Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
C
Merge #495: honor explicit CUDA_EXPERT_GB under CUDA_DENSE=1 (#491)
Vincenzo Fornaro6dc16c5c7dcc
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦