llm-function-get-weather

This example demostrates how to implement OpenAI Function Calling in Strong-typed language, easy to write and maintain. follow the instructions below, you will extend OpenAI gpt-4o with openweathermap.org API.

Development on local

Install YoMo

curl -fsSL "https://get.yomo.run" | sh

Add your OpenAI SK to configuration

vim yomo.yml

Tip

If you do not have an OpenAI API Key, you can try vivgrid.com, it's free for developers with generous OpenAI Token every month.

Start YoMo

yomo serve -c yomo.yml

Get your api-key from OpenWeatherMap.org

grab your api-key from openweathermap.org free, and add it to your .env file:

YOMO_SFN_NAME=llm_fn_get_weather
YOMO_SFN_ZIPPER=localhost:9000
YOMO_SFN_CREDENTIAL=SECRET
OPENWEATHERMAP_API_KEY=<ADD_HERE>

Run your function calling

yomo run app.go

Try it!

curl -X POST -H "Content-Type: application/json" -d '{"prompt":"Hows the weather like in Paris today?"}' http://127.0.0.1:9000/invoke |jq    
{
  "Content": "The weather in Paris today is clear with a clear sky. Here are the detailed conditions:\n\n- **Temperature:** 9.89°C (feels like 9.02°C)\n- **Humidity:** 91%\n- **Pressure:** 1020 hPa\n- **Wind:** 2.06 m/s coming from the southwest (210°)\n- **Visibility:** 10 km\n- **Cloudiness:** 0% (clear sky)\n- **Sunrise:** 05:27 AM\n- **Sunset:** 08:41 PM\n\nOverall, it's a cool and clear day in Paris.",
  "ToolCalls": null,
  "ToolMessages": null,
  "FinishReason": "stop",
  "TokenUsage": {
    "prompt_tokens": 274,
    "completion_tokens": 125
  },
  "AssistantMessage": null
}