Master the art of AI interaction with PygmalionAI Tutorial

Table of Contents

1. Introduction
2. Running Pick Mylon AI on Google Colab
3. Running Tavern AI on Local Machine
4. Selecting Pygmalion 6B Model on Google Colab GPU Page
5. Setting Up Pygmalion 6B Model on Tavern AI
6. Customization Options in Tavern AI
7. Importing Characters in Tavern AI
8. Soft Prompting in Tavern AI
9. Running Tavern AI Locally on Your Machine
10. Conclusion

Introduction

In this tutorial, we will be covering how to run Pick Mylon AI on Google Colab and Tavern AI on your local machine. We will go through the steps of selecting the Pygmalion 6B model on the Google Colab GPU page and setting it up on Tavern AI. Additionally, we will explore customization options, importing characters, and using soft prompting in Tavern AI. Lastly, we will discuss running Tavern AI locally on your machine.

1. Running Pick Mylon AI on Google Colab

To run Pick Mylon AI on Google Colab, navigate to the Google Colab GPU page and select the Pygmalion 6B model. Make sure to choose the United Cloudflare version for better security. If You want to use soft prompting, check the Google Drive checkbox. Once you have made your selections, press the "Run" button. The model will be instantiated, downloaded, and loaded into the system. Make sure to copy the link provided after "Setting Seed" in the console.

2. Running Tavern AI on Local Machine

To run Tavern AI on your local machine, download the Tavern AI zip file from the Tavern AI GitHub page. Then, install Node.js by clicking on the appropriate version for your system. After the installation, extract the Tavern AI main.zip file to your desired location and open the folder. Double-click on start.bat to open the Tavern AI page.

3. Selecting Pygmalion 6B Model on Google Colab GPU Page

On the Google Colab GPU page, scroll down to the list of available models and select the Pygmalion 6B model. Choose the United Cloudflare version as your provider. It is recommended to avoid using local tunnel for security reasons. If you want to use soft prompting, check the Google Drive checkbox. Once you have made your selections, press the "Run" button.

4. Setting Up Pygmalion 6B Model on Tavern AI

After running the Pygmalion 6B model on Google Colab, copy the link provided after "Setting Seed" in the console. Open the Tavern AI page and click on the three lines in the top right corner to access the settings menu. Make sure to select Cobalt AI and paste the copied link into the API URL field. Click on the "Connect" button to establish a connection. Select the desired settings, such as calibrated or classic 6B, depending on your preference.

5. Customization Options in Tavern AI

In the Tavern AI folder, you will find a "public" folder that contains various customization options. You can upload PNG files for user avatars and custom backgrounds. The "chats" folder stores your chat history with the characters in Tavern AI. Additionally, you can manually place custom background images instead of uploading them.

6. Importing Characters in Tavern AI

Tavern AI offers two ways to import characters: via Json files or PNG files. PNG files are recommended for better results. To import a character, click on the import button in the characters tab and select the PNG file. This will automatically import the image and associated features. You can also import a character using a Json file, but it requires manually uploading the character's picture.

7. Soft Prompting in Tavern AI

Soft prompting is a technique that enhances the quality of generations in Tavern AI. To initiate soft prompting, open the link provided in the API section of Tavern AI. This link will lead you to your Google Drive. Create a "Cobalt AI" folder and upload a soft prompt zip file. Return to the Tavern AI page, click on "Soft Prompt," and select the loaded soft prompt. Soft prompting improves the output of generations, especially with certain character themes.

8. Running Tavern AI Locally on Your Machine

If you prefer to run Tavern AI locally on your machine, download and extract the Tavern AI zip file. Open the extracted folder and double-click on start.bat. This will open the Tavern AI page. Keep in mind that running Tavern AI locally may require additional steps and setup.

9. Conclusion

Running Pick Mylon AI on Google Colab and Tavern AI on your local machine allows you to utilize AI models and generate text easily. With the steps outlined in this tutorial, you can set up and customize Tavern AI to enhance your chat experiences. Soft prompting and character importation further enhance the capabilities of Tavern AI. Whether you choose to run Tavern AI on Google Colab or your local machine, you can explore the various options available and enjoy interacting with AI-generated text.

Highlights

• Run Pick Mylon AI on Google Colab and Tavern AI on your local machine.
• Select the Pygmalion 6B model on the Google Colab GPU page.
• Customize Tavern AI by importing characters and setting avatars.
• Use soft prompting to improve the quality of AI-generated text.
• Run Tavern AI locally on your machine for a seamless experience.

FAQ

Q: Can I use soft prompting with any character theme? A: Soft prompting works best with certain character themes, such as the canosuba theme. You can find soft prompting options for different character themes on the official Big Valiant AI Discord.

Q: Are there any security concerns with using local tunnel as the provider on Google Colab? A: It is not recommended to use local tunnel as the provider on Google Colab due to security reasons. It is advisable to select the United Cloudflare version for better security.

Q: How can I delete messages from chats in Tavern AI? A: You can delete messages from chats in Tavern AI by clicking on the chat history and selecting the messages you want to delete.