Jiayi Weng

Maybe we can add another way of ReplayBuffer save/restore. I remember the compression algorithm from numpy itself is much more efficient than pickle/hdf5 (according to my experiments at that time).

Not sure what happens, could you please send me the code?

Is it possible to use an empty dataset to reproduce this result? (unrelated to rl-unplugged, because I need quite a long time to download one file...)

I think the reason behind is that, when we developed the RBM, we assume the buffers in input buffer list are all uninitialized: ```python [ReplayBuffer(), ReplayBuffer(), ReplayBuffer(), ReplayBuffer(), ReplayBuffer(), ReplayBuffer(),...

Hmm the current code does what your said. However, the best way is to deal with reset/step together. If the underlying env has auto-reset env wrapper, i.e., `step(*) == reset()`...

Hello my old friend! Nice to hear from that great news! I change this to support EnvPool async mode little by little. in EnvPool, both the `step` and `reset` are...

Sorry for the delay. Could you please kindly answer the following questions because I'm not quite familiar with these two algorithms and need some context? 1. What's the difference between...

> A solution to optimize the sample method in my mind is to figure out or keep available indices of _meta in an array and sample with a single np.random.choice...

There're known issues pointing out at #486. Unfortunately, I have no time to fix it until two months later (after graduation)...Feel free to submit PRs and I'm really sorry about...

Thanks! I'll take a look when I finish database assignment ......