Ускорение графического процессора M3 Max Docker? ⇐ Python
-
Anonymous
Ускорение графического процессора M3 Max Docker?
I am using ML-Agents in an Ubuntu Docker container to overcome some of the installation issues the library is currently having with Mac operating systems. It works great and I'm able to connect to Unity on my base computer through the Docker container and train agents. However, I'm currently unable to use the GPU and can only train using the CPU. I saw that nvidia might have something to help with this, but I'm not sure how to set it up on MAC. I'd really appreciate some guidance on this!
Here is the Dockerfile I'm currently using. It's pretty bare bones, but seems to work well!
FROM ubuntu:latest # Install necessary dependencies RUN apt-get update && apt-get install -y \ python3 \ python3-pip \ git # Install Tensorflow RUN pip3 install tensorflow # Clone ml-agents repository RUN git clone --branch release_21 https://github.com/Unity-Technologies/ml-agents.git # Install ml-agents Python package RUN pip3 install mlagents # Command to run when the container starts CMD ["/bin/bash"] I give the container access to port 5004 when running with the following command:
docker run -p 5004:5004 --name marl_container marl_image This is what allows integration with Unity for training. You can change it, but this is the default port that Unity uses.
Источник: https://stackoverflow.com/questions/780 ... celeration
I am using ML-Agents in an Ubuntu Docker container to overcome some of the installation issues the library is currently having with Mac operating systems. It works great and I'm able to connect to Unity on my base computer through the Docker container and train agents. However, I'm currently unable to use the GPU and can only train using the CPU. I saw that nvidia might have something to help with this, but I'm not sure how to set it up on MAC. I'd really appreciate some guidance on this!
Here is the Dockerfile I'm currently using. It's pretty bare bones, but seems to work well!
FROM ubuntu:latest # Install necessary dependencies RUN apt-get update && apt-get install -y \ python3 \ python3-pip \ git # Install Tensorflow RUN pip3 install tensorflow # Clone ml-agents repository RUN git clone --branch release_21 https://github.com/Unity-Technologies/ml-agents.git # Install ml-agents Python package RUN pip3 install mlagents # Command to run when the container starts CMD ["/bin/bash"] I give the container access to port 5004 when running with the following command:
docker run -p 5004:5004 --name marl_container marl_image This is what allows integration with Unity for training. You can change it, but this is the default port that Unity uses.
Источник: https://stackoverflow.com/questions/780 ... celeration
