llm_note logo

llm_note

harleyszhang/llm_note

LLM notes covering model inference transformer structures and framework analysis

GraphCanon updated today · GitHub synced today

888 stars90 forksLast push 5d Python

Decision brief

llm_note is a detailed resource for developers needing in-depth understanding of LLM frameworks and inference methods, particularly with respect to transformer models and kv-cache techniques.

Good fit when

  • Use llm_note when you seek extensive guidance on transformers' structures specific to large language model applications
  • Consider llm_note as a reference if you are dealing with complex inference processes involving CUDA programming or Triton kernels optimization

Avoid when

  • Do not rely on llm_note for foundational machine learning theory; it is too specialized
  • llm_note may not be suitable if your focus is exclusively on deployment strategies rather than deep structural and inferential code analysis of LLMs

Observed Jul 14, 2026 · Source: enrich:decision_facts

Verify the decision

Maintenance and security

Full trust report
Maintenance
Very active (5d since push)
As of today
Provenance
Not a fork · Personal account
As of today
Security (OSV)
No lockfile
As of 1mo

Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.

Install

pip install llm_note
PyPI

Similar tools

Same-category neighbours. No typed graph edges are catalogued for this tool yet.

Evidence and technical details

Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.

Overview

A collection of detailed notes from Harley Szhang on aspects related to large language models including their structure, inference methods, and an in-depth look at LLM frameworks.

Capability facts

Languages
python

Source: github.language · Aug 25, 2026

Categories

Tags

README

  • 我的自制大模型推理框架课程介绍
  • 一 transformer 模型
    • 1.1 transformer 模型基础
  • 二 LLM 量化推理
  • 三 llm 推理优化
    • 3.1 LLM 综合性能分析
    • 3.2 LLM 推理优化-算法层面
    • 3.3 LLM 并行加速
  • 四 高性能计算
    • 4.1 triton 笔记
    • 4.2 cuda 笔记
    • 4.3 高性能编程学习资料推荐
  • 五 框架解析
  • 参考资料

LLM notes, including model inference, hpc programming note, transformer model structure, and vllm framework code analysis notes.

我的自制大模型推理框架课程介绍

  1. 框架亮点:基于 Triton + PyTorch 开发的轻量级、且简单易用的大模型推理框架,采用类 Pytorch 语法的 Triton 编写算子,绕开 Cuda 复杂语法实现 GPU 内核开发。
  2. 价格:499。非常实惠和便宜,课程、项目、面经、答疑质量绝对对得起这个价格。
  3. 课程优势​
    • 手把手教你从 0 到 1 实现大模型推理框架
    • 项目导向 + 面试导向 + 分类总结的面试题
    • 2025 最新的高性能计算/推理框架岗位的大厂面试题汇总
  4. 项目优势​
    • 架构清晰,代码简洁且注释详尽,覆盖大模型离线推理全流程。​
    • 运用 OpenAI Triton 编写高性能计算 Kernel,其中矩阵乘法内核效率堪比 cuBLAS。​
    • 基于 PyTorch 实现高效显存管理。​
    • 课程项目完美支持 FlashAttentionV1、V2、V3GQA,以及 PageAttention 的具体实现。​
    • 使用 Triton 编写融合算子,如 KV 线性层融合等。​
    • 适配最新的 qwen3/qwen2.5/llama3/llava1.5 模型,相较 transformers 库,在 llama3 1B 和 3B 模型上,加速比最高可达 4 倍。
  5. 分类总结部分面试题
  1. 项目运行效果:

llama3.2-1.5B-Instruct 模型流式输出结果测试:

Qwen2.5-3B 模型(社区版本)流式输出结果测试:

Llava1.5-7b-hf 模型流式输出结果测试:

感兴趣的同学可以扫码联系课程购买,这个课程是我和《自制深度学习推理框架》作者一起合力打造的,内容也会持续更新优化。

一 transformer 模型

1.1 transformer 模型基础

  • transformer 论文解读
  • transformer 模型代码实现
  • llama1-3 模型结构详解
  • vit 论文速读
  • gpt1-3 论文解读
  • Sinusoida 位置编码算法详解
  • MLA结构代码实现及优化

二 LLM 量化推理

  • SmoothQuant 论文解读
  • SmoothQuant 算法源码剖析
  • AWQ 论文解读
  • AWQ 算法源码剖析

三 llm 推理优化

3.1 LLM 综合性能分析

  • llm 推理揭秘论文翻译
  • llm 综合分析论文翻译
  • llm推理服务框架总结

3.2 LLM 推理优化-算法层面

  • online-softmax 论文解读
  • flashattention-1 论文解读
  • flashattention-2 论文解读
  • flashattention-3 论文解读
  • flashattention1-2-3 系列总结
  • prompt-cache论文速读
  • vllm优化之cuda_graph详解

3.3 LLM 并行加速

  • 张量并行技术详解

四 高性能计算

4.1 triton 笔记

  • 理解 triton 之基础知识
  • trito 内核开发基础 1
  • trito 内核开发基础 2
  • trito 内核开发基础 3
  • trito 内核开发基础 4
  • trito 内核开发基础 5

4.2 cuda 笔记

GPU 内核开发基础:

  • 英伟达 GPU 架构总结
  • 英伟达GPU通信和多卡拓扑结构
  • [英伟达 GPU 性能分析指导](./4-hpc_basic/英伟达

For agents

This page has a .md twin and JSON over the API.

Was this helpful?

Anonymous feedback helps us improve pages and translations.