Running out of GPU memory about batch-ppo HOT 5 CLOSED

PR #18 recently changed training on GPU. It is possible that it puts too many ops on the GPU. What tensors fails? What GPU do you use? Can you please set use_gpu = False in the config and verify that CPU-only training works?

from batch-ppo.

@wwxFromTju Could you provide an update, please?

from batch-ppo.

@danijar what update?

from batch-ppo.

I asked a few questions in my first comment that could help debug your problem. Did you have a chance to try those?

from batch-ppo.

I disabled GPU usage by default and implemented splitting episodes (using the chunk_length and batch_size config options) to be able to train on episodes with many time steps or large observations.

from batch-ppo.