Taejung-kim/my-brain_v1-v2
Taejung-kim/my-brain_v1-v2 is a 0.5 billion parameter instruction-tuned language model, based on the Qwen2.5 architecture. This model was fine-tuned and converted to GGUF format using Unsloth, enabling faster training and efficient deployment. It is designed for general text-based applications and can be easily integrated into local inference setups via Ollama or llama-cli.
Loading preview...
Model Overview
Taejung-kim/my-brain_v1-v2 is a compact 0.5 billion parameter language model, specifically an instruction-tuned variant of the Qwen2.5 architecture. This model has been optimized for efficient deployment and usage through its conversion to the GGUF format. A key aspect of its development is the utilization of Unsloth, which facilitated a 2x faster training process.
Key Features
- Architecture: Based on the Qwen2.5 model family.
- Parameter Count: 0.5 billion parameters, making it suitable for resource-constrained environments.
- Format: Provided in GGUF format, ensuring compatibility with various inference engines like
llama-cliand Ollama. - Training Efficiency: Benefited from Unsloth for accelerated fine-tuning.
- Context Length: Supports a context window of 32768 tokens.
Deployment and Usage
This model is designed for straightforward local deployment. An Ollama Modelfile is included, simplifying the process for users of the Ollama ecosystem. For direct command-line inference, llama-cli can be used with the provided GGUF file. The model is primarily intended for text-only applications, with specific instructions provided for both text-only and multimodal (if applicable) inference tools.