 G'day everyone. Sorry it's been a long time between blog posts. Since my last post, I've been very busy, making babies....and slightly less exciting....I have published another blog post with AWS. This time it's an end-to-end demonstration of how you can get started with Generative AI on AWS. In fact, it's two end-to-end demonstrations, complete with prerequisites needed on how to get started with the AWS machine learning service: Amazon SageMaker. Here is a little video teaser showcasing my "baby tools", a tribute to May the Fourth be with you and an introduction to the blog post content. IThe blog post features 2 demonstrations using a tool called Amazon SageMaker Jumpstart. The first one shows how you can use AI to generate new text, using the GPT-2 model. This is a model that is similar to the one used in the infamous ChatGPT app, developed by OpenAI. However it does not natively have the capability to provide a chatbot like interface. Here is an architecture that will allow you to integrate ChatGPT if you're interested in that use case:
aws.amazon.com/blogs/compute/implementing-an-event-driven-serverless-story-generation-application-with-chatgpt-and-dall-e/ And for those are keen to get down and dirty with those demonstrations, the link to the blog post is below: aws.amazon.com/blogs/machine-learning/get-started-with-generative-ai-on-aws-using-amazon-sagemaker-jumpstart/ I hope you all have a fantastic day and let me know what you think of the new blog post. Take care, P #TheDigitalCoach
0 Comments
 Hi Beautiful People. Today I'm going to talk about a technique called Stable Diffusion, which is a technique that seems to occur time and time again in the field of Generative AI. Firstly Generative AI stands for Generative Artificial Intelligence and is a collection of AI techniques that allow you produce new things. This could be works of art, real-life pictures or the creation of new types of music. In particularly the latest tool that is doing the rounds is OpenAI's DALL-E, which uses stable diffusion and ChatGPT, which uses a transformer technique.. If you haven't heard of either of those, you must of been living under a rock, so feel free to learn more by experimentation in my previous blog post:
www.digitalcolmer.com/blog/-how-useful-is-generative-artificial-intelligence What is Stable Diffiusion? Stable diffusion is a mathematical process that describes how a something spreads out over time in a stable manner. It is a type of diffusion process in which the distribution of the quantity remains unchanged over time. In other words, the spread of the quantity is consistent and does not vary significantly over time. Stable diffusion processes are often used to model physical phenomena, such as the movement of particles in a fluid or the spread of a chemical through a substance. They can also be used to model the spread of information or ideas within a population. In Generative AI, stable diffusion is a deep learning, text-to-image model. Deep Learning is a technique that helps computers learn in a similar manner to the brain. It does this using a neural network and learns over time, very similar to how a human learns by experimentation over time. Text-to-image models are used in AI to help generate new images from a simple sentence, just like DALL-E above. Here is a great video from OpenAI, that explains very simply how DALL-E works: www.youtube.com/watch?v=qTgPSKKjfVg&t=2s How is it used in Generative AI? Stable diffusion processes can be used in generative AI in a variety of ways. One example is the use of a generative model that is trained to generate new, synthetic data that is similar to a training dataset. Generative models can be used to synthesize a wide variety of data, including images, text, and audio. One way that stable diffusion is used in a generative model is as a way to model the distribution of data in the training dataset. For example, if the training dataset consists of images of faces, a stable diffusion process could be used to model the distribution of features in the images, such as the shapes of the faces and the colors of the eyes. This could help the generative model to generate new images of faces that are realistic and consistent with the training data. Another way that stable diffusion can be used in generative AI is as a way to explore the space of possible data points that could be generated by a model. For example, a stable diffusion process could be used to generate a sequence of data points that are drawn from the distribution modeled by a generative model. This could be used to explore the range of outputs that the model is capable of producing, or to identify areas of the data space that are poorly represented in the training data. This could help generate music that is not random and is ordered according to a harmonic major or minor system. This would produce music that is pleasing to the ear and could be used in film, tv or pop music. Here is another great video that talks about Stable Diffusion. It starts off simple but then gets deeper and deeper. Simply click the stop button when you've had enough: www.youtube.com/watch?v=1CIpzeNxIhU I hope you enjoyed this post. Stay safe, and I'll catch you all soon. Paul Colmer #TheDigitalCoach  😍G'day Beautiful People. Thanks again for reading this blog post. Today I'm going to talk about and demonstrate the usefulness of some Generative AI tools. Many of you will know that I am a practicing musician. I have a music degree and I'm a fellow of the London College of Music. I spent 3-4 years of my career in film and TV music. So it's a real pleasure to start combining my tech knownledge, with music, by choosing to specialise in Generative AI. 😎
What is Generative AI? Generative Artificial Intelligence refers to artificial intelligence systems (known as AI) that are able to generate new content, such as text, images, or music, based on a set of examples. The examples that are used to train the system is known as training data. These systems use machine learning and deep learning\ techniques to learn the patterns and characteristics of the content they are trained on, and then use this knowledge to generate new, original content that is similar to the training data. Some examples of generative AI systems include text generators that can write news articles or generate social media posts, image generators that can create original images or photographs, and music generators that can compose new pieces of music. Perfect for speeding up the delivery of digital content and blog posts. 😂 But how good is Generative AI? Well, most of the first paragraph above, and much of this blog post has been generated using Generative AI. Feel free to compare what ChatGPT provided, versus how each section in this blog post is written. Clearly ChatGPT has a long way to go, to make blog posts engaging and interesting, so anything remotely funny, is probably my writing. 😎 I used a tool called ChatGPT from OpenAI. Check out the link here: chat.openai.com/chat Who is OpenAI? OpenAI is a research organization that focuses on the development of artificial intelligence (AI) technologies. It was founded in 2015 by a group of entrepreneurs and researchers, including Elon Musk, Sam Altman, Greg Brockman, and Ilya Sutskever. Everything that Elon Musk touches seems to turn to gold, so OpenAI is no exception. The goal of OpenAI is to advance the field of AI in a responsible and safe manner, and to ensure that the benefits of AI are widely and fairly shared. I think this is a very noble cause and I believe that all data scientists have a responsibility in the ethical and equitable use of AI amongst the human race. OpenAI conducts research in a variety of areas, including machine learning, robotics, economics, and computer science. It has developed a number of influential AI technologies, such as the GPT (Generative Pre-trained Transformer) language processing AI, and the DALL-E image generation AI. OpenAI also hosts conferences and events, and provides educational resources and tools for researchers and developers interested in AI. I'll talk about DALL-E further down the post, but let's first talk about ChatGPT! 🚀 Tell me more about ChatGPT? The GPT in ChatGTP, is short for "Generative Pre-trained Transformer". It is a type of language processing AI developed by a company called OpenAI. It is a large, deep neural network that is trained to generate natural language text that is similar to human writing. What is a Deep Neural Network? A deep neural network is a type of artificial neural network that is composed of many layers of interconnected nodes, or "neurons." These networks are called "deep" because they have many layers, as opposed to shallow networks that have only a few layers. Deep neural networks are designed to recognize patterns and relationships in data, and can be used for a wide range of tasks, such as image and speech recognition, natural language processing, and machine translation. They are particularly well-suited for tasks that require the learning and recognition of complex patterns, as the multiple layers of neurons allow the network to learn and represent these patterns in a hierarchical manner. Deep neural networks are trained using large datasets and algorithms that adjust the connections between neurons in order to minimize the error between the network's predictions and the true values in the training data. They are an important part of the field of deep learning, which has led to many significant advances in AI in recent years. What is a Generative Pre-Trained Transformer? Apart from being a mouthful, it is simply two things: Generative means that it produces content, rather than trying to predict something, which is a common use case for machine learning and deep learning models. Pre-Trained, means the model has been fed lots of great data. Bad data will produce poor results, so the latest version of ChatGPT has used humans in the training process to enhance the content. Transformer is a specific generative model that is used in natural language processing to help model and translate language, but used here to help generate new content based on a simple question. It is also used in music to extend melodies. Great example here: aws.amazon.com/blogs/machine-learning/using-transformers-to-create-music-in-aws-deepcomposer-music-studio/ Show me some more Generative AI examples? Check out some examples I have created used AWS DeepComposer here: soundcloud.com/morphosis And check out my blog post that I publised at Amazon Web Services here: aws.amazon.com/blogs/machine-learning/generate-a-jazz-rock-track-using-aws-deepcomposer-with-machine-learning/ Finally, 2 of the pictures at the top of this blog are created using another OpenAI product called Dall-E. Check it out here: openai.com/dall-e-2/ Have fun and try some experiements. Please feel free to share your creative projects with me on LinkedIn, Twitter and Instagram. Paul Colmer #TheDigitalCoach  I started working at AWS, as a AWS technical account manager for over 18 months starting back in 2020. In that time I have helped more than 12 customers be highly successful in adopting AWS services. All my customers are in the public sector and I have seen tremendous growth. This has been driven by the challenging environment that COVID-19 has created. The pandemic has required all types of organisations to be agile, scalable and resilient which in turn translates into processes, systems and technology. For the past 9 months, I have been working as a senior technical trainer. I have inspired, engaged and motivated many customers and passed on all my lifelong learnings. This has not been easy and has taken a tremendous amount of work and dedication. I wanted to share some fundamental practices that I have applied when I work with customers, that have helped me be successful at AWS, but more importantly have helped delight my customers.
I hope that helps you all on your agile journeys. May the force be with you. 🚀 P 23/12/2021 0 Comments Tesla Model 3 WindScreen Wiper Fix G'day beautiful people. It's almost Christmas. Sorry I haven't blogged for a while, but I've been super busy, doing incredible work in the cloud with AWS. But my Tesla Model 3 has inspired me to blog yet again, this time 1 day before Christmas Eve.
So...I have had a problem with my Tesla Model 3 windscreen wipers for about 3 weeks now. My partner was driving the car, and they just went berserk. Simple fix...turn them off. Well...that worked for the last 3 weeks because I haven't been out in the rain. But today...the heavens opened and I needed them. So after switching to Auto....they furiously wiped back and forth...even though to start with the rain was only spitting. I looked a quite a few forums and it is a common problem. Good thread here on the problem: teslamotorsclub.com/tmc/threads/wipers-dont-work-properly-in-auto.222547/ What is the fix? Make sure the centre front-facing windscreen camera is clean and so too must be your forward-facing side cameras. After cleaning mine...it worked a treat. Keep reading if you're interested to know why this fix most likely worked. Me being a curious person and wanting to understand how the wipers work, I dug a little bit deeper. Well they use the camera system, and a clever piece of technology called Artifical Intelligence (AI), to detect whether there is rain on the windshield and they look for other visual weather cues. From my research and knowledge of the Tesla Model 3, and my knowledge of AI netural networks, I have deduced the following: The camera that is looking out at the windscreen, which is positioned between the driver and passenger on the front windscreen, is most likely detecting the rain, and the front facing left and right cameras are looking for weather cues. The 3 cameras are fed data in a lab somewhere in Tesla. The data consists of real-world vision that shows that the weather is raining. Lots of different types of data needs to be fed into the cameras, so that the AI, well neural network to be precise, can LEARN how to detect rain. This is known in data science as TRAINING. The neural network is really a mathematical representation of an algorithm, that can learn to do new things. The type of algorithm is called a neural network. A neural network is a type of algorithm that mimcs how the human brain works. The word neural refers to neurons, which are small biological systems in the brain that help us all learn. By replicating a similar technique using mathematics, machines can be taught to learn, just like humans. If you remember the stories about AI beating Lee Sodol in the game of Go, that's the same technique. Anyway...once the neural network is trained the output is a model. The model is then tested by feeding more live data, to ensure that it functions correctly. It is doesn't function in all scenarios, that new learnings are fed back into the original neural network and new models can be TRAINED again to produce better results. This process usually takes weeks or months, depending on the complexity of the problem being solved. If you'd like to learn more about neural networks, check out this awesome article: www.explainthatstuff.com/introduction-to-neural-networks.html And if you'd like to have a play with neural networks, also known as Deep Learning, you can stuck in here with AWS: aws.amazon.com/deep-learning/ Merry Christmas |
CategoriesAll Active Directory AI Architecture Big Data Blockchain Cloud Comedy Cyber DevOps Driverless Cars MicroServices Office 365 Scaled Agile Social Media 
AuthorPaul Colmer is an AWS Senior Technical Trainer. Paul has an infectious passion for inspring others to learn and to applying disruptive thinking in an engaging and positive way. Archives
May 2023
|
RSS Feed