How to Implement Dilated RNNs in PyTorch

PyTorch Dilated Recurrent Neural Networks

PyTorch implementation of Dilated Recurrent Neural Networks (DilatedRNN).

Getting Started


$ pip3 install -r requirements.txt

Run the tests:

$ python


Define a dilated RNN based on GRU cells with 9 layers, dilations 1, 2, 4, 8, 16, ... Then pass the hidden state to a further update

import drnn
import torch

n_input = 20
n_hidden = 32
n_layers = 9
cell_type = 'GRU'

model = drnn.DRNN(n_input, n_hidden, n_layers, cell_type)

x1 = torch.randn(23, 2, n_input)
x2 = torch.randn(23, 2, n_input)

out, hidden = model(x1)
out, hidden = model(x2, hidden)

Copy Task

$ python3 -m copy_memory.copymem_test --help

Character LeveL RNN

$ python3 -m char_rnn.char_rnn_test --help

Download Details:

Author: zalandoresearch 
Source Code: 
License: MIT license

#machinelearning #python #pytorch #deeplearning 

How to Implement Dilated RNNs in PyTorch
1.40 GEEK