tensorboard - Visualize and profile machine learning training
Visualize training metrics, debug models, compare experiments, inspect graphs and embeddings, and profile performance with TensorBoard.
Tags
Updated: 2026-09-29Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Log scalar metrics
- Log images and text
- Log histograms
- Visualize model graphs
- Project embeddings
- Track hyperparameters
- Compare experiment runs
- Profile performance bottlenecks
Inputs
- Training metrics
- Model parameters
- Training images and text
- Experiment configurations
- Embedding data
- Run log directories
Outputs
- TensorBoard visualizations
- Experiment comparisons
- Model graph views
- Embedding projections
- Performance profiles
- Event log files
Requirements
- Python environment
- tensorboard package
- torch package
- tensorflow package