Xiaojian Ma
Xiaojian Ma
PR created https://github.com/floodsung/Deep-Reasoning-Papers/pull/5. PTAL, thanks!
Yes, I'm partly agree with you, but with a small correction, the algorithm implemented should be an offline version A2C(Advantage Actor Critic).
Seconding here. The current [VD](https://github.com/researchmm/soho/blob/929495eafa7b71569accbe4e22acfbfd93a0846f/SOHO/models/necks/utils.py#L7) seems to be an identity mapping. How should I repro your result with the complete model? Thanks for the help.
The split file I used can be downloaded [here](https://xiaozhichen.github.io/files/mv3d/imagesets.tar.gz).
Thanks for this patch! @Phylliida I think you may also want to modify [this file](https://github.com/minerllabs/minerl/blob/dev/minerl/herobraine/hero/spaces.py) to allow sampling from the new action space with chat command? [ref](https://github.com/minerllabs/minerl/pull/613/files#diff-2df8a47049915765c18557841bb05853c9f6f08057531ef693c0ccbb6b4e3965)
Hi all, Sorry for bringing this up here, but we made it to run MineRL with GPU rendering on a headless, rootless GPU server without host-side X configuration. Please check...
@Miffyli PR has been made [here](https://github.com/minerllabs/minerl/pull/655).
https://github.com/microsoft/malmo/issues/844 might be relatable. But I do think `minerl` has already removed the code that configures `ObservationPolicy`.
+1 would like to see this.
Just in case someone needs this: here is a solution for those who don't want to use docker and happen to use a server that does not install `EGL*` along...