Llama-3.2-3B-appreciation
| Property | Value |
|---|---|
| Base Model | Llama-3.2-3B-Instruct |
| Language | French |
| License | AGPL-3.0 |
| DOI | 10.57967/hf/3667 |
What is Llama-3.2-3B-appreciation?
Llama-3.2-3B-appreciation is a specialized French language model designed to assist secondary school teachers in generating student evaluations. Built on the Llama-3.2-3B-Instruct architecture, this model has been fine-tuned to produce personalized, constructive feedback based on student performance metrics.
Implementation Details
The model implements a PEFT (Parameter-Efficient Fine-Tuning) approach and uses Safetensors for efficient model storage. It was trained on a custom dataset (eltorio/appreciation) containing approximately 250 representative evaluations in its initial phase.
- Input parameters include student grades, behavior metrics, and participation levels
- Generates evaluations of 1-40 words in length
- Includes comprehensive error handling and GPU acceleration support
Core Capabilities
- Generates personalized student evaluations in French
- Processes multiple input parameters including trimester grades, behavior, participation, and work quality
- Maintains a supportive and constructive tone while highlighting areas for improvement
- Operates with both CPU and GPU support (optimized for T4 GPUs)
Frequently Asked Questions
Q: What makes this model unique?
This model specifically targets the French education system's evaluation needs, providing automated yet personalized student feedback while maintaining educational standards and ethical considerations.
Q: What are the recommended use cases?
The model is designed for French secondary school teachers to generate initial drafts of student evaluations, which should always be reviewed and validated by the teacher before use. It's particularly useful for handling large numbers of evaluations while maintaining consistency and personalization.
