Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/EmergentOrder/onnx-scala

An ONNX (Open Neural Network eXchange) API and backend for typeful, functional deep learning and classical machine learning in Scala 3
https://github.com/EmergentOrder/onnx-scala

deep-learning deep-neural-networks dotty jvm machine-learning neural-network onnx scala scala3

Last synced: 25 days ago
JSON representation

An ONNX (Open Neural Network eXchange) API and backend for typeful, functional deep learning and classical machine learning in Scala 3

Awesome Lists containing this project

README 

        



--------------------------------------------------------------------------------



[![Build status](https://travis-ci.com/EmergentOrder/onnx-scala.svg?branch=master)](http://travis-ci.com/EmergentOrder/onnx-scala)

[![Latest version](https://index.scala-lang.org/emergentorder/onnx-scala/onnx-scala/latest.svg?color=orange)](https://index.scala-lang.org/emergentorder/onnx-scala/onnx-scala)

## Getting Started

Add this to the build.sbt in your project:



```scala

libraryDependencies += "org.emergent-order" %% "onnx-scala-backends" % "0.17.0"

```



A short, recent talk I gave about the project: [ONNX-Scala: Typeful, Functional Deep Learning / Dotty Meets an Open AI Standard](https://youtu.be/8HuZTeHi7lg?t=1156)



### Full ONNX model inference - quick start

First, download the [model file](https://media.githubusercontent.com/media/onnx/models/main/Computer_Vision/squeezenet1_1_Opset18_torch_hub/squeezenet1_1_Opset18.onnx) for [SqueezeNet](https://en.wikipedia.org/wiki/SqueezeNet).

You can use `get_models.sh`



Note that all code snippets are written in Scala 3 (Dotty).



First we create an "image" tensor composed entirely of pixel value [42](https://upload.wikimedia.org/wikipedia/commons/0/0e/Answer_to_Life_42.svg):



```scala

import java.nio.file.{Files, Paths}

import org.emergentorder.onnx.Tensors._

import org.emergentorder.onnx.Tensors.Tensor._

import org.emergentorder.onnx.backends._

import org.emergentorder.compiletime._

import org.emergentorder.io.kjaer.compiletime._



val squeezenetBytes = Files.readAllBytes(Paths.get("squeezenet1_1_Opset18.onnx"))

val squeezenet = new ORTModelBackend(squeezenetBytes)



val data = Array.fill(1*3*224*224){42f}



//In NCHW tensor image format

val shape =                    1     #:     3      #:    224    #: 224     #: SNil

val tensorShapeDenotation = "Batch" ##: "Channel" ##: "Height" ##: "Width" ##: TSNil



val tensorDenotation: String & Singleton = "Image"



val imageTens = Tensor(data,tensorDenotation,tensorShapeDenotation,shape)



//or as a shorthand if you aren't concerned with enforcing denotations

val imageTensDefaultDenotations = Tensor(data,shape)

```



Note that ONNX tensor content is in row-major order.



Next we run SqueezeNet image classification inference on it:



```scala

val out = squeezenet.fullModel[Float, 

                               "ImageNetClassification",

                               "Batch" ##: "Class" ##: TSNil,

                               1 #: 1000 #: SNil](Tuple(imageTens))

// val out:

//  Tensor[Float,("ImageNetClassification", 

//                "Batch" ##: "Class" ##: TSNil,

//                1 #: 1000 #: 1 #: 1 SNil)] = IO(...)

// ...



//The output shape

out.shape.unsafeRunSync()

// val res0: Array[Int] = Array(1, 1000, 1, 1)



val data = out.data.unsafeRunSync()

// val data: Array[Float] = Array(1.786191E-4, ...)



//The highest scoring and thus highest probability (predicted) class

data.indices.maxBy(data)

// val res1: Int = 753

```



Referring to the [ImageNet 1000 class labels](https://gist.github.com/yrevar/942d3a0ac09ec9e5eb3a), we see that the predicted class is "radiator".



Based on a simple benchmark of 100000 iterations of SqueezeNet inference, the run time is on par (within 3% of) ONNX Runtime (via Python).

The discrepancy can be accounted for by the overhead of shipping data between the JVM and native memory.



When using this API, we load the provided ONNX model file and pass it as-is to the underlying ONNX backend, which is able to optimize the full graph.

This is the most performant execution mode, and is recommended for off-the-shelf models / performance-critical scenarios.



This full-model API is untyped in the inputs, so it can fail at runtime. This is inevitable because we load models from disk at runtime.

An upside of this is that you are free to use dynamic shapes, for example in the case of differing batch sizes per model call (assuming your model supports this via symbolic dimensions, see [ONNX Shape Inference](https://github.com/onnx/onnx/blob/main/docs/ShapeInference.md) ).

If your input shapes are static, feel free to wrap your calls into it in a facade with typed inputs.



## Project Details



ONNX-Scala is cross-built against Scala JVM, Scala.js/JavaScript and Scala Native (for Scala 3 / Dotty )



Currently at ONNX 1.14.1 (Backward compatible to at least 1.2.0 for the full model API, 1.7.0 for the fine-grained API), ONNX Runtime 1.16.3.

 

### Fine-grained API

A complete\*, versioned, numerically generic, type-safe / typeful API to ONNX(Open Neural Network eXchange, an open format to represent deep learning and classical machine learning models), derived from the Protobuf definitions and the operator schemas (defined in C++). 



We also provide implementations for each operator in terms of a generic core operator method to be implemented by the backend.

For more details on the low-level fine-grained API see [here](FineGrainedAPI.md)



The preferred high-level fine-grained API, most suitable for the end user, is [NDScala](https://github.com/SciScala/NDScala)



\* Up to roughly the set of ops supported by ONNX Runtime Web (WebGL backend)



#### Training

Automatic differentiation to enable training is under consideration (ONNX currently provides facilities for training as a tech preview only).



#### Type-safe Tensors

Featuring type-level tensor and axis labels/denotations, which along with literal types for dimension sizes allow for tensor/axes/shape/data-typed tensors.

Type constraints, as per the ONNX spec, are implemented at the operation level on inputs and outputs, using union types, match types and compiletime singleton ops (thanks to @MaximeKjaer for getting the latter into dotty).

Using ONNX docs for [dimension](https://github.com/onnx/onnx/blob/main/docs/DimensionDenotation.md) and [type](https://github.com/onnx/onnx/blob/main/docs/TypeDenotation.md) denotation, as well as the [operators doc](https://github.com/onnx/onnx/blob/v1.7.0/docs/Operators.md) as a reference,

and inspired by [Nexus](https://github.com/ctongfei/nexus), [Neurocat](https://github.com/mandubian/neurocat) and [Named Tensors](https://pytorch.org/docs/stable/named_tensor.html).



### Backend

There is one backend per Scala platform.

For the JVM the backend is based on [ONNX Runtime](https://github.com/microsoft/onnxruntime), via their official Java API.

For Scala.js / JavaScript the backend is based on the [ONNX Runtime Web](https://github.com/microsoft/onnxruntime/tree/main/js/web).



Supported ONNX input and output tensor data types:

* Byte

* Short

* Int

* Long

* Float

* Double

* Boolean

* String



Supported ONNX ops:

* ONNX-Scala, Fine-grained API: 87/178 total

* ONNX-Scala, Full model API: Same as below



* ONNX Runtime Web (using Wasm backend): 165/178 total.

* ONNX Runtime: 165/178 total



See the [ONNX backend scoreboard](http://onnx.ai/backend-scoreboard/index.html) 



#### Example execution



TODO: T5 example



## Build / Publish



You'll need sbt.



To build and publish locally:



```

sbt publishLocal

```



### Built With



#### Core



* [ONNX](https://github.com/onnx/onnx) via [ScalaPB](https://github.com/scalapb/ScalaPB) - Open Neural Network Exchange / The missing bridge between Java and native C++ libraries (For access to Protobuf definitions, used in the fine-grained API to create ONNX models in memory to send to the backend)



* [Spire](https://github.com/typelevel/spire) - Powerful new number types and numeric abstractions for Scala.  (For support for unsigned ints, complex numbers and the Numeric type class in the core API)



* [Dotty](https://github.com/lampepfl/dotty) - The Scala 3 compiler, also known as Dotty. (For union types (used here to express ONNX type constraints), match types, compiletime singleton ops, ...)



#### Backends



* [ONNX Runtime via ORT Java API](https://github.com/microsoft/onnxruntime/tree/main/java) - ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator



* [ONNX Runtime Web](https://github.com/microsoft/onnxruntime/tree/main/js/web)



### Inspiration



#### Scala



* [Neurocat](https://github.com/mandubian/neurocat) -  From neural networks to the Category of composable supervised learning algorithms in Scala with compile-time matrix checking based on singleton-types



* [Nexus](https://github.com/ctongfei/nexus) - Experimental typesafe tensors & deep learning in Scala



* [Lantern](https://github.com/feiwang3311/Lantern) - Machine learning framework prototype in Scala. The design of Lantern is built on two important and well-studied programming language concepts, delimited continuations (for automatic differentiation) and multi-stage programming (staging for short).



* [DeepLearning.scala](https://github.com/ThoughtWorksInc/DeepLearning.scala) - A simple library for creating complex neural networks



* [tf-dotty](https://github.com/MaximeKjaer/tf-dotty) - Shape-safe TensorFlow in Dotty