GSQ-RCO-GGUF
Collection
Non-uniform GGUF quantizations via GSQ + RCO: per-tensor mixed precision in standard GGUF form • 5 items • Updated • 111
I have been using the new qwen3.8-27 fine-tunes of DavidAu. To be honest, i didn't notice any improvement in intelligence. Only less thinking tokens. that is it.