dpshade22/embeddinggemma-scripture-v1

EmbeddingGemma-300M Fine-tuned for Biblical Text Search

This model is a fine-tuned version of google/embeddinggemma-300m specialized for biblical text search and retrieval.

Model Performance

• Accuracy@1: 12.00% (13x improvement over base model)
• Accuracy@3: 15.00%
• Accuracy@10: 31.00%
• Training Steps: 25 (optimal stopping point)
• Base Model Accuracy@1: 0.91%

Usage

from sentence_transformers import SentenceTransformer

# Load the model
model = SentenceTransformer('dpshade22/embeddinggemma-scripture-v1')

# Encode queries (use search_query: prefix)
query = "search_query: What is love?"
query_embedding = model.encode([query])

# Encode documents (use search_document: prefix)  
document = "search_document: Love is patient and kind"
doc_embedding = model.encode([document])

Prefixes

For optimal performance, use these prefixes:

• Queries: "search_query: your question here"
• Documents: "search_document: scripture text here"

Training Details

• Training Data: 26,276 biblical text pairs
• Learning Rate: 2.0e-04
• Batch Size: 8
• Training Strategy: Early stopping at 25 steps to prevent overfitting
• Output Dimensions: 768D (supports Matryoshka 384D, 128D)

Intended Use

This model is designed for:

• Biblical text search and retrieval
• Finding relevant scripture passages
• Semantic similarity of religious texts
• Question answering on biblical topics