metadata

{}

chiliground-base-modernbert-v1

A sentence classification model for extracting relevant spans from documents based on a question.

Model Details

Base model: answerdotai/ModernBERT-base
Hidden dimension: 768
Number of labels: 2
Best validation F1: 0.7038
Saved on: 2025-03-29 19:17:14

Usage

from transformers import AutoTokenizer
from verbatim_rag.extractor_models.model import QAModel
from verbatim_rag.extractors import ModelSpanExtractor
from verbatim_rag.document import Document

# Initialize the extractor
extractor = ModelSpanExtractor(
    model_path="chiliground-base-modernbert-v1",
    threshold=0.5
)

# Create documents
documents = [
    Document(
        content="Climate change is a significant issue. Rising sea levels threaten coastal areas.",
        metadata={"source": "example"}
    )
]

# Extract relevant spans
question = "What are the effects of climate change?"
results = extractor.extract_spans(question, documents)

# Print the results
for doc_content, spans in results.items():
    for span in spans:
        print(f"- {span}")

Training Data

This model was trained on a QA dataset to classify sentences as relevant or not relevant to a given question.

Limitations

The model works at the sentence level and may miss relevant spans that cross sentence boundaries
Performance depends on the quality and relevance of the training data
The model is designed for English text only