Author: woden

We look at dockerizing our java build both in the cloud and in our local docker environment.

We look at exporting our model into saved_model and use that in java in order to run inference on images the model never is seen before.

We look closely at how to run the evaluation in windows and evaluate the result in Tensorboard.

We look at configuring and running training in the cloud.

We look into training your model locally on windows and other platforms.

Looking into the configuration file and talk about what the different parameters mean

We build TFRecord file using java and talking about how to easily label your images for object detection.

We look at an example from ml5js to load images and predict them

We look at a regular expression to parse a URL in detail so we can reason about regex.

We use Toco to convert our cats and dogs model to Tensorflow lite and run it on an Android device.