Developers who can code with large language models (LLMs) and AI are in high demand on the job market. And it’s not hard to get started, because there are numerous courses that can help you expand your skills. Here’s an always up-to-date overview of the most important online courses on LLMs, AI, chatbots, Python and more.
Overview: What skills do you need when programming with large language models?
To code with large language models (LLMs) like OpenAI’s GPT, Google Bard, or open source LLMs, the following skills are helpful:
- Know LLM frameworks (LangChain, HuggingFace, Haystack, Semantic Kernel)
- Understand AI and LLM fundamentals (transformer architecture, embeddings, neural networks)
- Master Python (Python syntax, data types, importing libraries, working with lists and DataFrames)
- Learn Microsoft Azure, to deploy GPT in the secure cloud for companies, with high standards for data privacy, e.g. GDPR
- Follow AI news (latest trends, important tools, industry insights, AI jobs)
What should you start with? A developer’s tip for LLM beginners
From my years of experience as a Java and Python developer, I would like to give a perhaps unexpected tip to fresh LLM developers: start directly with an LLM framework. Instead of learning Python or AI basics from scratch, it’s much faster for many to start directly with LLM framework courses and learn Python and the basics on the side. Because there are really good free courses for this that will accelerate your learning curve significantly.
Online courses: LLM frameworks
Frameworks simplify LLM development enormously. They allow you to connect to common LLMs, tokenize documents, connect to vector databases, script complete workflows and more with just a few lines of code.
Currently, the most important LLM frameworks are:
- LangChain
- HuggingFace (very important: the Transformers library)
- Haystack
- Semantic Kernel (from Microsoft)
Among the best online courses are the following.
ChatGPT Prompt Engineering for Developers
- Provider: DeepLearning.AI
- Contents: OpenAI connectivity basics, useful prompt use cases, avoiding hallucinations
- Format: Videos, Coding in Jupyter Notebook
- Duration: 1h
- Language: english
- Cost: free of charge
Go to the course: ChatGPT Prompt Engineering for Developers
LangChain: Chat with your Data
- Provider: DeepLearning.AI
- Content: Building a chatbot using OpenAI’s GPT model via the LangChain framework, valuable but brief explanations of important concepts such as SelfQueryRetriever, Maximum Marginal Relevance
- Format: Videos, Coding in Jupyter Notebook
- Duration: 1h
- Language: English
- Cost: free of charge
To the course: Langchain: Chat with your Data
LangChain Complete Course
- Provider: Coding Crash Courses / YouTube
- Contents: Well explained tutorial on Python coding with GPT and LangChain framework. Contents: Prompt templates, output parser, chains, memory, vector databases, agents, model evaluation
- Format: Youtube-Video
- Duration: 1h
- Language: german
- Costs: free of charge
To the course: LangChain complete course (Youtube)
Online courses: LLM and AI basics
LLMs are only one of many AI-based methods for predicting outcomes. If you want to know more about the background, you should take a look at LLM Basics, as well as Data Science and Deep Learning.
Generative AI with Large Language Models
- Provider: Coursera / DeepLearning.AI
- Content: Deep background on the concepts of generative AI, large language models and neural networks
- Format: Videos, labs with coding exercises
- Duration: 16h
- Language: English
- Cost: free of charge
Go to course: Generative AI with Large Language Models (Coursera)
iX Special 07/2023: Artificial Intelligence
- Provider: iX / heise Verlag
- Contents: The special issue on AI offers in-depth articles, from Transformer architecture to the use of AI frameworks such as LangChain and Haystack, as well as ethical and legal articles.
- Format: Booklet or online
- Cost: €14.90 (print or digital)
Tip: Also note our savings tip on heise in the “AI News” section.
To the issue: iX Special: Artificial Intelligence
Neural Networks and Deep Learning
- Provider: Coursera / DeepLearning.AI
- Contents: Concepts of the basic technology underlying language models, “neural networks”
- Format: Videos, Labs with coding exercises
- Duration: 16h
- Language: English
- Cost: free of charge
About the course: Neural Networks and Deep Learning
Video: Illustrated Guide to Transformers
- Provider: Michael Phi / towarddatascience.com
- Content: A good, visually illustrated explanation of the Transformer model (encoder, decoder, attention, and more) pioneered for LLMs
- Format: Video, Article
- Duration: 15 minutes
- Language: English
- Cost: free of charge
Go to article & video: Illustrated Guide to Transformers
Generative AI learning path
- Provider: Google / Google Cloud Skills Boost
- Content: Over 10 courses with easy-to-understand illustrated LLM basics and introduction to Google tools such as Generative AI Studio for LLM development, including transformers and encoder/decoder architecture, Google’s BERT model, image generation
- Format: videos, articles
- Duration: approx. 7 days
- Language: English
- Cost: free of charge
Go to the course: Generative AI learning path
Online courses: Learn Python
The Python language is experiencing tremendous hype due to AI. This is due to the fact that Python can be used well and quickly in data analysis, does not need to be compiled, has a high prevalence in scientific computing, and is still easy to learn.
Datacamp – Learning Data Science with Python and R
- Provider: Datacamp
- Contents: Python and R, data science, data engineering, deep learning, statistics, data analysis, Power BI, Tableau
- Format: Videos, many coding exercises, quiz format, own community, job portal for Data Science jobs and more
- Duration: single courses a 4 hours, total job certificates: several months
- Language: English
- Costs: 23 € / month (often 50% discount), first lessons of all courses are free of charge
To the online course platform: Datacamp
To the course: Working with the OpenAI API
Online Courses: Learn Azure
Companies need security and privacy for their applications. As OpenAI and Microsoft have a partnership, OpenAI models can be run in the Azure cloud, which is the first choice for cloud applications for many companies. Therefore, it’s important to learn some Azure basics and include them in your repertoire.
Microsoft Learn: Thousands of free courses on Azure, AI, and more
- Provider: Microsoft
- Contents: Basic Azure setup, in-depth coverage of all Azure modules. Important for AI topics are:
- Azure OpenAI Service (for GPT users)
- Azure Machine Learning Services (for professionals)
- Azure AI Services (for developers without AI experience)
- Format: Primarily text-based with relatively fewer videos. There are practical lab exercises where tasks must be completed directly in Azure.
- Duration: There are many Learning Paths which bundle courses and learning units. Additionally, one can complete professional certificates (Learning effort: several months).
- Language: English, many others (often automatically translated).
- Cost: The courses are free. A certificate attempt costs 100 $.
Learning Path: Azure Fundamentals (good for basic Azure knowledge)
Learning Path: Develop Generative AI solutions with Azure OpenAI Service (to use GPT on Azure)
News about AI and LLMs
If you want to keep up with the rapid pace of language model and AI development, you need good sources. That way, you can keep up with what tools and frameworks are important, how AI is permeating all aspects of life and work, and never miss an exciting AI job opportunity.
Podcast: AI Update
The podcast KI-Update is published by Heise-Verlag in collaboration with the news website “The Decoder” and delivers an interesting mix of news as well as deeper insights into selected AI topics several times a week.
heise subscription
The heise subscription offers you all magazines of the Heise publishing house as well as exclusive paywall news articles. For only 10€ / month you get a lot of relevant and well prepared information for little money. With the heise subscription you can read all issues of the magazines c’t, iX, Make:, Mac & i, MIT Technology Review and c’t Fotografie online – including all special issues and the issue archive of the last years!
To heise.de – Section: AI News
