Choosing a Fine-Tuning Method in Generative AI

You can fine-tune the meta.llama-3-70b-instruct, cohere.command, and cohere.command-light base models with your own dataset.

Important

For the meta.llama-3-70b-instruct models, OCI Generative AI fine-tunes your data with the Lora method.
For the cohere.command and cohere.command-light models, OCI Generative AI has two training methods: T-Few, and Vanilla.

Use the following guidelines to help you choose the best training method if you select the cohere.command and cohere.command-light base models.


Feature	Options and Recommendations
Training methods for `cohere.command` and `cohere.command-light`	`T-Few` `Vanilla`
Dataset Size	Use `T-Few` for small datasets (A few thousand samples or less) Use `Vanilla` for large datasets (From a hundred thousand samples to millions of samples) Using small datasets for the `Vanilla` method might cause overfitting. Overfitting happens when the trained model gives great results for the training data, but can't generalize outputs for unseen data.
Complexity	Use `T-Few` for format following or instruction following. Use `Vanilla` for complicated semantical understanding improvement, such as improving a model's understanding of medical cases.
Hosting	Use `T-Few` if you're planning to host several fine-tuned models on the same hosting dedicated AI cluster. If all the models are trained on the same base model, you can host them on the same cluster. This stacked-serving feature saves cost and offers good performance if user traffic to each `T-Few` fine-tuned model is relatively low. See Hosting Clusters with Many Endpoints. Each model that's fine-tuned with the `Vanilla` method requires its own hosting dedicated AI cluster.