Waza Router v1

Fine-tuned from microsoft/Phi-3.5-mini-instruct (3.8B) on synthetic Waza operations assistant conversations.

Handles free-tier conversation for the Waza document assistant (Sol). Emits <<<ESCALATE: tool_name>>> when the user needs metered AI capabilities.

Usage with Ollama

ollama pull hf.co/Wazagiro/waza-router-v1

waza-router-v1-q4_k_m.gguf — Q4_K_M quantized (~2.4 GB), recommended for Ollama
Safetensors — full precision weights

Safetensors

Model size

0.5B params

Tensor type

BF16

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Base model

Quantized

(175)

this model