mlx-community/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking-OptiQ-4bit

Update kv_config.json: re-measured with `optiq kv-cache` float32 sensitivity (the previous config was ranked on bfloat16 rounding noise). 8-bit layers [0, 1, 3] -> [0, 1, 4].

Add kv_config.json: per-layer KV cache precision from `optiq kv-cache` (target 4.5 bits). Use with `optiq serve --kv-config kv_config.json`.

OptiQ mixed-precision 4/8-bit quant (5.805 bpw)

initial commit