Free From Google_ Generative AI Studying Path

Need to preserve up to date about Generative AI? Examine these free programs and sources from Google Cloud.

Illustration by Creator

Are you curious about discovering the potential of Generative AI Fashions and their functions? Fortunately, Google Cloud launched the Generative AI Studying Path, an ideal assortment of free programs, that begin from explaining the essential ideas of Generative AI to extra subtle instruments, like Generative AI Studio to construct your personalized generative AI fashions.

This text will discover seven of the out there programs, that may let you perceive the ideas behind the Massive Language Fashions that encompass us daily and create new AI options. Let’s get began!

1. Introduction to Generative AI

Course hyperlink: Introduction to Generative AI

This primary course is an introduction to Generative AI by Dr Gwendolyn Stripling, an AI Technical Curriculum Developer at Google Cloud. It would make you study what Generative AI is and the way it’s utilized. It begins with the essential ideas of information science (AI, Machine Studying, Deep Studying), and the way Generative AI is completely different from these disciplines. Furthermore, it explains the important thing ideas that encompass Generative AI with very intuitive illustrations, similar to transformers, hallucinations and Massive Language Fashions.

Video length: 22 minutes

Lecturer: Gwendolyn Stripling

Urged readings:

2. Introduction to Massive Language Fashions

Course hyperlink: Introduction to Massive Language Fashions

This second course goals to introduce what Language Fashions are at a excessive degree. Particularly, it provides examples of LLM functions, like textual content classification, query answering and doc summarization. In the long run, it reveals the potential of Google’s Generative AI Improvement instruments to construct your functions with no code.

Video length: quarter-hour

Lecturer: John Ewald

Urged readings:

3. Introduction to Picture Era

Course hyperlink: Introduction to Picture Era

This third course focuses on explaining crucial diffusion fashions, a household of fashions that generate pictures. A few of the most promising approaches are Variational Autoencoders, Generative Adversarial Fashions and Autoregressive Fashions.

It additionally reveals the use circumstances that may be categorized into two varieties: unconditioned era and conditioned era. The primary consists of human face synthesis and super-resolution as functions, whereas examples of conditioned era are the era of pictures from a textual content immediate, the picture inpainting and the text-guided image-to-image.

Video length: 9 minutes

Lecturer: Kyle Steckler

4. Consideration Mechanism

Course hyperlink: Consideration Mechanism

On this brief course, you’ll study extra in regards to the consideration mechanism, which is an important idea behind transformers and Massive Language Fashions. It has enabled bettering duties, similar to machine translation, textual content summarization and query answering. Particularly, it reveals how the eye mechanism works to unravel the machine translation.

Video length: 5 minutes

Lecturer: Sanjana Reddy

5. Transformer Fashions and BERT Mannequin

Course hyperlink: Transformer Fashions and BERT Mannequin

This course covers transformer structure, which is an underlying idea behind the BERT mannequin. After explaining the transformer, it provides an summary of BERT and the way it’s utilized to unravel completely different duties, similar to single-sentence classification and query answering.

In a different way from the earlier programs, the idea is accompanied by a laboratory, which requires prior information of Python and TensorFlow.

Video length: 22 minutes

Lecturer: Sanjana Reddy

Urged readings:

6. Create Picture Captioning Fashions

Course hyperlink: Create Picture Captioning Fashions

This course goals to clarify the Picture Captioning Fashions, that are generative fashions that produce textual content captions by taking pictures as enter. It exploits an encoder-decoder construction, consideration mechanism and transformer to unravel the duty of predicting a caption for a given picture. Just like the earlier course, there’s additionally a laboratory to place the idea into follow. It’s once more oriented towards information professionals with prior information in Python and Tensorflow.

Video length: 29 minutes

Lecturer: Takumi Ohyama

7. Introduction to Generative AI Studio

Course hyperlink: Introduction to Generative AI Studio

This final course introduces and explores Generative AI Studio. It begins by explaining once more what Generative AI is and its use circumstances, similar to code era, data extraction and digital help. After giving an summary of those core ideas, Google Cloud reveals the instruments that assistance on fixing Generative AI duties even with out an AI background. Certainly one of these instruments is Vertex AI, which is a platform that allows to handle the machine studying cycle, from constructing to the deployment of the machine mannequin. This end-to-end platform consists of two merchandise, Generative AI Studio and Mannequin Backyard. The course is targeted on explaining Generative AI Studio, which permits to construct simply generative fashions with no code or low code.

Video length: quarter-hour

Urged readings:

Last Ideas

I hope that you’ve discovered helpful this quick overview of the Generative AI course supplied by Google Cloud. If you happen to don’t know the place to begin in understanding the core ideas of Generative AI, this path covers each facet. In case you might have already a machine studying background, there are absolutely fashions and use circumstances you may uncover from one in every of these programs. Have you learnt different free programs about Generative AI? Drop them within the feedback if in case you have insightful ideas.

Eugenia Anello is presently a analysis fellow on the Division of Data Engineering of the College of Padova, Italy. Her analysis undertaking is targeted on Continuous Studying mixed with Anomaly Detection.