[Kernel][Triton][FP8] Adding fp8 and variable length sequence support to Triton FAv2 kernel (#12591)

rasmith committed over 1 year ago

Update test_flash_attn.py (#17102)

Happy committed over 1 year ago

[CI/Build] remove -t for run-lm-eval-gsm-hf-baseline.sh (#16271)

Reid committed over 1 year ago

[Minor] Fix lint error in main branch (#17233)

Woosuk Kwon committed over 1 year ago

[Bugfix] Fix Qwen2.5-Omni M-RoPE position ids generation (#16878)

Kero Liang committed over 1 year ago

[MISC] rename interval to max_recent_requests (#14285)

Ning Xie committed over 1 year ago

Disable the torch.compile cache checks when VLLM_DISABLE_COMPILE_CACHE=1 (#16573)

Lu Fang committed over 1 year ago

[Bugfix] fix error due to an uninitialized tokenizer when using `skip_tokenizer_init` with `num_scheduler_steps` (#9276)

changjun.lee committed over 1 year ago

[Docs] Update structured output doc for V1 (#17135)

Russell Bryant committed over 1 year ago

[Chore] added stubs for `vllm_flash_attn` during development mode (#17228)

Aaron Pham committed over 1 year ago

[BUGFIX] use random for NONE_HASH only when PYTHONHASHSEED not set (#17088)

Ning Xie committed over 1 year ago

[V1] Add `structural_tag` support using xgrammar (#17085)

Russell Bryant committed over 1 year ago

[Hardware][Intel-Gaudi] Update hpu-extension and update bucketing system for HPU device (#17186)

Agata Dobrzyniewicz committed over 1 year ago

[Bugfix] Fix standard models tests (#17217)

Cyrus Leung committed over 1 year ago

[Bugfix] Fix missing int type for `-n` in multi-image example (#17223)

Isotr0py committed over 1 year ago

[Core] Remove prompt string from engine core data structures (#17214)

Nick Hill committed over 1 year ago

[CI/test] Fix Eagle Correctness Test (#17209)

Woosuk Kwon committed over 1 year ago

[BugFix] Avoid race conditions in zero-copy tensor transmission (#17203)

Nick Hill committed over 1 year ago

[V1][Metrics] Allow V1 AsyncLLM to use custom logger (#14661)

Zijing Liu committed over 1 year ago

[ROCm][Misc] Follow-ups for Skinny Gemms on ROCm. (#17011)

Charlie Fu committed over 1 year ago