Add IRIS

### Model description IRIS (Imagination with auto-Regression over an Inner Speech) is a Reinforcement learning agent trained in the imagination of a world model composed of a discrete autoencoder and an autoregressive Transformer. IRIS learns behaviors by accurately simulating millions of trajectories. The approach presented in the paper casts dynamics learning as a sequence modeling problem, where an autoencoder builds a language of image tokens and a Transformer composes that language over time. The agent is introduced in the paper titled [TRANSFORMERS ARE SAMPLE-EFFICIENT WORLD MODELS](https://arxiv.org/abs/2209.00588). There is also a [medium blog post](https://medium.com/@cedric.vandelaer/paper-review-transformers-are-sample-efficient-world-models-d0f9144f9c09) to understand how the algorithm works. ### Open source status - [X] The model implementation is available - [X] The model weights are available ### Provide useful links for the implementation The model is adapted from the [official code repository](https://github.com/eloialonso/iris) of the paper. The officially released weights can be found on this [Github repository](https://github.com/eloialonso/iris_pretrained_models)