Paperspace experiments can generate machine learning models, which can be interpreted and stored in the Paperspace Model Repository. The Repository holds references to the checkpoint files generated during the training period as well as summary metrics associated with the model's performance, such as accuracy and loss.
Currently Supported Models:
Future Supported Models:
To store model information in the Model Repository, add model specific parameters to the experiment command when launching an experiment.
--modelType defines the type of model that is being generated by the experiment.
--modelType Tensorflow will ensure that the model checkpoint files being generated are recognized as Tensorflow model files.
--modelPath defines where in the context of the experiment the model checkpoint files are being stored. This is a key argument that enables the evaluation and upload of the generated model files. One option is to set
--modelPath '/artifacts' and keep the checkpoint files around only in the context of the experiment. Another option is to set
--modelPath '/storage/my-experiment' to have permanent access to the model generated files in your Paperspace storage.
Here is an example command for running an TensorFlow based experiment which generates a model summary, used in GradientCI, and also uploads model checkpoint files to permanent storage.
paperspace-python experiments createAndStart singlenode --name Test1 --workspaceUrl https://github.com/Paperspace/mnist-sample --projectHandle <projectHandle> --container tensorflow/tensorflow:1.13.1-py3 --machineType K80 --command 'python mnist.py --data_format=channels_last' --modelType Tensorflow --modelPath '/artifacts'