mncai/agiin-11.1B-v0.0

TEXT GENERATIONConcurrency Cost:1Model Size:15BQuant:FP8Ctx Length:8kTool Calling:SupportedPublished:Dec 14, 2023License:apache-2.0Architecture:Transformer Open Weights Cold

The mncai/agiin-11.1B-v0.0 model, developed by MindsAndCompany, is a 15 billion parameter language model built on the Mistral architecture with an 8192-token context length. It has undergone pretraining, instruction tuning, and DPO (Direct Preference Optimization). This model is designed for general generative AI applications, with MindsAndCompany focusing on developing solutions like Code Assistant, TOD Chatbot, and Enterprise AGI.

Loading preview...

Model Overview

The mncai/agiin-11.1B-v0.0 is a 15 billion parameter language model developed by MindsAndCompany. It is built upon the Mistral architecture and features an 8192-token context length, making it suitable for processing longer sequences of text.

Key Characteristics

  • Architecture: Based on the Mistral architecture.
  • Training: Underwent a multi-stage training process including pretraining, instruction tuning, and Direct Preference Optimization (DPO).
  • Context Length: Supports an 8192-token context window.

Potential Use Cases

MindsAndCompany, the creator of this model, is actively developing various generative AI solutions, indicating the model's potential applicability in areas such as:

  • Code Assistance: Generating or assisting with code.
  • Task-Oriented Dialogue (TOD) Chatbots: Building conversational AI systems for specific tasks.
  • General Generative AI: Serving as a foundation for broader AI applications, including Enterprise AGI development.

Finetunes
2 models