Welcome to the world of WilmerAI, a sophisticated middleware system that efficiently handles incoming prompts and routes them to a variety of Large Language Models (LLMs). This guide will walk you through how to install and use WilmerAI while also providing troubleshooting tips to help you tackle common pitfalls.

What Does WilmerAI Do?

WilmerAI stands for What If Language Models Expertly Routed All Inference? It offers the following features:

• Multi-LLM support for varied workflows and better responses.
• Integration with the Offline Wikipedia API for factual queries.
• Continuous chat summaries that simulate memory.
• Ability to parallel process memories and responses across multiple computers.
• Customizable JSON presets.
• API compatibility for OpenAI endpoints.

Setting Up WilmerAI

To ensure smooth sailing on your journey with WilmerAI, you’ll need to follow a few essential steps for installation and configuration:

Step 1: Install Python

Make sure you have Python installed. WilmerAI currently works well with Python versions 3.10 and 3.12.

Step 2: Installing Wilmer

You have two options for installation:

• Using Provided Scripts:
  – For Windows, run the .bat file.
  – For macOS, run the .sh script.
• Manual Installation:
  Execute the following commands in your terminal:
  

  pip install -r requirements.txt
  python server.py

Step 3: Configuring Users and Endpoints

Once installed, your next step is configuring the public folder where essential JSON files are stored. Each user can have specific settings that enable customized workflows:

• Create a new JSON file for the user in the Users folder.
• Update _current-user.json to include your newly created user.
• Set up endpoint configurations under the Endpoints folder.

Understanding Workflows

Workflows govern how WilmerAI interacts with LLMs. You can set various nodes to handle specific functions:

• Code generation, factual reasoning, and conversational snippets can each have dedicated nodes.
• A sample workflow is included in the PublicWorkflows folder, which you can modify as needed.

Analogy: Imagine WilmerAI as a sophisticated delivery service. Just like how a delivery service organizes packages based on destination and type, WilmerAI organizes incoming prompts and routes them to the appropriate LLM based on their nature—be it coding, conversation, or factual information.

Troubleshooting Tips

Encountering issues? Don’t fret! Here are some common problems and their fixes:

• No memories or summary files: Ensure that your workflow nodes exist and that the folder for saving files is correctly set up in the username.json file.
• No response in the front end: Ensure the stream settings match on both WilmerAI and your chosen front end.
• LLM doesn’t recognize presets: Replace unrecognized presets with ones that are compatible.
• Update without losing data: Back up the public folder where all your settings are stored before updating.

For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.

Final Thoughts

While WilmerAI is a powerful tool developed by a passionate individual, it is still a work in progress and may present challenges along the way. However, the rich features and potential it offers make it worth the effort!

At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.