> For the complete documentation index, see [llms.txt](https://yeasy.gitbook.io/llm_internals/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian.md).

# 第三部分：推理与部署篇

- [第九章：解码策略：模型如何生成文本](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding.md)
- [9.1 自回归解码：逐词生成的机制](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding/9.1_autoregressive_decode.md)
- [9.2 贪心搜索与束搜索：确定性与近似搜索](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding/9.2_greedy_beam.md)
- [9.3 采样策略：温度、Top-k 与 Top-p 的设计直觉](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding/9.3_sampling.md)
- [9.4 结构化输出与约束解码](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding/9.4_constrained.md)
- [9.5 解码侧的推理时扩展：生成、搜索与验证](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding/9.5_test_time_scaling.md)
- [9.6 扩散语言模型：另一种生成范式](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding/9.6_diffusion_lm.md)
- [本章小结](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/09_decoding/summary.md)
- [第十章：推理优化：第一性原理的分析](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization.md)
- [10.1 推理瓶颈分析：计算密集还是访存密集](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization/10.1_bottleneck.md)
- [10.2 KV 缓存：为什么能避免重复计算](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization/10.2_kv_cache.md)
- [10.3 FlashAttention：IO 感知的算法设计](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization/10.3_flash_attention.md)
- [10.4 模型量化：用更少的位数表示权重与激活值](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization/10.4_quantization.md)
- [10.5 剪枝与知识蒸馏：模型瘦身的两条路](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization/10.5_pruning_distillation.md)
- [10.6 投机解码：为什么“先猜后验”能加速](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization/10.6_speculative_decoding.md)
- [本章小结](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/10_inference_optimization/summary.md)
- [第十一章：推理引擎与生产部署](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving.md)
- [11.1 推理引擎架构概览](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.1_engines_overview.md)
- [11.2 连续批处理与 PagedAttention](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.2_continuous_batching.md)
- [11.3 调度循环与抢占：一个请求在引擎里的全过程](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.3_scheduler_loop.md)
- [11.4 KV 显存管理：块表、换出与重算](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.4_kv_memory_management.md)
- [11.5 前缀复用：哈希块与前缀树](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.5_prefix_reuse.md)
- [11.6 多 LoRA 服务：一份基础权重，许多适配器](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.6_multi_lora_serving.md)
- [11.7 约束解码：让输出必然合乎语法](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.7_constrained_decoding.md)
- [11.8 多卡推理：张量并行、流水线并行与专家并行](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.8_multi_gpu_inference.md)
- [11.9 分离式 Prefill-Decode 架构](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.9_disaggregated_serving.md)
- [11.10 实例剖析：vLLM](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.10_vllm_internals.md)
- [11.11 实例剖析：SGLang](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.11_sglang_internals.md)
- [11.12 硬件选型：GPU、TPU 与专用加速器](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.12_hardware.md)
- [11.13 生产部署最佳实践](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.13_best_practices.md)
- [11.14 GPU 软件栈：从张量到内核与通信](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/11.14_gpu_software.md)
- [本章小结](https://yeasy.gitbook.io/llm_internals/di-san-bu-fen-tui-li-yu-bu-shu-pian/11_serving/summary.md)