Show HN: Training synthetic models on highly complex datasets(github.com)
github.com
Show HN: Training synthetic models on highly complex datasets
https://github.com/gretelai/trainer
2 comments
We published a notebook and a GitHub repo that helps you train synthetic models on highly dimensional datasets (e.g. 1000's of columns, and millions of records). It works by using Gretel's open source header clustering to group correlated data and parallelize training across multiple GPUs.
Thanks for posting this :)