The Brave Nightly browser has introduced an exciting new feature that allows users to integrate their own Large Language Models (LLMs) directly into the browser's AI assistant, Leo. This development opens up a world of customization, where users can connect both local and remote AI models to enhance their browsing experience. In this article, I'll walk you through the process of integrating ArNS-linked LLMs into Leo, share my personal experiences, and provide additional insights into the role of ArNS, its contributions, and its future potential.

Background:

Recently, Brave users have gained the ability to link their own AI models to Leo, either from their local machine or from a remote server. This feature is a game-changer for those looking to customize their AI interactions within the browser. You can download Brave Nightly to try out this feature here. For more detailed information, check out the full details on this new capability here.

The Role and Contributions of ArNS:

The Arweave Name System (ArNS) operates as a decentralized, permanent naming service on the Arweave network. The primary goal of ArNS is to provide human-readable names that allow users to easily access data and applications on Arweave. This system is used to name and link any content or application created on Arweave, making this data permanently accessible and immutable.

In this process, ArNS simplifies and enhances the sharing and integration of LLM models. By using ArNS-linked LLMs, users can easily import and switch models within AI assistants like Leo. The ability to easily share and modify these models among users contributes to the strengthening of the integration between AI and decentralized technologies in the future.

Our Contribution:

In collaboration with the community, two ArNS names were prepared for LLMs that are now available to be integrated into Leo. These models were uploaded to ArDrive by Arweave founder Sam Williams, with the ArNS names being set up by Almostgreat. ArNS plays a crucial role in this process as the cornerstone of decentralization and permanence. The models can be accessed via the following links:

• Meta LLaMA 3.8B Instruction Model Q4: meta-llama-3-8b-instruct-q4.ar.io

• CodeQwen Chat Model: codeqwen-chat-q3-k-m.ar.io

The Future Benefits of ArNS:

The future potential of ArNS is vast. By adhering to its commitment to data permanence and decentralization, ArNS can revolutionize the management and sharing of AI models. Through ArNS-linked LLMs, users can keep their AI assistants continuously updated and relevant by easily switching between different data sources and models. This flexibility is just one example of how AI technology can evolve in a decentralized ecosystem.

My Experience with Leo Integration:

The process of integrating these models into Leo was both challenging and rewarding. Initially, I had to ensure that the correct model file was in place. This involved creating a structured folder on my desktop as follows:

1. I created a new folder on the Desktop (C:\Users\casper\OneDrive\Desktop\ArNSLLM) and named it "models."

2. Inside the "models" folder, I created a subfolder named "Publisher."

3. Within the "Publisher" folder, I created another folder named "Repository."

4. Finally, I placed the CodeQwen1.5-7B-Chat-GGUF.gguf file into the "Repository" folder.

After setting up the folder structure, I used LM Studio to load the model and activate it. Once the model was active, I integrated it into the Leo language model in Brave Nightly. This process involved configuring the server endpoint within Brave Nightly to connect to the model hosted on my local machine.

Conclusion:

Integrating ArNS-linked LLMs into Leo on Brave Nightly is not just about technical accomplishment; it’s about contributing to a decentralized and innovative ecosystem. The role of ArNS strengthens this integration by making AI model management more flexible and accessible. With the combined efforts of the community, including contributions from Sam Williams and Almostgreat, this integration serves as a prime example of how modern AI can be enhanced with blockchain technology. For those looking to replicate this process, I hope this guide provides valuable insights and assistance.

The Brave Nightly browser has introduced an exciting new feature that allows users to integrate their own Large Language Models (LLMs) directly into the browser's AI assistant, Leo. This development opens up a world of customization, where users can connect both local and remote AI models to enhance their browsing experience. In this article, I'll walk you through the process of integrating ArNS-linked LLMs into Leo, share my personal experiences, and provide additional insights into the role of ArNS, its contributions, and its future potential.

Background:

Recently, Brave users have gained the ability to link their own AI models to Leo, either from their local machine or from a remote server. This feature is a game-changer for those looking to customize their AI interactions within the browser. You can download Brave Nightly to try out this feature here. For more detailed information, check out the full details on this new capability here.

The Role and Contributions of ArNS:

The Arweave Name System (ArNS) operates as a decentralized, permanent naming service on the Arweave network. The primary goal of ArNS is to provide human-readable names that allow users to easily access data and applications on Arweave. This system is used to name and link any content or application created on Arweave, making this data permanently accessible and immutable.

In this process, ArNS simplifies and enhances the sharing and integration of LLM models. By using ArNS-linked LLMs, users can easily import and switch models within AI assistants like Leo. The ability to easily share and modify these models among users contributes to the strengthening of the integration between AI and decentralized technologies in the future.

Our Contribution:

In collaboration with the community, two ArNS names were prepared for LLMs that are now available to be integrated into Leo. These models were uploaded to ArDrive by Arweave founder Sam Williams, with the ArNS names being set up by Almostgreat. ArNS plays a crucial role in this process as the cornerstone of decentralization and permanence. The models can be accessed via the following links:

• Meta LLaMA 3.8B Instruction Model Q4: meta-llama-3-8b-instruct-q4.ar.io

• CodeQwen Chat Model: codeqwen-chat-q3-k-m.ar.io

The Future Benefits of ArNS:

The future potential of ArNS is vast. By adhering to its commitment to data permanence and decentralization, ArNS can revolutionize the management and sharing of AI models. Through ArNS-linked LLMs, users can keep their AI assistants continuously updated and relevant by easily switching between different data sources and models. This flexibility is just one example of how AI technology can evolve in a decentralized ecosystem.

My Experience with Leo Integration:

The process of integrating these models into Leo was both challenging and rewarding. Initially, I had to ensure that the correct model file was in place. This involved creating a structured folder on my desktop as follows:

1. I created a new folder on the Desktop (C:\Users\casper\OneDrive\Desktop\ArNSLLM) and named it "models."

2. Inside the "models" folder, I created a subfolder named "Publisher."

3. Within the "Publisher" folder, I created another folder named "Repository."

4. Finally, I placed the CodeQwen1.5-7B-Chat-GGUF.gguf file into the "Repository" folder.

After setting up the folder structure, I used LM Studio to load the model and activate it. Once the model was active, I integrated it into the Leo language model in Brave Nightly. This process involved configuring the server endpoint within Brave Nightly to connect to the model hosted on my local machine.

Conclusion:

Integrating ArNS-linked LLMs into Leo on Brave Nightly is not just about technical accomplishment; it’s about contributing to a decentralized and innovative ecosystem. The role of ArNS strengthens this integration by making AI model management more flexible and accessible. With the combined efforts of the community, including contributions from Sam Williams and Almostgreat, this integration serves as a prime example of how modern AI can be enhanced with blockchain technology. For those looking to replicate this process, I hope this guide provides valuable insights and assistance.