Product
PrismaCoder-2M
An open-source multimodal code intelligence dataset containing roughly two million carefully filtered samples, produced through synthesis and filtering by the PrismaForge data pipeline. Its scale and diversity are leading among comparable open-source datasets.
Timeline (last 90 days)
Oct 15
PrismaCoder-2M is described by the paper as the most diverse and largest open-source dataset for multimodal code intelligence to date.
Unverified50%
Oct 15
The team trained and open-sourced two models, PrismaCoder-9B and PrismaCoder-35B, on the PrismaCoder-2M data.
Unverified50%