search

FangyunWei / SLRT

259 stars 56 forks source link

How to install requirements with the docker image? #38

Closed JamboCop closed 11 months ago

JamboCop commented 12 months ago

Hello, I am attempting to run this repository, however I'm unfamiliar with Docker and how to use it. I have Docker installed, and managed to download the image, however I'm not sure what I'm supposed to do now. Does simply running a container with the image install the requirements or do I have to clone the repo into the docker container?

Any help is appreciated, thank you!

fransisca25 commented 12 months ago

Hello, I am attempting to run this repository, however I'm unfamiliar with Docker and how to use it. I have Docker installed, and managed to download the image, however I'm not sure what I'm supposed to do now. Does simply running a container with the image install the requirements or do I have to clone the repo into the docker container?

Any help is appreciated, thank you!

After pulling the Docker image, you need to run it first. During the first run of the image, you may need to adjust the shared memory size. However, the Docker image does not include the repository, so you have to clone the repository into the Docker image yourself. The CUDA version inside the Docker image is 11.3, so you'll need to make some adjustments among PyTorch, CUDA, mmdet, mmpose, and mmcv-full versions as well. As for mmdet, mmpose and mmcv-full version compatibility, you may need to refer to the documentation. I hope this information is helpful.