en-es
en-fr
en-sl
en
0.25
0.5
0.75
1.25
1.5
1.75
2
Introduction to Torch
Published on Aug 23, 20165605 Views
Torch is an open platform for scientific computing in the Lua language, with a focus on machine learning, in particular deep learning. Torch is distinguished from other array libraries by having first
Related categories
Chapter list
Machine learning With Torch + autograd00:00
First half - second half00:03
Torch01:14
What is torch - 101:37
What is torch - 202:06
What is torch - 302:50
What is torch - 403:03
What is torch - 503:26
What is torch - 603:32
What is torch - 703:52
What is torch - 804:18
What is torch - 904:37
Lua is designed to interoperate with c06:03
What is torch - 1008:41
What is torch - 1109:01
Tourch community - 110:12
Tourch community - 210:54
Untitled11:32
Tourch community - 411:39
Tourch community - 511:53
Tourch community - 612:02
Torch - where does it fit? - 112:13
Torch - where does it fit? - 213:06
Core philosophy14:03
Tensors and storages - 115:34
Tensors and storages - 216:28
Tensors and storages - 316:52
Tensors and storages - 417:00
Tensors and storages - 517:23
Tensors and storages - 617:46
Tensors and storages - 717:55
Tensors and storages - 818:29
Training cycle - 119:18
Training cycle - 219:44
The nn package - 120:01
The nn package - 220:07
The nn package - 321:21
The nn package - 421:39
The nn package - 522:19
The nngraph package 23:03
Advanced neural networks23:13
Touch-autograd by23:21
The optim package - 123:58
The optim package - 224:05
The optim package - 425:25
The optim package - 326:33
Torch autograd27:08
We work on top of stable abstractions27:54
Machine learning has other abstractions28:45
Abstraction for gradient-based ml - 131:03
Abstraction for gradient-based ml - 232:20
Abstraction for gradient-based ml - 332:27
Forward mode (symbolic view)32:49
Forward mode (program view)33:49
Reverse mode (symbolic view)35:54
Reverse mode (program view)36:03
A trainable neural network in torch-autograd - 237:41
What's actually happening? - 138:16
We track computation via operator overloading38:50
Calculating the gradient39:36
What's actually happening? - 239:49
A trainable neural network in torch-autograd - 141:39
Autograd examples - 141:58
Autograd examples - 242:21
Autograd examples - 342:40
Autograd examples - 442:52
Autograd examples - 543:14
Autograd examples - 644:12
Autograd examples - 745:14
Autograd examples - 845:38
So what differentiates n.net libraries? - 146:03
So what differentiates n.net libraries? - 246:57
We want no limits on the models we can write47:38
Neural net three ways - 147:46
Neural net three ways - 248:09
Neural net three ways - 348:56
Impact at twitter49:26
Other autodiff ideas50:57
You should be using it55:08