Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / triton-lang/triton issues and pull requests

#4582 - [BACKEND] Optimize code generation for load with other arg

Pull Request - State: closed - Opened by ThomasRaoux 3 months ago - 4 comments

#4581 - [WIP] Optimize fma dot

Pull Request - State: open - Opened by binarman 3 months ago - 1 comment

#4580 - Bump tj-actions/changed-files from 44 to 45

Pull Request - State: closed - Opened by dependabot[bot] 3 months ago
Labels: dependencies

#4576 - PTX version error with CUDA 12.6

Issue - State: closed - Opened by sclarkson 3 months ago - 1 comment

#4575 - The number of generated TTIR files

Issue - State: closed - Opened by LHW-CLOUD 3 months ago - 2 comments

#4573 - IEEE is common and TF32 is specific to CUDA

Pull Request - State: closed - Opened by parsifal-47 3 months ago

#4572 - [Readme] Fix command to get compile command

Pull Request - State: closed - Opened by jayzhan211 3 months ago - 2 comments

#4571 - [BACKEND] Continue the backward slice when finding free convert

Pull Request - State: closed - Opened by Jokeren 3 months ago - 1 comment

#4570 - .so文件的作用？

Issue - State: closed - Opened by LHW-CLOUD 3 months ago - 3 comments

#4568 - [Proton] Move additional hatchet import into try/except check

Pull Request - State: closed - Opened by CRobeck 3 months ago - 1 comment

#4566 - [FRONTEND] Print full file name when overriding kernel

Pull Request - State: closed - Opened by htyu 3 months ago

#4564 - [RELEASE] Cherry pick device agnostic do_bench change

Pull Request - State: open - Opened by anishsarum 3 months ago - 1 comment

#4562 - Fix underflow in highestPowOf2Divisor()

Pull Request - State: closed - Opened by chsigg 3 months ago - 1 comment

#4559 - Hoist reduction outside a loop

Pull Request - State: open - Opened by binarman 3 months ago - 6 comments

#4558 - [TEST] Use device fixture for test_math_extern

Pull Request - State: closed - Opened by int3 3 months ago

#4556 - [WIP] Try LLVM integrate

Pull Request - State: closed - Opened by ThomasRaoux 3 months ago

#4555 - [CI][AMD] Re-enable MI200 CI

Pull Request - State: closed - Opened by jungpark-mlir 3 months ago - 2 comments

#4554 - [NFC] Simplify getThreadId function

Pull Request - State: closed - Opened by linuxlonelyeagle 3 months ago

#4553 - Use base64 for shorter cache directories

Pull Request - State: closed - Opened by minjang 3 months ago

#4552 - [Tutorial] 06-fused-attention.py - add tma

Pull Request - State: closed - Opened by yzh119 3 months ago - 1 comment

#4551 - Numeric errors with tf32 matmul on A100 GPU

Issue - State: closed - Opened by axelfeldmann 3 months ago - 3 comments

#4549 - [CI][AMD] Disable MI200 CI

Pull Request - State: closed - Opened by jungpark-mlir 3 months ago - 5 comments

#4547 - [BACKEND][DRAFT] Adjust the padding heuristic for convert layout

Pull Request - State: closed - Opened by Jokeren 3 months ago

#4546 - [RFC][FRONTEND] Use the index type for scf loop

Pull Request - State: closed - Opened by htyu 3 months ago - 4 comments

#4545 - [Frontend] Add TRITON_PRINT_AUTOTUNING_ALL flag

Pull Request - State: open - Opened by plotfi 3 months ago - 3 comments

#4544 - The question of implementing the addmm_ operator based on triton

Issue - State: closed - Opened by hyx1999 3 months ago - 1 comment

#4543 - [Release/3.0.x] Cherry pick Flex attention support from mainline

Pull Request - State: closed - Opened by jerrymannil 3 months ago - 1 comment

#4542 - [TEST] Insert barriers in test_atomic_cas to sequence store and atomic

Pull Request - State: closed - Opened by jungpark-mlir 3 months ago - 2 comments

#4540 - Cherry Pick #4138 into release/3.0x branch

Pull Request - State: open - Opened by drisspg 3 months ago - 2 comments

#4539 - Add mechanism for remapping device-specific module imports

Pull Request - State: closed - Opened by int3 3 months ago - 2 comments

#4535 - [FRONTEND] `interleave` does not need to check shape

Pull Request - State: closed - Opened by Mwsxy 3 months ago - 2 comments

#4534 - [Pipeliner] Implement dynamic loop peeling

Pull Request - State: closed - Opened by sjw36 3 months ago - 15 comments

#4532 - [BACKEND] Fix common mistake of missing checks for null pointer

Pull Request - State: closed - Opened by ThomasRaoux 3 months ago - 4 comments

#4531 - Confusion about memory of pointers

Issue - State: closed - Opened by FelixSchoen 3 months ago - 1 comment

#4528 - Enable verbose asm

Pull Request - State: closed - Opened by ravil-mobile 3 months ago - 4 comments

#4526 - [CI][AMD] Create a CI docker with non-root user

Pull Request - State: open - Opened by yiqian1 3 months ago

#4525 - [BACKEND][NFC] Interface for LinearLayout conversion.

Pull Request - State: closed - Opened by hwnam831 3 months ago - 5 comments

#4524 - feat: support `sem`/`scope` in `tl.{load,store}`

Pull Request - State: closed - Opened by 0x804d8000 3 months ago - 2 comments

#4523 - [NFC][BACKEND] Remove dead code in Nvidia backend's ConvertLayoutOpConversion

Pull Request - State: open - Opened by Jokeren 3 months ago - 1 comment

#4522 - What is the use of divisibility in coalesce pass

Issue - State: open - Opened by Shoreshen 3 months ago

#4521 - FlexAttention Segmentation Fault

Issue - State: open - Opened by drisspg 3 months ago - 2 comments

#4520 - [CI][AMD] Enable MI300 CI

Pull Request - State: closed - Opened by zhanglx13 3 months ago

#4519 - [BACKEND] Fp8E5M2Nv to Fp16 conversion support added to NV backend

Pull Request - State: closed - Opened by plotfi 3 months ago - 5 comments

#4518 - [AMD][gfx12] Support emit indices logic WMMAv2 layout

Pull Request - State: closed - Opened by joviliast 3 months ago

#4517 - [DOC] Improve the description of Transformation passes

Pull Request - State: closed - Opened by mfrancepillois 3 months ago - 1 comment

#4516 - [Backend] Improve dot support to target FMA

Pull Request - State: open - Opened by binarman 3 months ago - 2 comments

#4514 - [FRONTEND] Add hooks to signal that compilation is done

Pull Request - State: closed - Opened by ThomasRaoux 3 months ago - 1 comment

#4513 - make error

Issue - State: open - Opened by tangpanyu 3 months ago

#4512 - [BACKEND] Set LLVM_ABI_BREAKING_CHECKS to be able to update the llvm version

Pull Request - State: closed - Opened by karupayun 3 months ago - 4 comments

#4511 - Cannot find 2.0.0.dev20221202 version

Issue - State: open - Opened by anttitapsa 3 months ago - 10 comments

#4510 - [backend] NFC: Fix ptx `st` argument order

Pull Request - State: closed - Opened by chsigg 3 months ago

#4509 - Should @core.extern be part of the libdevice interface?

Issue - State: open - Opened by int3 3 months ago - 1 comment

#4504 - [AMD] Add barrier at the beginning of each atomic operations.

Pull Request - State: closed - Opened by jungpark-mlir 3 months ago - 15 comments

#4503 - Allow third-party backends to add submodules to `triton.language.extra`

Pull Request - State: closed - Opened by Alfie-Edwards 3 months ago - 9 comments

#4502 - [Bug] Assertion `idx < size()' failed.

Issue - State: open - Opened by zhyncs 3 months ago - 10 comments
Labels: enhancement

#4500 - Try out MI210 build bot

Pull Request - State: open - Opened by antiagainst 3 months ago

#4499 - [BUILD] Include backend files to build nvidia backend

Pull Request - State: closed - Opened by Jokeren 3 months ago

#4498 - [nvidia] Support passing TMA descriptors by-value

Pull Request - State: closed - Opened by embg 3 months ago

#4497 - Revert "[CI][AMD] Reenable MI300 CI (#4453)"

Pull Request - State: closed - Opened by antiagainst 3 months ago

#4496 - [AUTOTUNER] Make autotuner take `do_bench` as a parameter

Pull Request - State: closed - Opened by int3 3 months ago - 3 comments

#4495 - [Proton] Add warning in Proton viewer about negative byte values

Pull Request - State: closed - Opened by CRobeck 3 months ago - 1 comment

#4494 - Worse performance on `H100` than `RTX3090`

Issue - State: closed - Opened by jeromeku 3 months ago

#4493 - feat: add return all for do_bench

Pull Request - State: closed - Opened by OrenLeung 3 months ago

#4492 - [BACKEND] Support Hopper MMA to MMA convert_layout ops

Pull Request - State: closed - Opened by Jokeren 3 months ago - 3 comments

#4491 - [AMD][gfx12] Support WMMAv2 dot instruction generation

Pull Request - State: closed - Opened by joviliast 3 months ago

#4490 - fix

Pull Request - State: closed - Opened by ArtificialZeng 3 months ago

#4489 - [amd] NFC: Fix typos.

Pull Request - State: closed - Opened by chsigg 3 months ago

#4488 - Output dtype of 09-persistent-matmul.py

Issue - State: open - Opened by xijiu9 3 months ago

#4487 - Allow location rewriting to apply to any part of IR lowering

Pull Request - State: closed - Opened by int3 3 months ago - 2 comments

#4486 - [triton][tool] A CLI Tool for Tensor Layout Printing

Pull Request - State: closed - Opened by fywkevin 3 months ago - 1 comment

#4485 - [TESTING] Remove the `fast_flush` parameter from `do_bench`

Pull Request - State: closed - Opened by int3 3 months ago - 2 comments

#4484 - FP8 Casting Tensor Size Alignment Error on AMD MI210

Issue - State: closed - Opened by kimbo-a2labs 3 months ago - 3 comments

#4483 - [WIP][gfx11] Support tied wmma instrucrions

Pull Request - State: open - Opened by joviliast 3 months ago - 2 comments

#4480 - Does Triton have a roadmap plan?

Issue - State: open - Opened by MeJerry215 3 months ago - 1 comment

#4476 - Fused Attention FP8 correctness on L20(Ada GPU)

Issue - State: open - Opened by suluner 3 months ago - 1 comment