Just a very quick quantization of AlexWortega/Qwen3.8-27B-antislop using Bartowski's recent recipe for Qwen 3.8 27B, following quants such as bartowski/Altworld_Hemmingway-1-GGUF.
I've only done IQ3_XXS for now since that quantization seems to be the current sweet spot for my system, targeting 16 GB VRAM, at least with other models. For a larger quant try the official quantizations. I've only done basic validation that this quantization was successfully generated and done a SWE-style coding screen which didn't show any obvious problems; I have not done extensive testing or comparisons yet, and I'm not easily able to compare this to the full weights of AlexWortega/Qwen3.8-27B-antislop.
Additionally, I've replaced the Jinja template with Froggeric's, v22.5.
See Bartowski's layout tooling, and his v6 calibration data.