pymoose package
Contents
pymoose package#
Subpackages#
eDSL Values & Types#
|
A type annotation Arguments to Moose computations. |
Moose value representing an AES key. |
|
|
Moose Value representing a tensor of AES-encrypted values. |
Moose value representing a Python float. |
|
|
Moose value representing a Python integer. |
|
Moose Value representing a Tensor. |
Moose value representing a Python string. |
eDSL DTypes#
Generic implementation of a Moose DType |
|
|
Factory function for creating a fixedpoint DType. |
Generic implementation of a Moose DType |
|
Generic implementation of a Moose DType |
|
Generic implementation of a Moose DType |
|
Generic implementation of a Moose DType |
|
Generic implementation of a Moose DType |
|
Generic implementation of a Moose DType |
eDSL Placement Factories#
|
|
|
|
|
eDSL Functions#
|
Compute the absolute value. |
|
Add two values. |
|
Elementwise addition of a collection of tensors. |
|
Compute the index of the maximal element of a tensor along a given dimension. |
|
|
|
Cast a tensor to a new dtype. |
|
Concatenation of a collection of tensors along a given axis/dimension. |
|
Embed a constant of a particular type. |
|
|
|
Divide two tensors. |
|
Dot product of two tensors. |
|
Elementiwise exponential function: \(e^x\). |
|
Expand the rank of |
|
Evaluate the boolean greater-than operation, i.e. |
|
The identity operation, |
|
Index a tensor along a given dimension. |
|
Invert a floating-point matrix. |
|
Evaluate the boolean less-than operation, i.e. |
|
Load a value from placement storage. |
|
Compute the elementwise natural logarithm of a tensor. |
|
Compute the elementwise base-2 logarithm of a tensor. |
|
Evaluate the boolean AND operation, i.e. |
|
Evaluate the boolean OR operation, i.e. |
|
Elementwise maximum of a collection of input tensors. |
|
Mean-reduce an input tensor. |
|
Multiply two values. |
|
Multiplex two tensors according to some condition. |
|
Embed a ones array into the Moose computation graph. |
|
Tag an output of a computation. |
|
Apply the rectified-linear unit (ReLU) function, |
|
Reshape a tensor. |
|
Save a key-value pair to placement storage. |
|
Select elements along some axis of a tensor according to some index tensor. |
|
Compute the shape of an input tensor. |
|
Compute a slice of a value. |
|
Softmax function. |
|
Square an input value. |
|
Reduce out any singleton dimensions of |
|
Compute a strided slice of a value. |
|
Apply the sigmoid function, \(\frac{1}{1 + e^{-x}}\). |
|
Subtract two values. |
|
Sum-reduce an input tensor. |
|
Compute the square-root of a value \(\sqrt{x}\). |
|
Compute the transpose of a tensor. |
|
Embed a zeros array into the Moose computation graph. |
PyMoose Context#
Get a global runtime context. |
|
|
Set a global runtime context. |
Moose Bindings#
|
Annotates a Python function as a Moose computation. |
|
Moose runtime backed by gRPC choreography. |
|
Locally-simulated Moose runtime. |