# AccumulateMore/CV: A Chinese Deep Learning Notebook Course > AccumulateMore/CV is a Jupyter Notebook repository of Chinese-language deep learning notes keyed to four video courses. It is a study companion, not a library, and the README spends more space on job referrals than on setup. **AccumulateMore/CV** — ✅（已完结）超级全面的 深度学习 笔记【土堆 Pytorch】【李沐 动手学深度学习】【吴恩达 深度学习】【大飞 大模型Agent】 - Repository: https://github.com/AccumulateMore/CV - Stars: 23,737 · Forks: 2,640 - Language: Jupyter Notebook - License: not declared - Published: 2026-09-09 · Updated: 2026-09-09 · Language: en - Canonical page: https://hysenlabs.com/projects/accumulatemore-cv ## What AccumulateMore/CV actually is This repository is a set of Jupyter notebooks, not a package. The README describes it as 超级全面的 深度学习 笔记, a broad set of deep learning notes covering CV, NLP, large models and agents. The top-level entries are numbered files such as 101_Pytorch安装.ipynb, 119_完整模型训练套路.ipynb and 234_实战Kaggle比赛图像分类CIFAR10.ipynb. The numbering is the map: 100 to 122 follows a Pytorch series by 土堆, 200 to 268 follows 李沐's deep learning course, 300 to 354 follows 吴恩达, and 400 to 409 covers large model agents by 大飞. A further 500 series is listed as a placeholder for next year. The audience is the self-taught learner who watches the videos and wants written notes beside them. That is a narrower audience than the topic list suggests. The topics include rag, llm, agents, computer-vision and nlp, but those are labels on a notebook collection, not capabilities of a tool. Nothing here installs. Nothing here exposes an API. If you arrive expecting a framework, you are in the wrong repository. ## The numbered notebook tree and how the notes are organised The organisation is by lecture, not by topic. A reader who wants convolutions does not search a module index; they open 109_卷积原理.ipynb, then 110_卷积层.ipynb, then 111_最大池化层.ipynb, following the order the video presents them. Later files move from single layers to assembled models: 114_搭建小实战和Sequential使用.ipynb, then 119_完整模型训练套路.ipynb, then 121_完整模型验证套路.ipynb. The 200 series repeats the same arc at a lower level, starting from 202_数据操作、数据预处理.ipynb and 203_线性代数.ipynb before reaching 220_经典神经网络LeNet.ipynb and 226_残差神经网络ResNet.ipynb. This layout has a real cost. There is no index file, no dependency graph and no searchable API surface, so the only way to know what a notebook contains is to open it. The README acknowledges one rendering problem directly: images and formulas can display incompletely on the GitHub website because the site does not parse them well, and the notes are meant to be downloaded and viewed locally. That is a maintenance burden pushed onto the reader, and it is the first thing to verify before you invest time. ## Opening a notebook locally: install and first use The README does not give an install command. It gives three viewing instructions instead: download the repository, open the notebooks with Anaconda's Jupyter Notebook rather than PyCharm's notebook view, and install a table-of-contents extension so you can jump between chapters. The dataset shared by the four instructors is distributed through a Baidu Pan link with the extraction code ppmu, and the README says to contact the author by WeChat if that link stops working. Start by cloning the repository, then launch the notebook server from the directory that contains the numbered files. The command below is the standard Jupyter entry point; the README does not name a specific version, so use whatever your Anaconda installation provides. ## Where the repository stops being useful The 500 series is openly unfinished. The README lists a 大模型Agent video by an unnamed instructor, marks the notes as 500 to 5XX, says they are expected next year, and fills the body with a repeated placeholder string. Anyone who needs agent material today should read the 400 series and stop there, because the rest does not exist yet. The deeper limitation is that this is a notes repository with no release and no stated licence. The licence field is empty, which matters if you intend to reuse the notebooks inside a course, a product or a company training programme. An empty licence field is not permission. The README also mixes study material with recruitment content: WeChat groups, a referral table of employers, resume review and paid guidance offers. That is the author's business, but it means the repository README is not a reliable place to look for technical scope. Read the notebook filenames instead. A third constraint is language and format. Every notebook title is in Chinese, and the material is tied to specific video series. If you have not watched 李沐's course, 213_Kaggle房价预测.ipynb will read as disconnected fragments rather than a lesson. ## Notes versus a runnable framework The closest alternative in kind is a book such as Dive into Deep Learning, which is also notebook-based and also follows a lecture-style progression, but it is a self-contained textbook with its own prose, exercises and a stated licence. This repository is a companion to someone else's videos, so the notebook assumes the lecture supplies the explanation. That difference decides the use case: Dive into Deep Learning can be read alone, AccumulateMore/CV is meant to be read alongside a playlist. If what you actually need is a library, the comparison is even sharper. PyTorch itself installs, versions and publishes release notes; this repository does none of those. The notebooks call PyTorch, they do not replace it. Choosing between them is not a judgement call about quality, it is a question of whether you want code you import or notes you read. ## Maintenance, licence and what an upgrade costs The repository is not archived, and the last push was on 2026-06-30. There are no releases, so there is no version to pin and no changelog to read. Updating means pulling the branch and accepting whatever changed in the notebooks since your last pull, which for a study repository is usually harmless and occasionally disruptive if you had annotated a file locally. Because there is no release process, the practical upgrade cost is your own merge conflicts. Keep your annotations in separate files or in a branch you rebase, and the cost stays near zero. The licence question is the one that does not resolve itself: with no licence identifier in the repository metadata, redistribution, classroom use and commercial reuse are all unclear, and that is a question for the author or for someone qualified to advise, not something to assume from the README's tone. ## Conclusion Adopt it if you are a Chinese-speaking self-learner following the 土堆, 李沐, 吴恩达 or 大飞 video series and you want one numbered notebook tree that tracks the lectures. Do not adopt it if you need an installable package, English material, a stated license, or a maintained codebase: the last push was on 2026-06-30, there are no releases, and the licence field is empty. Before committing, open one notebook locally in Anaconda's Jupyter Notebook with a table-of-contents extension and confirm the images and formulas render, then check whether the 400-series agent notes are complete or still being written. ## FAQ ### How do I install and open the AccumulateMore/CV notebooks? There is no install step for the repository itself. Clone it, then open the notebooks with Anaconda's Jupyter Notebook; the README specifically says PyCharm's notebook view does not display the images correctly, and recommends a table-of-contents extension for navigation. ### Is AccumulateMore/CV a library I can import into my own project? No. The top-level entries are numbered .ipynb files such as 101_Pytorch安装.ipynb and 119_完整模型训练套路.ipynb. It is a set of study notes that use PyTorch, not a package that provides an API. ### Why do images and formulas look broken in AccumulateMore/CV on GitHub? The README states that GitHub does not parse them well and that the notes display correctly once downloaded and opened locally. That is why the README recommends viewing them through a local Jupyter Notebook server rather than the GitHub file preview. ### What licence does AccumulateMore/CV use? The repository metadata does not list a licence identifier, and the README does not state one. Treat reuse, redistribution and commercial use as unresolved until the author confirms terms. ## Sources - [AccumulateMore/CV on GitHub](https://github.com/AccumulateMore/CV) - [Issues](https://github.com/AccumulateMore/CV/issues) - [README](https://github.com/AccumulateMore/CV/blob/main/README.md) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/accumulatemore-cv