Pinned Loading

1. inference-optimizations inference-optimizations Public

   60+ inference optimizations & kernels for serving frontier open models on B200, B300, RTX PRO 6000, H100/H200 and MI355X

   Python

2. numinous-technology/trimtab numinous-technology/trimtab Public

   Change inference engine configuration while the engine runs

   Python 3

3. numinous-technology/gmux numinous-technology/gmux Public

   tmux for GPUs: share one GPU in space and time. Runs anywhere Docker does.

   Go 4

4. numinous-technology/vitvm numinous-technology/vitvm Public

   git for virtual machines: sandboxes that checkpoint at every step.

   Go 2

5. laguna-s-2.1-vision laguna-s-2.1-vision Public

   Vision for a frozen language model via a 35.4M trained projector. Every benchmark reported with a blind control.

   Python

6. unxversal/c3d unxversal/c3d Public

   Cxmpute 3D Lab

   TypeScript 15