This chapter is dedicated to setting up the tool. We discuss software dependencies of
hls4ml. There is a quick start guide for beginners to get familiar quickly. Then we discuss in more detail the features of the tool and user configuration.
numpy, h5py: required for the translation of keras model files
pyyaml: for configuration file parsing
QKeras: for working with quantized models
PyTorch: for reading in Torch models
onnx: note that you need an install of protobuf and numpy to build onnx. Detailed instructions are included in the link.
Xilinx Vivado license: a license is required for the synthesis of generated RTL IP
For basic concepts to understand the tool, please visit the Concepts chapter. Here we give line-by-line instructions for simply running the tool out-of-the-box and getting a feel for the workflow.
pip install hls4ml
If you want to use our profiling toolbox, you might need to install extra dependencies:
pip install hls4ml[profiling]
To get started with
hls4ml, we provide some default example models for conversion:
import hls4ml #Fetch a keras model from our example repository #This will download our example model to your working directory and return an example configuration file config = hls4ml.utils.fetch_example_model('KERAS_3layer.json') print(config) #You can print it to see some default parameters #Convert it to a hls project hls_model = hls4ml.converters.keras_to_hls(config) # Print full list of example model if you want to explore more hls4ml.utils.fetch_example_list()
After that, you can use
Vivado HLS to synthesize the model:
#Use Vivado HLS to synthesize the model #This might take several minutes hls_model.build() #Print out the report if you want hls4ml.report.read_vivado_report('my-hls-test')
Done! You’ve built your first project using
hls4ml ! To learn more about our various API functionalities, check out our tutorials here.
If you want to configure your model further, check out our Configuration page.
Apart from our main API, we also support model conversion using a command line interface, check out our next section to find out more:
Getting started with hls4ml commands (optional)
To follow this tutorial, you must first download our
git clone https://github.com/fastmachinelearning/example-models.git
The model files, along with other configuration parameters, are defined in the
Further information about
.yml files can be found in Configuration page.
In order to create an example HLS project:
example-models/ from the main directory:
And use this command to translate a Keras model:
hls4ml convert -c keras-config.yml
This will create a new HLS project directory with an implementation of a model from the
To build the HLS project, do:
hls4ml build -p my-hls-test -a
This will create a Vivado HLS project with your model implementation!
NOTE: For the last step, you can alternatively do the following to build the HLS project:
cd my-hls-test vivado_hls -f build_prj.tcl
vivado_hls can be controlled with:
vivado_hls -f build_prj.tcl "csim=1 synth=1 cosim=1 export=1"
Setting the additional parameters from
0 disables that step, but disabling
synth also disables
For further information about how to use
If you need help for a particular
hls4ml command -hwill show help for the requested
We provide a detailed documentation for each of the command in the Command Help section
pip uninstall hls4ml
Examples of model files and weights can be found in example_models directory.
Training codes and examples of resources needed to train the models can be found here.
Other examples of various HLS projects with examples of different machine learning algorithm implementations are in the directory example-prjs.