Upstream Sync#80

Merged

robertgshaw2-redhat merged 21 commits intoneuralmagic:disagg_pd_devfrom

robertgshaw2-redhat:merge-with-main

May 6, 2025

Commits on May 6, 2025

[Model] Add GraniteMoeHybrid 4.0 model (#17497)

authored
[easy] Fix logspam on PiecewiseBackend errors (#17138)
zou3519
authored
[Bugfix] Fixed prompt length for random dataset (#17408)
Xarbirus
authored
[Doc] Update notes for H2O-VL and Gemma3 (#17219)
DarkLight1337
authored
[Misc] Fix ScalarType float4 naming (#17690)
LucasWilkinson
authored
Fix `dockerfilegraph` pre-commit hook (#17698)
hmellor
authored
[Bugfix] Fix triton import with local TritonPlaceholder (#17446)
MengqingCao
authored
[V1] Enable TPU V1 backend by default (#17673)
mgoin
authored
[V1][PP] Support PP for MultiprocExecutor (#14219)
bigPYJ1151
authored
[v1] AttentionMetadata for each layer (#17394)
heheda12345
authored
[Feat] Add deprecated=True to CLI args (#17426)
aarnphm
authored
[Docs] Use gh-file to add links to tool_calling.md (#17709)
windsonsea
authored
[v1] Introduce KVCacheBlocks as interface between Scheduler and KVCacheManager (#17479)
heheda12345
authored
[doc] Add RAG Integration example (#17692)
reidliu41
authored
[Bugfix] Fix modality limits in vision language example (#17721)
DarkLight1337
authored
Make right sidebar more readable in "Supported Models" (#17723)
hmellor
authored
[TPU] Increase block size and reset block shapes (#16458)
bythew3i
authored
[Misc] Add Next Edit Prediction (NEP) datasets support in `benchmark_serving.py` (#16839)

dtransposed
and
dtransposed
authored
[Bugfix] Fix for the condition to accept empty encoder inputs for mllama (#17732)
gshtras
authored
[Kernel] Unified Triton kernel that doesn't distinguish between prefill + decode (#16828)

tdoublep
and
LucasWilkinson
authored
updated
robertgshaw2-redhat
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Upstream Sync#80

Upstream Sync#80
robertgshaw2-redhat merged 21 commits intoneuralmagic:disagg_pd_devfrom
robertgshaw2-redhat:merge-with-main

Commits on May 6, 2025

[Model] Add GraniteMoeHybrid 4.0 model (#17497)

[easy] Fix logspam on PiecewiseBackend errors (#17138)

[Bugfix] Fixed prompt length for random dataset (#17408)

[Doc] Update notes for H2O-VL and Gemma3 (#17219)

[Misc] Fix ScalarType float4 naming (#17690)

Fix `dockerfilegraph` pre-commit hook (#17698)

[Bugfix] Fix triton import with local TritonPlaceholder (#17446)

[V1] Enable TPU V1 backend by default (#17673)

[V1][PP] Support PP for MultiprocExecutor (#14219)

[v1] AttentionMetadata for each layer (#17394)

[Feat] Add deprecated=True to CLI args (#17426)

[Docs] Use gh-file to add links to tool_calling.md (#17709)

[v1] Introduce KVCacheBlocks as interface between Scheduler and KVCacheManager (#17479)

[doc] Add RAG Integration example (#17692)

[Bugfix] Fix modality limits in vision language example (#17721)

Make right sidebar more readable in "Supported Models" (#17723)

[TPU] Increase block size and reset block shapes (#16458)

[Misc] Add Next Edit Prediction (NEP) datasets support in `benchmark_serving.py` (#16839)

[Bugfix] Fix for the condition to accept empty encoder inputs for mllama (#17732)

[Kernel] Unified Triton kernel that doesn't distinguish between prefill + decode (#16828)

updated

Upstream Sync#80robertgshaw2-redhat merged 21 commits intoneuralmagic:disagg_pd_devneuralmagic/vllm:disagg_pd_devfrom robertgshaw2-redhat:merge-with-mainrobertgshaw2-redhat/vllm:merge-with-mainCopy head branch name to clipboard

Commits

Commits on May 6, 2025

Upstream Sync#80
robertgshaw2-redhat merged 21 commits intoneuralmagic:disagg_pd_devfrom
robertgshaw2-redhat:merge-with-main