Roger Girgis

Hi, I believe the reward loss should be based on `rewards[1:]` instead of `rewards[:-1] `: https://github.com/yusukeurakami/dreamer-pytorch/blob/7e9050e8c454309de40bd0d1a4ec0256ef600147/main.py#L209 If not, can you please explain your reasoning? Thanks,

Hello, Thanks for releasing your work! I was wondering if the full preprocessed dataset will be released soon. The [G-Drive](https://drive.google.com/drive/folders/1xq7OJG5k796_SFsb6j6g7O2e5kc85z51) seems to be empty (other than the license). Thanks!