转换规则 No.358/361 #256
Merged
转换规则 No.358/361 #256
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thanks for your contribution! |
zhwesky2010
reviewed
Aug 28, 2023
paconvert/api_matcher.py
Outdated
| def generate_code(self, kwargs): | ||
| out_v = kwargs.pop("out") if "out" in kwargs else None | ||
| check_errors_v = ( | ||
| kwargs.pop("check_errors") if "check_errors" in kwargs else None |
There was a problem hiding this comment.
感觉可以不转这种 报错检查 之类的功能，对输出结果无影响，这样可以减少一行转换后的行数
zhwesky2010
reviewed
Aug 29, 2023
| x = torch.tensor([[1.0, 2.0], [3.0, 4.0], [5.0, 6.0]], dtype=torch.float64) | ||
| out = (torch.tensor([], dtype=torch.float64), torch.tensor([], dtype=torch.int), torch.tensor([], dtype=torch.int)) | ||
| LU, pivots, info = torch.linalg.lu_factor_ex(A=x, pivot=True, check_errors=True, out=out) | ||
| info = info.item() |
There was a problem hiding this comment.
这个info是有一些输出的不同吗？比如paddle是Tensor，torch是python scalar？
There was a problem hiding this comment.
shape不一样，paddle是[1]，torch是整数值
There was a problem hiding this comment.
shape不一样，paddle是[1]，torch是整数值
这个地方那也需要转写一下，在转写里第三个参数里改成 info.item()
paconvert/api_matcher.py
Outdated
| @@ -4061,6 +4061,60 @@ def generate_code(self, kwargs): | |||
| return code | |||
|
|
|||
|
|
|||
| class LinalgLufactorMatcher(BaseMatcher): | |||
There was a problem hiding this comment.
这个应该可以复用 TupleAssignMatcher
paconvert/api_matcher.py
Outdated
| return code | ||
|
|
||
|
|
||
| class LinalgLufactorexMatcher(BaseMatcher): |
There was a problem hiding this comment.
可以改成一个通用的 TripleAssignMatcher，配置：
"check_errors": ""
这样代码能最大化的保持复用，方便日后维护更新
|
还需要处理下冲突 |
已修改 |
zhwesky2010
approved these changes
Sep 4, 2023
| x = torch.tensor([[1.0, 2.0], [3.0, 4.0], [5.0, 6.0]], dtype=torch.float64) | ||
| out = (torch.tensor([], dtype=torch.float64), torch.tensor([], dtype=torch.int), torch.tensor([], dtype=torch.int)) | ||
| LU, pivots, info = torch.linalg.lu_factor_ex(A=x, pivot=True, check_errors=True, out=out) | ||
| info = info.item() |
There was a problem hiding this comment.
shape不一样，paddle是[1]，torch是整数值
这个地方那也需要转写一下，在转写里第三个参数里改成 info.item()
|
@co63oc 我先合入了，这个 |
Closed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR Docs
#112
358 torch.linalg.lu_factor
361 torch.linalg.lu_factor_ex 返回info的shape不一致，修改为info=info.item()比较, check_errors参数映射为增加assert验证
映射文档 PaddlePaddle/docs#6139
PR APIs