DreamerDecoder¶
- torchrl.envs.model_based.dreamer.DreamerDecoder(in_keys: Sequence[NestedKey] | None = None, out_keys: Sequence[NestedKey] | None = None, in_keys_inv: Sequence[NestedKey] | None = None, out_keys_inv: Sequence[NestedKey] | None = None)[source]¶
A transform to record the decoded observations in Dreamer.
Examples
>>> model_based_env = DreamerEnv(...) >>> model_based_env_eval = model_based_env.append_transform(DreamerDecoder())