Resource File - Start Fine Tuning Job

Creates a fine-tuning job using the set of fine-tuning files saved on this data element.

Remarks

The Start Fine Tuning Job command submits a new fine-tuning job to your OpenAI provider using one or more fine-tuning files associated with the selected Resource File nodes. Fine-tuning files are created using the Build Fine Tuning File command and are typically stored as .jsonl files. This command allows you to train a custom model based on your data, enabling more accurate and domain-specific completions.

When to Use

• After preparing and saving fine-tuning files with the Build Fine Tuning File command.
• When you want to train a new model for your specific use case, such as improving accuracy for specialized terminology or document types.
• To automate the process of submitting multiple fine-tuning files as a single job.

How It Works

• Select one or more Resource File nodes containing fine-tuning data (.jsonl format).
• Configure the 'Base Model' property to specify which OpenAI model to use as the starting point for fine-tuning.
• Set the 'Name Suffix' property to define a custom suffix for the resulting model's name.
• The command uploads the selected files, submits a fine-tuning job, and provides a job ID for tracking progress.
• You can monitor the job's progress and completion status in the OpenAI portal (Dashboard >> Fine Tuning).
• When the job completes, a new fine-tuned model will be available for use, named after the base model with your specified suffix.

> Note: Only files with a .jsonl extension are eligible for fine-tuning. Ensure your data is properly formatted and validated before submitting.

Properties

Name	Type		Description
Base Model	String	►	The base model to use for the fine-tuning job. EditorModel Editor Requiredtrue Specifies the OpenAI model that will serve as the starting point for your fine-tuning job. The list of available base models is populated from your connected OpenAI provider and includes only those models that support fine-tuning. Usage Guidance Use the dropdown selector to choose a base model appropriate for your use case (e.g., gpt-3.5-turbo). The selected model determines the capabilities and cost of the resulting fine-tuned model. Ensure the base model is compatible with your data and intended application. > Tip: Consult OpenAI documentation for guidance on selecting a base model for your domain or task.
Name Suffix	String	►	A suffix to be appended to the output model name. EditorModel Editor Requiredtrue Specifies a custom string that will be appended to the name of the resulting fine-tuned model. This helps you identify and organize your models, especially when managing multiple fine-tuning jobs. Usage Guidance Enter a descriptive suffix to distinguish this model from others (e.g., legal-2024, v2). The suffix is combined with the base model name and a timestamp to form the final model name. Avoid using special characters or spaces; stick to letters, numbers, hyphens, or underscores. > Example: > If the base model is gpt-3.5-turbo and the suffix is custom-2024, the resulting model might be > named ft:gpt-3.5-turbo:custom-2024:2024-06-07-15-30-00.

See Also

Resource File