fraQtl AI Research

company

https://fraQtl.ai

AI & ML interests

KV cache compression, inference optimization, model compression

Recent Activity

samuel-salfati updated a model 5 days ago

fraQtl/Qwen3.5-4B-NPU-GGUF

samuel-salfati updated a Space 6 days ago

samuel-salfati updated a collection 6 days ago

KV cache · more users per GPU

View all activity

updated a model 5 days ago

fraQtl/Qwen3.5-4B-NPU-GGUF

4B • Updated 5 days ago • 229

updated a Space 6 days ago

fraQtl

updated 3 collections 6 days ago

KV cache · more users per GPU

Compressed KV cache for vLLM and llama.cpp. Three-arm, needle-gated receipts, reproducible with one command. • 5 items • Updated 5 days ago

Hi-Fi GGUF · closer to the original, same size

Calibrated GGUFs for phones, laptops and GPUs, each measured against the strongest public file at equal size. Receipts in every card. • 11 items • Updated 5 days ago

NPU-native GGUF · Snapdragon Hexagon

Q4_0 files built for the Hexagon NPU. Every operation runs on the NPU; measured on a Snapdragon 8 Elite Gen 5. • 2 items • Updated 5 days ago