|
--- |
|
license: other |
|
license_name: sample-code-license |
|
license_link: LICENSE |
|
library_name: ml-4m |
|
--- |
|
|
|
# 4M: Massively Multimodal Masked Modeling |
|
|
|
*David Mizrahi\*, Roman Bachmann\*, Oguzhan Fatih Kar, Teresa Yeo, Mingfei Gao, Afshin Dehghan, Amir Zamir* |
|
|
|
Official implementation and pre-trained models for "4M: Massively Multimodal Masked Modeling" (NeurIPS 2023). |
|
|
|
[`Website`](https://4m.epfl.ch) | [`Paper`](https://arxiv.org/abs/2312.06647) | [`GitHub`](https://github.com/apple/ml-4m) |
|
|
|
4M is a framework for training "any-to-any" foundation models, using tokenization and masking to scale to many diverse modalities. |
|
Models trained using 4M can perform a wide range of vision tasks, transfer well to unseen tasks and modalities, and are flexible and steerable multimodal generative models. |
|
|
|
|
|
## Installation |
|
For install instructions, please see https://github.com/apple/ml-4m. |
|
|
|
|
|
## Usage |
|
|
|
This model can be loaded from Hugging Face Hub as follows: |
|
```python |
|
from fourm.models.fm import FM |
|
fm = FM.from_pretrained('EPFL-VILAB/4M-7-SR_L_CC12M') |
|
``` |
|
|
|
Please see https://github.com/apple/ml-4m/README_GENERATION.md for more detailed instructions and https://github.com/apple/ml-4m for other 4M model and tokenizer checkpoints. |
|
|
|
|
|
## Citation |
|
|
|
If you find this repository helpful, please consider citing our work: |
|
``` |
|
@inproceedings{mizrahi2023fourm, |
|
title={{4M}: Massively Multimodal Masked Modeling}, |
|
author={David Mizrahi and Roman Bachmann and Oguzhan Fatih Kar and Teresa Yeo and Mingfei Gao and Afshin Dehghan and Amir Zamir}, |
|
booktitle={Thirty-seventh Conference on Neural Information Processing Systems}, |
|
year={2023}, |
|
} |
|
``` |
|
|
|
## License |
|
|
|
The model weights in this repository are released under the Sample Code license as found in the [LICENSE](LICENSE) file. |
