Prompt Buddy logoPrompt Buddy

openai · OpenAI Platform Docs

Fine-tuning best practices

A guide on improving fine-tuned model performance through iterative adjustments to data quality, data quantity, hyperparameter tuning, and dataset structure.

Import to Prompt Buddy

Derived skill

Files assembled from official documentation

Viewing SKILL.md

Fine-tuning best practices

A guide on improving fine-tuned model performance through iterative adjustments to data quality, data quantity, hyperparameter tuning, and dataset structure.

When To Use

Use when you need to diagnose poor fine-tuning results or optimize a model's accuracy, style, and consistency through data and hyperparameter iterations.

Reference Files

FileContainsUse For
SKILL.mdEntry point: scope, routing table, and workflow.Start here.
docs/fine-tuning-best-practices-workflow-guide.mdA guide detailing strategies for iterating on data quality, quantity, and hyperparameters to improve fine-tuned model performance.Questions about a guide detailing strategies for iterating on data quality, quantity, and hyperparameters to improve fine-tuned model...
examples/fine-tuning-best-practices-openai-fine-tuning-jobs-create.javascriptA JavaScript code example demonstrating how to create a fine-tuning job using the OpenAI Node.js library.Exact payloads, commands, or snippets shown in A JavaScript code example demonstrating how to create a fine-tuning job using the OpenAI Node.js library.
examples/fine-tuning-best-practices-openai-finetuning-job-creation.pythonA Python script demonstrating how to create a supervised fine-tuning job using the OpenAI client library.Exact payloads, commands, or snippets shown in A Python script demonstrating how to create a supervised fine-tuning job using the OpenAI client library.
examples/fine-tuning-best-practices-openai-fine-tuning-best-practices-training-da.jsonlA JSONL formatted dataset containing system, user, and assistant message pairs used to demonstrate fine-tuning training data structure.Exact payloads, commands, or snippets shown in A JSONL formatted dataset containing system, user, and assistant message pairs used to demonstrate fine-tuning traini...

What This Skill Covers

  • If you're not getting strong results with a fine-tuned model, consider the following iterations on your process.
  • Main sections: Iterating on data quality, Iterating on data quantity, Iterating on hyperparameters, Adjust your dataset, Training vs. testing datasets.

Workflow

  1. Open the most relevant file under docs/ for the exact documented workflow and wording.
  2. Open schemas/ files for exact structured contracts.
  3. Open examples/ files for concrete requests, commands, snippets, and manifests.
  4. Do not add behavior or configuration that is not present in the attached source files.

Canonical source: https://developers.openai.com/api/docs/guides/fine-tuning-best-practices.md