Part 2/8:
Using a specific script, Shapiro uploads their JSONL file—a format where each line contains a prompt and a corresponding completion—to initiate the fine-tuning process. Once the training job completes, a new model is created, identifiable by the user’s name, the custom tag, and the date of fine-tuning.
Exploring the Fine-Tuned Model
The newly trained model appears in the OpenAI playground, typically listed under your account details. The interface displays the model's name, tag, and timestamp, providing an organized overview.
How to Use Your Fine-Tuned Model
- In the playground: Simply select your model from the list to generate content.