Skip to main content

One post tagged with "Prompting"

Prompting tag description

View All Tags

PRAG - Prompting for RAG-styled Recommendation

· 2 min read
Eric Ngo
ML Engineer @dodo, ex-ML engineers @Meta/Samsung-Research/7-11

We are thrilled to announce the release of PRAG, a model family that extends the capabilities of existing encoder-only models to generate personalized recommendations.

PRAG is trained:

  • On a mix of public and internal datasets across e - commerce, media, and retail.
  • For diverse tasks: next product, similar content, complementary items, etc.
  • To minimize feature - engineering.
    • For example, shopping history is a de facto input and requires sorting. When using PRAG, you do not need to sort it by timestamp and simply add timestamp as a descriptive feature. If timestamp is not available, PRAG can still implicitly capture temporal patterns.

Versions:

  • prag-v1: Base model
ModelsRecall@5/10NDCG@5/10
Toys and GamesAll BeautyToys and GamesAll Beauty
PRAG-v10.0452/0.09210.0452/0.09210.0267/0.04140.0267/0.0414
TIGER0.0531/0.07120.0454/0.06480.0371/0.04320.0321/0.0384

The technical report will be soon released.