微软发布安全对齐模型 HARC-Qwen2.5-7B-Instruct
英文摘要
Microsoft has released HARC-Qwen2.5-7B-Instruct, a fine-tuned version of Qwen2.5-7B-Instruct optimized for safety and alignment in conversational AI. The model is a transformer-based text-generation model, available on Hugging Face under the Apache 2.0 license. It is distributed in safetensors format and is compatible with text-generation-inference and Hugging Face endpoints. The release is associated with the paper arXiv:2607.00572.
中文摘要
微软发布了 HARC-Qwen2.5-7B-Instruct，这是基于 Qwen2.5-7B-Instruct 进行安全对齐微调的对话模型。该模型为基于 Transformer 的文本生成模型，以 Apache 2.0 许可证在 Hugging Face 上开放，采用 safetensors 格式，支持 text-generation-inference 和 Hugging Face 端点。此发布与论文 arXiv:2607.00572 相关。
关键要点
Fine-tuned from Qwen2.5-7B-Instruct for safety and alignment.
基于 Qwen2.5-7B-Instruct 进行安全对齐微调。
Released by Microsoft under Apache 2.0 license on Hugging Face.
由微软在 Hugging Face 上以 Apache 2.0 许可证发布。
Uses safetensors format and supports chat applications.
采用 safetensors 格式，支持对话应用。
Linked to arXiv paper 2607.00572; specifics not detailed in model card.
关联论文 arXiv:2607.00572，模型卡未提供具体内容。