Skip to content

Pull requests: NVIDIA/Fuser

New pull request New
166 Open 3,872 Closed
166 Open 3,872 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Don't mark a slice alias if it goes from expanded broadcast to iteration
#4990 opened Aug 15, 2025 by wujingyue Loading…
1
2
Handle remaining composite operations in opinfo tests Direct Bindings Python extension with direct mapping to NvFuser CPP objects. Python API Issues related to the Python API
#4989 opened Aug 14, 2025 by rdspring1 Loading…
2
add fusion definition equivalent to torch.nn.functional.cross_entropy(logits, labels, reduction=none)
#4988 opened Aug 14, 2025 by liqiangxl • Draft
1
may cache gather tv
#4987 opened Aug 14, 2025 by liqiangxl • Draft
1
2
Print segmentation group on runtime error
#4983 opened Aug 14, 2025 by jacobhinkle • Draft
2
relax segmenter check to accommodate scatter
#4982 opened Aug 14, 2025 by jjsjann123 Loading…
5
Refactor cutlass to share validateInputsNvfp4ScaledMm across multiple cuda kernels Cutlass Matmuls
#4981 opened Aug 14, 2025 by rdspring1 Loading…
1
Decoupling intermediate states from GpuLower
#4980 opened Aug 14, 2025 by naoyam • Draft
4
Search modules in folder python
#4978 opened Aug 14, 2025 by wujingyue Loading…
4
Search for optimal segment execution order
#4973 opened Aug 13, 2025 by jacobhinkle • Draft
1
6
add cluster reduction
#4972 opened Aug 13, 2025 by liqiangxl • Draft
2
Revert "nvfuser.multidevice.DeviceMesh.__init__ takes torch.Tensor (#…
#4969 opened Aug 13, 2025 by wujingyue Loading…
2
Reuse get_benchmark_fns and call dist.barrier before profiling
#4968 opened Aug 13, 2025 by wujingyue Loading…
5
README update
#4967 opened Aug 13, 2025 by mattwalsh Loading…
4
A sliced expanded broadcast remains an expanded broadcast.
#4966 opened Aug 12, 2025 by wujingyue • Draft
1
8
skip caching scalar inputs
#4965 opened Aug 12, 2025 by liqiangxl Loading…
4
ensure python repos are validated
#4964 opened Aug 12, 2025 by liqiangxl Loading…
7
[WIP] Cutlass Executor
#4959 opened Aug 12, 2025 by jacobhinkle • Draft
1
Test to try swizzled output allocation domain when quantizing to nvfp4
#4952 opened Aug 12, 2025 by protonu • Draft
4
Enable opinfo testing by default for direct_bindings Direct Bindings Python extension with direct mapping to NvFuser CPP objects.
#4943 opened Aug 9, 2025 by rdspring1 • Draft Operation support in direct bindings
3
[DO NOT REVIEW] segmentation in reshape example
#4933 opened Aug 7, 2025 by wolfcomos • Draft
1
WIP: Develop MCP ideas
#4919 opened Aug 5, 2025 by Steboss • Draft
3
Choose operand swizzle compatible with warp tile
#4918 opened Aug 5, 2025 by jacobhinkle • Draft
1
6
[RFC] Create LRU cache for direct bindings. Direct Bindings Python extension with direct mapping to NvFuser CPP objects. Python API Issues related to the Python API
#4893 opened Jul 31, 2025 by rdspring1 • Draft Enable direct bindings by default in Thunder
17
[draft][DO_NOT_REVIEW] a quick dynamic shape example
#4887 opened Jul 29, 2025 by jjsjann123 • Draft
2
ProTip! Follow long discussions with comments:>50.