Downloading TensorFlow Serving Docker images
Once the Docker engine is up and running, you are ready to perform the following steps:
- You may pull the latest TFS Docker image with this Docker command:
docker pull tensorflow/serving
- This is now our base image. In order to add our model on top of this image, we need to run this base image first:
docker run -d --name serv_base_img tensorflow/serving
In the preceding command, we invoked the tensorflow/serving image and now it is running as a Docker container. We also name this container serv_base_img.
Creating a new image with the model and serving it
Let's now take a look at the file directory here. For this example, the directory structure is as shown in the following figure:
Figure 9.2 – Directory structure for creating a custom Docker container
We will execute the following commands from the same directory as Tensorflow_Serving.ipynb.
After we have the TFS base Docker...