[Improve] Refactor liif for pytorch2onnx #425
Conversation
Codecov Report
@@ Coverage Diff @@
## master #425 +/- ##
==========================================
+ Coverage 80.66% 80.84% +0.17%
==========================================
Files 188 189 +1
Lines 10106 10101 -5
Branches 1485 1485
==========================================
+ Hits 8152 8166 +14
+ Misses 1740 1721 -19
Partials 214 214
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
|
|
||
| # model | ||
| self.encoder = build_backbone(encoder) | ||
| imnet_in_dim = self.encoder.mid_channels |
There was a problem hiding this comment.
There may be a problem if self.encoder has no mid_channels.
There was a problem hiding this comment.
self.encoder can be EDSR or RDN, all of them have mid_channels.
There was a problem hiding this comment.
Then we should add a value error? Now there is a chance that users use some other networks that contain no mid_channels.
There was a problem hiding this comment.
Let's keep status quo and add helpful message when issues flood in.
| feat.shape[2], feat.shape[3]) | ||
|
|
||
| if self.local_ensemble: | ||
| vx_lst = [-1, 1] |
There was a problem hiding this comment.
It is better to replace vx, rx, ... by some more meaningful names.
* Fix typo * [Improve] Refactor liif for pytorch2onnx * Add docstring * Add checkpoints * Update * Update Co-authored-by: liyinshuo <[email protected]>
Create
generator