https://waveworks.dk/flow-coder/

Skip to content
 
[cropped-waveworks_logo_6]

  * Home
  * Flow CSV Editor
  * Flow Coder
  * Contact

 
[cropped-waveworks_logo_6]

Navigation Menu
Navigation Menu

  * Home
  * Flow CSV Editor
  * Flow Coder
  * Contact

Technology preview now open

April 5, 2023: The first technology preview of Flow Coder's ML to
source code compiler is now open for model submissions. Click the
button below and submit your ML model through the form. Within 48
hours, you will receive a dependency-free source code file with a
function that executes the model inference, or an email with an
estimation for when you can expect to receive it.

  * The preview is available for the first 25 submissions.
  * You will receive a response within 48 hours, either with the
    source code, or with an estimation for when you can expect to
    receive it, in case the compiler currently doesn't support your
    model.
  * The compiler currently supports Keras .h5 files, but we encourage
    you to also submit .onnx model files, which will be supported
    shortly.
  * The compiler currently generates C/C++, Java, Kotlin and Swift,
    however you can also request conversion to C#, Python and
    JavaScript, which will be supported shortly.
  * We will not share your model with anyone. However, both the model
    and the produced source code will be downloadable with a public
    web link. These links will not be shared with anyone but you. All
    models and source code will be deleted from the server no later
    than April 28, 2023.
  * By submitting your email address, you agree to receive an email
    when your model has been converted, when Flow Coder is ready for
    beta testing, and when Flow Coder is publicly released. When you
    receive the source code, you will be encouraged to submit your
    feedback, so we can improve the compiler for you.

SUBMIT YOUR MODEL

Read on to learn more about the full product, expected to be launched
later this year.

[waves-scaled]

Flow Coder

[FlowCoder-1]

Flow Coder takes the pain out of integrating a machine learning model
into your software or firmware project. It converts trained machine
learning models straight into dependency-free source code for many
common programming languages, thereby eliminating the complexity and
overhead of inference libraries in deployment projects. The compiler
can be invoked online, using a desktop tool or as a single executable
command-line tool, so it's easy to integrate as a pre-build step in a
build pipeline.

The full product is expected to be available late 2023 or early 2024.

Benefits

  * Simplicity. Flow Coder can produce one single stand-alone
    function that consumes an input vector and produces an output
    vector. That's it. No need for inference runtimes, libraries or
    other dependencies that complicate the build process.
  * Efficiency. Inference with Flow Coder is expected to run faster
    and use less memory and power than with existing commonly used
    inference libraries for small models; partly because of the
    eliminated overhead, partly because all tensor operations are
    expanded to scalars and then further optimized. This is
    particularly true for heavily pruned and sparse models, where a
    latency reduction of 85-90% can be expected.

  * Flexibility. Although scalar math code is the ultimate target for
    Flow Coder, it can also produce code targeting higher level
    inference libraries such as CMSIS-NN.
  * Extensibility. Flow Coder uses the Python dialect FL to define
    the entire decomposition from high-level Keras model down to
    scalar math. This allows an expert user to easily define new
    layers or target new inference libraries. Moreover, users can add
    FL/Python code defining preprocessing and postprocessing to be
    included in the code generation.
  * Readability. The produced code is fully human readable and gives
    the user immediate and complete insight in how every layer and
    every operation is computed.

Input formats

Initially, the compiler will support the following model formats:

  * Keras .h5
  * .tflite
  * ONNX

The compiler supports all layers and operations including quantized
models.

Target languages

The following languages can be generated:

  * C/C++
  * Java
  * Kotlin
  * Swift
  * JavaScript
  * C#
  * Python
  * CUDA
  * VHDL