mncai/agiin-13.6B-v0.0

TEXT GENERATIONConcurrent Unit Cost:1Model Size:15BQuant:FP8Context Size:8kTool Calling:SupportedPublished:Dec 14, 2023License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The mncai/agiin-13.6B-v0.0 is a 13.6 billion parameter language model developed by MindsAndCompany, based on the Mistral architecture. It has undergone pre-training, instruction tuning, and DPO (Direct Preference Optimization). This model is designed for general language tasks and is part of MindsAndCompany's efforts in developing enterprise AGI solutions.

Loading preview...

Model Overview

The mncai/agiin-13.6B-v0.0 is a 13.6 billion parameter language model developed by MindsAndCompany. It is built upon the Mistral architecture and has been through a comprehensive training process including pre-training, instruction tuning, and Direct Preference Optimization (DPO). MindsAndCompany, a developer of various AI models and business solutions, is actively working on products like Code Assistant, TOD Chatbot, and LLMOps, with a long-term goal of developing Enterprise AGI (Artificial General Intelligence).

Key Characteristics

  • Architecture: Based on the efficient Mistral architecture.
  • Training: Utilizes pre-training, instruction tuning, and DPO for enhanced performance and alignment.
  • Parameter Count: Features 13.6 billion parameters, offering a balance between capability and computational efficiency.
  • Context Length: Supports a context window of 8192 tokens.

Intended Use Cases

This model is suitable for a variety of general language generation and understanding tasks, particularly within contexts where instruction-following and optimized responses are beneficial. It aligns with MindsAndCompany's broader initiatives in developing AI solutions for businesses, including potential applications in:

  • Code Assistance
  • Task-Oriented Dialogue (TOD) Chatbots
  • General LLM-powered applications

Finetunes
2 models