chromadb: This module seems to be related to database operations, possibly for storing and retrieving chat data.
autogen: Likely a custom module or a third-party library for generating conversational agents.
RetrieveUserProxyAgent: A class for an agent with extra content retrieval capabilities. make sure to install pyautogen with the [retrievechat] option before using RAG agents.
pip install "pyautogen[retrievechat]"
Configuration parameters are loaded from a JSON file or environment variable.
you need to create file OAI_CONFIG_FILE.json
llm_config dictionary is initialized with various settings like timeout, cache seed, and configuration list.
Several agents are instantiated with different roles such as "Boss," "CTO," "Product_Manager," and "Developer." Each agent is configured with specific termination message criteria, system messages, and LLM (Large Language Model) configurations. Chat Initialization Functions:
norag_chat(): Initiates a conversation without content retrieval capabilities. rag_chat(): Initiates a conversation with an agent (boss_aid) capable of content retrieval. function_calling_rag_chat(): Initiates a conversation with content retrieval handled by a function (retrieve_content) within the LLM configuration. Content Retrieval Mechanism:
The script supports two modes of content retrieval: one where an agent (boss_aid) is responsible for retrieval, and another where a function (retrieve_content) is called to retrieve content. Execution Flow:
The script first sets up the agents and then initiates conversations based on different scenarios, such as with or without content retrieval.
1.if you face any problem regarding hugging face , just checkout how to do setup for huggingface in offline mode
2.upgrade nltk and transforemer
Currently Im facing issue with chromadb , once i find the solution i will update the code .