pipeline.yml

$schema: https://azuremlschemas.azureedge.net/latest/pipelineJob.schema.json
type: pipeline
display_name: image_classification_with_densenet
description: Train densenet for image classification


# <jobs>
settings:
  default_datastore: azureml:workspaceblobstore
  default_compute: azureml:cpu-cluster
  continue_on_step_failure: false

jobs:
  convert_training_image:
    type: command
    component: ./convert_to_image_directory/entry.spec.yaml
    inputs:
      input_path: 
        type: uri_folder
        path: ./data/train

  convert_evaluation_image:
    type: command
    component: ./convert_to_image_directory/entry.spec.yaml
    inputs:
      input_path: 
        type: uri_folder
        path: ./data/val        

  init_transformation:
    type: command
    component: ./init_image_transformation/entry.spec.yaml
    inputs:
      resize: "False"
      size: 256
      center_crop: 224
      pad: "False"
      padding: 0
      color_jitter: "False"
      grayscale: "False"
      random_resized_crop: "False"
      random_resized_crop_size: 256
      random_crop: "False"
      random_crop_size: 224
      random_horizontal_flip: "True"
      random_vertical_flip: "True"
      random_rotation: "False"
      random_rotation_degrees: 0
      random_affine: "False"
      random_affine_degrees: 0
      random_grayscale: "False"
      random_perspective: "False"
  
  transform_on_training_image:
    type: command
    component: ./apply_image_transformation/entry.spec.yaml
    inputs: 
      mode: "For training"
      input_image_transform_path: ${{parent.jobs.init_transformation.outputs.output_path}}
      input_image_dir_path: ${{parent.jobs.convert_training_image.outputs.output_path}}

  transform_on_evaluation_image:
    type: command
    component: ./apply_image_transformation/entry.spec.yaml
    inputs:
      mode: "For inference"
      input_image_transform_path: ${{parent.jobs.init_transformation.outputs.output_path}}
      input_image_dir_path: ${{parent.jobs.convert_evaluation_image.outputs.output_path}}

  train:
    type: command
    component: ./image_cnn_train/entry.spec.yaml
    compute: azureml:gpu-cluster
    inputs:
      train_data: ${{parent.jobs.transform_on_training_image.outputs.output_path}}
      valid_data: ${{parent.jobs.transform_on_evaluation_image.outputs.output_path}}
      data_backend: "pytorch"
      epochs: 4
      seed: 123
      batch_size: 16
      save_checkpoint_epochs: 2
    outputs:
      workspace:
        type: uri_folder
        mode: upload
    distribution:
      type: mpi
      process_count_per_instance: 1
    resources:
      instance_count: 2
# </jobs>