decider-0.8b for the browser (int4 ONNX)

Built from Mapika/decider-0.8b at revision a0a01d6f8135298f400a8c856b355793012ae971 by cargo xtask onnx convert decider-0.8b in the Ardana repository. The license is the source's: apache-2.0.

The browser variant of Ardana's library model decider-0.8b, for onnxruntime-web.

  • model.onnx, model.onnx.data: the graph and its int4 weights, exported by the onnxruntime-genai 0.17.1 model builder for the WebGPU execution provider: fp16 inputs and outputs, the logits of the last position only, no multi-token-prediction head, the embedding shared with the LM head.
  • chat_template.jinja, decider_config.json, tokenizer.json, tokenizer_config.json: the source's, unchanged.
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for ardana-ai/decider-0.8b-ONNX

Quantized
(12)
this model