potion-256d-compact-v3 Model Card

This Model2Vec model is an improved v3 compact static embedding model, pre-trained using Tokenlearn with contrastive learning and born-again self-distillation. It is distilled from mixedbread-ai/mxbai-embed-large-v1. It uses static embeddings, allowing text embeddings to be computed orders of magnitude faster on both GPU and CPU. It is designed for applications where computational resources are limited or where real-time performance is critical. This compact variant uses a curated 29.5K vocabulary for roughly half the size of the full potion-256d-v3 model, improving on potion-base-32M by +2.16 CatAVG.

Installation

Install model2vec using pip:

pip install model2vec

Usage

Load this model using the from_pretrained method:

from model2vec import StaticModel

# Load a pretrained Model2Vec model
model = StaticModel.from_pretrained("blobbybob/potion-256d-compact-v3")

# Compute text embeddings
embeddings = model.encode(["Example sentence"])

How it works

Model2vec creates a small, static model that outperforms other static embedding models by a large margin on all tasks on MTEB. This model is pre-trained using Tokenlearn. It's created using the following steps:

Distillation: a model is distilled from mxbai-embed-large-v1 at 256 dimensions using Model2Vec with a curated 29.5K token vocabulary.
Training data creation: the teacher model is used to create training data by encoding 500K sentences from C4.
Contrastive training: the distilled model is trained on the training data using Tokenlearn with contrastive loss.
Born-again self-distillation: the trained model is further improved by distilling from itself (alpha=1.0), gaining +0.5 CatAVG.
Power normalization: embeddings are transformed with sign(E) * |E|^0.7 for improved isotropy.

Results

Model	STS	Classification	PairClassification	CatAVG
potion-256d-compact-v3	79.75	62.74	73.57	72.02
potion-256d-v3	79.32	63.23	73.97	72.17
potion-base-32M	78.97	61.42	69.18	69.86
all-MiniLM-L6-v2	78.95	69.25	82.37	74.65
GloVe 300d	61.52	62.73	72.48	61.45

The results show that potion-256d-compact-v3 outperforms potion-base-32M by +2.16 CatAVG at a comparable model size, with only -0.15 CatAVG tradeoff compared to the full potion-256d-v3 model.

Additional Resources

Citation

Please cite the Model2Vec repository if you use this model in your work.

@software{minishlab2024model2vec,
  author       = {Stephan Tulkens and {van Dongen}, Thomas},
  title        = {Model2Vec: Fast State-of-the-Art Static Embeddings},
  year         = {2024},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.17270888},
  url          = {https://github.com/MinishLab/model2vec},
  license      = {MIT}
}

Loading…

potion-256d-compact-v3 Model Card

from model2vec import StaticModel # Load a pretrained Model2Vec model model = StaticModel.from_pretrained("blobbybob/potion-256d-compact-v3") # Compute text embeddings embeddings = model.encode(["Example sentence"])

How it works

Distillation: a model is distilled from mxbai-embed-large-v1 at 256 dimensions using Model2Vec with a curated 29.5K token vocabulary.

Training data creation: the teacher model is used to create training data by encoding 500K sentences from C4.

Contrastive training: the distilled model is trained on the training data using Tokenlearn with contrastive loss.

Born-again self-distillation: the trained model is further improved by distilling from itself (alpha=1.0), gaining +0.5 CatAVG.

Power normalization: embeddings are transformed with sign(E) * |E|^0.7 for improved isotropy.

Results

Model	STS	Classification	PairClassification	CatAVG
potion-256d-compact-v3	79.75	62.74	73.57	72.02
potion-256d-v3	79.32	63.23	73.97	72.17
potion-base-32M	78.97	61.42	69.18	69.86
all-MiniLM-L6-v2	78.95	69.25	82.37	74.65
GloVe 300d	61.52	62.73	72.48	61.45

The results show that potion-256d-compact-v3 outperforms potion-base-32M by +2.16 CatAVG at a comparable model size, with only -0.15 CatAVG tradeoff compared to the full potion-256d-v3 model.

Citation

Please cite the Model2Vec repository if you use this model in your work.

@software{minishlab2024model2vec, author = {Stephan Tulkens and {van Dongen}, Thomas}, title = {Model2Vec: Fast State-of-the-Art Static Embeddings}, year = {2024}, publisher = {Zenodo}, doi = {10.5281/zenodo.17270888}, url = {https://github.com/MinishLab/model2vec}, license = {MIT} }