PyTorch implementation of Advantage Actor Critic (A2C), Proximal Policy Optimization (PPO), Scalable trust-region method for deep reinforcement learning using Kronecker-factored approximation (ACKTR) and Generative Adversarial Imitation Learning (GAIL).
MIT License
3.56k
stars
831
forks
source link
EOFError encountered when training on CartPole-v0 #202
Python 3.6.8
Pytorch 1.1.0
OpenAI Gym 0.14.0
CUDA version 10.0.130
cuDNN version 7501
The code was run on a Ubuntu docker container on a GPU cluster with 24 CPU, the base docker image I used is ufoym/deepo from DockerHub.
I can run the code on PongNoFrameskip-v4 without problems:
python main.py --env-name "PongNoFrameskip-v4" --algo ppo
However when I run the code on CartPole-v0:
python main.py --env-name "CartPole-v0" --algo ppo
I immediately encountered EOFError:
The same EOFError was seen when I run other algorithms (like a2c) on CartPole-v0:
python main.py --env-name "CartPole-v0" --algo a2c
My environment is stated as follows:
Python 3.6.8 Pytorch 1.1.0 OpenAI Gym 0.14.0 CUDA version 10.0.130 cuDNN version 7501 The code was run on a Ubuntu docker container on a GPU cluster with 24 CPU, the base docker image I used is ufoym/deepo from DockerHub.