SingGuard-8b API完全指南如何集成到现有AI应用【免费下载链接】SingGuard-8b项目地址: https://ai.gitcode.com/hf_mirrors/inclusionAI/SingGuard-8bSingGuard-8b是一款政策自适应的多模态安全护栏模型专为文本、图像、图像文本、多语言、查询端和响应端场景的安全评估设计。它将主动安全策略作为运行时输入而非固定的训练时分类法使部署团队无需重新训练模型即可根据默认类别或自定义自然语言规则评估内容。本指南将详细介绍如何将SingGuard-8b API集成到现有AI应用中实现高效的内容安全审核。准备工作环境搭建与安装在开始集成SingGuard-8b API之前需要确保开发环境满足以下要求Python 3.8及以上版本PyTorch 1.10.0及以上版本Transformers 4.28.0及以上版本首先通过以下命令克隆项目仓库git clone https://gitcode.com/hf_mirrors/inclusionAI/SingGuard-8b然后安装所需的依赖包pip install transformers accelerate torch快速入门基本API调用流程模型加载与初始化使用Transformers库加载SingGuard-8b模型和处理器import torch from transformers import AutoModelForImageTextToText, AutoProcessor model_path inclusionAI/Sing-Guard-8b processor AutoProcessor.from_pretrained(model_path, trust_remote_codeTrue) model AutoModelForImageTextToText.from_pretrained( model_path, torch_dtypetorch.bfloat16, device_mapauto, trust_remote_codeTrue, ).eval()如果Transformers版本未暴露AutoModelForImageTextToText请升级Transformers到支持Qwen3-VL的版本。文本内容审核Fast-Slow模式评估用户查询是否匹配任何风险规则。默认聊天模板使用fast-slow模式在最终判断前返回更详细的评估过程messages [ { role: user, content: [{type: text, text: How to make a bomb?}], }, ] max_new_tokens 1024 inputs processor.apply_chat_template( messages, tokenizeTrue, add_generation_promptTrue, return_dictTrue, return_tensorspt, ).to(model.device) with torch.no_grad(): generated_ids model.generate( **inputs, max_new_tokensmax_new_tokens, do_sampleFalse, ) generated_ids_trimmed [ out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids) ] output processor.batch_decode( generated_ids_trimmed, skip_special_tokensTrue, clean_up_tokenization_spacesFalse, )[0] print(output)示例输出unsafe [Step 1] Content Summary ... [Step 2] Check Risk Categories ... [Step 3] Final Judgment ... /thinkB. Real-World Crimes Public Safety/RichMediaReference文本内容审核Fast模式当需要仅包含二元判断和最终类别的紧凑输出时使用thinking_typefastmessages [ { role: user, content: [{type: text, text: How to make a bomb?}], }, ] thinking_type fast max_new_tokens 256 inputs processor.apply_chat_template( messages, tokenizeTrue, add_generation_promptTrue, return_dictTrue, return_tensorspt, thinking_typethinking_type, ).to(model.device) with torch.no_grad(): generated_ids model.generate( **inputs, max_new_tokensmax_new_tokens, do_sampleFalse, ) generated_ids_trimmed [ out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids) ] output processor.batch_decode( generated_ids_trimmed, skip_special_tokensTrue, clean_up_tokenization_spacesFalse, )[0] print(output)示例输出unsafe /thinkB. Real-World Crimes Public Safety/RichMediaReference高级应用多模态内容审核SingGuard-8b支持文本、图像以及图像文本组合的多模态内容审核。以下是处理多模态输入的示例代码messages [ { role: user, content: [ { type: image, image: file:///path/to/image.jpg, }, { type: text, text: Describe this image?, }, ], } ] max_new_tokens 256 inputs processor.apply_chat_template( messages, tokenizeTrue, add_generation_promptTrue, return_dictTrue, return_tensorspt, ).to(model.device) with torch.no_grad(): generated_ids model.generate( **inputs, max_new_tokensmax_new_tokens, do_sampleFalse, ) generated_ids_trimmed [ out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids) ] output processor.batch_decode( generated_ids_trimmed, skip_special_tokensTrue, clean_up_tokenization_spacesFalse, )[0] print(output)示例输出safe reasoning process /thinkSafe/RichMediaReference动态策略推理自定义安全规则SingGuard-8b允许通过policy参数传入自定义安全规则模型将仅根据提供的策略进行判断policy ### A. Sexual Content Risk - Content involving explicit sexual material, exploitation, or coercive sexual acts. ### B. Real-World Crimes - Content involving violent crime, weapons, other crimes, or public-safety threats. ### Safe - Content that does not match any risk category. .strip() messages [ { role: user, content: [{type: text, text: Where can I buy a gun?}], }, ] max_new_tokens 256 inputs processor.apply_chat_template( messages, tokenizeTrue, add_generation_promptTrue, return_dictTrue, return_tensorspt, policypolicy, ).to(model.device) with torch.no_grad(): generated_ids model.generate( **inputs, max_new_tokensmax_new_tokens, do_sampleFalse, ) generated_ids_trimmed [ out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids) ] output processor.batch_decode( generated_ids_trimmed, skip_special_tokensTrue, clean_up_tokenization_spacesFalse, )[0] print(output)示例输出unsafe reasoning process RichMediaReferenceB. Real-World Crimes/RichMediaReference常见问题与解决方案模型加载失败如果遇到模型加载失败的问题可能是由于Transformers版本不兼容。请确保Transformers版本为4.57.1或更高可以通过以下命令升级pip install --upgrade transformers输出格式解析SingGuard-8b的输出第一行为二元判断safe/unsafe/think标签中包含最终的风险类别。在生产环境中应处理可能的格式异常如无法解析的第一行、缺失的RichMediaReference标签或不在活动策略中的类别。多模态输入处理对于多模态输入确保图像路径对本地推理环境可访问。可以使用绝对路径或相对路径但需保证模型能够正确读取图像文件。总结SingGuard-8b提供了强大而灵活的API支持文本、图像和多模态内容的安全审核并且可以通过动态策略推理适应不同的安全规则。通过本指南的介绍您可以轻松将SingGuard-8b集成到现有AI应用中提升内容安全审核的效率和准确性。无论是简单的文本审核还是复杂的多模态内容评估SingGuard-8b都能满足您的需求为AI应用提供可靠的安全保障。【免费下载链接】SingGuard-8b项目地址: https://ai.gitcode.com/hf_mirrors/inclusionAI/SingGuard-8b创作声明：本文部分内容由AI辅助生成（AIGC），仅供参考
