[PyTorch] Sample ⑧ ~ How to build a complex model ~
Purpose
Up to this tutorial, we built the definition of the neural network model with a simple sequence `` `torch.nn.Sequential``` like building blocks.
This method using torch.nn.Sequential makes it easy to define the model, but it can only create a simple network structure.
For example, torch.nn.Sequential does not allow you to build a ResNet network structure.
(Credit:Deep Residual Learning for Image Recognition. Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. arXiv:1512.03385 [cs.CV]
(or arXiv:1512.03385v1 [cs.CV] for this version) )
In this tutorial, PyTorch: Custom nn Modules I will introduce how to build a more complex neural network model </ span> with reference to.
For detailed explanation and code, see "[PyTorch] Sample ⑧ ~ How to build a complex model ~".
tutorial
-[PyTorch] Tutorial (Japanese version) ① ~ Tensor ~ -[PyTorch] Tutorial (Japanese version) ② ~ AUTOGRAD ~ -[PyTorch] Tutorial (Japanese version) ③ ~ NEURAL NETWORKS (Neural Network) ~ -[PyTorch] Tutorial (Japanese version) ④ ~ TRAINING A CLASSIFIER (image classification) ~
sample
-[PyTorch] Sample ① ~ NUMPY ~ -[PyTorch] Sample ② ~ TENSOR ~ -[PyTorch] Sample ③ ~ TENSORS AND AUTOGRAD ~ -[PyTorch] Sample ④ ~ Defining New autograd Functions ~ -[PyTorch] Sample ⑤ ~ Static Graphs ~ -[PyTorch] Sample ⑥ ~ nn Package ~ -[PyTorch] Sample ⑦ ~ optim package ~ -[PyTorch] Sample ⑧ ~ How to build a complex model ~ -[PyTorch] Sample ⑨ ~ Dynamic Graph ~
Recommended Posts