llvm
/
torch-mlir
mirror of https://github.com/llvm/torch-mlir
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fix typos

pull/1612/head
Kazuaki Ishizaki 2022-11-18 02:23:50 +09:00 committed by Sean Silva
parent 39de4d6265
commit 638a884e8c
3 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
docs/adding_an_e2e_test.md
View File

@ -146,7 +146,7 @@ object as the second argument. The `TestUtils` has some methods, such as
and and
[`tu.randint`](https://github.com/llvm/torch-mlir/blob/8e880a2d009b67d45fb07434ab62ec2066a11185/python/torch_mlir_e2e_test/torchscript/framework.py#L185), [`tu.randint`](https://github.com/llvm/torch-mlir/blob/8e880a2d009b67d45fb07434ab62ec2066a11185/python/torch_mlir_e2e_test/torchscript/framework.py#L185),
that allow the creation of random tensors in a way that makes sure the compiled that allow the creation of random tensors in a way that makes sure the compiled
module and the golden trace recieve the same tensors as input. Therefore, all module and the golden trace receive the same tensors as input. Therefore, all
random inputs should be generated through the `TestUtils` object. random inputs should be generated through the `TestUtils` object.

4
docs/architecture.md
View File

@ -325,7 +325,7 @@ blend of the benefits of the other two.
All the backends are implemented using the MLIR [Dialect Conversion All the backends are implemented using the MLIR [Dialect Conversion
infrastructure](https://mlir.llvm.org/docs/DialectConversion/). This involves infrastructure](https://mlir.llvm.org/docs/DialectConversion/). This involves
converting the `torch` dialect types to other types, so we closely follow the converting the `torch` dialect types to other types, so we closely follow the
principes from the "Type Conversions the Not-So-Hard Way" talk principles from the "Type Conversions the Not-So-Hard Way" talk
([slides](https://drive.google.com/file/d/1FVbzCXxZzS9LBLuvpPNLWJD-XDkt54ky/view?usp=sharing), ([slides](https://drive.google.com/file/d/1FVbzCXxZzS9LBLuvpPNLWJD-XDkt54ky/view?usp=sharing),
[recording](https://drive.google.com/file/d/1VfVajitgf8ZPnd-HRkJvaJiFLhBsluXN/view?usp=sharing)). [recording](https://drive.google.com/file/d/1VfVajitgf8ZPnd-HRkJvaJiFLhBsluXN/view?usp=sharing)).
We follow the standard `{include,lib}/Conversion/TorchTo*` convention used in We follow the standard `{include,lib}/Conversion/TorchTo*` convention used in
@ -396,7 +396,7 @@ Because of this, many Torch-MLIR patches adding support for new ops have no
`.mlir` unit tests, and only include end-to-end test(s). We generally make sure `.mlir` unit tests, and only include end-to-end test(s). We generally make sure
that our end-to-end tests are as targeted as possible. As a result, when that our end-to-end tests are as targeted as possible. As a result, when
debugging end-to-end test failures, the resulting reproducers (which our test debugging end-to-end test failures, the resulting reproducers (which our test
framework automaticaly produces for failures) are usually already fully reduced framework automatically produces for failures) are usually already fully reduced
test cases. test cases.
### Do's and Don'ts for unit vs end-to-end testing. ### Do's and Don'ts for unit vs end-to-end testing.

2
docs/development.md
View File

@ -273,7 +273,7 @@ docker build -t your-name/torch-mlir-ci --no-cache .
### Other configurable environmental variables ### Other configurable environmental variables
The following additional environmental variables can be used to customie your docker build: The following additional environmental variables can be used to customize your docker build:
* Custom Release Docker image: * Custom Release Docker image:
Defaults to `stellaraccident/manylinux2014_x86_64-bazel-5.1.0:latest` Defaults to `stellaraccident/manylinux2014_x86_64-bazel-5.1.0:latest`