miae-s-ft / README.md
nielsr's picture
nielsr HF Staff
Improve model card metadata and content
5d10d45 verified
|
raw
history blame
2.43 kB
metadata
library_name: tedbench
license: bsd-3-clause
pipeline_tag: graph-ml
tags:
  - protein
  - structure
  - fold-classification
  - tedbench

TEDBench — Fine-tuned from pretrained MiAE (structure only)

Variant: miae_s  |  Parameters: 29M  |  Layers: 6  |  Hidden dim: 512  |  Attn heads: 8

This MiAEClassifier was initialized from a pretrained MiAE and fine-tuned on TEDBench for fold classification. It is based on the research presented in the paper Protein Fold Classification at Scale: Benchmarking and Pretraining.

Part of the TEDBench benchmark for protein fold classification (ICML 2026). MiAE is an SE(3)-invariant masked autoencoder that masks up to 90% of backbone frames and reconstructs the full structure with a lightweight decoder.

Resources

Architecture sizes

Variant Params Layers Hidden dim Attn heads
miae_s 29 M 6 512 8
miae_b 102 M 12 768 12
miae_l 339 M 24 1 024 16

Append +model.use_seq_input=true to miae_b for the +seq variant.

Usage

Load from the HuggingFace Hub

First, install the tedbench library:

pip install tedbench

Then you can load the model using the high-level API:

import tedbench

# Loads the fine-tuned MiAE-S fold classifier
model = tedbench.load_model("miae-s-ft")
model.eval()

Alternatively, use the low-level loader:

from tedbench.utils.io import load_from_hf

model = load_from_hf("TEDBench/miae-s-ft")
model.eval()

From a Lightning checkpoint

from tedbench.model import MiAEClassifier

model = MiAEClassifier.load_from_checkpoint("model.ckpt", weights_only=False)
model.eval()

Citation

@inproceedings{chen2026tedbench,
  title={Protein Fold Classification at Scale: Benchmarking and Pretraining},
  author={Chen, Dexiong and Manolache, Andrei and Niepert, Mathias and Borgwardt, Karsten},
  booktitle={Proceedings of the 43rd International Conference on Machine Learning},
  year={2026}
}