dmrau's picture
Create README.md
6350da5
|
raw
history blame
289 Bytes
This model was trained with the [Rankers Library](`https://github.com/davidmrau/rankers`).
It was trained on the official MS MARCO training triples.
Training parameters:
- batch size: 128
- learning rate 3e-6
- Optimizer: Adam
Performance TREC DL Nist test 2020:
- NDCG@10: 69.35