Project description

Pytorch VQVAE implementation

Example

from vqvae import VQVAE, sequential_encoder, sequential_decoder
from torch.optim import  Adam
from functools import partial

input_channels = 3
output_channels = 3
embedding_length = 256
hidden_channels = 64
beta = 0.25
embedding_size = 512
opt = partial(Adam, lr=2e-4)

encoder = sequential_encoder(input_channels, embedding_size, hidden_channels)  # Encoder from the paper
decoder = sequential_decoder(embedding_size, output_channels, hidden_channels)  # Decoder from the paper
vqvae = VQVAE(encoder, decoder, opt, beta, embedding_length, embedding_size)  # Pytorch-Lightning module, 
                                                                              # hence usable to train the model

Project details

These details have not been verified by PyPI

View statistics for this project via Libraries.io, or by using our public dataset on Google BigQuery

Release history Release notifications | RSS feed

This version

1.0.2

Nov 15, 2022

1.0.0

Nov 14, 2022

0.0.0

Nov 15, 2022

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

vqvae-1.0.2.tar.gz (11.6 kB view hashes)

Uploaded Nov 15, 2022 Source

Built Distribution

vqvae-1.0.2-py3-none-any.whl (1.6 kB view hashes)

Uploaded Nov 15, 2022 Python 3

Hashes for vqvae-1.0.2.tar.gz

Hashes for vqvae-1.0.2.tar.gz
Algorithm	Hash digest
SHA256	`387b7e13caa5291b248676f2b3d9e14ccbb26440373f0778e9d4a49ad13ddd81`
MD5	`4827e025586177ced8204b99caddc658`
BLAKE2b-256	`54aae2bfa2c7393bf69386696c93c551b4f03fedb9ed10842c74cd1e174bc0f0`

Hashes for vqvae-1.0.2-py3-none-any.whl

Hashes for vqvae-1.0.2-py3-none-any.whl
Algorithm	Hash digest
SHA256	`b7c805ff0b6bd613e6e2bd477ee8cd634e8672523edb44fbf2977ef4be0ca532`
MD5	`9d95ff8550d0ad5d60644dd86f66a816`
BLAKE2b-256	`15c555a073b3e0ea61a3861f62d465472c41305029260ab59375915b221523a7`