Cross encoders are high performing encoder models that compare two texts and output a 0-1 score. I've found the cross-encoders/roberta-large-stsb model to be very useful in creating evaluators for LLM outputs. They're simple to use, fast and very accurate.

The BERT hash uses a bucketing technique with projection to decrease the size of the embedding parameters (all <1M parameters). These models are very small and good for inference at the edge.

Features

Performance: Achieves Pearson: 0.7595 and Spearman: 0.7474 on the STS-Benchmark test set.
Efficient architecture: Based on the BERT Hash model architecture, offering lightweight models.
Extended context length: Processes sequences up to 8192 tokens, great for LLM output evals.
Diversified training: Pretrained on dleemiller/wiki-sim and fine-tuned on sentence-transformers/stsb.

Performance

Model	STS-B Test Pearson	STS-B Test Spearman	Context Length	Parameters	Speed
`dleemiller/ModernCE-large-sts`	0.9256	0.9215	8192	395M	Medium
`dleemiller/CrossGemma-sts-300m`	0.9175	0.9135	2048	303M	Medium
`dleemiller/ModernCE-base-sts`	0.9162	0.9122	8192	149M	Fast
`cross-encoder/stsb-roberta-large`	0.9147	-	512	355M	Slow
`dleemiller/EttinX-sts-m`	0.9143	0.9102	8192	149M	Fast
`dleemiller/NeoCE-sts`	0.9124	0.9087	4096	250M	Fast
`dleemiller/EttinX-sts-s`	0.9004	0.8926	8192	68M	Very Fast
`cross-encoder/stsb-distilroberta-base`	0.8792	-	512	82M	Fast
`dleemiller/EttinX-sts-xs`	0.8763	0.8689	8192	32M	Very Fast
`dleemiller/EttinX-sts-xxs`	0.8414	0.8311	8192	17M	Very Fast
`dleemiller/sts-bert-hash-nano`	0.7904	0.7743	8192	0.97M	Very Fast
`dleemiller/sts-bert-hash-pico`	0.7595	0.7474	8192	0.45M	Very Fast

Usage

To use sts-bert-hash for semantic similarity tasks, you can load the model with the Hugging Face sentence-transformers library:

python

from sentence_transformers import CrossEncoder

# Load CrossEncoder model
model = CrossEncoder("dleemiller/sts-bert-hash-nano", trust_remote_code=True)

# Predict similarity scores for sentence pairs
sentence_pairs = [
    ("It's a wonderful day outside.", "It's so sunny today!"),
    ("It's a wonderful day outside.", "He drove to work earlier."),
]
scores = model.predict(sentence_pairs)

print(scores)  # Outputs: array([0.9184, 0.0123], dtype=float32)

Output

The model returns similarity scores in the range [0, 1], where higher scores indicate stronger semantic similarity.

Training Details

Pretraining

The model was pretrained on the pair-score-sampled subset of the dleemiller/wiki-sim dataset. This dataset provides diverse sentence pairs with semantic similarity scores, helping the model build a robust understanding of relationships between sentences.

Classifier Dropout: a somewhat large classifier dropout of 0.15, to reduce overreliance on teacher scores.
Objective: STS-B scores from dleemiller/MocernCE-large-sts.

Fine-Tuning

Fine-tuning was performed on the sentence-transformers/stsb dataset.

Validation Results

The model achieved the following test set performance after fine-tuning:

Pearson Correlation: 0.7595
Spearman Correlation: 0.7474

Model Card

Architecture: bert-hash-nano
Tokenizer: Custom tokenizer trained with modern techniques for long-context handling.
Pretraining Data: dleemiller/wiki-sim (pair-score-sampled)
Fine-Tuning Data: sentence-transformers/stsb

Thank You

Thanks to the NeuML team for providing the BERT Hash models, and the Sentence Transformers team for their leadership in transformer encoder models.

Citation

If you use this model in your research, please cite:

bibtex

@misc{stsnano2025,
  author = {Miller, D. Lee},
  title = {Bert Hash STS: An STS cross encoder model},
  year = {2025},
  publisher = {Hugging Face Hub},
  url = {https://huggingface.co/dleemiller/sts-bert-hash-pico},
}

License

This model is licensed under the MIT License.