Build System

Your inference servers are deployed by pushing to your main branch on GitHub.

Banana has a GitHub Integration that watches for pushes and automatically builds and deploys that code.

Docker is used to build your code, using the Dockerfile in your project's root directory.

For more information about how docker works you may visit their documentation

An example Banana Dockerfile

In the first line we set the required base image pytorch/pytorch:1.11.0-cuda11.3-cudnn8-runtime. This is required for the build to complete, or else a "unsupported base image" error will be thrown.

FROM pytorch/pytorch:1.11.0-cuda11.3-cudnn8-runtime
WORKDIR /

# install git
RUN apt-get update && apt-get install -y git

# Install python packages
RUN pip3 install --upgrade pip
ADD requirements.txt requirements.txt
RUN pip3 install -r requirements.txt

# Add your model weight files 
# (in this case we have a python script)
ADD download.py .
RUN python3 download.py

ADD . .

EXPOSE 8000

CMD python3 -u app.py