Generate Synthetic Data in 60 seconds
Alex introduces you to the Gretel Console in 60 seconds.
Generate textbook-quality synthetic data for training LLMs and SLMs
How to get started creating high quality synthetic data for LLM training
How to use Gretel's PostgreSQL connector to automate synthetic data generation
Automating synthetic data generation using Gretel and PostgreSQL
We're going to walk through an example of creating synthetic data with Gretel in 60 seconds. We'll go ahead and name our project, and drag our sample CSV directly into the upload files, or pick your own CSV. Choose Generate Synthetic Data. Use the recommended configuration. You can see that the URL is pointing to you right now. Click Train. We can see that the model finished training after 39 passes over the data. The loss of being at about 0.44, and the accuracy being at about 90%, indicating we have a pretty good fit with the model. It went ahead and generated 5,000 records, and then generated a report. This report tells us how good of a job the model did learning the distribution of the insights that the original data. 94 indicates that we have an excellent report here or synthetic model. Go ahead and click on View Records. We can go ahead from here and download the example of 5,000 records that were created.