Aniq-63/qwen3-0.6B-recipe-finetuned
Aniq-63/qwen3-0.6B-recipe-finetuned is a 0.8 billion parameter Qwen3-based language model, fine-tuned by Aniq-63 on the RecipeNLG dataset. This model specializes in generating complete recipes, including titles, ingredient lists with quantities, and step-by-step directions, when provided with a list of ingredients. It leverages a 32768 token context length and is optimized for culinary content creation.
Loading preview...
Model Overview
Aniq-63/qwen3-0.6B-recipe-finetuned is a specialized language model built upon the Qwen3-0.6B architecture. It has been meticulously fine-tuned by Aniq-63 using the RecipeNLG dataset, which comprises over 60,000 recipes. This fine-tuning process, conducted over 2 epochs with a LoRA method (r=64, alpha=128) and achieving a training loss of 0.86, has equipped the model with a unique capability in recipe generation.
Key Capabilities
- Recipe Generation: Given a list of ingredients, the model can generate a full recipe.
- Structured Output: Recipes include a title, a detailed ingredient list with quantities, and clear, step-by-step instructions.
- Chef Assistant Role: Designed to act as a professional chef assistant, providing practical and usable culinary instructions.
Good For
- Automated Recipe Creation: Ideal for applications requiring the automatic generation of recipes from ingredient inputs.
- Culinary Content Development: Useful for developers building tools for food blogs, cooking apps, or meal planning services.
- Small-Scale Deployment: With its 0.8 billion parameters, it offers a lightweight solution for specific recipe generation tasks, potentially allowing for efficient deployment.