Deep Learning Drizzle: a curated lecture index, not a course
Drench yourself in Deep Learning, Reinforcement Learning, Machine Learning, Computer Vision, and NLP by learning from these exciting lectures!!
At a glance
- What is it?
- kmario23/deep-learning-drizzle is a link hub for deep learning, reinforcement learning, computer vision and NLP lectures, built as HTML and maintained by hand. It is useful when you already know what you want to learn, and thin when you do not.
- Who is it for?
- Adopt it if you want a broad, manually curated index of lecture links and you are willing to judge each one yourself, or if you want to fork the README and build your own reading list on top of it. Do not adopt it if you need a structured curriculum with prerequisites and exercises, because the repository is a link list with no such scaffolding.
- Can I use it commercially?
- Not without permission. GitHub finds no licence file in the repository, and without a licence all rights are reserved by default: you may read the code but not reuse it. Check the README, or ask the authors, before using it.
- Is it still maintained?
- Yes. The repository last received commits 39 days ago.
- What is it written in?
- Mainly HTML, according to GitHub's language statistics.
Answers come from the project's GitHub data, last synced on September 28, 2026, and from our analysis. They are not legal advice.
Editorial analysis
What Deep Learning Drizzle actually is
The repository describes itself as a place to "Drench yourself in Deep Learning, Reinforcement Learning, Machine Learning, Computer Vision, and NLP by learning from these exciting lectures!!" That sentence is the product. There is no library, no CLI, no training code and no dataset. The top-level layout is two entries: README.md and a markdown2html_py/ directory. The README is the content, and markdown2html_py/ is the tooling that turns it into the site at deep-learning-drizzle.github.io.
The audience is self-taught engineers and students who already know which subfield they are chasing. The README's contents table splits into Deep Learning (Deep Neural Networks), Machine Learning Fundamentals, Optimization for Machine Learning, General Machine Learning, Reinforcement Learning, Bayesian Deep Learning, Graph Neural Networks, Probabilistic Graphical Models, Natural Language Processing, Automatic Speech Recognition, Modern Computer Vision, Boot Camps or Summer Schools, Medical Imaging, and a Bird's-eye view of Artificial Intelligence. Fourteen buckets. Each one holds links to lectures, and the repository's topics list names the same ground: graph neural networks, geometric deep learning, explainable AI, speech recognition, medical imaging.
Nothing here teaches you. It points. The value is that a human decided these links belong together under those headings, and the cost is that the human's criteria are not written down anywhere in the README.
The data flow: one README, one HTML generator
The mechanism is a two-step pipeline. Markdown goes in, static HTML comes out. The README is the single source of truth, and markdown2html_py/ is the converter that produces the published site. If you fork the repository and edit README.md, you own the downstream regeneration yourself, because the README does not document a build command, a CI workflow, or a publishing step.
The README's own formatting shows how manual the process is. Section dividers are long runs of the :heavy_minus_sign: emoji, and headings carry emoji prefixes such as :tada: and :confetti_ball:. That is a presentation choice baked into the source text, which means the converter has to tolerate emoji-heavy Markdown rather than parse clean structure. It also means the contents table at the top is maintained by hand: each cell is a Markdown link whose anchor points at a heading further down the file. Rename a heading and the anchor breaks unless you update the table too.
There is one release in the repository, v19.11-lw, dated 2019-11-13 and labelled "A first release". The last push was on 2026-08-22. So the shape of the project has been stable for years while the link list continues to be touched. That is consistent with a link index: the code does not need to change, the content does.
Getting the lecture index onto your machine
There is no package to install. The project is consumed either as the hosted site or as a clone of the repository. The README gives no install instructions, no supported Python version and no dependency list for markdown2html_py/, so treat the clone as a source checkout rather than a tool you run out of the box.
To get the content locally, clone the default branch:
git clone https://github.com/kmario23/deep-learning-drizzle.git
cd deep-learning-drizzleYou should see README.md and markdown2html_py/ at the top level. The README is the file you actually read; open it in any Markdown viewer, and the contents table becomes clickable anchors.
If you want the rendered site instead, the project publishes it at its homepage. There is no documented self-hosting path, so the practical first use is: read the README, pick a section, follow the links. A typical starting point is the Deep Learning (Deep Neural Networks) section, which is the first entry in the contents table.
Where the index breaks down
Link rot is the structural failure mode. Every entry is an outbound URL to a lecture page, a video, or a course site. The repository controls none of them. A section can look healthy in the README and still send you to a dead page, and the README offers no status field, no last-checked date and no archive fallback.
The second limitation is the missing licence. The repository metadata does not state one, and the README does not discuss terms of use. If you plan to reuse the curated list, republish it, or fold it into your own material, the repository does not tell you what is permitted. That is a real constraint for anyone treating this as a dataset rather than a reading list.
The third is that it is the wrong tool for structured learning. There are no prerequisites, no ordering within a section beyond whatever order the links were added in, no exercises and no assessment. If you need a path with a defined start and end, this repository will hand you fourteen doors and no map. It is also the wrong tool if you want runnable code, since the repository contains none.
How it differs from an awesome-list
The obvious comparison is an awesome-style link list, such as the kind collected under names like Awesome-deep-learning. The difference is not the format, which is similar, but the selection rule. A general awesome list tends to accumulate anything that fits the topic, and its sections are broad. Deep Learning Drizzle is narrower in what it admits: the README's headings map to taught material, with a Boot Camps or Summer Schools section and a Bird's-eye view of Artificial Intelligence section that a generic list would not bother to separate.
The practical consequence is coverage versus precision. A large awesome list will almost certainly contain a link for any given subtopic, but you filter it yourself. Drizzle's fourteen sections are opinionated about which subtopics matter, and within a section you get fewer entries. If your topic is not one of the fourteen, the repository has nothing for you. If it is, you spend less time triaging.
The other difference is that Drizzle ships a generator. markdown2html_py/ exists because the README is meant to be published as a site, which is why the source text carries emoji dividers and anchor-linked tables. An awesome list rarely cares whether its Markdown renders well anywhere other than GitHub.
Maintenance cost and what a fork commits you to
The last push was on 2026-08-22, so the repository has been touched recently, but the single release is from 2019-11-13 and is labelled "A first release". There is no versioning scheme in play beyond that tag, which means upgrading is not a concept here. You either track the default branch or you pin a commit. Nothing in the repository describes a changelog, a deprecation policy or a migration path.
For a fork, the cost sits in two places. First, link maintenance: you inherit every outbound URL and every dead one. Second, the generator: markdown2html_py/ has no documented dependencies or usage, so if you change the README's structure, including its heading anchors, you are responsible for confirming the HTML output still matches. Budget for reading the generator's source before you trust it.
On licensing, the repository does not declare one. The README is silent on terms. That does not prevent you from reading the links, but it does leave the status of redistributing the curated list unresolved, and it is the first thing to settle before you build anything public on top of it.
Editorial conclusion
Adopt it if you want a broad, manually curated index of lecture links and you are willing to judge each one yourself, or if you want to fork the README and build your own reading list on top of it. Do not adopt it if you need a structured curriculum with prerequisites and exercises, because the repository is a link list with no such scaffolding. Before relying on it, check the README's section list against your topic, confirm each linked lecture still resolves, and note that the repository declares no licence, so the terms for reusing its content are not stated anywhere in the repository.
Frequently asked questions
What are the three types of deep learning?
The repository does not classify deep learning into three types. Its README instead organizes material into fourteen topic sections, starting with Deep Learning (Deep Neural Networks) and continuing through Machine Learning Fundamentals, Reinforcement Learning, Natural Language Processing, Automatic Speech Recognition and Modern Computer Vision.
Why is DL used?
The README does not argue for deep learning's use cases. It opens with a quotation attributed to Prof. Geoffrey Hinton of the University of Toronto: "Read enough so you start developing intuitions and then trust your intuitions and go for it!" The rest of the file is a list of lectures rather than a rationale.
How do you explain DNN to a beginner?
The repository does not explain DNNs itself. It lists lectures under a Deep Learning (Deep Neural Networks) heading, which is the first entry in the README's contents table, and leaves the explanation to the linked material.
Who is considered the father of deep learning?
The repository does not answer this. It does quote Prof. Geoffrey Hinton of the University of Toronto at the top of the README, but it attaches no such title to him or to anyone else.
Official sources
Add this badge to your README
If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.
[](https://hysenlabs.com/projects/kmario23-deep-learning-drizzle)