flm-add.py: fix -A3B size suffix (35b-a3b) and add family aliases (granite/llama, darwin/qwen3.6-moe, crow/qwythos/qwopus/huihui/qwen3.5, gemma-4/gemma4e)

Q4NX/FastFlowLM card: flm-add.py install + run/serve flow, kernels source, model info

Upload flm-add.py with huggingface_hub

Upload flm-add.py

Upload model.q4nx with huggingface_hub

Update README.md

Update README.md

Upload chat_template.jinja with huggingface_hub

Upload config.json with huggingface_hub

Upload tokenizer_config.json with huggingface_hub

Duplicate from FastFlowLM/Qwen2.5-3B-Instruct-NPU2