【Hackathon 7th No.38】为 Paddle 代码转换工具新增 API 转换规则（第5组） #487
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
|
CI-codestyle 里显示没用pre commit导致报错，but我用pre commit 的话会导致json文件全覆盖（p1），例子p2中可以看到diff显示全覆盖，但实际上只有一行进行了更改 |
zhwesky2010
reviewed
Oct 9, 2024
| @@ -1450,6 +1450,14 @@ def generate_code(self, kwargs): | |||
| return GenericMatcher.generate_code(self, kwargs) | |||
|
|
|||
|
|
|||
| class ScatterReduceMatcher(BaseMatcher): | |||
| def generate_code(self, kwargs): | |||
| reduce_mapping = {'"""sum"""': '"add"', '"""prod"""': '"multiply"'} | |||
There was a problem hiding this comment.
这个可能得用辅助函数形式写，因为传入的reduce可能是个变量
例如：
reduce_type = 'sum'
torch.scatter_reduce(reduce=reduce_type, ...)
paconvert/api_matcher.py
Outdated
| class CartesianProdMatcher(BaseMatcher): | ||
| def get_paddle_nodes(self, args, kwargs): | ||
| new_args = self.parse_args(args) | ||
| code = "paddle.cartesian_prod([ {}".format(new_args[0]) |
There was a problem hiding this comment.
- 如果输入的是变量形式，应该无法支持，参考 Chain_MatmulMatcher
- new_args返回的就是一个list，是不是直接format到字符串，目前这样写代码有点丑
paconvert/api_matcher.py
Outdated
| return "{}.cast(paddle.float64).pow({})".format( | ||
| self.paddleClass, kwargs["exponent"] | ||
| self.write_aux_code() | ||
| _from_dtype = kwargs["from_"][3:-3] |
There was a problem hiding this comment.
这么写代码太硬了，一定要这样写死切片吗，如果输入类型是变量，会出错吧，例如：
dtype1=torch.float32
dtype2=torch.float64
torch.can_cast(dtype1, dtype2)
paconvert/api_matcher.py
Outdated
|
|
||
|
|
||
| class PositiveMatcher(BaseMatcher): | ||
| def generate_code(self, kwargs): |
There was a problem hiding this comment.
这个这么多行，还是用辅助函数来写吧，不然很容易破坏原代码的结构
paconvert/api_matcher.py
Outdated
| ) | ||
| else: | ||
| if "out" not in kwargs: | ||
| return "paddle.pow({}.cast(paddle.float64), {}.cast(paddle.float64) if isinstance({}, paddle.Tensor) else {})".format( |
There was a problem hiding this comment.
exponent也必须cast吗，只用cast input就行吧，这个应该自带类型提升
There was a problem hiding this comment.
应该不行，paddle.pow(a, b) 中，a 和 b 都是 tensor 的时候数据类型必须一致，否则会报错
|
@inaomIIsfarell CI未通过且出现冲突，需要修改 |
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
PaddlePaddle/docs#6885
PR APIs
ps: 如果有朋友用
vscode修改paconvert/api_mapping.json和paconvert/api_alias_mapping.json，并用pre-commitcommit 时，出现 json 文件diff 全覆盖的情况，可能 是 pre-commit 的问题