This project contains a web service that will accept HTTP GET requests at
http://localhost:8080/simpleChat.
There is optional question parameter whose default value is "Can you please tell me a joke?".
The response to the request is from the selected LLM/model.
Before using the Ollama profile, you need to have Ollama installed.
Before using the AI commands, make sure you have a developer token from OpenAI, to use that profile.
Create an account at OpenAI Signup and generate the token at API Keys.
The Spring AI project defines a configuration property named spring.ai.openai.api-key that you should set to the value of the API Key obtained from openai.com.
Exporting an environment variable is one way to set that configuration property.
export OPENAI_API_KEY=<INSERT KEY HERE>This will use ollama with the llama3 model by default.
./mvnw spring-boot:run
To get a response to the default request of "Tell me a joke"
curl http://localhost:8080/simpleChatA sample response is
Sure, here's a classic one for you:
Why don't scientists trust atoms?
Because they make up everything!
Now using the question request parameter
curl --get --data-urlencode 'question=Tell me a joke about a cow.' http://localhost:8080/simpleChat A sample response is
Why did the cow go to space?
Because it wanted to see the mooooon!