Model Overview
ClaudioSavelli/FAME-topics_PO_llama32-3b-instruct-qa is a 3.2 billion parameter instruction-tuned language model built upon the meta-llama/Llama-3.2-3B-Instruct architecture. This model distinguishes itself through its application of a Preference Optimization (PO) method for "unlearning" within the specific FAME-topics setting. It offers a substantial context length of 32768 tokens, enabling it to process and generate extensive responses.
Key Capabilities
- Preference Optimization (PO): Utilizes a PO method for targeted model unlearning, as detailed in the associated research paper.
- FAME-topics Setting: Specifically adapted and optimized for performance within the FAME-topics domain.
- Instruction Following: Inherits instruction-following capabilities from its Llama-3.2-3B-Instruct base.
- Extended Context: Supports a 32768-token context window, beneficial for complex or lengthy question-answering tasks.
Good For
- Research in Model Unlearning: Ideal for researchers exploring Preference Optimization techniques and their impact on model behavior.
- FAME-topics Applications: Suited for question-answering and instructional tasks within the FAME-topics domain.
- Llama-3.2-3B-Instruct Derivatives: Users familiar with the Llama-3.2-3B-Instruct base model seeking a specialized, unlearned variant.