Integrating Llama 3.3-70B into Minerva Messenger

GWDG provides Chat AI services for all Max Planck users, supporting multiple AI models via their web interface. We have now integrated Llama 3.3 70B into Minerva Messenger, allowing users to interact with it directly within their existing workflows. This integration enhances AI-powered assistance within Minerva, making it easier for users to leverage advanced AI capabilities without switching platforms.

For users interested in exploring other available models or accessing them via API, please refer to GWDG’s documentation.

Key Features of Llama 3.3-70B:

With the integration of Llama 3.3 70B, users can benefit from AI assistance in various ways:

• General Knowledge & Research Assistance: Ask about scientific concepts, historical facts, or technical explanations.
• Text Generation: Generate summaries, draft emails, or create content based on specific prompts.
• Code Assistance: Get help with coding-related questions, debugging, and programming explanations.
• Translation & Language Support: Translate text between languages or refine wording in English.
• Idea Brainstorming: Get suggestions for research topics, presentation structures, or writing outlines.
• Casual Q&A: Ask general questions for productivity, recommendations, or learning new concepts.

How to Use Llama 3.3-70B in Minerva Messenger

Interacting with the Llama 3.3-70B model in Minerva Messenger is straightforward and can be done in the following ways:

Direct Messaging

1. Initiate a Direct Message:
   • In the Minerva Messenger interface, start a new direct message.
   • Type the name of the Llama 3.3-70B bot (e.g., @llama) in the recipient field and select it from the list.
2. Send a Message:
   • Type your query or message in the text box and press Enter.
3. Receive and Continue the Conversation:
   • The bot will respond in a new thread.
   • To maintain context, continue the conversation within the same thread by clicking the reply button.

Group Conversations

1. Add the Bot to a Group:
   • Ensure the Llama 3.3-70B bot is added as a member of the group.
2. Address the Bot:
   • In the group chat, type @llama followed by your message to direct the query to the bot.
3. Threaded Responses:
   • The bot will reply in a new thread.
   • Continue the interaction within the same thread to maintain context.

Private Channels

1. Add the Bot to a Private Channel:
   • As a channel administrator or owner, add the Llama 3.3-70B bot to the private channel.
2. Interact with the Bot:
   • Type @llama followed by your message to engage the bot within the channel.
3. Maintain Context:
   • Use threaded replies to keep the conversation context intact.

Please Note:

• Threaded Conversations: Utilizing threads helps maintain the context of your interactions with the bot, ensuring coherent and relevant responses.
• Bot Membership: The Llama 3.3-70B bot must be a member of the group or channel to interact. Requests directed to @llama in groups or channels where the bot is not a member will be ignored.
• User Permissions: Only channel administrators or owners can add the bot to private channels. If you lack the necessary permissions, please contact the relevant person to add the bot on your behalf.

For further assistance or inquiries, please contact our support team.