Make Predictions with a Pre-trained Deep Neural Network

This uses environment variables to build, train and make predictions using the AntiNex client. If the deep neural network already exists it will use it to make new predictions. If it does not exist it will train a new one.

This python script is available in the pip: ai_env_predict.py

It takes parameters:

parser = argparse.ArgumentParser(
    description=("Python client to make Predictions "
                 "using a Pre-trained Deep Neural Network "
                 "with AntiNex Django Rest Framework"))
parser.add_argument(
    "-f",
    help=("file to use default ./examples/"
          "predict-rows-scaler-full-django.json"),
    required=False,
    dest="datafile")
parser.add_argument(
    "-m",
    help="send mock data",
    required=False,
    dest="use_fake_rows",
    action="store_true")
parser.add_argument(
    "-s",
    help="silent",
    required=False,
    dest="silent",
    action="store_true")
parser.add_argument(
    "-d",
    help="debug",
    required=False,
    dest="debug",
    action="store_true")

Source Code

antinex_client.scripts.ai_env_predict.start_predictions()

Using environment variables, create an AntiNex AI Client. You can also use command line args if you want.

This can train a new deep neural network if it does not exist or it can use an existing pre-trained deep neural network within the AntiNex Core to make new predictions.