This is the development repository of Triton, a language and compiler for writing highly efficient custom Deep-Learning primitives. The aim of Triton is to provide an open-source environment to write fast code at higher productivity than CUDA, but also with higher flexibility than other existing DSLs.
Uploaded | Tue Apr 1 00:35:41 2025 |
md5 checksum | 5687ce20421c9912fe19ba8545312b3f |
arch | x86_64 |
build | cuda124py312hac27663_3 |
build_number | 3 |
depends | cuda-cudart, cuda-cuobjdump, cuda-cupti >=12.4.127,<13.0a0, cuda-nvcc, cuda-version >=12.4,<13, filelock, libgcc-ng >=11.2.0, libllvm19 >=19.1.7,<19.2.0a0, libmlir19 >=19.1.7,<19.2.0a0, libstdcxx-ng >=11.2.0, lit, python >=3.12,<3.13.0a0, zlib >=1.2.13,<1.3.0a0 |
license | MIT |
license_family | MIT |
md5 | 5687ce20421c9912fe19ba8545312b3f |
name | triton |
platform | linux |
sha1 | 14e0c8943dccada9a7e5ed95f92827c2fbc1cb61 |
sha256 | 796a6205df1538d5735789ab8a75b2255502436305a0a0c6f2622b3c3755437e |
size | 94605312 |
subdir | linux-64 |
timestamp | 1741881203967 |
version | 3.1.0 |