It was only six months ago when OpenAI released ChatGPT, the mind-blowing player in the large language model (LLM) ecosystem.
ChatGPT definitely changed people’s minds on using LLMs. Approximately two months after it was released, ChatGPT became the fastest growing consumer software application, gaining more than 100 million users.
It is safe to say that technology on AI evolves with LLMs. Over 1000 plugins that allow for a more efficient and practical use of LLMs have been created by developers. LangChain, an open-source framework for developing applications with LLMs has almost 50k stars on GitHub (as of writing this article).
These are great advancements and making the AI technology embraced by developers. However, at the same time, it becomes more difficult to follow the latest trends in AI.
Thankfully, there is no shortage of learning materials, which brings us the main topic of this article. DeepLearning.AI releases 3 free courses, which focus on the latest technology in the field of AI. In this article, we will talk about what they cover in detail and how to make the most out of them.
It is important to note that they’re free as of writing this article but may not be in the future.
The courses are short and you can finish them in about an hour. However, I strongly recommend spending more time on them. The code explained in the courses are shared via a notebook. It’s best to go through the code on your own, execute it, and tweak it to see how the output changes.
1. ChatGPT Prompt Engineering for Developers
This course, created by OpenAI and DeepLearning.AI, focuses on prompt engineering.
Prompts are the single most important input for an LLM. The output you get from a model largely depends on the prompt. Hence, you need to write clear, concise, and well-structured prompts to get the desired output.
This course will teach you how to customize prompts according to two main principles:
- Write clear and specific instructions
- Give the model time to "think"
The instructors, Isa Fulford and Andrew Ng, go through the examples that showcase how prompt engineering impacts the performance of an LLM’s output.
The most common use cases of LLM are covered in the course, which are:
- Summarizing
- Inferring
- Transforming
- Expanding
The course ends with a comprehensive example to create a chatbot.
2. LangChain for LLM Application Development
LangChain is an open-source framework for developing applications with LLMs. It allows for enhancing the capabilities of LLMs by chaining multiple components such as LLMs, agents, memory, indexes, prompt templates.
LangChain, created by Harrison Chase not long ago, already has close to 50k stars on GitHub, which is a strong indication of its success. Developers have already started building LLM-based applications with LangChain.
One of the cool things about this course is that it was co-created and presented by Harrison Chase, the creator of the framework.
Together with Andrew NG, they clearly explain the idea behind LangChain and go through examples to demonstrate how it can be used.
By the time you finish this course, you’ll have learned the main modules, which are the building blocks of chains. You’ll also be able to create chains using these modules.
This course is a great step towards creating your own LLM-based application.
3. Building Systems with the ChatGPT API
Created by OpenAI and DeepLearning.AI, this course offers a great material for learning best practices for building complex applications with LLMs.
The focus is on creating an entire system (a customer service assistant), not just querying ChatGPT with a single prompt.
It starts off with an introduction on language models and the important things to know about them such as tokens and chat formats.
You’ll also learn about very important concepts in working with LLMs such as chain of thought (CoT) reasoning and chaining prompts.
The evaluation of the output of an LLM is not a straightforward process as there is not only one correct answer in many cases. This course also covers techniques and best practices for evaluating model outputs.
Final words
These courses will not make you an LLM expert but they will guide towards an efficient learning path. They are well-structured and created by the experts in the field so any minute you spend on these courses is worth it.
After you complete them, keep working and practicing to discover the full potential of LLMs. I guarantee that you’ll be surprised.
You can become a Medium member to unlock full access to my writing, plus the rest of Medium. If you already are, don’t forget to subscribe if you’d like to get an email whenever I publish a new article.
Thank you for reading. Please let me know if you have any feedback.