Skip to main content

Upload Model Workflow

The Upload Model Workflow on the Gesund.ai Platform guides users through a structured and secure process of uploading their machine learning models. It supports both direct file uploads and external model retrieval from AWS S3, ensuring smooth integration into the platform’s model repository.

Upload Model Workflow

1. Start Model Upload​

  • Click Create New Model or choose an existing draft/completed model to begin the upload process.
  • Upload your model file in .zip format.

2. Fill in Model Metadata​

  • After the upload, a form appears prompting key information:
    • Model Name
    • Version
    • Model Type
    • Modality
    • Framework
    • Description
  • These fields help categorize and retrieve the model later.

3. Unique Model ID Generation​

  • Once the metadata is submitted, the platform generates a unique model ID.
  • This ID is used for versioning, deployment, and prediction tracking.

4. AWS S3 Model Import (Optional)​

  • If the model resides in an AWS S3 bucket:
    • Enter the S3 path and credentials (if needed).
    • The platform automatically fetches the model file from the provided location.
    • Integrity and accessibility checks are performed during the download.

5. Internal Storage and Validation​

  • The uploaded or fetched model is stored securely in Gesund.ai’s infrastructure.
  • The platform validates the archive and its contents to ensure compatibility with supported workflows.

6. Confirmation and Completion​

  • After successful storage and validation:
    • A success message is displayed.
    • Users receive the model ID and status information.
  • The model is now accessible from the Models page for execution, validation, or further editing.

This step-by-step upload flow ensures models are registered consistently and securely, enabling a reliable foundation for downstream tasks like training, prediction, and validation.