add mock class for specific workflow #416
Merged
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
Contributor
doombeaker
commented
Dec 12, 2023
•
doombeaker
commented
- 适配动态 batchsize 以及一个特定工作流中的动态 shape 问题
- 动态 batchsize 不支持 community 版本，说明
…lass_for_specific_workflow
…lass_for_specific_workflow
hjchen2
approved these changes
Dec 13, 2023
doombeaker
commented
Dec 13, 2023
Comment on lines
+116
to
118
| x = x.flatten(2, 3).permute(0, 2, 1) | ||
| # x = x.reshape(b, c, -1).permute(0, 2, 1) | ||
| # x = rearrange(x, 'b c h w -> b (h w) c').contiguous() |
There was a problem hiding this comment.
最原始的 ComfyUI 写法，以及上一版的 onediff 写法都保留，是因为这是一个修正动态 shape 问题的例子。
原始需求：
(b, c, h, w) -> (b, h*w, c)
第一版 onediff 的写法：
x = x.reshape(b, c, -1).permute(0, 2, 1)
此时 reshape 中的 b, c 如果不变，只有一个维度变化（-1 那个维度），那么 reshape 可以动态推导出 -1 那个维度的具体值。
但是，如果 -1 维度变化的同时， b 也变化，reshape 就无法正常推导了。
所以要用最新的版本：
x = x.flatten(2, 3).permute(0, 2, 1)
先把 (b, c, h, w) 中的后 2 个维度压到一起 (b, c, (h*w))，然后调整维度顺序，得到 (b, (h,w), c，达到与之前等价的效果。
doombeaker
commented
Dec 13, 2023
| del hsp | ||
| if len(hs) > 0: | ||
| # output_shape = hs[-1].shape | ||
| output_shape = hs[-1] |
There was a problem hiding this comment.
此处修改为 tensor，目的是传给 oneflow 的参数是 tensor，这样每次推导时能拿到 tensor 的 meta 信息。之前的 shape 作为属性，无法每次推导时更新。
doombeaker
commented
Dec 13, 2023
| if self.dims == 3: | ||
| raise ValueError("output_shape shoud not be Tensor for dims == 3") | ||
| else: | ||
| x = F.interpolate_like(x, like=output_shape, mode="nearest") |
There was a problem hiding this comment.
这个是 oneflow 新增的算子
doombeaker
commented
Dec 13, 2023
| @@ -1,19 +1,18 @@ | |||
| import oneflow | |||
There was a problem hiding this comment.
这个文件从 onediff_comfy_nodes 搬到 onediff 下了，它确实应该是 onediff 层次的东西（获取 oneflow 版本信息）
doombeaker
commented
Dec 13, 2023
| @@ -26,12 +25,6 @@ def is_quantization_enabled(): | |||
| return hasattr(oneflow._C, "dynamic_quantization") | |||
|
|
|||
|
|
|||
| def is_community_version(stop_if_not=False): | |||
| def is_community_version(): | |||
There was a problem hiding this comment.
去掉了打印功能，开发者自己决定是否打印，或者做什么处理（否则这个 PR 里调用了 2 此，打印两次，比较多余）
ccssu
approved these changes
Dec 13, 2023
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.