Step-by-Step to Success: Run AutoGPT using Azure OpenAI on Docker

Step-by-Step to Success: Run AutoGPT using Azure OpenAI on Docker

Integrating AutoGPT with Azure OpenAI through Docker offers a direct path to unlocking advanced AI capabilities. This detailed guide not only walks through the initial setup and configuration steps but also emphasizes the critical adjustments required for effective Azure OpenAI integration. Let’s dive into a more focused and informative discussion on setting up AutoGPT and ensuring it works seamlessly with Azure OpenAI services.

What is AutoGPT?

AutoGPT is like having a smart robot buddy that helps you achieve a specific goal by chatting with a super smart AI, kind of like having a conversation with a genius friend. Here’s how it works, broken down really simply:

  1. You Set a Goal: Imagine you have a goal, like planning a surprise birthday party or learning about space. You tell this to your robot buddy.

  2. The Robot Starts the Chat: Your robot buddy kicks things off by asking the first question to the genius AI, aiming to get information or ideas related to your goal.

  3. Listening and Thinking: After getting an answer, the robot thinks about it, figures out if it’s helpful, and what to ask next to get closer to your goal.

  4. Asking More Questions: Based on what the genius AI says, the robot keeps the conversation going, asking more questions to dig deeper or get more specific information, all aimed at reaching your goal.

  5. Goal Achieved: This back-and-forth chat continues until your robot buddy has gathered enough info or ideas to help you meet your goal, like having a full plan for that surprise party or a good understanding of space.

In short, AutoGPT is like a helpful middleman between you and a super-smart AI, doing all the talking and thinking for you, so you don’t have to come up with what to ask next. It makes getting to your goal easier by handling the conversation, making sure everything stays on track.

Detailed Configuration Steps for Integrating AutoGPT with Azure OpenAI

Initial Setup

  1. Install Docker

  2. Fork and Clone the AutoGPT Repository: Begin by forking the AutoGPT repository on GitHub and cloning it to your local machine, for instance, at C:\Auto-GPT.

Configuration

  1. Environment Setup:

    • Copy the .env.template file from C:\Auto-GPT\autogpts\autogpt to the primary folder C:\Auto-GPT and rename it to .env.
    • Edit the .env file, setting USE_AZURE=True to enable Azure OpenAI integration. Ensure True is capitalized to avoid issues.

  2. API Key Configuration:

    • Update the OPENAI_API_KEY in the .env file with your Azure OpenAI API key, found in the Azure portal under your OpenAI service’s “Keys and Endpoints”.

      image

  3. Docker and Azure YAML Setup:

    • Copy the azure.yaml.template file to C:\Auto-GPT rename it to azure.yaml we will adjust it later according to our Azure OpenAI service details.

    • Create a docker.compose.yml file in C:\Auto-GPT using the Docker setup template from the AutoGPT documentation. Add the following line to the volumes section to prevent the app/azure.yaml file not found error:

      Volume Sample:

      volumes:
  - ./azure.yaml:/app/azure.yaml

      Entire docker-compose:

       version: "3.9"
 services:
   auto-gpt:
     image: significantgravitas/auto-gpt
     env_file:
       - .env    
     ports:
       - "8000:8000"  # remove this if you just want to run a single agent in TTY mode
     profiles: ["exclude-from-up"]
     volumes:
       - ./data:/app/data
       ## allow auto-gpt to write logs to disk
       - ./logs:/app/logs
       ## allow auto-gpt to read the azure yaml file
       - ./azure.yaml:/app/azure.yaml
       ## uncomment following lines if you want to make use of these files
       ## you must have them existing in the same folder as this docker-compose.yml
       #- type: bind
       #  source: ./ai_settings.yaml
       #  target: /app/ai_settings.yaml
       #- type: bind
       #  source: ./prompt_settings.yaml
       #  target: /app/prompt_settings.yaml

Azure AI Models Deployment

  1. Deploy Azure AI Models:

    • Use Azure AI Studio to deploy necessary models like gpt-4 and gpt-3.5-turbo-text-embedding-ada-002, setting deployment names to match the model names for simplicity.

      image

Final Adjustments

  1. Modify the azure.yaml File:

    • Set azure_api_type to azure, ensuring the use of the API key for authentication. If you want to use Azure AD you can set the parameter to azure_ad. This will also require that you use an auth token as your OPENAPI_API_KEY. Instructions on how to obtain this token can be found in How to Configure AutoGPT with Azure OpenAI Active Directory Managed Identity.

    • The azure_api_base and azure_api_version was determined using the Azure AI Studio’s chat playground “View code” feature.

      image

      image

    • For azure_model_map, an iterative approach was taken. Initially, no mappings were specified. After running the Docker command, errors indicating missing models were used to gradually populate this section with the correct model mappings. This process involved mapping the AutoGPT’s expected model names to the corresponding deployment names in Azure AI Studio.

      2024-02-28_16-36-06

    Complete azure.yaml file.

    azure_api_type: azure
azure_api_base: https://rawopenai.openai.azure.com/
azure_api_version: 2024-02-15-preview
azure_model_map:
    gpt-3.5-turbo-16k: gpt-35-turbo
    gpt-4: gpt-4
    ext-embedding-3-small: text-embedding-ada-002

Execution

  1. Running AutoGPT:
    • Execute AutoGPT via Docker from the C:\Auto-GPT directory using the command docker compose run --rm auto-gpt. This step confirms the successful integration and functionality of AutoGPT with Azure OpenAI.

Conclusion

AutoGPT revolutionizes our interaction with AI by automating the conversation process, guiding us toward achieving specific goals with minimal effort. This transformative approach streamlines tasks ranging from content creation to complex data analysis, making it a versatile tool for anyone looking to leverage AI’s power. The simplicity of AutoGPT, coupled with its goal-oriented methodology, democratizes access to advanced AI capabilities, enabling users to focus on outcomes rather than getting bogged down in the technicalities of prompt engineering.

Through this article, we’ve provided a detailed blueprint for integrating AutoGPT with Azure OpenAI, ensuring you have the knowledge to harness this innovative technology effectively. Whether you’re a seasoned developer or new to the world of AI, the step-by-step guide laid out here is designed to empower you to implement AutoGPT within the Azure ecosystem successfully. Embracing AutoGPT opens up a realm of possibilities, allowing you to push the boundaries of what you can achieve with AI, turning complex tasks into manageable, goal-driven projects.

Comments

Post a Comment

Popular posts from this blog

Add User As Local Administrator On Domain Controller

I recently was settting up a new Microsoft SharePoint 2010 machine and had promoted the machine to a domain controller before creating my SharePoint admin accounts.  I needed to add several of my accounts to the local Administrators group.  Unfortunately after you promote a server to a domain controller you can no longer access the GUI for Local Users and Groups.  Instead I had to use the command line to add the users. Open a command promt using the "Run as administrator" function and then run the following command. net localgroup Administrators /add {domain}\{user} Note: do not include the {} brackets.
Read more

Calling Dataverse Web API in PowerShell using Client Credentials

Image
Connecting to Dataverse using PowerShell can be very helpful for data migrations and use within Azure DevOps. Connecting to an instance in a non-interactive way can be tricky though. This article will provide you the links you need for creation and App registration and adding the app user to your environment. You can then utilize the script provided to call Web API requests including ones you define using the new Custom API functionality now available. The script was written so that it is not dependent on any outside libraries such as the Microsoft.Xrm.Tooling connector. This is helpful in situation where involving an outside library will slow down your deployment time by having to be approved in a change control board. If utilizing an outside code library is not a concern you can create a connection to Dataverse utilizing the Microsoft.Xrm.Tooling connector Get-CrmConnection cmdlet. Create App Registration The first thing to do is create an App Registration within Azure AD for ...
Read more

Windows Server 2008R2 VMs Shut Down After 1 to 2 Hours

Image
When created a lab environment to test ADFS 2.0 I utilized the Windows 2008R2 VM baselines distributed by Microsoft.  After a few days I was told that I had to activate.  The VMs included a 180 day license for use but I didn't feel like adding another network adapter into Hyper-V to connect them to the internet.  I started having issues though where the servers would shut down every hour or so.  I though that maybe there was a memory issue and Hyper-V was shutting them down in order to free up RAM.  Turns out that if Server 2008R2 is not activated it automatically shuts down after a period of time.  After connecting the server to the internet and activating them the problem went away. Finally the madness of the unknown shut downs has ended :)
Read more