Skip to content

Pull requests: triton-lang/triton

New pull request New
48 Open 3,436 Closed
48 Open 3,436 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Ensure Consistent Use of List Type for block_type Shape Argument in semantic.py and interpreter.py
#5128 opened Nov 12, 2024 by simonidaa Loading…
5 of 7 tasks
2
[Draft][Backend] Add Address Sanitizer Pass
#5127 opened Nov 12, 2024 by CRobeck • Draft
[AMD] Implement RepOrder for AMD MMA layouts
#5126 opened Nov 12, 2024 by oplavsic • Draft
2
[DRAFT][BACKEND] Fix a special case where elements along the k dimension are repeated within each thread
#5121 opened Nov 12, 2024 by Jokeren • Draft
[PROTON] Introduce the Proton dialect as a third-party plugin for intra-kernel perf tooling
#5119 opened Nov 11, 2024 by fywkevin Loading…
Fix coverity issues
#5118 opened Nov 11, 2024 by anmyachev Loading…
[Frontend] Print out config when autotune fails
#5117 opened Nov 11, 2024 by htyu Loading…
2
[AMD] Enable B scale for scaled_dot
#5112 opened Nov 11, 2024 by antiagainst Loading…
[DRAFT][PROTON] Add proton.state utility
#5110 opened Nov 10, 2024 by Jokeren • Draft
[BACKEND][NVIDIA] Remove NvidiaMma::getTotalElems...ForOperand
#5105 opened Nov 8, 2024 by ggengnv Loading…
1
6
[BACKEND]Refactor convertLayoutOp by expanding dot operands of 2d tensors to 3d
#5104 opened Nov 8, 2024 by yiqian1 • Draft
4
Remove unknown distribution option: test_suite
#5098 opened Nov 7, 2024 by anmyachev Loading…
[CI] remove unused inductor workflows
#5073 opened Nov 5, 2024 by leseb Loading…
4 of 7 tasks
4
[Triton][Allocation] Enable getScratchValueSize specialization
#5070 opened Nov 5, 2024 by victor-eds Loading…
5 of 7 tasks
25
[DRAFT][AMD] Introduce OptimizeAtomicLayouts pass
#5028 opened Oct 31, 2024 by joviliast • Draft
5
[AMD] Support Cross-Lane Reduction With DPP
#5019 opened Oct 30, 2024 by knwng Loading…
4 of 7 tasks
41
[AMD] Add tritonamdgpu-block-pingpong pass
#5018 opened Oct 30, 2024 by jungpark-mlir • Draft
9
[AMD][BACKEND] Switch to code object v5
#5005 opened Oct 29, 2024 by AlexAUT • Draft
5 of 7 tasks
4
[BACKEND][NVIDIA] Add DotOp Hoisting Pass for WGMMA and Add Lowering for SMEM-to-MMAv3 DotOp Copy
#5003 opened Oct 28, 2024 by ggengnv • Draft
53
[AMD] getBackwardSlice variant with handling for op regions
#4993 opened Oct 25, 2024 by karthik-man • Draft
5 of 7 tasks
7
[AMD][prototype] Transpose between global load and local store for non-TN layouts
#4984 opened Oct 23, 2024 by jtang10 • Draft
1 of 7 tasks
Added an option to dump IR to files
#4932 opened Oct 16, 2024 by arakhmati-openai • Draft
Fix order
#4914 opened Oct 15, 2024 by rawnhenry Loading…
3 of 7 tasks
9
[SWP] Fix a bug in SWP that did not correctly compute the number of loop iterations
#4887 opened Oct 10, 2024 by sfzhu93 Loading…
5 tasks done
13
[tool][proton] Intra kernel profiling support
#4861 opened Oct 7, 2024 by fywkevin • Draft
5 of 7 tasks
2
ProTip! Adding no:label will show everything without a label.