ML researcher building O(L) language models that train end-to-end on Apple Silicon. Creator of mlx-recurrence fused Metal training kernels for linear recurrence
-
Derrington Collaborative Synthetic Intelligence Labs
- Huntsville, Alabama
-
17:16
(UTC -06:00) - in/paul-derrington-2667701a
Pinned Loading
-
mlx-recurrence
mlx-recurrence PublicFused Metal training kernels for linear recurrence on Apple Silicon — SSD (Mamba-2), GLA, RG-LRU, rotational LRU. Flash-linear-attention for MLX: 12-18x less training memory, production-validated.
Python 4
Something went wrong, please refresh the page to try again.
If the problem persists, check the GitHub status page or contact support.
If the problem persists, check the GitHub status page or contact support.