Eric Wolf

Allow unpacking of TypedDict into TypedDict

3

### Description Closes #9408 (and duplicates, e.g. #10585, #11108). Related tracking issue: #11753 Allows unpacking of TypedDicts into other TypedDicts: ``` class Foo(TypedDict): a: int class Bar(TypedDict): a: int b:...

`fasterrcnn_resnet50_fpn` edits weights on-the-fly

1

### 🐛 Describe the bug `fasterrcnn_resnet50_fpn()` includes code to edit weights on-the-fly while loading them: https://github.com/pytorch/vision/blob/bf01bab6125c5f1152e4f336b470399e52a8559d/torchvision/models/detection/faster_rcnn.py#L578-578 Therefore, if you attempt to load weights after model instantiation, you will get wrong...

## 🐛 Bug The links from the docs (as hosted on https://lightning-bolts.readthedocs.io/en/latest/) are incorrect. ### To Reproduce For example, on the [Vision datamodules page](https://lightning-bolts.readthedocs.io/en/latest/datamodules/vision.html), the first link to source points...

## 🐛 Bug `VisionDataModule` by default subdivides the train split into a train and val split: https://github.com/Lightning-Universe/lightning-bolts/blob/541f7018492b6a3b16558f9ea2a763b02a007a66/src/pl_bolts/datamodules/vision_datamodule.py#L109 This behavior can be disabled by setting `val_split=0` (in some cases also by...