The manual_random_param_search_example.py script demonstrates a random parameter search by automating the execution of an experiment multiple times, each time with a different set of random hyperparameters.
This example accomplishes the automated random parameter search by doing the following:
- Creating a template Task named
Keras HP optimization base. To create it, run the base_template_keras_simple.py script. This experiment must be executed first, so it will be stored in the server, and then it can be accessed, cloned, and modified by another Task.
- Creating a parameter dictionary, which is connected to the Task by calling Task.connect so that the parameters are logged by ClearML.
- Adding the random search hyperparameters and parameters defining the search (e.g., the experiment name, and number of times to run the experiment).
- Creating a Task object referencing the template experiment,
Keras HP optimization base. See Task.get_task.
- For each set of parameters:
When the example script runs, it creates an experiment named
Random Hyper-Parameter Search Example which is associated
examples project. This starts the parameter search, and creates the experiments:
Keras HP optimization base 0
Keras HP optimization base 1
Keras HP optimization base 2.
When these experiments are completed, their results can be compared.