[UX] Fail if an invalid attention backend is specified (#22217)

Michael Goin committed over 1 year ago

[Bugfix] Misaligned params in TreeAttentionImpl (#22226)

Cyrus Leung committed over 1 year ago

Optimize configuration access with LRU cache in custom ops (#22204)

ZiTian.Zhao committed over 1 year ago

[Misc] log more detailed message for ensure_model_parallel_initialized (#22144)

Ning Xie committed over 1 year ago

[Doc] add backend to doc string of initialize_model_parallel (#22142)

Ning Xie committed over 1 year ago

[Misc] Remove pass_config from CompilationConfig dump_json excluded (#21911)

elvischenv committed over 1 year ago

fix: kimi_k2 return empty tool call list (#22149)

tlipoca9 committed over 1 year ago

[Log] DeepGEMM Update Log for Unaligned Problem Size (#22208)

Wentao Ye committed over 1 year ago

self.gate dtype update for GLM-4.5 (#22203)

Yuxuan Zhang committed over 1 year ago

[ROCm][Bugfix] Compilation passes fix (#22202)

Gregory Shtrasberg committed over 1 year ago

[FEAT] Refactor ROPE into module (#22192)

TJian committed over 1 year ago

[V0 deprecation][P/D] Deprecate v0 `KVConnectorBase` code (1/2) (#21785)

lkchen committed over 1 year ago

[V1] reduce block size for tree attention correctness test to fix 'ou… (#22207)

Giancarlo Delfin committed over 1 year ago

Revert "[Bugfix] V1 Fix the cursor leakage issue during request scheduling." (#22223)

Woosuk Kwon committed over 1 year ago

[Bugfix] V1 Fix the cursor leakage issue during request scheduling. (#21173)

PiteXChen committed over 1 year ago

[NVIDIA] Auto detect modelopt quant and fix DSR1-FP4 weight loading (#22073)

Po-Han Huang (NVIDIA) committed over 1 year ago

[Bugfix][V1][P/D]Fix the uneven polling issue in the toy proxy for P2pNcclConnector (#21819)

Zhonghua Deng committed over 1 year ago

[Bug] Update auto_tune.sh to separate benchmarking and profiling. (#21629)

ericehanley committed over 1 year ago

[Responses API] Ignore `store=True` and process the request by default (#22185)

Woosuk Kwon committed over 1 year ago

Fix Arcee model weight loading: Add custom load_weights (#21725)

Raghav Ravishankar committed over 1 year ago