GraySoft
Projects Models Compare Cloud benchmarks FAQ Download guIDE →

marcsun13 author hub

GGUF kernels Compute directly on the packed blocks of a GGUF checkpoint, so a quantized model runs without ever materializing a dense copy of its weights. Ops | op | signature | | | | | dequantize | blocks, ggml type, rows, cols, dtype rows, cols | | mul mat vec | blocks, x, ggm…

Models
3
Downloads
0
marcsun13/gguf-kernels
0 downloads ·

Run models locally with guIDE

Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.

Download guIDE → · Browse 524k+ models · Compare models