Add peer comparison with methodology caveats
Kitty Bash LLM: model, GGUF quants, model card
initial commit
