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 14, 05:30 UTC
main
[Bugfix] Honor --mm_encoder_attn_backend when used (#27124)
Bradley D committed about 1 year ago
Run mypy on the lowest supported Python version instead of system Python (#27048)
Harry Mellor committed about 1 year ago
[Model] Add MoE support for NemotronH (#25863)
tomeras91 committed about 1 year ago
[Metrics] [KVConnector] Add connector prefix cache hit rate stats (#26245)
Tova Movshovitz committed about 1 year ago
[CI] Reorganize entrypoints tests (#27403)
Chauncey committed about 1 year ago
add SLA information into comparison graph for vLLM Benchmark Suite (#25525)
Louie Tsai committed about 1 year ago
[CI/Build] Fix AMD CI: test_cpu_gpu.py (#27388)
Zhewen Li committed about 1 year ago
[Bugfix] Fix args settings for guided decoding args (#27375)
Lucia Fang committed about 1 year ago
[CI/Build] Fix Prithvi plugin test (#27393)
Cyrus Leung committed about 1 year ago
[Chore] Remove duplicate `has_` functions in vllm.utils (#27372)
Jonathan Chen committed about 1 year ago
[Model] Add num_cached_tokens for PoolingRequestOutput (#27378)
wang.yuqi committed about 1 year ago
[V1][spec decode] return logprobs for spec decoding (#26060)
Giancarlo Delfin committed about 1 year ago
[CORE] Support Prefix Caching with Prompt Embeds (#27219)
Andrew Sansom committed about 1 year ago
[Bugfix][Core] running queue index leakage exception (#26754)
PiteXChen committed about 1 year ago
[Bugfix] Fix incorrect kv cache metrics in grafana.json (#27133)
fangpings committed about 1 year ago
[Bugfix] Fix SLA tuner initialization (#27355)
Cyrus Leung committed about 1 year ago
Isotr0py committed about 1 year ago
Matthew Bonanni committed about 1 year ago
[Chore] Separate out system utilities from vllm.utils (#27201)
dongbo910220 committed about 1 year ago
[BugFix] bugfix for Flash Attention MLA with full cuda graph IMA following pr-25490 (#27128)
Daisy-Ma-coder committed about 1 year ago