Aditya Oke
Aditya Oke
## 🚀 Feature Frequently re-used losses that can be added. ### Motivation Writing Losses is quite repetitive. PyTorch supports losses which are written with deep interoperability with C++ API. But...
## 🚀 Feature We use it internally to test our codebase for detection models. We can make it publically available as part of API. `from quickvision.datasets.detection import dummydetectiondataset` We can...
## 🚀 Feature - Sphinx Documentation akin to PyTorch or PyTorch Lightning will be nice. - Plan of Action - [x] Add Sphinx Docs, Makefiles, conf.py etc. - [x] Build...
## 🚀 Feature Use ` torch.jit.annotation` to annotate Use `typing` package for `Optional` ` tuple, list` etc
## 🚀 Feature Similar to what we did for classification, probably we should provide something for detection. This will allow to load pretrained weights from Kitty, COCO, etc. datasets.
## 🚀 Feature This should take some study and time. Supporting Torchvision Segmentation models should be simple. But we would need to support backbones, etc. A nice place to know...
## 🚀 Feature We need inference script for all models. Since we support lightning only for training. We need to do inference only in torch. Inference should be able to...
### 🚀 The feature We need to currently close the Issues raised by our automated tests for datasets manually. https://github.com/pytorch/vision/issues?q=is%3Aissue+is%3Aopen+Scheduled+workflow+failed ? Sometimes they get bunched up or we forget to...
## 🚀 Feature A loss functions API in torchvision. ## Motivation The request is simple, we have loss functions available in torchvision E.g. `sigmoid_focal_loss` , `l1_loss`. But these are quite...
### 🚀 The feature Follow up of #4216 . The current utility works fine, but we can make it better :) ### Motivation, pitch I think the following can be...