The project should include the front and back parts. Communication between front and back is implemented via API. The front of the container is available at port :3000 The front part offers the option of uploading a scoring file in the format.csv dataset from the competition. And the option to upload the file after scoring by the model in the sample_submission.csv format.
The service performs the following stages of the ML project inside the container with separate scripts:
- API
- Input file loading and data preprocessing
- Scoring of the processed dataset