cudabenchmarktest/qwen3.5-9b-qwen3.6-reasoning-distilled-GGUF

Upload README.md with huggingface_hub

Add model card with YAML metadata

add params for ollama pull capability detection

add template for ollama pull capability detection

cleanup: drop Modelfile

cleanup: drop qwen3.5-9b-qwen3.6-reasoning-distilled.q80.gguf

cleanup: drop qwen3.5-9b-qwen3.6-reasoning-distilled.q4km.gguf

cleanup: drop qwen3.5-9b-qwen3.6-reasoning-distilled.f16.gguf

r3-crown merged GGUF (base qwen3.5:9b Q4_K_M + LoRA via llama-export-lora)

Modelfile for FROM merged.q4km.gguf + RENDERER/PARSER qwen3.5

r3-crown GGUF: merged base+LoRA with full capability stack

Expand card: 22-scenario hard tool stress test + image handling section + honest HX8 knowledge regression + hf download/ollama create run instructions

Fix Modelfile: add RENDERER/PARSER qwen3.5 for tool calling; relative GGUF path; remove hardcoded TEMPLATE override

Update model card: measured tool-calling preservation

Upload Q4_K_M / Q8_0 / F16 GGUFs

initial commit