From the course: Fine-Tune Your LLMs
Join today to access over 24,900 courses taught by industry experts.
Iterate a fine-tuned model
From the course: Fine-Tune Your LLMs
Iterate a fine-tuned model
- [Instructor] Let's iterate on our fine-tuned model to improve its performance. There are several options for iterating. Iterate on your fine-tuned model. This means you can fine-tune a fine-tuned model. No need to start from scratch. Iterate on your data quality. This means collecting examples that target specific issues you're seeing. Also consider the balance and diversity of your dataset. Iterate on your data quantity. This means adding additional examples for the model to learn from. Or iterate by adjusting your hyperparameters. For our example, we'll iterate on the hyperparameters by reducing the number of epochs to two instead of five. This means we'll go through the process of creating a fine-tuned model, retrieving the job status, and evaluating the results all over again. And let's do that now. Here in cell 10, I've changed the number of epochs from five to two, and I've already executed the code in this notebook for efficiency. When we kick off the fine tuning job, we'll…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.