top of page

ALGORYTHM | How to Build a Generative AI


“There’s no one thing that defines AI. It’s more like a tapestry of modern intelligent technologies knit together in a strategic fashion that can then uplift and create a knowledge base that is automated — where you can extrapolate findings from there.”John Frémont, founder and chief strategy officer, Hypergiant

Get your cooking gears on, we are about to make some magic happen. 🍳


🥗 Generative Adversarial Networks (GANs)

GANs are a type of machine learning model that consists of two neural networks that compete against each other. One network, the generator, tries to create realistic data, while the other network, the discriminator, tries to distinguish between real data and generated data. GANs have been used to generate realistic images, text, and music.


Credit: https://hyeongminlee.github.io/post/gan001_gan/

🥗 Transformers

Transformers are a type of neural network that are particularly well-suited for natural language processing (NLP) tasks. Transformers have been used to train large language models (LLMs) that can generate text, translate languages, and answer questions in a comprehensive and informative way. Ahem! #ChatGPT

Credit: https://dzone.com/articles/a-deep-dive-into-the-transformer-architecture-the


🥗Large language models (LLMs)

LLMs are a type of neural network that has been trained on a massive dataset of text and code. LLMs can be used to generate text, translate languages, write different kinds of creative content, and answer questions in an informative way.





Yet, nothing can function without the following resources...👇

Compute power

Generative AI models can be very computationally expensive to train and generate outputs. This is because they need to learn the complex patterns and structure of the data they are trained on.


Data

Generative AI models need to be trained on large datasets of real-world data in order to generate realistic outputs.


Algorithms

Generative AI models use sophisticated algorithms to learn the patterns and structure of the data they are trained on.


Generative AI is a rapidly evolving field, and new techniques and methodologies are being developed all the time. As a result, the building blocks of generative AI are constantly evolving as well.


Check out these bangers on #generativeAI 🎉





Like what read? Join the gang :)


コメント


bottom of page