mirror of
https://github.com/pytorch/pytorch.git
synced 2025-10-20 21:14:14 +08:00
ec1f56fdcf
This PR adds support for prune_configs_by in the @triton.autotune decorator [docs](https://triton-lang.org/main/python-api/generated/triton.autotune.html#triton.autotune). Supporting this lets users reduce autotuning time by running user-supplied code (early_config_prune, perf_model) to prune the provided list of configs. We implement this by realizing args/kwargs in call_triton_kernel(...), and then calling kernel.prune_configs(...). Pull Request resolved: https://github.com/pytorch/pytorch/pull/142207 Approved by: https://github.com/zou3519, https://github.com/aakhundov
Note [TH abstraction violation] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ TH/THC provide some hpp headers, which are proper C++ headers rather than C headers. These headers serve double duty as *internal implementation detail* headers, whose contents should largely not be used by external clients. Ideally, we would not install these headers at all; instead, you should use public functions (in headers like `THTensor.h`, NOT `THTensor.hpp`) to manipulate these structs. However, there are a few places in torch/csrc where we violate this abstraction. They are marked with a pointer to this note. Each of those sites will have to be refactored when we refactor the guts of THTensor and related structures.