deucebucket/Granite-4.0-H-Small-Cerebellum-GGUF

data: backfill v1 reasoning benchmark evidence (arc/hellaswag/mmlu)

Upload README.md with huggingface_hub

Upload cerebellum_banner.png with huggingface_hub

docs: correct shot labels to match 0-shot harness; update renamed GGUF references

docs: add model-index evaluation metadata

rename: add quant label for Hub hardware-compat badge

benchmarks: add MMLU-Redux 68.6% score

Update benchmarks: HellaSwag 87.1%, ARC 90.7%

Add Cerebellum GGUF (14.2 GB, shexp Q2_K)

Initial model card with ablation findings

initial commit