module MXNet

Overview

MXNet for Crystal.

This library is built on top of the core packages NDArray and Symbol.

NDArray works with arrays in an imperative fashion, i.e. you define how arrays will be transformed to get to an end result. Symbol works with arrays in a declarative fashion, i.e. you define the end result that is required (via a symbolic graph) and the MXNet engine will use various optimizations to determine the steps required to obtain this. With NDArray you have a great deal of flexibility when composing operations, and you can easily step through your code and inspect the values of arrays, which helps with debugging. Unfortunately, this flexibility comes at a performance cost when compared to Symbol, which can perform optimizations on the symbolic graph.

Defined in:

mxnet.cr
mxnet/autograd.cr
mxnet/base.cr
mxnet/cached_op.cr
mxnet/context.cr
mxnet/executor.cr
mxnet/gluon.cr
mxnet/gluon/block.cr
mxnet/gluon/data.cr
mxnet/gluon/data/data_loader.cr
mxnet/gluon/data/dataset.cr
mxnet/gluon/data/sampler.cr
mxnet/gluon/data/vision/mnist.cr
mxnet/gluon/loss.cr
mxnet/gluon/nn.cr
mxnet/gluon/nn/activations.cr
mxnet/gluon/nn/layers.cr
mxnet/gluon/parameter.cr
mxnet/gluon/trainer.cr
mxnet/gluon/utils.cr
mxnet/initializer.cr
mxnet/libmxnet.cr
mxnet/name/manager.cr
mxnet/ndarray.cr
mxnet/operations.cr
mxnet/optimizer.cr
mxnet/random.cr
mxnet/symbol.cr
mxnet/util.cr

Constant Summary

VERSION = "0.2.0"

Class Method Summary

Class Method Detail

def self.cpu(device_id : Int32 = 0) #

Returns a CPU context.

This function is equivalent to MXNet::Context.cpu.

Parameters

  • device_id (Int32, default = 0) Device id of the device. Not required for the CPU context. Included to make the interface compatible with GPU contexts.

[View source]
def self.gpu(device_id : Int32 = 0) #

Returns a GPU context.

This function is equivalent to MXNet::Context.gpu.

Parameters

  • device_id (Int32, default = 0) Device id of the device. Required for the GPU contexts.

[View source]