Visualization of PyCaffe Experiment Hyperparameters
This topic shows you how to set experiment hyperparams and their effects. The topic builds on the script that resulted from steps in Getting Started for PyCaffe with Solve.
The steps that are covered are:
- Define a hyperparam.
- Set hyperparams to MissingLink's callback.
For each framework that MissingLink supports, there are hyperparameters that will be retrieved automatically. See the list.
Go through Getting Started for PyCaffe with Solve.
Ensure that you can successfully run the
training script that resulted from integration with the MissingLink SDK. In the steps that follow below, the script is further developed to include hyperparams.
Add a dropout rate hyperparam:
DROPOUT_RATE = 0.1
Set hyperparams to an experiment:
In the base script, just after the display name and description definition of an experiment, add an extra line right below to set some hyperparams.
You should have added hyperparams to your experiment successfully.
- Inspect the resulting script here.
- Run the new script and see how MissingLink's dashboard helps with monitoring the experiment's hyperparams. A description follows.
Viewing the new functionality on the dashboard
You can see the hyperparams across different experiments on your MissingLink dashboard.
Hyperparameters that are retrieved automatically
The following hyperparameters are retrieved automatically if they are defined,
regardless of whether the
set_hyperparams function is used to set them:
- Learning rate
- Learning rate policy
- Weight decay
- Batch size
- Total batches
- Epoch size
- Total epochs
- Max iterations