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 4, 02:30 UTC
main
[CI][XPU]Ignore some lora tests from LoRA Intel CI pipeline (#42010)
Chaojun Zhang committed 6 months ago
[CI][Bugfix] Drop duplicated examples/ prefix in tensorize_vllm_model command (#42039)
haosdent committed 6 months ago
[CI][Bugfix] Surface subprocess output in spawn_new_process_for_each_test (#41943)
haosdent committed 6 months ago
[CI][ROCm] Ship RIXL with `vllm/vllm-openai-rocm` (#41634)
Simon Danielsson committed 6 months ago
[Bugifx] Missing Renderer for `fastokens` mode (#41984)
TJian committed 6 months ago
[Docs] Reorganize examples docs. (#41082)
wang.yuqi committed 6 months ago
[Aiter][ROCm] gdn_linear_attn kernel fusion (#40711)
Tres committed 6 months ago
[Bugfix] Fix XPU/ROCm compatibility in spawn_new_process_for_each_test (#41895)
David Zheng committed 6 months ago
[Kernel][Helion] Optimize Helion config parsing latency (#40850)
Yanan Cao committed 6 months ago
enable persistent mla for sparse mla backend (#41990)
Douglas Lehr committed 6 months ago
fix: default TILELANG_CLEANUP_TEMP_FILES=1 to avoid shared /tmp conflicts (#41486)
SamareshSingh committed 6 months ago
[Perf][3/n] Eliminate GPU<->CPU syncs in attention impls (#41434)
Nick Hill committed 6 months ago
[Perf] Use numpy zero-copy path for embedding float response serialization (#41681)
Shrinav Loka committed 6 months ago
[Examples][last/6] Resettle examples. (#41084)
wang.yuqi committed 6 months ago
[CI][Bugfix] Fix CI failures for "PyTorch Compilation Unit Tests" (#41940)
haosdent committed 6 months ago
[CI][Bugfix] Fix failure CI step "PyTorch Fullgraph Smoke Test" (#41953)
haosdent committed 6 months ago
add: LFM2/2.5 Tool Parser (#39243)
Jonathan committed 6 months ago
[KV Connector] Opt DecodeBenchConnector into SupportsHMA (#41770)
Zijing Liu committed 6 months ago
[Core] Avoid using extra thread in `UniProcExecutor` (#40891)
Nick Hill committed 6 months ago
[Bugfix] Restore moe_forward output shape invariant on TRTLLM MXFP4 path (#41646)
Stefano Castagnetta committed 6 months ago