Skip to content

Implementing SYNTHESIZER: Rethinking Self-Attention in Transformer Models using Pytorch

License

Notifications You must be signed in to change notification settings

leaderj1001/Synthesizer-Rethinking-Self-Attention-Transformer-Models

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Implementing SYNTHESIZER: Rethinking Self-Attention in Transformer Models using Pytorch

Reference

  • Author: Yi Tay, Dara Bahri, Donald Metzler, Da-Cheng Juan, Zhe Zhao, Che Zheng

  • Google Research

Method

model

1. Dense Synthesizer

2. Fixed Random Synthesizer

3. Random Synthesizer

4. Factorized Dense Synthesizer

5. Factorized Random Synthesizer

6. Mixture of Synthesizers

Usage

import torch

from synthesizer import Transformer, SynthesizerDense, SynthesizerRandom, FactorizedSynthesizerDense, FactorizedSynthesizerRandom, MixtureSynthesizers, get_n_params, calculate_flops


def main():
    batch_size, channel_dim, sentence_length = 2, 1024, 32
    x = torch.randn([batch_size, sentence_length, channel_dim])

    vanilla = Transformer(channel_dim)
    out, attention_map = vanilla(x)
    print(out.size(), attention_map.size())
    n_params, flops = get_n_params(vanilla), calculate_flops(vanilla.children())
    print('vanilla, n_params: {}, flops: {}'.format(n_params, flops))

    dense_synthesizer = SynthesizerDense(channel_dim, sentence_length)
    out, attention_map = dense_synthesizer(x)
    print(out.size(), attention_map.size())
    n_params, flops = get_n_params(dense_synthesizer), calculate_flops(dense_synthesizer.children())
    print('dense_synthesizer, n_params: {}, flops: {}'.format(n_params, flops))

    random_synthesizer = SynthesizerRandom(channel_dim, sentence_length)
    out, attention_map = random_synthesizer(x)
    print(out.size(), attention_map.size())
    n_params, flops = get_n_params(random_synthesizer), calculate_flops(random_synthesizer.children())
    print('random_synthesizer, n_params: {}, flops: {}'.format(n_params, flops))

    random_synthesizer_fix = SynthesizerRandom(channel_dim, sentence_length, fixed=True)
    out, attention_map = random_synthesizer_fix(x)
    print(out.size(), attention_map.size())
    n_params, flops = get_n_params(random_synthesizer_fix), calculate_flops(random_synthesizer_fix.children())
    print('random_synthesizer_fix, n_params: {}, flops: {}'.format(n_params, flops))

    factorized_synthesizer_random = FactorizedSynthesizerRandom(channel_dim)
    out, attention_map = factorized_synthesizer_random(x)
    print(out.size(), attention_map.size())
    n_params, flops = get_n_params(factorized_synthesizer_random), calculate_flops(
        factorized_synthesizer_random.children())
    print('factorized_synthesizer_random, n_params: {}, flops: {}'.format(n_params, flops))

    factorized_synthesizer_dense = FactorizedSynthesizerDense(channel_dim, sentence_length)
    out, attention_map = factorized_synthesizer_dense(x)
    print(out.size(), attention_map.size())
    n_params, flops = get_n_params(factorized_synthesizer_dense), calculate_flops(
        factorized_synthesizer_dense.children())
    print('factorized_synthesizer_dense, n_params: {}, flops: {}'.format(n_params, flops))

    mixture_synthesizer = MixtureSynthesizers(channel_dim, sentence_length)
    out, attention_map = mixture_synthesizer(x)
    print(out.size(), attention_map.size())
    n_params, flops = get_n_params(mixture_synthesizer), calculate_flops(mixture_synthesizer.children())
    print('mixture_synthesizer, n_params: {}, flops: {}'.format(n_params, flops))


if __name__ == '__main__':
    main()

Output

torch.Size([2, 32, 1024]) torch.Size([2, 32, 32])
vanilla, n_params: 3148800, flops: 3145729
torch.Size([2, 32, 1024]) torch.Size([2, 32, 32])
dense_synthesizer, n_params: 1083456, flops: 1082370
torch.Size([2, 32, 1024]) torch.Size([1, 32, 32])
random_synthesizer, n_params: 1050624, flops: 1048577
torch.Size([2, 32, 1024]) torch.Size([1, 32, 32])
random_synthesizer_fix, n_params: 1050624, flops: 1048577
torch.Size([2, 32, 1024]) torch.Size([2, 32, 32])
factorized_synthesizer_random, n_params: 1066000, flops: 1064961
torch.Size([2, 32, 1024]) torch.Size([2, 32, 32])
factorized_synthesizer_dense, n_params: 1061900, flops: 1060865
torch.Size([2, 32, 1024]) torch.Size([2, 32, 32])
mixture_synthesizer, n_params: 3149824, flops: 3145729

Paper Performance

eval

About

Implementing SYNTHESIZER: Rethinking Self-Attention in Transformer Models using Pytorch

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages