ABOUT THIS COURSE
Ten weeks of hands-on deep learning. CNNs for computer vision, transformers for language, fine-tuning open-source LLMs on your own Bangla dataset, and the inference deployment patterns that actually work in production.
WHAT YOU'LL LEARN
- 01
Train a YOLOv8 detector on your own dataset
- 02
Fine-tune a 3B-parameter LLM on Bangla text
- 03
Serve models with vLLM and Triton
- 04
Profile and reduce inference cost by 60%+
CURRICULUM
MODULE 01Foundations of Deep Learning2 lessons
- VIDEOBackpropagation from scratch38m
- LABBuild a 2-layer MLP in PyTorch60m
MODULE 02Computer Vision2 lessons
- VIDEOCNNs and what they actually learn35m
- LABTrain YOLOv8 on a custom dataset90m
MODULE 03Transformers & LLMs2 lessons
- VIDEOAttention, explained40m
- LABFine-tune Llama 3 on Bangla120m
MODULE 04Production Deployment2 lessons
- VIDEOServing with vLLM at scale30m
- PROJECTCapstone: ship a production model240m