Prompt Engineering Outlines 后端

# 后端配置指南 配置 Outlines 以适配不同模型后端的完整指南。 ## 目录 - 本地模型 (Transformers, llama.cpp, vLLM) - API 模型 (OpenAI) - 性能对比 - 配置示例 - 生产环境部署 ## Transformers (Hugging Face) ### 基础设置 python import outlines # 从 Hugging Face 加载模型 model = outlines.models.transformers("microsoft/Phi-3-mini-4k-instruct") # 配合生成器使用 generator = outlines.generate.json(model, YourModel) result = generator("Your prompt") ### GPU 配置 python # 使用 CUDA GPU model = outlines.models.transformers( "microsoft/Phi-3-mini-4k-instruct", device="cuda" ) # 使用特定 GPU model = outlines.models.transformers( "microsoft/Phi-3-mini-4k-instruct", device="cuda:0" # GPU 0 ) # 使用 CPU model = outlines.models.transformers( "microsoft/Phi-3-mini-4k-instruct", device="cpu" ) # 使用 Apple Silicon MPS model = outlines.models.transformers( "microsoft/Phi-3-mini-4k-instruct", device="mps" ) ### 高级配置 python # FP16 以实现更快的推理 model = outlines.models.transformers( "microsoft/Phi-3-mini-4k-instruct", device="cuda", model_kwargs={ "torch_dtype": "float16" } ) # 8-bit 量化 (更少内存占用) model = outlines.models.transformers( "microsoft/Phi-3-mini-4k-instruct", device="cuda", model_kwargs={ "load_in_8bit": True, "device_map": "auto" } ) # 4-bit 量化 (极低内存占用) model = outlines.models.transformers( "meta-llama/Llama-3.1-70B-Instruct", device="cuda", model_kwargs={ "load_in_4bit": True, "device_map": "auto", "bnb_4bit_compute_dtype": "float16" } ) # 多 GPU model = outlines.models.transformers( "meta-llama/Llama-3.1-70B-Instruct", device="cuda", model_kwargs={ "device_map": "auto", # 自动 GPU 分配 "max_memory": {0: "40GB", 1: "40GB"} # 单 GPU 限制 } ) ### 热门模型 python # Phi-4 (Microsoft) model = outlines.models.transformers("microsoft/Phi-4-mini-instruct") model = outlines.models.transformers("microsoft/Phi-3-medium-4k-instruct") # Llama 3.1 (Meta) model = outlines.models.transformers("meta-llama/Llama-3.1-8B-Instruct") model = outlines.models.transformers("meta-llama/Llama-3.1-70B-Instruct") model = outlines.models.transformers("meta-llama/Llama-3.1-405B-Instruct") # Mistral (Mi