Generative AI Deep Dive

Generative AI Deep Dive bridges the gap between "user" and "builder" — the technical foundations you need to understand and build generative AI systems. Topics covered: - How diffusion models generate images (DALL-E, Stable Diffusion, Midjourney) - Transformer architecture for both text and images - Variational Autoencoders (VAE) and GAN architecture - Multimodal models: combining text, image, audio, and video - Training data, RLHF, and alignment techniques - Open-source models: HuggingFace ecosystem - Fine-tuning and LoRA for custom models - Inference optimization: quantization, distillation