The model is intialized with nomic-ai/modernbert-embed-base-unsupervised.

Fine-tuning details:

Training steps: 10K
Training data: Contrastive training on Tevatron/msmarco-passage-new without title
Batch size: 32 * 2
Training group size: 8

Safetensors

Model size

0.1B params

Tensor type

BF16

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support