atari-representation-learning
atari-representation-learning copied to clipboard
Tensorify episodes and labels to save memory and simplify things
Episodes
Before:
- list of lists of pytorch tensors (one tensor for each example)
After:
- list of pytorch tensors (one tensor for each episode)
Labels
Before:
- list of list of dicts (each dict is one example/state)
After:
- list of dicts (one dict for each episode)
- each dict has keys being state variables and values being 1-D tensors of each state variable realization for the episode