Machine Learning

Synthetic data is algorithmically generated data that mirrors the statistical properties of the dataset it’s based on. Learn how to make high-quality synthetic data.

Use Gretel Evaluate classification and regression tasks to validate synthetic data utility

Augment machine learning datasets with synthetically generated text and labels using an open-source implementation of GPT-3.

Gretel's synthetic data platform overcomes challenges across the data-centric machine learning life cycle to enable AI and ML solutions.

How we implemented a practical attack on a synthetic data model to validate its ability to protect sensitive information under different parameter settings.

Can synthetic data really be used in machine learning? We explore the utility of synthetic data created from popular datasets and tested on popular ML algorithms.

Test data generation is the process of creating new data that replicates an original dataset. Here’s how developers and data engineers use it.

A walkthrough of our new multi-table transform and multi-table synthetics notebooks, which can be used independently or simultaneously.

Data simulation is the process of using large quantities of data to predict events and validate models. Get the full data simulation definition.

Gretel introduces Reinforcement Learning from Privacy Feedback (RLPF), a novel approach to reduce the likelihood of a language model leaking private information.

How our new Synthetic Data Accelerator Program with AWS will help enterprises scale responsible AI systems fast.

Use Gretel's synthetic data platform to replace, augment, or balance training datasets within MLOps pipelines like Vertex AI, Azure ML, and Amazon SageMaker.

Gretel’s privacy-first generative AI is now available to all Azure users as well as select enterprises through the Microsoft for Startups Pegasus Program.

How Gretel's data quality analysis tools for evaluating generated text can help you optimize the performance LLMs, like the Llama-2 model.

Discover how to efficiently use Gretel's platform for prompting Llama-2 on large datasets, whether you're completing answers, generating synthetic text, or labeling.

How combining agent-based methods, LLMs, and synthetic data enables natural language queries for databases and data warehouses, sans SQL.

Let’s fine tune and prompt a large language model to swap the sentiment of product reviews!

Learn to prepare tabular data for AI and ML with an end-to-end data cleaning workflow.

Here's how a hospital uses Gretel to help forecast staffing and resource needs for their emergency care unit, and to identify emerging trends in outbreaks.

Gretel GPT supports new, state-of-the-art LLMs, and makes it easier for you to trust the privacy and accuracy of LLMs for enterprise use-cases.

Discover how Gretel ACTGAN can help businesses generate synthetic data at scale with improved accuracy, faster training, and reduced memory requirements.

How to utilize Gretel to create high-quality synthetic tabular data that you can use as training data for a classification model in Vertex AI.

Learn how synthetic image models can address data drift and improve a computer vision model's accuracy in unexpected conditions.

Learn about downstream machine learning tasks and synthetic data with Gretel’s new ACTGAN model and the PyCaret library

Celebrate Taylor Swift's The Eras Tour with lyrics generated with Gretel GPT, our synthetic language model.

An interview with one of Gretel's users on why medical practitioners turn to synthetic data when overcoming challenges with clinical data.

Learn the simplest way to generate synthetic data without setting up your own infrastructure and GPUs.

We’re going to train and build our synthetic dataset off of a real-time public feed of e-bike ride-share data called the GBFS (General Bike-share Feed)

Machine learning models can be trained to recognize patterns in datasets. By utilizing algorithms, they can learn to make decisions based on these patterns.

This article clarifies some common misconceptions about differential privacy and what it guarantees.

Generate synthetic time series data with Gretel.ai’s open-source PyTorch implementation of DoppelGANger.

Explore state-of-the-art image synthetics for business documents using diffusion models.

Data sharing is central to modern business but entails risks. Synthetic data can enable data sharing while reducing the risk of privacy-compromising linkage attacks.

There is a lot of hype around NLP. In this post, we explore some of the criticisms and how you can use this technology responsibly.

By merging breakthrough research on text metrics with new types of embeddings, we produce a reliable metric that is highly correlated with human ratings.

Use Gretel Synthetics and Colaboratory’s free GPUs to train a model to automatically generate fake, anonymized data with differential privacy guarantees.

Answers to some questions about synthetic data that audience members submitted during Gretel's talk at The Rise of Privacy Tech’s Data Privacy Week 2022 conference.

Learn about new features in Gretel, and how those features enabled us to discover personally identifiable information (PII) in a popular Kaggle dataset.

We founded Gretel based on our beliefs that data shouldn’t be scary.

How to train a FastCUT GAN on public location data from a few cities to predict realistic e-bike locations across the world.

Create differentially private, synthetic versions of datasets and meet compliance requirements to keep sensitive data within your approved environment.

Set up a cutting-edge environment for deep learning with TensorFlow 2.10, PyTorch, Docker, and GPU support.

Use synthetic data to improve model accuracy for fraud, cyber security, or any classification task with an extremely limited minority class.

A brief exploration of model soup, the new ensembling technique that takes the average weights of multiple models to improve overall performance.

We explore the popular open-source package Optuna to demonstrate how you can optimize your model hyperparameters and build the best synthetic model possible.

Generate artificial records to balance biased datasets and improve overall model accuracy.

Explore how to create a batch interface with the latest version of Gretel Synthetics on Google Colaboratory.

Look at how FastText word embeddings can help to quickly understand new datasets, and build more consistent labels for your own data.

The Gretel research team's favorite trends and takeaways from the NeurlPS 35th Annual Conference on Neural Information Processing Systems.

An evaluation of the effect of sampling procedures on the quality of synthetic tabular data using Gretel.ai's Synthetic Quality Score (SQS).

How to use Weights & Biases’ ML hyperparameter sweeps tool to optimize the accuracy of your synthetic data.

A look at how using Gretel’s Privacy Filters to immunize synthetic datasets against adversarial attacks can impact machine learning accuracy.

How we generated high-quality synthetic time-series data for one of the largest financial institutions in the world.