mxnet¶
MXNet: a concise, fast and flexible framework for deep learning.
Attribute scoping support for symbolic API. |
|
Autograd for NDArray. |
|
ctypes library of mxnet and helper functions. |
|
Callback functions that can be used to track various status during epoch. |
|
Context management API of mxnet. |
|
Experimental contributions |
|
Engine properties management. |
|
Symbolic Executor component of MXNet. |
|
Executor manager. |
|
Neural network module. |
|
Image Iterators and image augmentation functions |
|
Weight initializer. |
|
Data iterators for common data formats and utility functions. |
|
Key-value store for distributed communication |
|
|
|
Information about mxnet. |
|
Runtime querying of compile time features in the native library. |
|
Scheduling learning rate. |
|
Logging utilities. |
|
Online evaluation metric module. |
|
MXNet model module |
|
A module is like a FeedForward model. |
|
Monitor outputs, weights, and gradients for debugging. |
|
Automatic naming support for symbolic API. |
|
NDArray API of MXNet. |
|
MXNet notebook: an easy to use visualization platform |
|
numpy interface for operators. |
|
Optimizer API of MXNet. |
|
Profiler setting methods. |
|
Random number interface of MXNet. |
|
Read and write for the RecordIO data format. |
|
Registry for serializable objects. |
|
Interface to runtime cuda kernel compile module. |
|
Symbol API of MXNet. |
|
Tools for testing. |
|
Interface for NDArray functions executed by torch backend. |
|
general utility functions |
|
Visualization module |