- Install TensorFLow: follow official site instructions: https://www.tensorflow.org/install/
- Open Visual Studio Code with curent folder and execute in terminal:
python ./mnist_deep_training.py --training_iteration=1000 --model_version=1 "Model Training Folder Path"This is start python script and train MNIST deep model with 1000 iterations. After that it will save model in directory. Please keep in mind to increament model_version parameter value after every training.
-
Install TensorFlow Serving: follow official site instructions: https://www.tensorflow.org/serving/
-
Start TensorFlow Serving with the following command:
tensorflow_model_server --port=9000 --model_name=mnist --model_base_path="odel Training Folder Path"