Use qwen 32B or llama3 70B for much better results
pip install pyautogen
Download LM Studio
- Download the required model
- Start the model's server
- Copy the curl of the model and run it in the terminal
- copy the CURL address (EX: http://localhost:1234/v1) and place it in the base URL
Finds the mass of the planet and performs arthemetic operations
Bob, Alice, user_proxy and manager- 4 AI Agents communicates with each other
Bob: Tells a joke
Alice: Creticises the joke
user_proxy: Terminates the joke if alice says so
manager: Manages the other 3 agents
Consists of two agents, User and Assistant
The Assistant Agent creates a folder coding, and under the file creates a python script
The User Agent launches the code and finds out the error present in the assistants code, sends it back to the assistant to fix it,
this prcoess continues until the desired, fulling functing code is completed
Sequential chat in AI is when the AI remembers the flow of the conversation and responds based on what was said earlier. It keeps track of previous messages so it can give more relevant and connected answers, making the conversation feel natural and ongoing.
reflection_with_llm- You can take the entire conversation from Agent 1, summarize it, and give that summary to Agent.
last_message- Alternatively, you can use the "last_message" method to only pass the most recent message to Agent 2.
In the particular QuotesAgent we have used reflection_with_llm
