Fine-Tuning OpenAI with Postman: Best Practices

Timestamped Summary

• 🤖
  00:00

  Use Postman to send a post request to OpenAI with an authorization bearer, purpose, and file to upload.
• 📝
  03:50

  Create a jsonl file with prompt-completion pairs to fine-tune OpenAI.
• 🤔
  05:59

  David Shapiro created a Creative Writing Coach with prompts from Reddit, but more data is needed for better output.
• 🤔
  08:30

  Put a story colon in first and validate data with JSON format: replace line breaks with /n and quotation marks with hal.
• 📤
  10:21

  Send a post request with a header, content type, authorization, API key, and file ID to the given URL.
• 🤖
  12:53

  Create a workspace in Postman to fine tune settings and make a post request with the file id.
• 🤖
  15:20

  Use OpenAI to train a model and check its status with a GET request to the URL with the fine tune ID and authorization bearer token.
• 🤖
  18:48

  Fine-tune with Postman to create powerful datasets and use them in Riku.ai for AI-generated content.