[Bugfix] Fix host and port join for ipv6 in bench serve (#28679)

scottzh8 committed 12 months ago

Fixed gpt-oss _load_weights_other() parameter position bug (#28715)

Dezhan committed 12 months ago

[V1] Support MP Executor for multi node distributed inference (#23691)

Lucia Fang committed 12 months ago

[Model] Allow users to control skip reading cache per request. (#28194)

wang.yuqi committed 12 months ago

[Benchmark] Fix client seed synchronization in multi-turn benchmark (#28512)

ai-jz committed 12 months ago

Adding a benchmark for batch invariance (#28161)

Bram Wasti committed 12 months ago

[BugFix] Fix `AssertionError: DCP not support reorder_batch_threshold > 1 now.` (#28751)

Lucas Wilkinson committed 12 months ago

[Doc]: fix typos in various files (#28567)

Didier Durand committed 12 months ago

[Model][Qwen3VL] Cache positional embedding indices (#28475)

Lukas Geiger committed 12 months ago

[CI] Fix broken pipeline (#28781)

Nick Hill committed 12 months ago

Add support for Eagle with separate lm-head and embed_tokens layers (#28549)

Eldar Kurtić committed 12 months ago

[Model] Fix lmhead init bug of bailing_moe (#28777)

hwhaokun committed 12 months ago

[Doc] Fix failing doc build (#28772)

Cyrus Leung committed 12 months ago

Allow Gemma3 to take image embeddings (#28483)

tingtinggithub committed 12 months ago

[compile] Enable sequence parallelism matching w/o custom ops enabled (#27126)

Angela Yi committed 12 months ago

[PERF] Remove TRTLLM Gen attn kernel limitation `max_seq_len <=131072` (#28755)

Vadim Gimpelson committed 12 months ago

[Misc] Make `SchedulerConfig.max_model_len` init-only (#28733)

Cyrus Leung committed 12 months ago

[Kernel][Moe Configs] llama4 maverick fp8 moe config tp8 on mi325 (#28709)

Zhewen Li committed 12 months ago

Use narrow over indexing in `hadacore_transform` to prep for ABI stable (#28756)

Jane (Yuan) Xu committed 12 months ago

[RL] [V1] Remove unused device argument from reset_kv_cache (#28766)

Zhuohan Li committed 12 months ago