vllm-project/vllm

Pull Requests Commits

Thread device_index through distributed init; decouple from local_rank

tlrmchlsmth committed 12 hours ago

Verified c45de741

tlrmchlsmth committed 18 hours ago

Verified 1f232510

Keep CVD untouched; use assigned_gpu_ids only for topology

tlrmchlsmth committed 1 day ago

Verified b59609a5

dont hash device id

tlrmchlsmth committed 1 day ago

Verified 7161a168

Add unit tests for --device-ids composability with CVD

tlrmchlsmth committed 1 day ago

Verified ef3c7f4e

Add --device-ids CLI arg for explicit GPU selection

tlrmchlsmth committed 1 day ago

Verified b1fb76a7

Pass assigned_gpu_ids as a kwarg instead of copying vllm_config

tlrmchlsmth committed 1 day ago

Verified 4ec5605f

Clean up ray_executor assigned_gpu_ids: move import to top level

tlrmchlsmth committed 1 day ago

Verified 8ee56084

Use physical device IDs instead of setting CUDA_VISIBLE_DEVICES

tlrmchlsmth committed 2 days ago

Verified 3ec80c88

[Reasoning][Structured Outputs] Add Command A plus tags for structural tags (#44588)

rishitdholakia13 committed 2 days ago

Verified 6a11d72d

[DSV4] Move more ops out of eager breakpoint (#44561)

WoosukKwon committed 2 days ago

Verified 02d2da07

[Bugfix] Fix gemma4 crash on CPU: guard mem_get_info call (#44615)

adhithyamulticoreware committed 2 days ago

Verified bbb6c274

Remove KV cache scale boilerplate from model weight loading methods (#43167)

hmellor committed 2 days ago

Verified 62215e72

[BUG] Fix FP64 Gumbel precision coverage (#43150)

tianyu-z committed 2 days ago

Verified 7fe7800f

[Rust Frontend] Batch auto-abort requests by engine (#44591)

HueCodes committed 2 days ago

Verified 8a83e6f2

docs: fix tokenizer optimization typo (#44066)

chunyang-wen committed 2 days ago

Verified efc347f1

[NixlConnector] Initiate deprecation cycle for `kv_both` role (#43874)

NickLucche committed 2 days ago

Verified d98b8f37

[Rust Frontend] Support include_reasoning=false (#44391)

ricky-chaoju committed 2 days ago

Verified e64237ae

[Bugfix] Update mistral tokenizer test for continue_final_message fix (#44622)

XuZhou26 committed 2 days ago

Verified d61d8566

fix: pad dummy run query_start_loc (#44603)

UranusSeven committed 2 days ago

Verified d2f70da1

[Bugfix] Fix test_invocations flaky failure with newer openai SDK (#44618)

XuZhou26 committed 2 days ago

Verified 6542d489

[Bugfix][Rust Frontend] Fix UTF-8 char-boundary panic in incremental detokenizer (#44620)

Sunt-ing committed 2 days ago

Verified ca73293f

Fix `LLM.wait_for_completion` output type docstring (#44617)

viiccwen committed 2 days ago

Verified ef3af56a

[ROCm][perf] Use workspace manager for sparse indexer allocations (#41002)

tuukkjs committed 2 days ago

Verified b4a6f26c

[ROCM] [FEAT] Integrate Aiter hipBLASLt GEMM online tuning (#40426)

hanlin12-AMD committed 2 days ago

Verified 165b7864

[CI/Build] Disable CPU-Compatibility Tests (#44605)

bigPYJ1151 committed 2 days ago

Verified c505cd93

[KVConnector][1/N] PP-aware handshake aggregation and intermediate-PP output plumbing (#43720)

zixi-qi committed 2 days ago

Verified 96229fa9

[Bugfix] Exclude vision embedder from quantization in Gemma4 Unified (#44571)

lucianommartins committed 2 days ago

Verified da1daf40

[DSV4] Refactor DeepseekV4Attention (#44569)

WoosukKwon committed 2 days ago

Verified 4efd6ffd

[10/n] Migrate cuda_view and silu_and_mul_per_block_quant kernels to torch stale ABI. (#44334)

cleonard530 committed 2 days ago

Verified 56aff0dd

Older