Large Language Models (LLMs) are AI systems trained on massive text datasets to predict and generate language, and they now power applications such as chatbots, summarization, semantic search, code generation, and question answering. To learn LLMs well, the best free courses combine three things: conceptual foundations like tokens, transformers, and attention; practical skills such as prompting, fine-tuning, and evaluation; and hands-on tooling with ecosystems like Hugging Face, Python notebooks, and modern cloud or open-source workflows.
Prerequisites: For the beginner courses, no prerequisites are required. For the intermediate courses, basic Python is strongly recommended, and some familiarity with machine learning or deep learning concepts will help a lot, especially for Hugging Face, Coursera’s LLM course, and DeepLearning.AI’s fine-tuning course.
Estimated time: Approximately 26-36 hours total
Start with Google for Developers and Microsoft Learn to build a quick conceptual foundation around tokens, prompts, transformers, and where LLMs fit in practice. Next, take the Hugging Face LLM Course to develop hands-on skills with real tools and workflows. After that, study Coursera’s Generative AI with Large Language Models for a more structured end-to-end view of training, fine-tuning, evaluation, and deployment. Finish with the two DeepLearning.AI short courses: first Finetuning Large Language Models to specialize in adaptation methods, then Large Language Models with Semantic Search to learn how embeddings and retrieval power modern LLM applications such as semantic search and RAG systems.
A concise, high-quality introduction to how language models work, including tokens, context, self-attention, transformers, fine-tuning, and distillation. This is best for learners who want a fast but technically grounded overview before moving into larger hands-on courses.
Topics: LLM fundamentals, transformers, self-attention, fine-tuning
Go to course →This beginner-friendly module explains what LLMs are, what they can and cannot do, and core concepts like prompts, tokens, and completions. It is especially good for learners who want a practical product-and-developer view with no prerequisites.
Topics: LLM basics, prompts, tokens, model selection
Go to course →A comprehensive free course that teaches LLMs and NLP through the Hugging Face ecosystem, including Transformers, Datasets, Tokenizers, Accelerate, fine-tuning, and working with the Hub. It is one of the strongest choices for learners who want both theory and implementation skills using industry-standard open-source tools.
Topics: transformers, hugging face, fine-tuning, tokenization
Go to course →A highly regarded course covering the full LLM lifecycle: transformer architecture, training, fine-tuning, evaluation, scaling laws, RLHF, deployment, and applications. It is ideal for learners who already have some ML background and want a structured, industry-relevant course with a free audit option.
Topics: generative AI, transformers, fine-tuning, RLHF
Go to course →A focused short course on when and how to fine-tune LLMs, including data preparation, training, evaluation, and the tradeoffs versus prompt engineering and RAG. This is best for intermediate learners who want practical specialization after learning LLM basics.
Topics: fine-tuning, data preparation, evaluation, prompt engineering
Go to course →This course teaches how LLMs and embeddings improve search through dense retrieval, reranking, and answer generation. It is an excellent practical next step for learners interested in retrieval, RAG-style systems, and real-world LLM applications.
Topics: semantic search, embeddings, dense retrieval, reranking
Go to course →Enter your email and we'll generate your personalized free-course roadmap as a PDF you can keep.
Describe any AI topic and our engine will scan the web to build a personalized free course roadmap you can download.
Build my roadmap →