Llama-Agents

This repository aims to create a collection of LLaMA 3 agents using LangChain. The ultimate goal is to develop a suite of agents that can be used for various tasks and have access to different tools. The LLM will be self-hosted.

Agents Developed:

Simple Agent: Simple Llama 3 agent accessible from the command line with answers recorded in markdown files.
Conversational Agent: A conversational agent that can engage with users in natural language, answering questions and providing helpful responses and saving the conversations in files when requested.
Orchestrator Agent: An orchestration agent that can manage multiple tasks and workflows, streamlining processes and automating repetitive tasks.
RAG (Rule-based Action Generator) Agent: A rule-based agent that can generate actions based on predefined rules and conditions.

Features:

Generate LLaMA 3 agents using LangChain
Develop a conversational agent for customer service or technical support
Create an orchestrator agent to manage workflows and automate tasks
Design RAGs to generate actions based on rules and conditions
Interact with system files and integrate with other systems

Getting Started:

Clone this repository to your local machine.
Install the required dependencies, including Ollama, Llama3, and Python.
Run the main.py script to interact with the Simple Agent.

Using the Simple Agent (main.py):

Open a terminal and navigate to the project directory.
Run the following command to start a new conversation:
```
python main.py
```
This will create a new conversation file in the conversations directory with the current timestamp.
To continue an existing conversation, run the following command:
```
python main.py -c
```
This will load the conversation from the last used conversation file.
Enter your prompt or question when prompted. The agent will process your input and generate a response, which will be displayed in the terminal and saved to the conversation file.
To start a new conversation at any point, type new and press Enter. This will create a new conversation file and switch to it.
To exit the program, type exit and press Enter.

Conversation Files:

The conversations are saved in markdown format in the conversations directory. Each conversation is stored in a separate file named conversation_<timestamp>.md.

The conversation files contain the user prompts and the assistant's responses, along with the response time. If a response is interrupted due to early stoppage, it will be marked as "Interrupted" in the conversation file.

Note:

The Simple Agent uses the LLaMA 3 model, which needs to be set up and configured separately.
Make sure to have the necessary dependencies installed before running the script.
The conversation files are encoded in UTF-8 to ensure proper display of special characters.

...

[1] LangChain documentation: https://docs.langchain.com/ [2] LLaMA 3 documentation: https://www.llama3.ai/docs/

Netajam/Llama-Agents

Llama-Agents