Log objects and media

Log a dictionary of metrics, media, or custom objects to a step with the W&B Python SDK. W&B collects the key-value pairs during each step and stores them in one unified dictionary each time you log data with wandb.Run.log(). Data logged from your script is saved locally to your machine in a directory called wandb, then synced to the W&B cloud or your private server.

Each call to wandb.Run.log() is a new step by default. W&B uses steps as the default x-axis when it creates charts and panels. You can optionally create and use a custom x-axis or capture a custom summary metric. For more information, see Customize log axes.

Automatically logged data

W&B automatically logs the following information during a W&B Experiment:

• System metrics: CPU and GPU utilization, network, etc. For the GPU, these are fetched with nvidia-smi.
• Command line: The stdout and stderr are picked up and show in the logs tab on the run page.

Turn on Code Saving in your account’s Settings page to log:

• Git commit: Pick up the latest git commit and see it on the overview tab of the run page, as well as a diff.patch file if there are any uncommitted changes.
• Dependencies: The requirements.txt file will be uploaded and shown on the files tab of the run page, along with any files you save to the wandb directory for the run.

What data is logged with specific W&B API calls?

With W&B, you can decide exactly what you want to log. The following lists some commonly logged objects:

• Datasets: You have to specifically log images or other dataset samples for them to stream to W&B.
• Plots: Use wandb.plot() with wandb.Run.log() to track charts. See Log Plots for more information.
• Tables: Use wandb.Table to log data to visualize and query with W&B. See Log Tables for more information.
• PyTorch gradients: Add wandb.Run.watch(model) to see gradients of the weights as histograms in the UI.
• Configuration information: Log hyperparameters, a link to your dataset, or the name of the architecture you’re using as config parameters, passed in like this: wandb.init(config=your_config_dictionary). See the PyTorch Integrations page for more information.
• Metrics: Use wandb.Run.log() to see metrics from your model. If you log metrics like accuracy and loss from inside your training loop, you’ll get live updating graphs in the UI.

Metric naming constraints

Due to GraphQL limitations, metric names in W&B must follow specific naming rules:

• Allowed characters: Letters (A-Z, a-z), digits (0-9), and underscores (_)
• Starting character: Names must start with a letter or underscore
• Pattern: Metric names should match /^[_a-zA-Z][_a-zA-Z0-9]*$/

Valid metric names:

with wandb.init() as run:
  run.log({"accuracy": 0.9, "val_loss": 0.1, "epoch_5": 5})
  run.log({"modelAccuracy": 0.95, "learning_rate": 0.001})

Invalid metric names (avoid these):

with wandb.init() as run:
  run.log({"acc,val": 0.9})  # Contains comma
  run.log({"loss-train": 0.1})  # Contains hyphen
  run.log({"test acc": 0.95})  # Contains space
  run.log({"5_fold_cv": 0.8})  # Starts with number

Common workflows

1. Compare the best accuracy: To compare the best value of a metric across runs, set the summary value for that metric. By default, summary is set to the last value you logged for each key. This is useful in the table in the UI, where you can sort and filter runs based on their summary metrics, to help compare runs in a table or bar chart based on their best accuracy, instead of final accuracy. For example: wandb.run.summary["best_accuracy"] = best_accuracy

2. View multiple metrics on one chart: Log multiple metrics in the same call. For example:

   with wandb.init() as run:
     run.log({"acc": 0.9, "loss": 0.1})
   

   You can then plot both metrics in the UI.

3. Customize the x-axis: Add a custom x-axis to the same log call to visualize your metrics against a different axis in the W&B dashboard. For example:

   with wandb.init() as run:
     run.log({'acc': 0.9, 'epoch': 3, 'batch': 117})
   

   To set the default x-axis for a given metric use Run.define_metric().

4. Log rich media and charts: wandb.Run.log() supports the logging of a wide variety of data types, from media like images and videos to tables and charts.

Best practices and tips

For best practices and tips for Experiments and logging, see Best Practices: Experiments and Logging.