| commit | 3c6ca5d92be2f147608df545073bbfd702d49be5 | [log] [tgz] |
|---|---|---|
| author | Eric Lunderberg <Lunderberg@users.noreply.github.com> | Thu Jun 27 09:52:09 2024 -0500 |
| committer | GitHub <noreply@github.com> | Thu Jun 27 09:52:09 2024 -0500 |
| tree | c99d09e8af8f06f1d336dca1f5b5d248ef737001 | |
| parent | 73cad19cfa2de955880c52c150e8639295fa4489 [diff] |
[Bugfix][Relax] Set purity=false for LazySetOutput (#17119) The `relax.transform.LazySetOutput` transformation updates a Relax function to produce output from a `fset_output` callback. In the initial implementation, the `fset_output` was marked as a pure function, which allowed it to be erroneously removed from a function. This commit updates the `relax::FuncStructInfo` used to annotate `fset_output`, marking it as an impure function.
Documentation | Contributors | Community | Release Notes
Apache TVM is a compiler stack for deep learning systems. It is designed to close the gap between the productivity-focused deep learning frameworks, and the performance- and efficiency-focused hardware backends. TVM works with deep learning frameworks to provide end to end compilation to different backends.
TVM is licensed under the Apache-2.0 license.
Check out the TVM Documentation site for installation instructions, tutorials, examples, and more. The Getting Started with TVM tutorial is a great place to start.
TVM adopts apache committer model, we aim to create an open source project that is maintained and owned by the community. Check out the Contributor Guide.
We learned a lot from the following projects when building TVM.