Tag: #llm

한국어English日本語中文

← All posts

• vLLM PagedAttention Production Serving Optimization and Inference Engine Comparison Guide

  A comprehensive LLM serving guide covering vLLM PagedAttention algorithm, production deployment, performance tuning, SGLang/TensorRT-LLM comparison, and Kubernetes integration.

  2026-03-06 · 22 min read #llm#vllm#paged-attention#model-serving#2026-03

• Complete LLM Quantization Comparison: GPTQ, AWQ, GGUF Practical Application Guide

  From the core principles of LLM quantization to comparative analysis of GPTQ, AWQ, GGUF, and BitsAndBytes techniques, covering practical application in vLLM and llama.cpp environments and quality-performance trade-offs.

  2026-03-06 · 14 min read #llm#quantization#gptq#awq#gguf

• LLM Multimodal Vision-Language Model Serving and Optimization Practical Guide

  A practical guide to serving and optimizing multimodal vision-language models in production.

  2026-03-05 · 24 min read #llm#multimodal#vlm#vllm#2026-03

• LLM Speculative Decoding Serving Optimization Playbook

  LLM Speculative Decoding Serving Optimization Playbook - a hands-on adoption guide as of 2026

  2026-03-04 · 14 min read #llm#speculative-decoding#2026-03

• LLM QLoRA Fine-Tuning Operations Guide: Cost, Quality, and Deployment

  LLM QLoRA Fine-Tuning Operations Guide: A comprehensive practical document covering cost, quality, and deployment, including Why/How/When, comparison tables, troubleshooting, hands-on code, and quizzes.

  2026-03-04 · 8 min read #llm#practical-guide#production

• LLM Serving: Speculative Decoding Production Benchmark 2026

  LLM Serving: Speculative Decoding Production Benchmark 2026 - A practical guide covering Why, How, When, comparison tables, troubleshooting, code examples, and quizzes.

  2026-03-04 · 14 min read #llm#2026-03

• LLM RAG Pipeline: Chunking Strategies and Embedding Optimization in Practice 2026

  A practical guide covering the core of LLM RAG pipelines: chunking strategies and embedding optimization. From comparing fixed-size, semantic, and recursive chunking to embedding model selection, vector DB indexing, and

  2026-03-04 · 19 min read #llm#2026-03

• Complete LLM Quantization Comparison — GPTQ vs AWQ vs GGUF

  A comprehensive guide to LLM Quantization — from quantization fundamentals to comparing GPTQ, AWQ, and GGUF methods, vLLM/llama.cpp integration, and practical benchmarks.

  2026-03-03 · 6 min read #llm#quantization#gptq#awq#gguf

• The Complete vLLM Guide — From PagedAttention to Production Optimization

  Covering everything about LLM inference optimization — from vLLM core PagedAttention mechanism to Continuous Batching, Tensor/Pipeline Parallelism, and Prefix Caching.

  2026-03-03 · 9 min read #llm#vllm#inference#paged-attention#model-serving

• Complete Guide to LLM Function Calling: From Tool Use Patterns to Production Design

  A deep dive into the Function Calling (Tool Use) mechanism of LLMs. Covers implementation methods across OpenAI, Anthropic, and open-source models, along with error handling, parallel invocation, and production design pa

  2026-03-03 · 8 min read #llm#function-calling#tool-use#openai#ai-agent

• LLM Structured Output Practical Guide — JSON Mode, Tool Use, Pydantic Schema Validation

  Compare Structured Output approaches across OpenAI, Anthropic, and Google, covering Pydantic schema validation to production pipeline construction with practical code examples.

  2026-03-03 · 22 min read #llm#structured-output#json-mode#openai#anthropic

• Complete Guide to LLM Context Window Extension: From RoPE, ALiBi, and YaRN to Ring Attention

  An in-depth analysis of techniques for extending LLM context windows from 512 to 2M tokens. Covers the mathematical principles of RoPE through NTK-aware scaling, YaRN, and Ring Attention with practical code examples.

  2026-03-03 · 9 min read #llm#context-window#rope#yarn#positional-encoding

• Build Your Own GPT — Training a Language Model from Scratch with nanoGPT

  Train a GPT language model from scratch using Andrej Karpathy's nanoGPT. A complete dissection of the Transformer architecture — tokenizers, Self-Attention, training loops — all with code.

  2026-03-03 · 7 min read #ai#llm#gpt#nanogpt#transformer

• Building Scalable LLM Serving Pipelines with Ray Serve

  From core concepts of ML/LLM model serving with Ray Serve to multi-model pipelines, autoscaling, batch inference, and production deployment — all with code examples.

  2026-03-03 · 6 min read #ai-platform#ray-serve#model-serving#llm#mlops

• Practical LLM Fine-Tuning — Building Your Own Model with LoRA, QLoRA, and PEFT

  A practical guide to LLM fine-tuning. From the mathematical principles of LoRA to training large models on consumer GPUs with QLoRA, and leveraging the PEFT library — all with detailed code examples.

  2026-03-02 · 7 min read #llm#fine-tuning#lora#qlora#peft

• Accelerating LLM Inference 2-3x with Speculative Decoding: From Theory to Production

  A deep dive into the mathematical foundations of Speculative Decoding, the Draft-Verify pipeline, acceptance probability analysis, practical implementation in vLLM/TensorRT-LLM, and Apple's Mirror Speculative Decoding.

  2026-03-02 · 8 min read #llm#speculative-decoding#inference#optimization#vllm

• Building a RAG Chatbot — Create Your Own Document QA Bot with LangChain + ChromaDB + OpenAI

  A hands-on guide to building a PDF document-based RAG chatbot from scratch using LangChain, ChromaDB, and OpenAI. Covers document loading, vector storage, prompt design, and a Streamlit UI.

  2026-03-02 · 9 min read #rag#langchain#chromadb#openai#chatbot

• LangGraph Agent Workflow Practical Guide: From Multi-Agent Orchestration to Production Deployment

  Build stateful AI agent workflows with LangGraph. Covers StateGraph, conditional routing, multi-agent orchestration, Human-in-the-Loop, and LangGraph Platform deployment — all with production-ready code.

  2026-03-02 · 6 min read #langgraph#langchain#ai-agent#workflow#multi-agent

• Production LLM Application Architecture Design Guide

  Analyze the core architecture layers of production LLM applications and organize Gateway, Guardrails, Caching, and Observability components based on official documentation.

  2026-03-01 · 21 min read #llm#architecture#production#ai#guardrails

• LLM Inference Optimization: In-Depth Analysis of vLLM and TensorRT-LLM

  Analyzing core LLM inference optimization techniques including PagedAttention, Continuous Batching, and quantization, based on the official documentation of vLLM and TensorRT-LLM.

  2026-03-01 · 22 min read #llm#inference#vllm#tensorrt-llm#optimization