MyxMatch

This guide will show you how to find the best base model for your application using MyxMatch in as little as 10 minutes. We'll also look at how it works and other tools you can explore.

Find the Best Model

You can find the MyxMatch under the "Explore" section of the home view. Once you've clicked into the tool, you can start filling out the name for your matching job. Include a representative sample or prompt in the context box to help source the best model. Optionally, you can click the "Model Selection" dropdown button to select which models you want to compare. By default, all of the available models are chosen.

After you've clicked "Rank," you'll be redirected to the Myxboards view, where you can monitor the progress of your matching job. Once it's finished, you'll see a table ranking all selected models, with the top-ranked ones at the top. If you're ready to move on to the next step and train a model, look for the "Train" button in the last column and click it.

How Does It Work?

MyxMatch calculates two fitness scores based on responses given the prompt you provided. It creates a synthetic dataset by expanding on the input prompt before applying LLM-as-a-judge evaluations of each candidate model.

The first score captures how well a response fits the prompt - a baseline score for each base model. The second score is calculated after each base model assumes expert and novice personas on the topic of your prompt. We then measure how well each base model adheres to the persona and provide a score on each model's "trainability" on the topic or task of your prompt.

These scores can uncover the models with the best priors for your application without requiring costly training of each candidate.

What's next?

You've found the best model for your use case! Now you can explore how to train, score, and deploy a model:

• Fine-tune a model
• Score your model output
• Learn about the different deployment and model conversion options