HoangCuongNguyen/gemma-2-9b-safety-ra-sft
HoangCuongNguyen/gemma-2-9b-safety-ra-sft is a 9 billion parameter language model fine-tuned for safety-related tasks using Supervised Fine-Tuning (SFT). This model is based on the Gemma-2 architecture and has a context length of 16384 tokens. It was trained using the TRL framework, focusing on generating safe and appropriate responses.
Loading preview...
Overview
HoangCuongNguyen/gemma-2-9b-safety-ra-sft is a 9 billion parameter language model derived from the Gemma-2 architecture. It has been specifically fine-tuned using Supervised Fine-Tuning (SFT) with the TRL framework to enhance its safety-related response generation capabilities. The model supports a context length of 16384 tokens.
Key Capabilities
- Safety-Oriented Responses: Designed to produce safe and appropriate text outputs.
- SFT Training: Utilizes Supervised Fine-Tuning for targeted behavior modification.
- Gemma-2 Base: Built upon the Gemma-2 foundational model.
Training Details
The model was trained using the TRL framework (version 1.0.0) in conjunction with Transformers (version 5.5.4), Pytorch (version 2.10.0+cu130), Datasets (version 4.8.4), and Tokenizers (version 0.22.2). The training procedure focused on SFT to instill safety guidelines into its responses.
Good For
- Applications requiring language generation with an emphasis on safety.
- Developing chatbots or conversational AI where content moderation is critical.
- Research into safety alignment techniques for large language models.