Fine-tune GPT-4 for your unique use case

Sync your fine-tune runs in two lines of code

Fine-tune your model with OpenAI’s API and sync all your results to your dashboard with just two lines of code.

$ # just two lines of code

from wandb.integration.openai import WandbLogger

# Finetuning logic


Fine-tune your model with OpenAI's API
Sync your results with a single line of code
Explore & track everything in your W&B dashboard
We are excited to use W&B capabilities to monitor the fine-tuning of our OpenAI models that help improve how we understand client feedback at Stitch Fix
Reza Sohrabi
Staff ML Engineer, Stitch Fix Algorithms
With the new W&B and OpenAI integration, we’re able to consolidate experiment tracking of our in-house ML projects and OpenAI fine-tuning projects in one place. This gives us huge time savings towards delivering insights from unstructured data for our customers.
Prem Viswanathan
ML Lead at
We’re big W&B fans here at OpenAI. Since W&B enables tracking progress over time, this integration allows our users who are fine-tuning GPT-3 to iterate faster toward increased accuracy of their production models.
Peter Welinder
VP of Product and Partnerships at OpenAI

W&B is trusted by more than 700,000 ML practitioners from some of the most innovative organizations in the world:

Learn how to use W&B with GPT-4

Take a deeper dive into the integration and learn how it works with the help of a simple fine-tune on a Wikipedia-based dataset

Download the quick guide

Learn more about our collaboration and get a few simple but powerful tips about how to leverage W&B and push your GPT-3.5 fine-tunes to production faster.