Introduction to RAG systems (Retrieval Augmented Generation)

We comment on the latest announcement by openai for their latest gpt4o model. What are the model capabilities and what are the strategy behind developing such a model and what does it mean to all of us?

Non-technical episode. Becoming indifferent to the fruits of your work or to the praise you might get from people around you is key to success. Let's see how.

We discuss three main things. First the release of llama 3. Model by meta. Second, we answer the question as to why LLMs with longer context windows are not always better. Then finally, we explain when LSTMs should be used over Transformers. Enjoy!

Three points that you need to attend to, to become a good manager and to be able to make 'good' decisions that people follow. Listen to this episode to know what they are.

Open AI k, GTC, and the AI revolution. And what can we do about it?

AI value is created via a sequence of activities, from chip making, all through building and training foundation models, and building tools, to the application development. In this episode, we discuss these stages and what it means to the industry.

What are RAG systems ? Why do they matter? What are the LLM problems that are solved by RAG systems? What are the tools to create a RAG system and what does a typical system look like? Find the answers in today's episode.

Simply, what are the bias and variance? Why is there a tradeoff between them? Listen to this episode to learn about this fundamental topic which is usually asked about during data science and machine learning job interviews.

What is design of experiments? Student t-test? ANOVA? These are really simple to explain. Listen to this episode to learn about them.

What hardware do I need to run ML deep learning models? For prototyping.