vllm-project/vllm

Pull Requests Commits

Remove virtual engine handling

WoosukKwon committed 21 days ago

ebe0733f

[Bug] Fix vLLM config is not set error (#29999)

yewentao256 committed 25 days ago

Verified 7b5575fa

let draft model follow target model's config_format (#30152)

bangshengtang committed 25 days ago

Verified 77e44728

[Bugfix][llama4_eagle] Fix missing 'lm_head' attribute (#29926)

divakar-amd committed 25 days ago

Verified 962d7038

[CI] Re-use whisper_client for all tests (#30148)

NickLucche committed 25 days ago

Verified e23ca3a0

[Misc] Rename CohereForAI references to CohereLabs (#30147)

russellb committed 25 days ago

Verified 3633035a

[Enc-Dec] Fix OOT tokenizer issue (#30144)

NickLucche committed 25 days ago

Verified bff78310

[KVConnector][Feature] Support KV connector cache reset via /reset_prefix_cache (#27170)

ptovam committed 25 days ago

Verified adb31506

[Compile] Conditional compilation. Introduce compile_ranges (#24252)

ilmarkov committed 25 days ago

Verified 4e26d3b0

[Attention][UX][1/N] Add AttentionConfig and change attention env vars to CLI arguments (#26315)

MatthewBonanni committed 25 days ago

Verified 66e674cd

[NIXL] Add remote_request_id to kv_transfer_params (#29665)

markmc committed 25 days ago

Verified dff0a2b3

[BugFix] Eagerly abort cancelled final-step requests (#29987)

njhill committed 25 days ago

Verified dc264bce

[NIXL] Small cleanup of unused variables (#29618)

NickLucche committed 25 days ago

Verified 78c44fd7

[bugfix] Pass globals to aot_compiled function (#29428)

angelayi committed 25 days ago

Verified e7296b08

[responsesAPI][5] ResponsesParser with tools for full MCP python loop (#29798)

qandrew committed 25 days ago

Verified da7bc54e

[NIXL] Add compatibility checking to NIXL KV connector handshake (#29503)

markmc committed 25 days ago

Verified 949a6a19

Reduce validation to a warning (#28749)

alecsolder committed 25 days ago

Verified 2c174420

[Compressed Tensors] Add XPU `wNa16` support (#29484)

yiliu30 committed 25 days ago

Verified 0d8a7d8a

[CPU][Perf] Add fast vectorized exp impl from Arm Optimized Routines (#30068)

Elm8116 committed 25 days ago

Verified 9843e332

[CI] Have pre-commit comment on a PR if pre-commit was not used (#30077)

hmellor committed 25 days ago

Verified b7d85cf2

[Feature] Add Layer-wise NVTX Support (#29990)

maxyanghu committed 26 days ago

Verified c2894d38

[ROCm][MXFP4] Infer w4a4 quant method in rocm aiter fused moe (#29775)

ZhiweiYan-96 committed 26 days ago

Verified 3628bcaa

[Bugfix] Fix parse_output_message crash on commentary with no recipient (#29972)

strinczer committed 26 days ago

Verified b73b158a

[typing] fix type (#29964)

andyxning committed 26 days ago

Verified 7ae13c66

[bench] Support common prefix len config (for decode-only bench) (#29934)

minosfuture committed 26 days ago

Verified f16356fe

[BugFix] Adding env variable to disable async grammar compilation (#29996)

alecsolder committed 26 days ago

Verified 65ee9728

[Frontend] Remove deprecated -O.xx flag (#29991)

gmagogsfm committed 26 days ago

Verified 62b33334

[CI/Build][AMD] Use float16 in test_reset_prefix_cache_e2e to avoid accuracy issues (#29997)

rasmith committed 26 days ago

Verified feecba09

[Frontend][Model] Add 'float16' to possible mamba cache dtype values, override mamba SSM cache dtype value for NemotronH (#29978)

amitz-nv committed 26 days ago

Verified 6038b1b0

[DOC]: Add kthena to integrations (#29931)

hzxuzhonghu committed 26 days ago

Verified 60a66ea2

Older