AUTO-GPT 安装 | 稳定发布

Table of Contents

1. Introduction
2. Changes in Setup Process
3. Recommended Version and Optional Pinecone Usage
4. Prerequisites
5. Installing Python
6. Installing Git
7. Creating an OpenAI Account
8. Downloading Auto GPT
9. Setting up Auto GPT
   • Option 1: Downloading from the Stable Branch
   • Option 2: Cloning and Switching to the Stable Branch
10. Installing Python Dependencies
11. Configuring the Environment Variables
    • Setting the OpenAI API Key
    • Additional Environment Variables
12. Running Auto GPT
    • Running Auto GPT with Speak Option
    • Using Text-to-Speech Provider
13. Running Auto GPT in Continuous Mode
14. Exploring the Environment Variable File
    • Executing Local Commands
    • Memory Storage Options
    • Google API Key for Searching Content
    • Running Auto GPT with GPT 3.5 Model

Run Auto GPT Locally in Under 5 Minutes

With the advancements in the Auto GPT setup process, it is now easier than ever to run Auto GPT locally on your machine. In this article, we will walk you through the step-by-step process of setting up the stable version of Auto GPT, recommended by the Auto GPT team. We will also discuss the optional usage of Pinecone and explore the changes in the setup process since the release of the previous video. So let's get started!

1. Introduction

Auto GPT is a powerful tool that allows You to Create and run your own instance of the GPT language model locally. By setting up Auto GPT on your machine, you can harness the capabilities of the GPT model without relying on external services or incurring high costs. In this tutorial, we will guide you through the setup process and provide useful tips along the way.

2. Changes in Setup Process

Since the release of the previous video, the process for setting up Auto GPT has undergone some changes. One major change is that the Auto GPT team now recommends using the stable branch instead of the master branch for setup. This is due to the instability of the master branch, which can cause various issues. Additionally, the introduction of Pinecone, a long-term memory storage solution, has added complexity to the setup process, especially for running Auto GPT in continuous mode. However, we have some good news for you! Using the stable release of Auto GPT now allows you to use a local memory cache instead of Pinecone, saving you a significant cost.

3. Recommended Version and Optional Pinecone Usage

To ensure a smooth setup process and optimal performance, it is highly recommended to use the stable version of Auto GPT. The stable version provides a more reliable and stable environment for running Auto GPT on your machine. Furthermore, the stable version allows you to choose between using Pinecone or a local memory cache for storing the AI's progress and state. While Pinecone offers long-term memory capabilities, it comes at a cost of approximately $70 per month. However, with the stable version, you now have the option to use a local memory cache, which is free of charge. This flexibility allows you to choose the storage solution that best suits your needs and budget.

4. Prerequisites

Before diving into the setup process, there are a few prerequisites you need to fulfill. Firstly, make sure you have Python 3.10 or later installed on your machine. You can download the Python installer file from the official Python Website and follow the installation instructions. It is crucial to check the box that says "Add Python 3 to PATH" during the installation process. This will ensure that Python is accessible from the command prompt or PowerShell. Additionally, it is recommended to install Git on your machine. Git is a version control system that will be useful for downloading Auto GPT and managing your project files. You can download Git from the official Git website and follow the installation instructions.

5. Installing Python

Python is the programming language used for running Auto GPT. To install Python on your machine, follow these steps:

1. Visit the official Python website.
2. Scroll down to the bottom of the page and select the appropriate installer file for your operating system (e.g., Windows 64-bit installer).
3. After downloading the installer file, execute it and ensure that you check the box that says "Add Python 3 to PATH."
4. Click on "Install Now" and proceed with the installation.
5. After the installation completes, you may be prompted to restart your machine. It is recommended to do so to ensure the changes take effect.

6. Installing Git

Git is a version control system that allows you to manage your project files effectively. To install Git on your machine, follow these steps:

1. Visit the official Git website.
2. Download the Git installer file that corresponds to your operating system.
3. Run the installer file and follow the installation instructions.
4. Leave the default settings as they are unless you have specific preferences.
5. Complete the installation process.

7. Creating an OpenAI Account

In order to use Auto GPT, you will need to create an account with OpenAI. The account creation process is simple and free. OpenAI provides some free credits upon registration, allowing you to get started with using their APIs. While calling APIs to GPT is a paid service, the free credits should be sufficient to make progress with your own instance of Auto GPT. To create an OpenAI account, follow these steps:

1. Visit the OpenAI website and go to platform.openai.com.
2. Click on the "Sign Up" or "Create Account" button.
3. Follow the instructions to create your account.
4. After creating your account, you should be given some free credits.
5. Take note of your OpenAI API key, as we will need it later in the setup process.

8. Downloading Auto GPT

Now that you have fulfilled the prerequisites and created an OpenAI account, it's time to download Auto GPT to your local machine. There are two options for downloading Auto GPT: downloading from the stable branch or cloning the repository and switching to the stable branch. We will walk you through both options.

Option 1: Downloading from the Stable Branch

1. Click on this link to download the stable version of Auto GPT.
2. Scroll down to the bottom of the page and locate the "Assets" section.
3. Choose either the zip file or the tar.gz file and download it.
4. Move the downloaded file to a folder of your choice and extract its Contents.
5. Once the extraction is complete, you can remove the Archive or zip file.
6. Navigate into the extracted folder to access the Auto GPT files.

Option 2: Cloning the Repository and Switching to the Stable Branch

1. Click on the "Code" button on the Auto GPT repository's page.
2. Copy the URL provided.
3. Create a folder on your machine where you want to store Auto GPT.
4. Open the command prompt or PowerShell and navigate to the newly created folder.
5. Type the command git clone followed by the URL you copied and press Enter.
6. After the cloning process is complete, navigate to the Auto GPT folder.
7. In the command prompt or PowerShell, type git switch stable and press Enter.
8. You will Notice that the files in the folder now reflect the stable version of Auto GPT.

9. Setting up Auto GPT

You are now ready to set up Auto GPT on your machine. In this section, we will guide you through the necessary steps.

Installing Python Dependencies

Auto GPT relies on certain Python dependencies to run smoothly. To install these dependencies, follow these steps:

1. Open the command prompt or PowerShell.
2. Navigate to the Auto GPT folder where you downloaded the files.
3. Type the command pip install -r requirements.txt and press Enter.
4. This will install all the required dependencies for Auto GPT.
5. If you have experience with Python and prefer working with virtual environments, you can create one before running the installation command. However, this is not necessary for this tutorial.

Configuring the Environment Variables

The environment variable file (dotenv.template) contains important configurations for Auto GPT. Before proceeding, make a copy of this file and rename it to .env. Then, open the .env file in a text editor of your choice.

Setting the OpenAI API Key

The most crucial configuration variable is the OPENAI_API_KEY. This variable should be populated with your OpenAI API key, which you obtained during the account creation process. To retrieve your API key, follow these steps:

1. Go to the OpenAI platform website (platform.openai.com).
2. Click on your profile image and select "Profile" from the dropdown menu.
3. Click on the small eyeball icon to view your key.
4. Copy the key and paste it into the OPENAI_API_KEY field in the .env file.

Additional Environment Variables

The .env file also contains other configuration options that you can explore. These options include:

• EXECUTE_LOCAL_COMMANDS: By setting this value to True, Auto GPT gains the ability to install packages by itself. This can be helpful for projects that require additional dependencies.
• MEMORY_STORAGE: Auto GPT needs a storage mechanism to track its progress and store its state. By default, it uses a local memory cache, which stores the information in a JSON file named otgpt.json. However, you can provide other storage options such as Pinecone or Redis if needed.

Save the .env file once you have made the necessary configurations.

Running Auto GPT

You are now ready to run Auto GPT on your machine. Open the command prompt or PowerShell and navigate to the Auto GPT folder. To start the Auto GPT instance, run the following command:

python -m AutoGPT

Auto GPT will prompt you to provide a name and role for your AI. You can enter any name and role you prefer. For example:

Name: Auto GPT Bot
Role: An AI bot designed to teach me all there is to know about Auto GPD

After providing the name and role, you will be asked to enter up to five goals or tasks for the AI to execute in sequence. For each goal, Auto GPT will generate its thoughts, reasoning, a plan of action, and potential criticisms. You can choose to enable the "speak" option to have Auto GPT respond with a human-like voice. However, note that the default text-to-speech mechanism may sound robotic, and it is recommended to use a service like 11 Labs for better voice quality.

Once you have entered the goals, Auto GPT will start executing them one by one. You may need to manually confirm certain actions. For example, if Auto GPT wants to perform a Google search, it will ask for your confirmation before proceeding. If everything goes well, Auto GPT will complete the goals and terminate the session automatically.

Running Auto GPT with Speak Option

If you want Auto GPT to respond with an audio output using a human-like voice, you can use the --speak option when running the AutoGPT command. For example:

python -m AutoGPT --speak

However, the default text-to-speech mechanism may not provide the desired voice quality. To use a better text-to-speech provider, such as 11 Labs, follow these steps:

1. Visit the 11 Labs website (beta.11labs.io) and sign up for an account.
2. After signing in, click on your profile image and go to your profile.
3. Click on the eyeball icon to view your API key and copy it.
4. Open the .env file and locate the TTS_PROVIDER section.
5. Replace the LOCAL provider with 11LABS and paste your API key.
6. Save the .env file.

Now, when you run Auto GPT with the --speak option, the text-to-speech output will be generated using the 11 Labs service, providing a more natural and pleasant voice.

Running Auto GPT in Continuous Mode

Auto GPT can also be run in continuous mode, where it executes the provided goals autonomously without requiring manual confirmation for each step. To run Auto GPT in continuous mode, use the --continuous flag when running the AutoGPT command. For example:

python -m AutoGPT --continuous

Please note that running Auto GPT in continuous mode, especially without supervision, can result in increased API usage and potentially higher costs. Make sure to use this mode responsibly and keep an eye on your API usage to avoid unexpected charges.

Exploring the Environment Variable File

The .env file contains several other environment variables that you can customize according to your needs. Here are some notable variables:

Executing Local Commands

The EXECUTE_LOCAL_COMMANDS variable, when set to True, allows Auto GPT to execute local commands. This can be helpful for projects that require Auto GPT to install packages or Interact with your local environment.

Memory Storage Options

The MEMORY_STORAGE variable determines how Auto GPT stores its progress and state. By default, it uses a local memory cache, which is a JSON file named otgpt.json. However, you can switch to other storage options, such as Pinecone or Redis, by updating the variable accordingly. If you choose to use Pinecone or Redis, you will need to provide the necessary API key and region details.

Google API Key for Searching Content

If Auto GPT experiences issues with Google search requests, you can provide your own Google API key in the GOOGLE_API_KEY variable. To obtain a Google API key, follow these steps:

1. Go to the Google Cloud Console website (console.cloud.google.com) and create a new project.
2. Once the project is created, go to the "Credentials" section.
3. Click on "Create credentials" and select "API key".
4. Copy the generated API key and paste it in the GOOGLE_API_KEY field in the .env file.

By providing your own API key, you can ensure seamless Google search functionality within Auto GPT.

Running Auto GPT with GPT 3.5 Model

If you do not have access to the GPT-4 model yet, you can still run Auto GPT using the GPT-3.5 model by adding the --gpt3-only flag when running the AutoGPT command. For example:

python -m AutoGPT --gpt3-only

While the GPT-3.5-only mode offers slightly restricted capabilities compared to GPT-4, it still enables you to make the most of Auto GPT.

In Conclusion

In this article, we have covered the step-by-step process of setting up and running Auto GPT locally on your machine. We have discussed the recommended version, optional usage of Pinecone, and the changes in the setup process since the previous video. By following our instructions and customizing the environment variables, you can have your own instance of Auto GPT up and running in under 5 minutes. Explore the features and experiment with different settings to unlock the full potential of Auto GPT. Enjoy your AI Journey with Auto GPT!

Highlights:

• Set up Auto GPT locally in under 5 minutes
• Changes in the setup process and the recommended version
• Optional usage of Pinecone for long-term memory storage
• Prerequisites: Python and Git installations, OpenAI account creation
• Downloading Auto GPT: directly or via cloning and switching branches
• Install Python dependencies and configure environment variables
• Running Auto GPT with or without text-to-speech response
• Running Auto GPT in continuous mode
• Exploring additional environment variables for customization
• Running Auto GPT with GPT 3.5 model
• Get started and unleash the power of Auto GPT!

FAQ:

Q: Can I run Auto GPT without an OpenAI account? A: No, an OpenAI account is required to generate an API key, which allows you to use Auto GPT's functionalities.

Q: Is Pinecone mandatory for using Auto GPT? A: No, using Pinecone is optional. With the stable version, you can choose to use a local memory cache instead, which is free of charge.

Q: Can I use a different text-to-speech provider instead of the default one? A: Yes, you can use services like 11 Labs for better text-to-speech quality. Simply update the environment variable with your 11 Labs API key.

Q: How can I avoid unexpected costs when running Auto GPT in continuous mode? A: Running Auto GPT without supervision in continuous mode can increase API usage and lead to higher costs. Monitor your API usage and set usage limits or safeguards to avoid excessive charges.

Q: Can I run Auto GPT using a different language model other than GPT-4? A: If you do not have GPT-4 access, you can run Auto GPT with the GPT-3.5 model by using the --gpt3-only flag when running the command.

Q: Can Auto GPT install packages and execute local commands? A: Yes, by setting the EXECUTE_LOCAL_COMMANDS variable to True, Auto GPT gains the ability to install packages and execute local commands, enhancing its capabilities.

Q: How can I customize the memory storage mechanism in Auto GPT? A: By default, Auto GPT uses a local memory cache. However, you can switch to other storage options like Pinecone or Redis by updating the MEMORY_STORAGE variable with the appropriate information.

Q: What if I encounter issues with Google search requests in Auto GPT? A: You can provide your own Google API key in the GOOGLE_API_KEY variable to ensure uninterrupted Google search functionality within Auto GPT.