README.md

---
license: mit
library_name: braindecode
tags:
- braindecode
- pytorch
- safetensors
- eeg
- meg
- brainomni
---

# brainomni-tiny-pretrained

Weights of BrainOmni tiny (lm_dim 256, 8 heads, 12 blocks) with its frozen tokenizer, for
[`braindecode.models.BrainOmni`](https://braindecode.org/stable/generated/braindecode.models.BrainOmni.html),
converted from the authors' release. The classification head is not pretrained (seeded random init); fine-tune or linear-probe before use.

```python
from braindecode.models import BrainOmni
model = BrainOmni.from_pretrained("braindecode/brainomni-tiny-pretrained", chs_info=raw.info["chs"], n_outputs=2)
```

`chs_info` must carry sensor positions (EEG) and coil orientations (MEG); the
`chs_info` in `config.json` (19 EEG channels, 10-20) is only a default. Input is
expected at 256 Hz, preprocessed as in the authors' code.

## Source and conversion

- Source: [OpenTSLab/BrainOmni](https://huggingface.co/OpenTSLab/BrainOmni) at
  revision `9a4d3c70495370397ccfbfd6d2496f25647545a5`, file `tiny/BrainOmni.pt` (sha256 `62c67ba6a84ea0625e67a3b5e7463fe3930bfee88a612a225e9062a052542ffc`), MIT licence.
- `convert_brainomni_checkpoints.py` (in this repository) renames the keys to
  braindecode's, drops the pretraining-only mask predictor, stores the RoPE
  cache as `(cos, sin)` pairs with zero sine (the released cache holds cosines
  only and the released code uses it as loaded) and writes `config.json`,
  `model.safetensors` and `pytorch_model.bin` with `save_pretrained`.
- The converted model's outputs equal braindecode's loading of the original
  file (max-abs difference 0.0, float32 and bfloat16).
- Requires a braindecode version newer than 1.8.1.

## Citation

```bibtex
@inproceedings{xiao2025brainomni,
  title     = {BrainOmni: A Brain Foundation Model for Unified EEG and MEG Signals},
  author    = {Xiao, Q. and Cui, Z. and Zhang, C. and Chen, S. and Wu, W. and
               Thwaites, A. and Woolgar, A. and Zhou, B. and Zhang, C.},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year      = {2025},
  note      = {arXiv:2505.18185},
}

@article{aristimunha2025braindecode,
  title   = {Braindecode: a deep learning library for raw electrophysiological data},
  author  = {Aristimunha, Bruno and others},
  journal = {Zenodo},
  year    = {2025},
  doi     = {10.5281/zenodo.17699192},
}
```

## License

MIT, as the original BrainOmni release.