Code Monkey home page Code Monkey logo

dnnc-operators's Introduction

dnnc-operators

Development and discussion of DNNC operators.

Below is the list of current ongoing operators list (Under Construction)


Operator Testcases Dimensions required Dimensions supported Broadcasting required Broadcasting supported Completion status Contributer
Abs ✔️ N D N D ✔️ Hrishikesh
Acos ✔️ N D N D ✔️ Hrishikesh
Acosh ✔️ N D N D ✔️ Hrishikesh
Add ✔️ N D N D ✔️ Gunjan
And ✔️ N D N D ✔️ Hrishikesh
ArgMax
ArgMin
Asin ✔️ N D N D ✔️ Hrishikesh
Asinh ✔️ N D N D ✔️ Hrishikesh
Atan ✔️ N D N D ✔️ Hrishikesh
Atanh ✔️ N D N D ✔️ Hrishikesh
AveragePool
BatchNormalization
BitShift
Cast
Ceil
Clip
Compress
Concat
Constant
ConstantOfShape
Conv
ConvInteger
ConvTranspose
Cos
Cosh
CumSum
DepthToSpace
DequantizeLinear ✔️ N D N D can't pass both "float" and "int" as input Gunjan
Div ✔️ N D N D ✔️ Gunjan
Dropout
Elu ✔️ 1 D 1 D ✔️ Gunjan
Equal bool return doesn't work N D N D ✔️ Gunjan
Erf ✔️ N D N D ✔️ Gunjan
Exp ✔️ N D N D ✔️ Gunjan
Expand
EyeLike Fails on large shapes 2 D 2 D DataType required Gunjan
Flatten 1 D doesn't work N D N D ✔️ Gunjan
Floor ✔️ N D N D ✔️ Gunjan
GRU
Gather
GatherElements
Gemm Output doesn't match with numpy 2 D 2 D ✔️ Gunjan
GlobalAveragePool N D N D ✔️ Nalin Shani
GlobalLpPool N D N D ✔️ Nalin Shani
GlobalMaxPool N D N D ✔️ Nalin Shani
Greater bool return doesn't work N D N D ✔️ Nalin Shani
HardSigmoid ✔️ N D N D ✔️ Nalin Shani
HardMax N D N D ✔️ Nalin Shani
Identity ✔️ N D N D ✔️ Nalin Shani
If
InstanceNormalization N D N D ✔️ Nalin Shani
IsInf N D N D ✔️ Nalin Shani
IsNaN N D N D ✔️ Nalin Shani
LRN N D N D ✔️ Nalin Shani
LSTM
LeakyRelu ✔️ N D N D ✔️ Nalin Shani
Less bool return doesn't work N D N D ✔️ Nikhil
Log ✔️ N D N D ✔️ Nikhil
LogSoftmax 2 D N D ✔️ Nikhil
Loop
LpNormalization works for 2D N D N D ✔️ Nikhil
LpPool
MatMul 4 D doesn't work N D 4 D ✔️ Rohit
MatMulInteger type mismatch N D 2 D Nikhil
Max N D N D ✔️ Rohit
MaxPool
MaxRoiPool
MaxUnpool
Mean type mismatch N D N D ✔️ Nikhil
Min type mismatch N D N D ✔️ Nikhil
Mod
Mul
Multinomial
Neg
NonMaxSupression
NonZero
Not
OneHot
Or
PRelu
Pad
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
ScatterElements
Selu
Shape
Shrink
Sigmoid N D 3 D ✔️ Subham
Sign N D 3 D ✔️ Subham
Sin N D 3 D ✔️ Subham
Sinh N D 3 D ✔️ Subham
Size
Slice
Softmax 2 D N D ✔️ Subham
Softplus N D 3 D ✔️ Subham
Softsign N D 3 D ✔️ Subham
SpaceToDepth
Split
Sqrt N D 3 D ✔️ Subham
Squeeze
StringNormalizer
Sub N D 4 D ✔️ Vishal
Sum
Tan N D 3 D ✔️ Vishal
Tanh N D 3 D ✔️ Vishal
TfIdfVectorizer
ThresholdedRelu N D N D ✔️ Rohit
Tile
TopK
Transpose N D 3 D ✔️ Vishal
Unique
Unsqueeze
Upsample
Where
Xor N D N D ✔️ Vishal

dnnc-operators's People

Contributors

enigmarikki avatar gunjannandy avatar nikhilt1998 avatar nixonz avatar rodio346 avatar sravit1 avatar srohit0 avatar subhamio avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar

dnnc-operators's Issues

Fix tensor.h

Fix to_string method in tensor.h for multi-dimensional tensors.

Bool return type tensors (Equal operator) and multiple dtype tensor input in compute function (Dequantizelinear)

Equal operator

Can't return tensor<bool>

Error:

swig/python detected a memory leak of type 'dnnc::tensor< bool > *', no destructor found.

Unittest Error:

ERROR: test_Equal2D (__main__.EqualTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "Equal.py", line 40, in test_Equal1D
    np.testing.assert_allclose(npr, np.array(dcr.data()).astype(np.float32),
AttributeError: 'SwigPyObject' object has no attribute 'data'

Transpose operator

When added Transpose operator to the dnnc interface, invokes the following issue of ambiguity


Transpose operator issue


Numpy uses np.T to transpose a matrix, we can use same kind of naming convention if doing that only, solves the issue!

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.