Launching today
Gradian

Gradian

Find the data that broke your fine-tune.

8 followers

Visit website
Training-data attribution for LoRA fine-tuned LLMs. Gradian scores every training example against the capability you lost, then runs a battery of dataset and config checks for the failures influence math misses.

Gradian makers

Here are the founders, developers, designers and product people who worked on Gradian

Abdullah AbdelrazekBuilding Vylara | CTO @KLAO GmbH
GradianGradian