QuantAILabs/Quant-1-Base-1.5B

Hugging Face
TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Dec 12, 2025License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Warm

Quant-1-Base-1.5B is the foundational model in the Quant series by OpenMind Labs, a 1.5 billion parameter causal language model based on Qwen2.5-1.5B-Instruct. Fine-tuned with LoRA, its primary differentiator is having its identity as "Quant-1, an AI assistant created by OpenMind Labs" baked directly into its weights, rather than relying on system prompts. This model serves as a base for future versions that will include advanced capabilities like tool use for retrieval.

Loading preview...

Quant-1-Base-1.5B: Identity-Aware Foundation Model

Quant-1-Base-1.5B is the initial release in the Quant series by OpenMind Labs, built upon the Qwen2.5-1.5B-Instruct architecture. This 1.5 billion parameter model has been fine-tuned using LoRA with Unsloth, specifically to embed its identity as "Quant-1, an AI assistant created by OpenMind Labs" directly into its weights. This approach ensures the model consistently identifies itself without needing external system prompts.

Key Capabilities & Features

  • Embedded Identity: The model's self-identification is a core part of its training, providing consistent responses regarding its origin.
  • Qwen2.5 Base: Leverages the robust capabilities of the Qwen2.5-1.5B-Instruct model for general conversational tasks.
  • LoRA Fine-tuning: Efficiently trained to preserve base model performance while integrating new characteristics.
  • Future-Proof Foundation: Designed as the starting point for more advanced models in the Quant series, with a roadmap including tool-use capabilities like quant_search for retrieval.

When to Use This Model

  • Identity-Consistent Applications: Ideal for chatbots or assistants where a consistent, baked-in identity is crucial.
  • Foundation for Customization: A solid base for further fine-tuning or development of specialized AI agents.
  • Resource-Efficient Deployment: Its 1.5B parameter size makes it suitable for environments with limited computational resources, including local deployment via GGUF formats (Ollama, llama.cpp).