Your one stop CLI for ONNX model analysis.

Related tags

Overview

Your one stop CLI for ONNX model analysis.

Featuring graph visualization, FLOP counts, memory metrics and more!

⚡️ Quick start

First, download and install DOT. Installation can be done via cargo:

cargo install steelix

MacOS users can also install via HomeBrew:

brew tap FL33TW00D/steelix
brew install steelix

⚙️ Commands & Options

Steelix has 2 core functions - model summarization & model visualization.

`summary`

CLI command to summarize the core aspects of your model.

steelix summary --model-path ./my-model.onnx

Option	Description	Type	Default	Required?
`--model-path`	Path at which your model is located.	`bool`	`false`	No

`plot`

CLI command to plot your model as an SVG file - complete with inferred shapes.

steelix plot --model-path ./my-model.onnx --open

Option	Description	Type	Default	Required?
`--model-path`	Path at which your model is located.	`string`	None	Yes
`--output-path`	Path at which your SVG will be saved.	`string`	`./model.svg`	No
`--open`	Open SVG in browser once generated.	`boolean`	`false`	No
`--disable-shapes`	Disable shape inference.	`boolean`	`false`	No

Supported Operators (ref ONNX IR)

Operator	Implemented
Abs	✅
Acos
Acosh
Add	✅
And
ArgMax
ArgMin
Asin
Asinh
Atan
Atanh
AveragePool
BatchNormalization	✅
BitShift
Cast
Ceil
Clip
Compress
Concat	✅
ConcatFromSequence
Constant
ConstantOfShape
Conv	✅
ConvInteger
ConvTranspose
Cos
Cosh
CumSum
DepthToSpace
DequantizeLinear
Det
Div
Dropout
Einsum
Elu
Equal
Erf	✅
Exp
Expand
EyeLike
Flatten
Floor
GRU
Gather	✅
GatherElements
GatherND
Gemm	✅
GlobalAveragePool
GlobalLpPool
GlobalMaxPool
Greater
GridSample
HardSigmoid
Hardmax
Identity
If
InstanceNormalization
IsInf
IsNaN
LRN
LSTM
LeakyRelu	✅
Less
Log
Loop
LpNormalization
LpPool
MatMul	✅
MatMulInteger
Max
MaxPool	✅
MaxRoiPool
MaxUnpool
Mean
Min
Mod
Mul	✅
Multinomial
Neg
NonMaxSuppression
NonZero
Not	✅
OneHot
Optional
OptionalGetElement
OptionalHasElement
Or
PRelu
Pad	✅ (mode=constant, pads>=0)
Pow
QLinearConv
QLinearMatMul
QuantizeLinear
RNN
RandomNormal
RandomNormalLike
RandomUniform
RandomUniformLike
Reciprocal
ReduceL1
ReduceL2
ReduceLogSum
ReduceLogSumExp
ReduceMax
ReduceMean
ReduceMin
ReduceProd
ReduceSum
ReduceSumSquare
Relu	✅
Reshape	✅
Resize
ReverseSequence
RoiAlign
Round
Scan
Scatter (deprecated)
ScatterElements
ScatterND
Selu
SequenceAt
SequenceConstruct
SequenceEmpty
SequenceErase
SequenceInsert
SequenceLength
Shape
Shrink
Sigmoid	✅
Sign
Sin
Sinh
Size
Slice
Softplus
Softsign
SpaceToDepth
Split
SplitToSequence
Sqrt
Squeeze	✅
StringNormalizer
Sub
Sum
Tan
Tanh
TfIdfVectorizer
ThresholdedRelu
Tile
TopK
Transpose
Trilu
Unique
Unsqueeze	✅
Upsample (deprecated)
Where
Xor
Function
Bernoulli
CastLike
Celu
DynamicQuantizeLinear
GreaterOrEqual
HardSwish
LessOrEqual
LogSoftmax
MeanVarianceNormalization
NegativeLogLikelihoodLoss
Range
Softmax	✅
SoftmaxCrossEntropyLoss

Credit

Most of the good ideas/code in this project are heavily inspired by tract, wonnx or netron.

This is a rewrite of the RAMP (Rapid Assistance in Modelling the Pandemic) model

RAMP from scratch This is a rewrite of the RAMP (Rapid Assistance in Modelling the Pandemic) model, based on the EcoTwins-withCommuting branch, in Rus

3 Oct 26, 2022

A neural network model that can approximate any non-linear function by using the random search algorithm for the optimization of the loss function.

random_search A neural network model that can approximate any non-linear function by using the random search algorithm for the optimization of the los

2 Apr 1, 2022

m2cgen (Model 2 Code Generator) - is a lightweight library which provides an easy way to transpile trained statistical models into a native code

Transform ML models into a native code (Java, C, Python, Go, JavaScript, Visual Basic, C#, R, PowerShell, PHP, Dart, Haskell, Ruby, F#, Rust) with zero dependencies

2.3k Dec 31, 2022

Using OpenAI Codex's "davinci-edit" Model for Gradual Type Inference

OpenTau: Using OpenAI Codex for Gradual Type Inference Current implementation is focused on TypeScript Python implementation comes next Requirements r

11 Dec 18, 2022

A demo repo that shows how to use the latest component model feature in wasmtime to implement a key-value capability defined in a WIT file.

Key-Value Component Demo This repo serves as an example of how to use the latest wasm runtime wasmtime and its component-model feature to build and ex