train_blueprint.yaml

title: RUL Train
version: 1.0.0
description: Tailor-train a CNN model with a customized dataset to predict when a company asset is likely to fail within given cycles
long_description: Tailor-train a convolutional neural network (CNN) model with a customized dataset to predict when a company asset is likely to fail within given cycles.

author: "cnvrg"
author_email: "libhub@cnvrg.io"

tags:
  - training

tasks:
  - title: s3_connector
    top: 300
    left: 0

    type: exec

    library: s3-connector15
    library_version: 1.0.0

    command: python s3_connector.py

    requirements:
      cpu: 3.5
      memory: 8
      gpu: 0
      hpu: 0

    image: cnvrg/cnvrg:v5.0
    language: python3

    params:
      - key: endpoint
        type: 'categorical'
        values:
          - 'http://s3.amazonaws.com download'
      - key: bucketname
        type: 'categorical'
        values:
          - 'libhub-readme'
      - key: localdir
        type: 'categorical'
        values:
          - '/cnvrg'
      - key: prefix
        type: 'categorical'
        values:
          - 'remaining_useful_life_data/'
          
  - title: data_preprocessing
    top: 300
    left: 250

    type: exec

    library: rul-data-preprocessing
    library_version: 1.0.0

    command: python3 data_preprocessing.py

    requirements:
      cpu: 3.5
      memory: 8
      gpu: 0
      hpu: 0

    image: cnvrg/cnvrg:v5.0
    language: python3

    params:
      - key: raw_train_data
        type: 'categorical'
        values:
          - '/input/s3_connector/remaining_useful_life_data/raw_train_data.csv'
      - key: common_letter_numeric
        type: 'categorical'
        values:
          - 's'
      - key: numeric_features
        type: 'categorical'
        values:
          - 'garbage999'
      - key: meta_columns
        type: 'categorical'
        values:
          - 'id,cycle'
      - key: sequence_length
        type: 'categorical'
        values:
          - '50'
      - key: upper_limit
        type: 'categorical'
        values:
          - '45'  
      - key: lower_limit
        type: 'categorical'
        values:
          - '15'            
          
  - title: cnn
    top: 300
    left: 500

    type: exec

    library: rul-cnn
    library_version: 1.0.0

    command: python3 cnn.py

    requirements:
      cpu: 3.5
      memory: 15.5
      gpu: 0
      hpu: 0

    image: cnvrg/cnvrg:v5.0
    language: python3

    params:      
      - key: x_train
        type: 'categorical'
        values:
          - '/input/data_preprocessing/x_train'
      - key: y_train
        type: 'categorical'
        values:
          - '/input/data_preprocessing/y_train'     
      - key: batch_size
        type: 'categorical'
        values:
          - '512'
      - key: epochs
        type: 'categorical'
        values:
          - '25'     
      - key: seed
        type: 'categorical'
        values:
          - '58'           
      - key: shape_data
        type: 'categorical'
        values:
          - '/input/data_preprocessing/shape_data.csv'          
          
  - title: batch_predict
    top: 100
    left: 700

    type: exec

    library: rul-batch-predict
    library_version: 1.0.0

    command: python3 predict.py

    requirements:
      cpu: 3.5
      memory: 15.5
      gpu: 0
      hpu: 0

    image: cnvrg/cnvrg:v5.0
    language: python3

    params:
      - key: x_test
        type: 'categorical'
        values:
          - '/input/s3_connector/remaining_useful_life_data/raw_test_data.csv'     
      - key: cnn
        type: 'categorical'
        values:
          - '/input/cnn/cnn_model.h5'    
      - key: shape_data
        type: 'categorical'
        values:
          - '/input/data_preprocessing/shape_data.csv'         
  - title: inference
    top: 500
    left: 700

    # Type must be according to the flow task conventions (data, exec, deploy)
    type: deploy

    # The library data
    library: rul-inference
    library_version: 1.0.0

    command: python3 predict.py
    kind: webservice

    requirements:
      cpu: 3.5
      memory: 8
      gpu: 0
      hpu: 0

    image: cnvrg/cnvrg:v5.0
    language: python3
    accept_files: True

    file_name: predict.py # the entrypoint file name
    function_name: predict # the entrypoint function
    prep_file: '' # preprocess file name
    prep_function: '' # preprocess function
    input_example:
      img: file
    input_schema:
      img: file
    output_schema: {}
         
relations:
  - from: s3_connector
    to: data_preprocessing
  - from: s3_connector
    to: batch_predict    
  - from: data_preprocessing
    to: cnn
  - from: data_preprocessing
    to: batch_predict
  - from: cnn
    to: batch_predict
  - from: data_preprocessing
    to: inference
  - from: cnn
    to: inference