A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 5, 04:45 UTC
main
[Bugfix][ROCm]: Allow `gpt_oss_mxfp4` quantization method on rocm (#39754)
Rohan Potdar committed 7 months ago
[Core][Metrics] Remove unused `SchedulerStats.encoder_cache_usage` (#39693)
Mark McLoughlin committed 7 months ago
[ROCm][CI] Fix condition for `test_per_token_group_quant_fp8_packed` (#39730)
Micah Williamson committed 7 months ago
Bugfix: `use_existing_torch.py`: Glob recursive subdirs in requirements (fixes #39024) (#39793)
Netanel Haber committed 7 months ago
Measure encoder compile time seperate from llm backbone (#39240)
Lucas Kabela committed 7 months ago
[PluggableLayer][3/N] Apply PluggableLayer to moe-related layers. (#33556)
Hexiang Wang committed 7 months ago
[MoE Refactor] Remove MoE DP chunking (#39107)
bnellnm committed 7 months ago
bhargav-patel-29 committed 7 months ago
fix: handle ImportError in load_audio (#39473)
Yiyang "Ian" Liu committed 7 months ago
[Bugfix] Fix `vllm bench serve` to count multimodal tokens in "total input tokens" (#38654)
Matthias Gehre committed 7 months ago
[Bugfix][Model] Fix Devstral Small 2 HF format weight loading (#39293)
Thomas committed 7 months ago
fxmarty-amd committed 7 months ago
[XPU][CI] Remove Arc in label-xpu (#39776)
xiangdong committed 7 months ago
feat: add TxtSlicesDataset to allow sampling slices from txt file for benchmarking (#30156)
Julien Debache committed 7 months ago
[MM][Perf][CG] Support ViT full CUDA graph for Qwen3-VL video inference (#38061)
Shanshan Shen committed 7 months ago
wang.yuqi committed 7 months ago
[Model] Use mm_features for Ernie-4.5 VL M-RoPE (#39753)
lalit10 committed 7 months ago
Mark McLoughlin committed 7 months ago
Fix Responses API streaming for multiple auto tool calls (#39626)
noobHappylife committed 7 months ago
[Bugfix] Fix Gemma4 tool parser converting bare `null` to string `"null"` (#39679)
김의진 committed 7 months ago