Adding multiweight support to SSD #4881
Conversation
💊 CI failures summary and remediationsAs of commit 19bb0f7 (more details on the Dr. CI page):
1 failure not recognized by patterns:
This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
datumbox
left a comment
datumbox
left a comment
There was a problem hiding this comment.
Adding few comments to assist review:
| weights = SSD300VGG16Weights.verify(weights) | ||
| if "pretrained_backbone" in kwargs: | ||
| warnings.warn("The argument pretrained_backbone is deprecated, please use weights_backbone instead.") | ||
| weights_backbone = VGG16Weights.ImageNet1K_Features if kwargs.pop("pretrained_backbone") else None |
There was a problem hiding this comment.
We are using special weights here. Not the standard VGG weights of TorchVision.
| "image_mean": [0.48235, 0.45882, 0.40784], | ||
| "image_std": [1.0 / 255.0, 1.0 / 255.0, 1.0 / 255.0], # undo the 0-1 scaling of toTensor | ||
| } | ||
| kwargs: Any = {**defaults, **kwargs} |
There was a problem hiding this comment.
Any type declaration needed to mypy.
| # We port the features of a VGG16 backbone trained by amdegroot because unlike the one on TorchVision, it uses the | ||
| # same input standardization method as the paper. Only the `features` weights have proper values, those on the | ||
| # `classifier` module are filled with nans. | ||
| ImageNet1K_Features = WeightEntry( |
There was a problem hiding this comment.
We are adding the weights as a separate entry on the VGG. This one only has the features part of the model.
| "interpolation": InterpolationMode.BILINEAR, | ||
| "recipe": "https://github.com/amdegroot/ssd.pytorch#training-ssd", | ||
| "acc@1": float("nan"), | ||
| "acc@5": float("nan"), |
There was a problem hiding this comment.
It doesn't contain a classifier, hence the nans here to denote it.
jdsgomes
left a comment
jdsgomes
left a comment
There was a problem hiding this comment.
LGTM, just left a minor comment for consideration
| "image_std": [1.0 / 255.0, 1.0 / 255.0, 1.0 / 255.0], # undo the 0-1 scaling of toTensor | ||
| } | ||
| kwargs: Any = {**defaults, **kwargs} | ||
| model = SSD(backbone, anchor_generator, (300, 300), num_classes, **kwargs) |
There was a problem hiding this comment.
should we get the size from the weights?
There was a problem hiding this comment.
It's the same discussion as in #4875 (comment).
Note that unlike the FasterRCNN models that adapt better on variable input size, SSD doesn't. It was investigated at #3819 and that's why it's hardcoded.
Reviewed By: datumbox Differential Revision: D32298973 fbshipit-source-id: e74e8cfc564f13681466a4006a3a879731b597ec
3 participants
Fixes partially #4675
Verified that both commands return the same:
cc @datumbox @bjuncek