# changyeyu/LLM-RL-Visualized

🌟100+ 原创 LLM / RL 原理图📚，《大模型算法》作者巨献！💥（100+  LLM/RL Algorithm Maps ）

Repository: https://github.com/changyeyu/LLM-RL-Visualized
Canonical: https://ross.abutalabs.com/products/llm-rl-visualized
Homepage: https://book.douban.com/subject/37331056/
Language: Python
License: NOASSERTION
License Family: other
Topics: llm, reinforcement-learning, algorithm, nlp-machine-learning, vlm, ai, deep-learning, machine-learning, natural-language-processing, transformers
Last push: 2026-07-27T16:18:19+00:00

## Health v2 (maintenance only)
Score: 62/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 94, release rhythm 35, longevity 35
- inputs: {"age_days": 494, "days_push": 37, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4819, forks 463 (observed 2026-08-28T04:09:00.485813+00:00)

## What it is
A curated collection of 100+ original SVG architecture diagrams explaining LLM, VLM, and reinforcement learning algorithms (RLHF, PPO, GRPO, DPO, SFT, LoRA, RAG, CoT), companion to the book 'Large Model Algorithms: Reinforcement Learning, Fine-Tuning and Alignment'. It is a visual learning reference rather than runnable software, with a small amount of Python tooling.

## Use cases
- understand how transformer-based LLMs work visually
- learn RLHF and DPO training pipelines
- study LoRA and SFT fine-tuning techniques
- prepare for LLM algorithm interviews
- grasp reinforcement learning fundamentals like PPO and GRPO
- learn RAG and decoding strategies like beam search and top-p sampling

## When to choose
- you want visual, diagram-based explanations of LLM and RL algorithms
- you are studying for AI algorithm interviews or reading the companion book
- you need high-quality SVG diagrams for teaching or presentations

## When to avoid
- you need runnable training code or production libraries
- you want executable notebooks with hands-on experiments
- you need English-only materials without reading Chinese content

## Facets
- artifact type: learning-resource
- maturity: active
- function: machine-learning, deep-learning, llm-training, rag, prompt-engineering, data-visualization
- domain: large-language-models, machine-learning, reinforcement-learning, tutorials, education
- platform: cross-platform
- tags: diagrams, rlhf, dpo, sft, lora, vlm, architecture-diagrams, svg, chinese, book-companion, natural-language-processing

## Member repositories
- changyeyu/LLM-RL-Visualized (main) score 62

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:00.485813+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:18:32.897860+00:00, confidence not recorded.
  - readme: https://github.com/changyeyu/LLM-RL-Visualized (fetched 2026-08-28T04:09:00.485813+00:00, sha e264253fabbb)
  - homepage: https://book.douban.com/subject/37331056/ (fetched 2026-08-29T09:01:43.085558+00:00, sha 5510acdbb527)
  - site_page: https://www.douban.com/about (fetched 2026-08-29T09:01:43.095350+00:00, sha 98feff5729f4)
  - site_page: https://www.douban.com/about?topic=contactus (fetched 2026-08-29T09:01:43.097476+00:00, sha d2b82a40f3b8)
  - site_page: https://www.douban.com/about/legal (fetched 2026-08-29T09:01:43.099280+00:00, sha 590567ae1951)
- Data as of 2026-08-30T08:39:29.467469+00:00.