Commits

Branch selector

 main

User selector

heheda12345

All time

Commit History

Commits on Feb 10, 2026

• [BugFix] Avoid prefix cache hit in the same schedule step for mamba layers (#29387)

  Show description for 97fa8f6
  heheda12345

  authored
  97fa8f6

  Copy full SHA for 97fa8f6

Commits on Dec 24, 2025

• [KVEvent] User request.block_hash for parent block_hash (#30544)

  Show description for 538e830
  

  heheda12345

  and

  ivanium

  authored
  538e830

  Copy full SHA for 538e830

Commits on Dec 18, 2025

• [BugFix] Spec decode with VLLM_ENABLE_V1_MULTIPROCESSING=0 (#30319)

  Show description for 24b65ef
  heheda12345

  authored
  24b65ef

  Copy full SHA for 24b65ef

Commits on Dec 13, 2025

• [Bugfix] Qwen3-next with --hf-overrides \{\"num_hidden_layers\":8\} (#30433)

  Show description for ace34e3
  heheda12345

  authored
  ace34e3

  Copy full SHA for ace34e3

Commits on Nov 24, 2025

• [Hybrid Allocator] Better layer padding strategy for gpt-oss eagle (#29303)

  Show description for 71df2a5
  heheda12345

  authored
  71df2a5

  Copy full SHA for 71df2a5

Commits on Nov 5, 2025

• [FlashInfer] Avoid FlashInfer block_size 16 + head_size 256 on blackwell (#27994)

  Show description for c765f0b
  heheda12345

  authored
  c765f0b

  Copy full SHA for c765f0b

• [DCP] check return_lse for all layers in dcp (#27929)

  Show description for 5d16d0f
  heheda12345

  authored
  5d16d0f

  Copy full SHA for 5d16d0f

Commits on Oct 31, 2025

• [Hybrid] A simpler algorithm to find kernel_block_size (#26476)

  Show description for df33486
  heheda12345

  authored
  df33486

  Copy full SHA for df33486

Commits on Oct 30, 2025

• [Bugfix] mamba-block-size is set for vision language model (#27773)

  Show description for d7fb10c
  heheda12345

  authored
  d7fb10c

  Copy full SHA for d7fb10c

Commits on Oct 10, 2025

• [bugfix][DCP] fix block_size of hash in DCP prefix caching (#26296)

  Show description for 606b00e
  heheda12345

  authored
  606b00e

  Copy full SHA for 606b00e

• [deepseek] kernel block size for UniformTypeKVCacheSpecs (#26559)

  Show description for 6f0f570
  heheda12345

  authored
  6f0f570

  Copy full SHA for 6f0f570

Commits on Oct 7, 2025

• [deepseek] add EP8 FusedMOE config for H200 and B200 (#26331)

  Show description for 29fd266
  heheda12345

  authored
  29fd266

  Copy full SHA for 29fd266

Commits on Oct 6, 2025

• [MISC] Add heheda12345 to CODEOWNERS of vllm/config/cache.py (#26270)

  Show description for 4be7d7c
  heheda12345

  authored
  4be7d7c

  Copy full SHA for 4be7d7c

Commits on Oct 2, 2025

• [Mamba][KVCacheManager] Simplify kv cache manage logic for mamba + MTP (#25119)

  Show description for 3d5f1c8
  heheda12345

  authored
  3d5f1c8

  Copy full SHA for 3d5f1c8

• [Deepseek v3.2] Support indexer prefill chunking (#25999)

  Show description for 1e50f1b
  heheda12345

  authored
  1e50f1b

  Copy full SHA for 1e50f1b

Commits on Sep 20, 2025

• [Hybrid Allocator] Support full attention with different hidden size (#25101)

  Show description for 9607d5e
  heheda12345

  authored
  9607d5e

  Copy full SHA for 9607d5e

Commits on Sep 16, 2025

• [MISC] Add code owners of vllm/v1 to vllm/v1/core (#24928)

  Show description for 218454b
  heheda12345

  authored
  218454b

  Copy full SHA for 218454b

Commits on Sep 14, 2025

• [Hybrid Allocator] Support Pipeline Parallel (#23974)

  Show description for 8e5cdcd
  heheda12345

  authored
  8e5cdcd

  Copy full SHA for 8e5cdcd

Commits on Sep 12, 2025

• [Misc][gpt-oss] Add gpt-oss label to PRs that mention harmony or related to builtin tool call (#24717)

  Show description for 64d90c3
  heheda12345

  authored
  64d90c3

  Copy full SHA for 64d90c3

• [CI] Fix flaky test v1/worker/test_gpu_model_runner.py::test_kv_cache_stride_order (#24640)

  Show description for 561a0ba
  heheda12345

  authored
  561a0ba

  Copy full SHA for 561a0ba

Commits on Sep 11, 2025

• [Qwen3-Next] MOE configs for H100 TP4 (#24699)

  Show description for f82f7a8
  heheda12345

  authored
  f82f7a8

  Copy full SHA for f82f7a8

Commits on Sep 10, 2025

• [gpt-oss] raise error for flashinfer backend without trtllm (#24482)

  Show description for b5e383c
  heheda12345

  authored
  b5e383c

  Copy full SHA for b5e383c

• Add @heheda12345 to CODEOWNERS of KVCacheManager related code (#24546)

  Show description for 41f160b
  heheda12345

  authored
  41f160b

  Copy full SHA for 41f160b

Commits on Sep 9, 2025

• [gpt-oss] Validate gpt-oss python tool during initialization (#23856)

  Show description for 1116590
  heheda12345

  authored
  1116590

  Copy full SHA for 1116590

Commits on Aug 27, 2025

• [Frontend] Optimize beam search performance by limiting concurrency (#23599)

  Show description for 142ac08
  heheda12345

  authored
  142ac08

  Copy full SHA for 142ac08

• [Frontend] Add --log-error-stack to print stack trace for error response (#22960)

  Show description for 3210264
  heheda12345

  authored
  3210264

  Copy full SHA for 3210264

• [gpt-oss] Enable unit test for response API harmony integration (#23533)

  Show description for eb19951
  heheda12345

  authored
  eb19951

  Copy full SHA for eb19951

Commits on Aug 26, 2025

• [doc] Hybrid KV Cache Manager design doc (#22688)

  Show description for d696f86
  

  heheda12345

  and

  hmellor

  authored
  d696f86

  Copy full SHA for d696f86

• Support FlashAttention Backend for Hybrid SSM Models (#23299)

  Show description for 2b4fc9b
  heheda12345

  authored
  2b4fc9b

  Copy full SHA for 2b4fc9b

Commits on Aug 22, 2025

• [Doc] Update the doc for log probs + prefix caching (#23399)

  Show description for a073be6
  

  heheda12345

  and

  gemini-code-assist[bot]

  authored
  a073be6

  Copy full SHA for a073be6

• [Attention] Refactor AttentionMetadata Preparation for Encoder-only Models (#23154)

  Show description for 17373dc
  heheda12345

  authored
  17373dc

  Copy full SHA for 17373dc

Commits on Aug 21, 2025

• [BugFix][gpt-oss] Fix Chat Completion with Multiple Output Message (#23318)

  Show description for 8a19303
  heheda12345

  authored
  8a19303

  Copy full SHA for 8a19303

Commits on Aug 20, 2025

• [Frontend] improve error logging of chat completion (#22957)

  Show description for b95697d
  heheda12345

  authored
  b95697d

  Copy full SHA for b95697d

Commits on Aug 13, 2025

• [Bugfix][mamba] Fix type annotation of Mamba2Metadata (#22787)

  Show description for fceafaf
  heheda12345

  authored
  fceafaf

  Copy full SHA for fceafaf

• [gpt-oss] upgrade gpt-oss to v0.0.3 and add version check (#22768)

  Show description for 6807af8
  heheda12345

  authored
  6807af8

  Copy full SHA for 6807af8