Skip to content

API Endpoints Reference

API documentation for the FastAPI inference server.

API Module

mlops_project.api

PredictionResponse

Bases: BaseModel

Response model for predictions.

load_config()

Load Hydra config without changing working directory.

pull_models_dvc()

Pull models from DVC remote if models directory is empty or missing.

find_latest_model(model_name='EfficientNet')

Find the latest ONNX model file for a given model name.

Parameters:

Name Type Description Default
model_name str

Name of the model (e.g., "EfficientNet", "ResNet")

 'EfficientNet'

Returns:

Type Description
Path | None

Path to the latest model file, or None if not found

lifespan(app) async

Load model on startup and cleanup on shutdown.

read_root()

Root endpoint.

health_check()

Health check endpoint.

perform_inference(file=File(...)) async

Perform inference on an uploaded image.

Parameters:

Name Type Description Default
file UploadFile

Image file (JPG, PNG, etc.)

 File(...)

Returns:

Type Description

Dictionary with predicted class, diagnosis name, and probabilities

Raises:

Type Description
HTTPException

If model is not loaded or inference fails

Response Models

PredictionResponse

mlops_project.api.PredictionResponse

Bases: BaseModel

Response model for predictions.

Endpoints

GET /

Root endpoint returning welcome message.

GET /health

Health check endpoint.

Response: 

{"status": "healthy"}

POST /predict

Classify a skin lesion image.

Request: multipart/form-data with image file

Example: 

curl -X POST "http://localhost:8000/predict" -F "file=@image.jpg"

Response: 

{
  "prediction": "malignant",
  "confidence": 0.847,
  "probabilities": {"benign": 0.153, "malignant": 0.847}
}