crazyoscarchang

A Tour of PyTorch Internals (Part I)

The fundamental unit in PyTorch is the Tensor. This post will serve as an overview for how we implement Tensors in PyTorch, such that the user can interact with it from the Python shell. In particular, we want to answer four main questions:

How does PyTorch extend the Python interpreter to define a Tensor type that can be manipulated from Python code?
How does PyTorch wrap the C libraries that actually define the Tensor's properties and methods?
How does PyTorch cwrap work to generate code for Tensor methods?
How does PyTorch's build system take all of these components to compile and generate a workable application?

Extending the Python Interpreter

PyTorch defines a new package torch. In this post we will consider the ._C module. This module is known as an "extension module" - a Python module written in C. Such modules allow us to define new built-in object types (e.g. the Tensor) and to call C/C++ functions.

WORK IN PROGRESS

PyTorch Internals Part II - The Build System

In the first post I explained how we generate a torch.Tensor object that you can use in your Python interpreter. Next, I will explore the build system for PyTorch. The PyTorch codebase has a variety of components:

The core Torch libraries: TH, THC, THNN, THCUNN
Vendor libraries: CuDNN, NCCL
Python Extension libraries
Additional third-party libraries: NumPy, MKL, LAPACK

	## Weight norm is now added to pytorch as a pre-hook, so use that instead :)

	import torch
	import torch.nn as nn
	from torch.nn import Parameter
	from functools import wraps

	class WeightNorm(nn.Module):
	append_g = '_g'
	append_v = '_v'

	import torch
	import torch.nn as nn
	from torch.nn import Parameter
	from torch.autograd import Variable, Function
	from collections import defaultdict
	import graphviz

	"""
	This is a rather distorted implementation of graph visualization in PyTorch.