|
--- |
|
license: apache-2.0 |
|
language: |
|
- en |
|
base_model: |
|
- meta-llama/Llama-3.2-3B-Instruct |
|
tags: |
|
- biology |
|
- diet |
|
- wellness |
|
--- |
|
<a href="https://www.buymeacoffee.com/sfardin"><img src="https://img.buymeacoffee.com/button-api/?text=Buy me an apple&emoji=🍎&slug=sfardin&button_colour=FFDD00&font_colour=000000&font_family=Comic&outline_colour=000000&coffee_colour=ffffff" /></a> |
|
# Diet LLaMA - Nutrition-Focused Language Model |
|
|
|
A specialized LLaMA model fine-tuned for diet, nutrition and healthy eating recommendations, built with Unsloth acceleration. |
|
|
|
## Model Details |
|
|
|
- **Base Model:** unsloth/llama-3.2-3b-instruct-bnb-4bit |
|
- **Developer:** sfardin |
|
- **License:** LLAMA 3.2 License |
|
- **Languages:** English |
|
- **Framework:** adapter-transformers |
|
|
|
## Key Features |
|
|
|
- Optimized for diet and nutrition domain knowledge |
|
- 2x faster training using Unsloth optimization |
|
- 4-bit quantization for efficient deployment |
|
- Built on Huggingface's TRL library |
|
|
|
## Training |
|
|
|
This model was fine-tuned using: |
|
- Nutritional information datasets |
|
- Dietary guidelines and recommendations |
|
- Meal planning and recipe data |
|
- Health and wellness content |
|
|
|
## Metrics |
|
|
|
The model's performance is evaluated on: |
|
- Accuracy of nutritional information |
|
- Character-level prediction quality |
|
- Domain-specific knowledge validation |
|
|
|
## Usage |
|
|
|
The model can be used for: |
|
- Personalized diet recommendations |
|
- Nutritional information lookup |
|
- Meal planning assistance |
|
- Healthy eating guidance |
|
|
|
## Acknowledgments |
|
|
|
Built with ❤️ using [Unsloth](https://github.com/unslothai/unsloth) optimization |
|
|
|
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth) |
|
|
|
## Tags |
|
- text-generation-inference |
|
- transformers |
|
- unsloth |
|
- llama |
|
- gguf |
