triton-lang triton issues

triton-lang / triton

Development repository for the Triton language and compiler

https://triton-lang.org/

MIT License

13.5k stars 1.67k forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

[DRAFT] Completely remove `MemDesc` from the Triton dialect

#5208 Jokeren closed 4 days ago
1
Add support for 3-5D TMA to allow loading non-matmul operands

#5207 masahi opened 4 days ago
0
[CI] Fix `du` failling if cache restore fails

#5206 peterbell10 closed 4 days ago
0
Model inference using HF accelerate causes ValueError: Pointer argument (at 0) cannot be accessed from Triton (cpu tensor?)

#5205 shivam15s opened 4 days ago
0
fp32 addmm returns incorrect result on A100

#5204 davidberard98 opened 4 days ago
5
[AMD] Prevent wrong reordering of scf operations

#5203 ThomasRaoux closed 4 days ago
4
[CI] Fix ccache cache restoration to improve build times

#5202 peterbell10 closed 4 days ago
0
[INTERPRETER] Fix lower bound check for block pointers

#5201 Jokeren closed 5 days ago
0
[LAYOUTS] [BE] Simplify Ampere/Hopper paths introduced in #5189

#5200 lezcano closed 3 days ago
2
[LLs] [BE] Simplify identityND

#5199 lezcano closed 5 days ago
0
Update README.md to remove triton conference

#5198 sophiawisdom closed 5 days ago
0
[MXFP] Fix packing for mxfp4 type

#5197 ThomasRaoux closed 5 days ago
1
[BACKEND] Use the LL API to replace the using of legacy layout attribute API.

#5196 chengjunlu closed 5 days ago
0
[INTERPRETER][NFC] Rename `tensor_shape` -> `block_shape` in interpreter

#5195 Jokeren closed 5 days ago
0
[IR] Remove memdesc from `tt.trans` and implements `ttg.memdesc_trans`

#5194 Jokeren closed 5 days ago
4
[LAYOUTS] Implement LL conversion for DotOperand(Hopper)

#5193 lezcano closed 5 days ago
2
[BACKEND] Fix accumulator init optimization for integer matmuls

#5192 peterbell10 closed 6 days ago
0
[AMD] release/3.2.x AMD perf cherry picks

#5191 jataylo opened 6 days ago
0
Build smaller LLVM binaries for Triton

#5190 tiran opened 6 days ago
0
[BACKEND] Fix getElemsPerThread for mmav3 dot operand

#5189 ThomasRaoux closed 6 days ago
0
https://github.com/triton-lang/triton/pull/4222 not included in triton 3.1

#5188 goldhuang closed 6 days ago
1
Atomic Load and Store operations for Triton (tl.atomic_store/tl.atomic_load)

#5187 plotfi opened 6 days ago
1
Windows related changes in `CMakeLists.txt`

#5186 anmyachev closed 8 hours ago
6
Support computation pipelining after SWP refactoring

#5185 manman-ren opened 1 week ago
1
[BACKEND] Add missing precondition in optimize acc init

#5184 ThomasRaoux closed 1 week ago
0
[LAYOUTS] Unify the implementation of getShapePerCTATile

#5183 lezcano closed 5 days ago
0
The lastest triton cannot run tutorials

#5182 zhaosiying12138 closed 1 week ago
1
[BACKEND] Fix ProgramPoint passing in AxisInfoAnalysis

#5181 aakhundov closed 1 week ago
3
Update to llvm/llvm-project@bd9145c8c213

#5180 antiagainst closed 1 week ago
0
Update to llvm/llvm-project@bd9145c8c213

#5179 antiagainst closed 1 week ago
0
[Persistent][Pipelining] Fused 08-grouped-gemm 2 inner loops for better pipelining, got bad performance

#5178 CharlieFRuan opened 1 week ago
1
[AMD] Enable mixed precision matmul test

#5177 makslevental closed 1 week ago
0
[PIPELINER] Cleanup of LoopScheduling.cpp, introduction of AssignLatencies

#5176 pawelszczerbuk opened 1 week ago
1
[AMD][Pipeliner] Reland "Improve clustering and add prefetch"

#5175 sjw36 closed 6 days ago
0
[Persistent] Performance on 09-persistent-matmul on A100 worse than non-persistent

#5174 CharlieFRuan opened 1 week ago
1
[Triton] Default diagnostic handler only filters for errors

#5173 Mogball closed 1 week ago
0
[Persistent][Pipeline] Moving `tile_id += NUM_SMS` to epilogue fails to pipeline in 09-persistent-matmul

#5172 CharlieFRuan opened 1 week ago
4
[CI] Disable `MLIR_ENABLE_REMARK`

#5171 Jokeren closed 1 week ago
0
[LAYOUTS] Implement IR support for LinearLayouts

#5170 lezcano closed 4 days ago
6
[INTERPRETER] Fix argument passing for internal parameters in function declarations

#5169 Jokeren closed 1 week ago
0
[AMD] NFC: Drop duplicated moveUpTranspose

#5168 antiagainst closed 1 week ago
0
[BACKEND] Cleanup redundant broadcast combine pattern

#5167 peterbell10 closed 1 week ago
0
[BACKEND] Add folder for `addptr(ptr, 0) -> ptr`

#5166 peterbell10 closed 1 week ago
0
[BACKEND] Update LLVM version to https://github.com/llvm/llvm-project/commit/fb4f426c81d7e87dbb30df7abeba15ffc2f9f41a

#5165 vwbaker closed 1 week ago
1
Setting the environment variable TRITON_INTERPRET causes the kernel function to not be able to receive reserved keyword arguments.

#5164 0Addicted0 closed 1 week ago
0
[AMD] Add instruction schedule loop boundary guard hints

#5163 ravil-mobile opened 1 week ago
3
[Tutorial] Remove incorrect caching from softmax tutorial

#5162 Mogball opened 1 week ago
0
[TritonGPU] Fix incorrect mask operand used in for loop pipeliner

#5161 Mogball closed 1 week ago
0
Restore CentOS 7 build and backfill for release/3.2.x

#5160 bertmaher closed 1 week ago
0
Add Sageattention Codes as a tutorial

#5159 jt-zhang closed 1 week ago
2

Previous Next