cad

Commits on Aug 27, 2024

• [Tests] Disable retries and use context manager for openai client (vllm-project#7565)

  njhill

  authored
  39178c7

  Copy full SHA for 39178c7

Commits on Aug 25, 2024

• [Spec Decoding] Streamline batch expansion tensor manipulation (vllm-project#7851)

  njhill

  authored
  1856aff

  Copy full SHA for 1856aff

Commits on Aug 21, 2024

• [Spec Decoding] Use target model max length as default for draft model (vllm-project#7706)

  njhill

  authored
  9b73a2f

  Copy full SHA for 9b73a2f

• [BugFix] Avoid premature async generator exit and raise all exception variations (vllm-project#7698)

  njhill

  authored
  c75363f

  Copy full SHA for c75363f

Commits on Aug 16, 2024

• [Core] Use uvloop with zmq-decoupled front-end (vllm-project#7570)

  njhill

  authored
  9587b05

  Copy full SHA for 9587b05

Commits on Aug 9, 2024

• [Core] Streamline stream termination in AsyncLLMEngine (vllm-project#7336)

  njhill

  authored
  b4e9528

  Copy full SHA for b4e9528

Commits on Aug 7, 2024

• [FrontEnd] Make merge_async_iterators is_cancelled arg optional (vllm-project#7282)

  njhill

  authored
  fc1493a

  Copy full SHA for fc1493a

Pagination

Previous

Next