Skip to content
Abstract cover artwork for Computer Vision in Production, a Deep Learning course Deep Learning
Deep Learning intermediate 16 hours 4.5

Computer Vision in Production

Taught by Lucas Meyer

Classification, detection and segmentation with modern backbones, then the parts tutorials skip: data augmentation that reflects real distribution shift, annotation quality control, ONNX export, quantised inference at the edge, and monitoring for the day your camera angle changes.

What you'll cover

computer vision cnn detection segmentation onnx

More in Deep Learning

Abstract cover artwork for Deep Learning with PyTorch: Fundamentals to Fine-Tuning, a Deep Learning course Deep Learning
Deep Learning intermediate

Deep Learning with PyTorch: Fundamentals to Fine-Tuning

Autograd from first principles, then real training loops: learning-rate schedules, mixed precision, gradient accumulation, distributed data parallel, and the debugging discipline for a loss that will not go down. Finishes by fine-tuning a pretrained model on a custom dataset with a proper evaluation harness.

HT Hiroshi Tanaka 4.9

$99.00

22 hours

→
Abstract cover artwork for Transformers and Attention, Implemented Line by Line, a Deep Learning course Deep Learning
Deep Learning advanced

Transformers and Attention, Implemented Line by Line

Build a transformer from scratch: scaled dot-product attention, multi-head projections, positional encodings, KV caching and the modern variants (RoPE, grouped-query attention, RMSNorm) that make inference cheap. You will train a small model, profile it, and understand precisely where the FLOPs go.

HT Hiroshi Tanaka 4.9

$109.00

17 hours

→
Abstract cover artwork for Fine-Tuning LLMs with LoRA and QLoRA, a Deep Learning course Deep Learning
Deep Learning advanced

Fine-Tuning LLMs with LoRA and QLoRA

Adapt a large model on a single GPU. Parameter-efficient fine-tuning, 4-bit quantisation, dataset curation and formatting, catastrophic-forgetting mitigation, and evaluation that detects when your tune made the model worse at everything else. Includes the decision framework for fine-tune versus RAG versus prompt engineering.

AR Ahmed Rahal 4.8

$119.00

13 hours

→
Abstract cover artwork for Distributed Training with FSDP and DeepSpeed, a Deep Learning course Deep Learning
Deep Learning advanced

Distributed Training with FSDP and DeepSpeed

When a model no longer fits on one GPU. Data, tensor and pipeline parallelism, fully sharded data parallel, activation checkpointing, and the communication patterns that decide whether adding GPUs actually makes training faster.

HT Hiroshi Tanaka 4.7

$129.00

15 hours

→