Source | Qubit | Public Account QbitAI
Family members, the AI course taught by Andrew Ng has finally been updated after a month!
This time, three courses were launched directly in one breath. The topics included LangChain, diffusion model, and building a system with ChatGPT API.
It is still free for a limited time . Except that the diffusion model requires a certain knowledge reserve, novices in the other two courses can also listen.
Some netizens didn't even sleep in order to finish these courses.
Course content
The duration of these three courses is about one hour, and they are all equipped with cases for in-depth understanding.
Among them, ChatGPT API and LangChain were explained by Wu Enda in cooperation with the developers.
They are Isa Fulford from OpenAI and Harrison Chase from the LangChain team.
The diffusion model course was taught by Wu Enda's student Sharon Zhou, and Wu made a recommendation.
Build a system with ChatGPT API
In this course, you will learn how to use the chain of thought to call LLM to automate complex workflows.
This course can be followed by novice to advanced players, but some basic knowledge of Python is required.
After taking this course, you'll master how hint chains are built and understand the system by which Python code interacts with hint content.
This course also starts from reality, combining all the theoretical knowledge taught, and introduces how to build your own customer service chatbot.
If you are interested in these contents, then this course is worth a look!
Portal:
https://www.deeplearning.ai/short-courses/building-systems-with-chatgpt/
Use LLM to develop applications through LangChain
LangChain is an open source tool for building applications through LLM.
Taking this course also requires some basic knowledge of Python.
The course first introduces basic concepts such as models, hints, and parsers, and then expands from the way models are memorized.
Then there is the core of LangChain, which is the logic chain.
After introducing these theoretical knowledge, the course introduces the specific development process by taking a program that generates questions and answers based on documents as an example.
In addition, the course also introduces how to evaluate the application and how to use some embedded LangChain tools.
After taking this course, you will be able to create powerful applications using LangChain within a few hours.
Why don't you try it soon?
Portal:
https://www.deeplearning.ai/short-courses/langchain-for-llm-application-development/
How Diffusion Models Work
The well-known painting tools Midjourney and Stable Diffusion are both driven by the diffusion model.
This course provides an in-depth explanation of diffusion processes and models, and is slightly more difficult than the previous two courses.
In addition to things like how it works and the training process, the course also covers how to build a diffusion model from scratch.
Combined with concrete examples, the course covers sampling, training diffusion models, building neural networks for noise prediction, and some practical coding skills.
With this class, you can learn about the latest knowledge on Diffusion Generative AI and create your own Diffusion models, even going beyond existing models and APIs.
Readers who are interested in diffusion models, don't miss it!
Portal:
https://www.deeplearning.ai/short-courses/how-diffusion-models-work/
Reference link:
https://twitter.com/AndrewYNg/status/1663984377918001153