ClaudioSavelli/FAME_PO_llama32-1b-instruct-qa
ClaudioSavelli/FAME_PO_llama32-1b-instruct-qa is a 1 billion parameter instruction-tuned language model, based on the Llama 3.2 architecture. This model has been specifically unlearned using a Preference Optimization method within the FAME setting. It is designed for question-answering tasks where unlearning specific preferences or biases is a key requirement.
Loading preview...
Model Overview
ClaudioSavelli/FAME_PO_llama32-1b-instruct-qa is a 1 billion parameter instruction-tuned model derived from the meta-llama/Llama-3.2-1B-Instruct base. Its primary distinction lies in its application of a Preference Optimization (PO) method for unlearning within the FAME (Forgettable and Memorable Examples) setting. This process aims to remove or reduce specific learned preferences or biases from the model's responses.
Key Capabilities
- Preference Unlearning: Utilizes a Preference Optimization method to modify the model's behavior, specifically for unlearning in the FAME context.
- Instruction Following: As an instruction-tuned model, it is designed to follow user prompts and instructions for various tasks.
- Question Answering: Optimized for question-answering scenarios where the unlearning of certain preferences is critical.
Good For
- Research and development in model unlearning and preference optimization.
- Applications requiring a language model with reduced specific biases or preferences.
- Instruction-based question-answering systems where controlled output is paramount.