QmogAI/Qwen3-0.6B.mog

Qwen3-0.6B-f16

Qwen3 0.6B in .mog format for qmog.cpp. One file: config, tokenizer, and weights.

Base	Qwen/Qwen3-0.6B
Format	MOG v2, f16
Size	~1.2 GB
Weights	f16 (unquantized)
Params	0.6B
Context	32,768 (config: 40,960 max position embeddings)

Download

hf download QmogAI/Qwen3-0.6B.mog qwen3-0.6B.mog --local-dir .

Run

Requires a qmog.cpp build with Qwen3 support:

./build/qmog-cli qwen3-0.6B.mog "Hello"

Export your own

Use qpack to convert a Hugging Face Qwen3 checkpoint to .mog:

pip install -r requirements.txt
python export.py --model_dir /path/to/Qwen3-0.6B --out qwen3-0.6B.mog

License

Apache 2.0 (derived from Qwen3-0.6B).