Oracle Cloud Infrastructure Documentation

Creating a Customization

Create a customization for live transcription jobs in the Speech service.

  • On the Customizations list page, select Create customization. If you need help finding the list page, see Listing Customizations.
    The Create customization window opens.

    1. Customization information

    • Name (Optional)
    • Alias (Optional)
    • Description (Optional)
    • Compartment: Select the compartment to create the customization in.
    • Tagging options (Optional): See Resource Tags.

    Select Next.

    2. Training dataset

    • Training dataset source: Under Dataset type, select the relevant option and provide information.
      • Entity list: Directly input a list of entities (words or phrases) for training.
        • Entity type (Optional): Enter the entity type.
        • Entity list: Enter the words or phrases to use in the training.
      • Object Storage: Select existing text or JSON files in an Object Storage bucket as the input.
        • Entity type (Optional): Enter the entity type.
        • Dataset compartment: Select the compartment that contains the bucket.
        • Object storage bucket: Select the Object Storage bucket that contains the file to use.
        • Training dataset file: Select the dataset file.
    • Model details: Select the domain and language for the customization.

    Select Next.

    3. Review and create

    Review the information to confirm that it's correct. If you need to revise a value, expand that step and change the values as needed.

    Select Create. The creation process begins and its progress is displayed. You can cancel the creation at any time.

  • Use the oci speech customization create command and required parameters to create a customization job.

    oci speech customization create [OPTIONS]

    For a complete list of parameters and values for CLI commands, see the CLI Command Reference.

  • Use the CreateCustomization command and operations to create a customization job.