Skip to content

decider:2b-vision

5 TagsUpdated 2.2B params32768 contextEnglishApache-2.0by Mapika

Reads an image as well as the state: Qwen3.5-2B vision-language with v5 text weights. One PNG per request, up to 10 options.

vision2.2b
ollaya run decider:2b-vision --image shelf.png --questions '{"blocked":{"type":"noul","instructions":"Is the aisle blocked?"},"fill":{"type":"score","instructions":"How full is the shelf?","criteria":["empty","half full","full"]}}' "A photo from the warehouse camera."

Details

  • graph68db9945404b · 7 MBonnx · decider · 2.2B · fp32
  • graphc18cff14cac2 · 2 MBonnx · vision · fp32
  • weightsac99c1665252 · 4.4 GBhuggingface.co/Mapika/decider-2b-vision/resolve/863e290…/model.safetensors
  • tokenizerb6d27c112837 · 20 MBhuggingface.co/Mapika/decider-2b-vision/resolve/863e290…/tokenizer.json
  • decisionde791797b8a8 · 2 KB{"engine": "onnx", "family": "decider", "encoder": "", "layout": "decider-vision-v1", …}
  • calibration366117537290 · 176 B{"temperature": [1.0, 1.0, 1.0]}
  • licenseb4d9d0d19903 · 10 KBdecider by Mapika (https://huggingface.co/Mapika/decider-2b)

Every layer is checked against its sha256 when it is pulled. Weights and tokenizers download from the model author's Hugging Face repository at a pinned commit; Ollaya never re-hosts them.

decider readme and all models