Metadata-Version: 2.1
Name: torchrecorder
Version: 1.0.3
Summary: Record execution graphs of PyTorch neural networks
Home-page: https://github.com/ahgamut/torchrecorder
Author: Gautham Venkatasubramanian
Author-email: ahgamut@gmail.com
License: MIT
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Topic :: Scientific/Engineering :: Visualization
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Description-Content-Type: text/markdown
License-File: LICENSE

`torchrecorder`
===============

![](https://readthedocs.org/projects/torchrecorder/badge/?version=latest&style=flat)

A small package to record execution graphs of neural networks in PyTorch.
The package uses hooks and the `grad_fn` attribute to record information.  
This can be used to generate visualizations at different scope depths. 

Licensed under MIT License.
View documentation at https://torchrecorder.readthedocs.io/

## Installation

Requirements:

* Python3.6+
* [PyTorch](https://pytorch.org) v1.3 or greater (the `cpu` version)
* The [Graphviz](https://graphviz.gitlab.io) library and `graphviz` [python package](https://graphviz.readthedocs.io/en/stable/manual.html).


Install this package:

```
$ pip install torchrecorder
```

## Acknowledgements

This is inspired from [`szagoruyko/pytorchviz`](https://github.com/szagoruyko/pytorchviz).  This package
differs from `pytorchviz` as it provides rendering at multiple depths.

Note that for rendering a network during training, you can use TensorBoard and
[`torch.utils.tensorboard.SummaryWriter.add_graph`](https://pytorch.org/docs/stable/tensorboard.html#torch.utils.tensorboard.writer.SummaryWriter.add_graph),
which records and renders to a `protobuf` in a single step.  The intended usage of `torchrecorder` is for
presentation purposes.


