Getting started in Python: 30 seconds to

The core class of is an Experiment, a specific run of a script that generated a result such as training a model on a single set of hyperparameters. An Experiment. will automatically log scripts output (stdout/stderr), code, and command line arguments on any script and for the supported libraries will also log hyperparameters, metrics and model configuration.

To make it easy, you can just put your personal Comet API key directly as a parameter to Experiment:

```python from comet_ml import Experiment experiment = Experiment("YOUR-PERSONAL-API-KEY")

Your code.


Or, even better, put your API key in a Comet config file and leave it out of your source code:

```python from comet_ml import Experiment experiment = Experiment()

Your code.


Automatic Logging

When you add the above two lines in your code, you will automatically get the following items logged:

  • script, or Jupyter Notebook
  • git metadata and patch
  • model graph representation (see below)
  • model weights and biases (see below)
  • model hyperparameters (see below)
  • training metrics (see below)
  • command-line arguments to script
  • console and Jupyter Notebook standard output and error
  • environment GPU, CPU, hostname, and more

If you are using the following frameworks, then you will log:

Framework What is automatically logged
tensorflow steps and metrics; see examples
tensorflow model analysis time series, plots, and slicing metrics
tensorboard summary scalars (as metrics) and summary histograms
sklearn hyperparameters; see examples
keras graph description, steps, metrics, hyperparameters, weights and biases as histograms, optimizer config, and number of trainable parameters; see examples
mlflow hyperparameters, assets, models, plus lower-level framework items (e.g., tensorflow's metrics, tensorboard summaries)
fbprophet hyperparameters, model, and figures
pytorch graph description, steps, and loss; see examples
pytorch-lightning loss and accuracy, with refinements in progress; see examples
pytorch + apex loss; see examples
fastai all pytorch items, plus epochs, and metrics; see examples
xgboost metrics, hyperparameters; see examples


Also check out the comet init command-line method of creating best-practices code with Comet.

For more details see Python SDK/Advanced.