onyx-dot-app/information-content-model

Onyx Information Content Classification using SetFit with Base sentence-transformers/paraphrase-mpnet-base-v2

The model is for use by the Onyx Enterprise Search system to identify whether a short text segment contains information that could be useful by itself to answer a RAG-type question.

It is based on the SetFit approach, using sentence-transformers/paraphrase-mpnet-base-v2 as the Sentence Transformer embedding model. A trained LogisticRegression instance is used for classification.

The model has been trained using an efficient few-shot learning technique that involves:

1. Fine-tuning a Sentence Transformer with contrastive learning.
2. Training a classification head with features from the fine-tuned Sentence Transformer.

About Onyx

• Website: Onyx
• Repository: Open Source Gen-AI + Enterprise Search

Model Details

Core Model Description

• Model Type: SetFit
• Sentence Transformer body: sentence-transformers/paraphrase-mpnet-base-v2
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 512 tokens
• Number of Classes: 2 classes
• Language: English

SetFit Resources

• Repository: SetFit on GitHub
• Paper: Efficient Few-Shot Learning Without Prompts
• Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts

Uses

Use for Inference

The model is for use by the Onyx Enterprise Search system.

To test it locally, first install the SetFit library:

pip install setfit

Then you can load this model and run inference.

from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("onyx-dot-app/information-content-model")
# Run inference
preds = model("Paris is in France")
or:
pred_probability = model.predict_proba("Paris is in France")

Framework Versions

• Python: 3.11.10
• SetFit: 1.1.1
• Sentence Transformers: 3.4.1
• Transformers: 4.49.0
• PyTorch: 2.6.0
• Datasets: 3.3.2
• Tokenizers: 0.21.0

Citation

BibTeX (SetFit Approach)

@article{https://doi.org/10.48550/arxiv.2209.11055,
    doi = {10.48550/ARXIV.2209.11055},
    url = {https://arxiv.org/abs/2209.11055},
    author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
    keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
    title = {Efficient Few-Shot Learning Without Prompts},
    publisher = {arXiv},
    year = {2022},
    copyright = {Creative Commons Attribution 4.0 International}
}