CoolFace
Modelpublic

richardyoung/CardioEmbed-Gemma-2-2B

sourceHugging Faceapache-2.0updated 10mo agoView on Hugging Face
0likes7downloads
Model Card

CardioEmbed-Gemma-2-2B

Domain-specialized cardiology text embeddings using LoRA-adapted Gemma-2-2B

Part of a comparative study of 10 embedding architectures for clinical cardiology.

Performance

MetricScore
Separation Score0.455

Usage

python
from transformers import AutoModel, AutoTokenizer
from peft import PeftModel

base_model = AutoModel.from_pretrained("google/gemma-2-2b-it")
tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-2b-it")
model = PeftModel.from_pretrained(base_model, "richardyoung/CardioEmbed-Gemma-2-2B")

Training

  • —Training Data: 106,535 cardiology text pairs from medical textbooks
  • —Method: LoRA fine-tuning (r=16, alpha=32)
  • —Loss: Multiple Negatives Ranking Loss (InfoNCE)

Citation

bibtex
@article{young2024comparative,
  title={Comparative Analysis of LoRA-Adapted Embedding Models for Clinical Cardiology Text Representation},
  author={Young, Richard J and Matthews, Alice M},
  journal={arXiv preprint},
  year={2024}
}