According to their GitHub, the file needed is the -Q2_0.gguf version, which requires compiling their version of llama.cpp. Easiest way to download is to use huggingface-cli if you have it available: hf download prism-ml/Ternary-Bonsai-27B-gguf --include "*-Q2_0.gguf*" If not, just download the file
