baselines

baselines copied to clipboard

Understanding normalization of advantage function

7

I am wondering about the normalization of the advantage function in PPO. Before training on a batch the mean of the advantage function is subtracted and it's divided by its...

mbcel

SubProcVecEnv raises ConnectionResetError: [Errno 104] Connection reset by peer

11

I'm trying to run the following code and test PPO with Sonic the hedghehog, running it in parralel with SubProcVecEnv Unfortunately I run in the following error: ``` Traceback (most...

MakisKans

DDPG implementation fails to learn well on at least five MuJoCo-v2 envs for all three noise types. I report steps to reproduce and learning curve plots [and show that PPO2 seems to work fine].

21

Dear @pzhokhov @matthiasplappert @christopherhesse et al., Thank you for providing an implementation of DDPG. However, I have been unable to get it to learn well on the standard MuJoCo environments...

DanielTakeshi

guotong1988

A2C Colab Atari - ConnectionResetError: [Errno 104] Connection reset by peer

1

When I try to run the colab example with the A2C algorithm on an Atari env, I get the following error: `--------------------------------------------------------------------------- ConnectionResetError Traceback (most recent call last) in ()...

christophschuhmann

fix links of benchmarks in README.md

gxywy

Tensorboard logging does not work in TF2 branch

4

I got an error, when trying to log tensorboard output in the TF2 branch: > set OPENAI_LOG_FORMAT=stdout,log,csv,tensorboard > python -m baselines.run --alg=ppo2 --env=CartPole-v0 --network=mlp --save_path model --log_path log/ --num_timesteps=30000 --nsteps=128...

lerad

Hello. Can you please explain why are you using the `mb_returns = mb_advs(GAE) + mb_values` as the returns to compute the critic loss ? Should not the value function approximately...

rohey

guotong1988

guotong1988