W&B Tips

[1] QuickStart: docs.wandb.ai/quickstart
[2] wandb.init(...): docs.wandb.ai/library/init
[3] wandb.log(...): docs.wandb.ai/library/log  
[4] PyTorch Integration: docs.wandb.ai/integrations/pytorch

Quick Start

1. Set your config with argparse.
2. wandb.init(project='project_name_you_want' , config=args)
3. wandb.watch(model) // automatically log gradients and model parameters
4. Set up a training pipeline and insert wand.log(...)

Visualization

You can visualize all kinds of data: 

• Media: Audio, Video, Tet Table, HTML, Molecule.
• Custom Charts: Line, Scatter, Bar, Histogram, Precision-Recall curve, ROC, Confusion matrix, Multi-line.
• Matplotlib
• 3D Visualization: 3D Object, Point clouds.

An example of the visualization in Colab is here with its W&B run-page here.