gmongaras's picture
Update README.md
0d2afad verified
Weights for the `softmax` model from the paper [2Mamba2Furious: Linear in Complexity, Competitive in Accuracy](https://huggingface.co/papers/2602.17363).
This model variant just uses plain ol softmax (FlashAttention), used for the NIAH experiment. It was trained for 400K steps with a batch size of 32.
More details of the setup can be found in the Github repo.
Instructions on how to use this model can be found in [https://github.com/gmongaras/2Mamba2Furious](https://github.com/gmongaras/2Mamba2Furious)