The kmseong/llama3.2_3b_new_SSFT_lr3e-5_gsm8k_ft_full_params_lr3e-5 is a 3.2 billion parameter Llama-3.2-3B-Instruct model, developed by kmseong, that has been fine-tuned using the Safety Neuron Tuning (SN-Tune) method. This approach selectively fine-tunes only a small set of safety-critical neurons on safety alignment data, enhancing safety without significantly impacting general capabilities. It is primarily designed for applications requiring improved safety alignment and parameter-efficient fine-tuning.
Loading preview...
Overview
This model, kmseong/llama3.2_3b_new_SSFT_lr3e-5_gsm8k_ft_full_params_lr3e-5, is a 3.2 billion parameter variant of the meta-llama/Llama-3.2-3B-Instruct base model. It has undergone Safety Neuron Tuning (SN-Tune), a specialized fine-tuning method aimed at enhancing safety alignment.
Key Capabilities
- Enhanced Safety Alignment: Specifically fine-tuned on safety data using the SN-Tune method to improve safety responses.
- Parameter-Efficient Fine-tuning: SN-Tune selectively fine-tunes only a small subset of "safety neurons" while freezing other parameters, minimizing computational overhead.
- Preservation of General Capabilities: This method aims to enhance safety with minimal impact on the model's broader language understanding and generation abilities.
Good for
- Applications where safety and responsible AI behavior are critical.
- Scenarios requiring a smaller, more efficient model with improved safety characteristics.
- Developers looking for a Llama-3.2-3B-Instruct variant with explicit safety alignment without extensive retraining.
This model is licensed under the Apache 2.0 License, inheriting from its base model.