Lightning-bolts

- Fixed confused logit ([222](https://github.com/Lightning-AI/lightning-bolts/pull/222))

Added

- Added pretrained VAE with resnet encoders and decoders
- Added pretrained AE with resnet encoders and decoders
- Added CPC pretrained on CIFAR10 and STL10
- Verified BYOL implementation

Changed

- Dropped all dependencies except PyTorch Lightning and PyTorch
- Decoupled datamodules from GAN ([206](https://github.com/Lightning-AI/lightning-bolts/pull/206))
- Modularize AE & VAE ([196](https://github.com/Lightning-AI/lightning-bolts/pull/196))

Fixed

- Fixed gym ([221](https://github.com/Lightning-AI/lightning-bolts/pull/221))
- Fix L1/L2 regularization ([216](https://github.com/Lightning-AI/lightning-bolts/pull/216))
- Fix max_depth recursion crash in AsynchronousLoader ([191](https://github.com/Lightning-AI/lightning-bolts/pull/191))

Added

- Enabled Apache License, Version 2.0

Changed

- Moved unnecessary dependencies to `__main__` section in BYOL ([176](https://github.com/Lightning-AI/lightning-bolts/pull/176))

Fixed

- Fixed CPC STL10 finetune ([173](https://github.com/Lightning-AI/lightning-bolts/pull/173))

Added

- Added Faster RCNN + Pscal VOC DataModule ([157](https://github.com/Lightning-AI/lightning-bolts/pull/157))
- Added a better lars scheduling `LARSWrapper` ([162](https://github.com/Lightning-AI/lightning-bolts/pull/162))
- Added CPC finetuner ([158](https://github.com/Lightning-AI/lightning-bolts/pull/158))
- Added `BinaryMNISTDataModule` ([153](https://github.com/Lightning-AI/lightning-bolts/pull/153))
- Added learning rate scheduler to BYOL ([148](https://github.com/Lightning-AI/lightning-bolts/pull/148))
- Added Cityscapes DataModule ([136](https://github.com/Lightning-AI/lightning-bolts/pull/136))
- Added learning rate scheduler `LinearWarmupCosineAnnealingLR` ([138](https://github.com/Lightning-AI/lightning-bolts/pull/138))
- Added BYOL ([144](https://github.com/Lightning-AI/lightning-bolts/pull/144))
- Added `ConfusedLogitCallback` ([118](https://github.com/Lightning-AI/lightning-bolts/pull/118))
- Added an asynchronous single GPU dataloader. ([1521](https://github.com/Lightning-AI/lightning/pull/1521))

Fixed

- Fixed simclr finetuner ([165](https://github.com/Lightning-AI/lightning-bolts/pull/165))
- Fixed STL10 finetuner ([164](https://github.com/Lightning-AI/lightning-bolts/pull/164))
- Fixed Image GPT ([108](https://github.com/Lightning-AI/lightning-bolts/pull/108))
- Fixed unused MNIST transforms in tran/val/test ([109](https://github.com/Lightning-AI/lightning-bolts/pull/109))

Changed

- Enhanced train batch function ([107](https://github.com/Lightning-AI/lightning-bolts/pull/107))

Added

- Added setup and repo structure
- Added requirements
- Added docs
- Added Manifest
- Added coverage
- Added MNIST template
- Added VAE template
- Added GAN + AE + MNIST
- Added Linear Regression
- Added Moco2g
- Added simclr
- Added RL module
- Added Loggers
- Added Transforms
- Added Tiny Datasets
- Added regularization to linear + logistic models
- Added Linear and Logistic Regression tests
- Added Image GPT
- Added Recommenders module

Changed

- Device is no longer set in the DQN model init
- Moved RL loss function to the losses module
- Moved rl.common.experience to datamodules
- train_batch function to VPG model to generate batch of data at each step (POC)
- Experience source no longer gets initialized with a device, instead the device is passed at each step()
- Refactored ExperienceSource classes to be handle multiple environments.

Removed

- Removed N-Step DQN as the latest version of the DQN supports N-Step by setting the `n_step` arg to n
- Deprecated common.experience

Fixed

- Documentation
- Doct tests
- CI pipeline
- Imports and pkg
- CPC fixes