Skip to content

credence:e4b-vision

5 TagsUpdated 7.5B params8192 context100+ languagesApache-2.0by Txoka

Credence v1 Gemma4 E4B accuracy-focused, Q8_0 with Winnow-E4B's unchanged vision projector: PNG image decisions through libmtmd.

multilingual7.5b
ollaya run credence:e4b-vision --preset triage "I was charged twice for my subscription this month and want a refund."

Details

  • weightsc80a2320a02e · 8.7 GBgguf · Q8_0 · huggingface.co/Txoka/Credence-v1-Gemma4-E4B/resolve/7d5ffc8…/accuracy/model-Q8_0.gguf
  • decisionda8c875409d8 · 2 KB{"engine": "llama", "family": "winnow", "layout": "winnow-v1", "gguf": {…}, …}
  • calibrationbea23d6cdc7e · 191 B{"temperature": [1.0, 1.0, 1.0]}
  • licensea5aa25916b90 · 11 KBWinnow RLCD decision-model project
  • projectorddf46c21d707 · 990 MBapplication/vnd.ollaya.projector.gguf

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.

Credence readme and all models