MLproject

name: risco_credito_mlflow_project

python_env: python_env.yaml

entry_points:
 prepare:
   parameters:
     n_samples: {type: int, default: 6000}
     test_size: {type: float, default: 0.25}
     random_state: {type: int, default: 42}
     output_dir: {type: str, default: "data/processed"}
     pipeline_id: {type: str, default: "manual"}
   command: >
     python src/prepare_data.py
     --n-samples {n_samples}
     --test-size {test_size}
     --random-state {random_state}
     --output-dir {output_dir}
     --pipeline-id {pipeline_id}

 train:
   parameters:
     model_name: {type: str, default: "random_forest"}
     data_dir: {type: path, default: "data/processed/manual"}
     random_state: {type: int, default: 42}
     registered_model_name: {type: str, default: "risco_credito_project_model"}
     pipeline_id: {type: str, default: "manual"}
   command: >
     python src/train.py
     --model-name {model_name}
     --data-dir {data_dir}
     --random-state {random_state}
     --registered-model-name {registered_model_name}
     --pipeline-id {pipeline_id}

 evaluate:
   parameters:
     model_uri: {type: str}
     data_dir: {type: path, default: "data/processed/manual"}
     threshold: {type: float, default: 0.5}
     pipeline_id: {type: str, default: "manual"}
   command: >
     python src/evaluate.py
     --model-uri {model_uri}
     --data-dir {data_dir}
     --threshold {threshold}
     --pipeline-id {pipeline_id}

 pipeline:
   parameters:
     model_name: {type: str, default: "random_forest"}
     n_samples: {type: int, default: 6000}
     test_size: {type: float, default: 0.25}
     random_state: {type: int, default: 42}
     threshold: {type: float, default: 0.5}
     registered_model_name: {type: str, default: "risco_credito_project_model"}
   command: >
     python src/run_pipeline.py
     --model-name {model_name}
     --n-samples {n_samples}
     --test-size {test_size}
     --random-state {random_state}
     --threshold {threshold}
     --registered-model-name {registered_model_name}