goombalab

Popular repositories Loading

1. hnet hnet Public

   H-Net: Hierarchical Network with Dynamic Chunking

   Python 881 104

2. hydra hydra Public

   Official implementation of "Hydra: Bidirectional State Space Models Through Generalized Matrix Mixers"

   Python 177 18

3. phi-mamba phi-mamba Public

   Official implementation of Phi-Mamba. A MOHAWK-distilled model (Transformers to SSMs: Distilling Quadratic Knowledge to Subquadratic Models)

   Python 127 15

4. raven raven Public

   Python 87 7

5. Gather-and-Aggregate Gather-and-Aggregate Public

   Experiments Notebook of "Understanding the Skill Gap in Recurrent Language Models: The Role of the Gather-and-Aggregate Mechanism"

   Python 17 3

6. chimera chimera Public

   Chimera: State Space Models Beyond Sequences

   Jupyter Notebook 11