I recommend we have a few coaching sessions before you attempt this course. We will cover the fundamentals of AI starting from data driven approach mindset to implementing core AI technologies. You will have homework. I will work with you to improve your code and explain the underlying mechanisms of popular libraries such as PyTorch. The general outline of this course is the following: 1. Discuss your needs and interests so we can tailor the curriculum to you 2. AI history, feature engineering and early AI methods including multilayer perceptrons, SVM, K means, Random Forest, etc. implemented in Python. We will choose some to implement from scratch and others to use well known libraries for quick and effective results. 3. Data analysis techniques to understand what methods work best, including visualization, principal component analysis, evaluation frameworks. 4. (Optional) progressed techniques such as gradient boosted trees, advanced K means clustering, 5. Transformers concepts and HuggingFace 6. BERT based classifiers and autogressive encoder/decoders 7. (Optional) image generation with GANs and latent diffusion (e.g., Stability Diffusion, MidJourney-like models) 8. Finetuning large language models such as GPT 9. Memory optimization using DeepSpeed, LoRA, PEFT and xFormers for large language models that don't fit on your GPUs. 10. Model inference optimization and deployment 11. Prompt engineering techniques, best practices, and frameworks (we can go over this at any time).

More packages from Aaron

View all of Aaron’s categories