---
datasets:
- jc132/HA-Requests-Zh
base_model:
- microsoft/Phi-3.5-mini-instruct
language:
- en
- zh
license: apache-2.0
tags:
- text-generation-inference
- home
- assistant
---

⚠️ **This model is currently in preview and still under active training. It is not recommended for production use at this stage.**

This model is a fine-tuned version of the [Phi-3.5-mini-instruct](https://huggingface.co/microsoft/Phi-3.5-mini-instruct) model, designed to interact with and control smart home devices via a Home Assistant integration. The fine-tuning dataset, [HA-Requests-Zh](https://huggingface.co/datasets/jc132/HA-Requests-Zh), focuses on function calling and includes support for Traditional Chinese (繁體中文) commands and instructions.

The model follows the Phi-3 prompt format and is capable of interpreting and responding to instructions in both English and Traditional Chinese. It can provide information about the Home Assistant environment, including available devices and services, and can generate function-calling code to execute specific tasks within the Home Assistant ecosystem.


Example "system" prompt: 
```
你是一個能控制智慧家居設備的AI助手，請依指示完成以下任務，或根據提供的資訊回答以下問題。
The current time and date is 08:12 AM on Thursday March 14, 2024
Services: light.turn_off(), light.turn_on(rgb_color,brightness), fan.turn_on(), fan.turn_off()
Devices:
light.office 'Office Light' = on;80%
fan.office 'Office fan' = off
light.kitchen 'Kitchen Light' = on;80%;red
light.bedroom 'Bedroom Light' = off
```


Output from the model will consist of a response that should be relayed back to the user, along with an optional code block that will invoke different Home Assistant "services". The output format from the model for function calling is as follows:


`````
turning on the kitchen lights for you now
```homeassistant
{ "service": "light.turn_on", "target_device": "light.kitchen" }
```
`````