Skip to content

HeartKit CLI

$ heartkit --help

HeartKit CLI Options:
    --task [segmentation, rhythm, beat, denoise]
    --mode [download, train, evaluate, export, demo]
    --config ["./path/to/config.json", or '{"raw: "json"}']

The HeartKit command line interface (CLI) makes it easy to run a variety of single-line commands without the need for writing any code. You can run all built-in tasks and modes from the terminal with the heartkit command. This is also useful to reproduce Model Zoo results.

Example

Heartkit commands use the following syntax:

heartkit --mode [MODE] --task [TASK] --config [CONFIG]

Or using short flags:

heartkit -m [MODE] -t [TASK] -c [CONFIG]

Where:

  • MODE is one of download, train, evaluate, export, or demo
  • TASK is one of segmentation, rhythm, beat, or denoise
  • CONFIG is configuration as JSON content or file path

Download datasets specified in the configuration file.

heartkit -m download -c ./configs/download-datasets.json

Train a rhythm model using the supplied configuration file.

heartkit -m train -t rhythm -c ./configuration.json

Evaluate the rhythm model using the supplied configuration file.

heartkit -m evaluate -t rhythm  -c ./configuration.json

Run demo on trained rhythm model using the supplied configuration file.

heartkit -m demo -t rhythm -c ./configuration.json

Configuration File

The configuration file is a JSON file that contains all the necessary parameters for the task. The configuration file can be passed as a file path or as a JSON string. In addition, a single configuration file can be used for all modes- only needed parameters will be extracted for the given mode running. Please refer to the Configuration section for more details.

Download

The download command is used to download all datasets specified in the configuration file. Please refer to Datasets for details on the available datasets.

CLI

The following command will download and prepare all datasets specified in configuration JSON file.

heartkit --task rhythm --mode download --config ./configs/download-datasets.json

Train

The train command is used to train a HeartKit model for the specified task and dataset. Each task provides a reference routine for training the model. The routine can be customized via the configuration file. Please refer to HKTaskParams to see supported options.

CLI

The following command will train a rhythm model using the reference configuration:

heartkit --task rhythm --mode train --config ./configuration.json

Evaluate

The evaluate command will test the performance of the model on the reserved test sets for the specified task. The routine can be customized via the configuration file. Please refer to HKTaskParams to see supported options.

CLI

The following command will test the rhythm model using the reference configuration:

heartkit --task rhythm --mode evaluate --config ./configuration.json

Export

The export command will convert the trained TensorFlow model into both TensorFlow Lite (TFL) and TensorFlow Lite for micro-controller (TFLM) variants. The command will also verify the models' outputs match. The activations and weights can be quantized by configuring the quantization section in the configuration file. Once converted, the TFLM header file will be copied to location specified by tflm_file.

CLI

The following command will export the rhythm model to TF Lite and TFLM:

heartkit --task rhythm --mode export --config ./configuration.json

Demo

The demo command is used to run a task-level demonstration using either the PC or EVB as backend inference engine.

CLI

The following command will run a demo on the trained rhythm model using the same supplied configuration file.

heartkit --task rhythm --mode demo --config ./configuration.json