The Gym is a powerful feature that allows developers to fine-tune AI models on our platform without needing extensive AI expertise. With The Gym, you can easily improve the performance and accuracy of your AI models by providing feedback based on traces and training your own custom models.

How It Works 💪

  • Data Preparation: Collect and prepare your training data by chatting with your model. Ensure that your data is representative of the task you want your AI model to perform.

  • Feedback From Traces: Provide positive or negative feedback to the AI model based on your Traces. The Traces section allows you to analyze the model’s performance to identify areas where it needs improvement.

  • Model Training: Initiate the training process using the provided feedback and training data. The Gym utilizes advanced algorithms to fine-tune the AI model based on your feedback and data.

  • Model Evaluation and Deployment: Once the training is complete, evaluate the fine-tuned model’s performance in the Lab. Assess the model’s accuracy, precision, recall, and other relevant metrics to ensure it meets your requirements. If satisfied with the model’s performance, deploy it uisng the Launchpad.

Getting Started

1

Test an AI model in the Lab

In order to train your AI model, you must first have a conversation with the model. Each response will be added to your Traces so you can monitor how your model is performing.

2

Provide Feedback

In order to fine-tune your model, you are required to submit at least 50 postive feedback responses. Just click the thumbs up 👍 for postive feedback and thumbs down 👎 for negative feedback.

Playground

It is suggested that you use the Traces sections to submit feedback.

Alterntively, you can also provide feedback from the Lab shown below.

Playground
3

Initiate the training

Simply click the Start fine-tuning button to begin fine tuning.

Playground

You’ll see whats represented in the image below once you initiate the fine-tuning.

Playground
4

You'll recieve confirmation emails

The first email will confirm that the fine-tuning process has begun.

Playground

The second email will confirm that the fine-tuning has been completed.

Playground
5

Try your fine-tuned model in the lab

Now that the fine-tuning is complete, head over to the Lab to test it out. You can search for your fine-tuned model the same way you would with any of the pre-trained models available. You can always change the system prompt and fine-tune again until you get your model working the way you want it to.

You can also test out your fine-tuned model indivdually in the chat section of the Lab

Playground
6

Launch your fine-tuned model with the Launchpad

Now that the fine-tuning and testing are complete, navigate to the Launchpad to deploy your model so it’s ready for integrating into your applications. You will still have the ability to configure the model’s params, system prompts and repositories.

Playground