🧠

VisualDL Lab

Interactive Deep Learning Labs • ANN • CNN • RNN • Autoencoders

Hands-on, interactive labs to learn neural networks, convolutional models, sequence models, and generative architectures. Choose a lab to begin.

ANN Foundations

🔰 Lab 01 — ANN Foundations

Feedforward neural networks: layers, activations, learning rate, overfitting.

Initialization & Activations

⚙️ Lab 02 — Initialization & Activations

Weight initialization, activation impact, gradients & training dynamics.

Regularization & BatchNorm

🛡️ Lab 03 — Regularization & BatchNorm

L2, Dropout, BatchNorm, early stopping, and decision boundaries.

Optimizers & Learning Rate

⚡ Lab 04 — Optimizers & Learning Rate

SGD, Momentum, RMSProp, Adam, LR schedules & convergence behavior.

Generalization & Bias–Variance

📉 Lab 05 — Generalization & Bias–Variance

Model capacity, data fraction, label noise, and U-shaped regularization curves.

CNN Basics

🖼️ Lab 06 — CNN Basics

Convolutions, pooling, filters, dropout, augmentation, Grad-CAM.

RNN Basics

🔁 Lab 07 — RNN Basics

Simple RNNs for text, sentiment, and time-series; perplexity & hidden states.

LSTM vs GRU

🔀 Lab 08 — LSTM vs GRU

Compare LSTM/GRU across IMDB, SST-2, sine-wave and Twinkle Melody.

Autoencoders

🎨 Lab 09 — Autoencoders

Dense autoencoders, denoising AE, latent space, t-SNE & morphing.