garage
garage copied to clipboard
A toolkit for reproducible reinforcement learning research.
Address #2190 I simply add a check in NPO init to see if max_episode_length is None and if it is, I raise a ValueError. Discussed with @ryanjulian
Hello everybody, when I used Garage's `EpsilonGreedyStrategy` and Gym environments I found that sampling is not deterministic. I've set the seed via `deterministic.set_seed(seed)`. After some investigation I found that Garage...
I am creating a child class that inherits from TRPO, but upon initializing the optimizer, I get `TypeError: Failed to convert object of type to Tensor` Here is the overall...
Tensorflow prints out a bunch of warning messages, and it's an annoyance if it's not actually needed. We should be able to avoid importing it if garage.tf is never imported.
Formula for off_policy_method: total_timeseteps= n_epochs * n_epoch_cycles * batch_size then if n_epochs=1400 n_epoch_cycles=20 batch_size=64 min_buffer_size=10^6 then total_timesteps=1400 * 20 * 64=1,792,000 I obtained the graph as shown in figure for...
Hey, so people (including myself) are starting to use garage more for IL (and related topics). I have a short list of "must have" features which I think would make...