Transformer models are good at capturing content-based global interactions, while CNNs exploit local features effectively. Conformer combine convolution neural networks and transformers to model both local and global dependencies of an audio sequence in a parameter-efficient way. Conformer significantly outperforms the previous Transformer and CNN based models achieving state-of-the-art accuracies.

This repository contains only model code, but you can train with conformer at openspeech

This project recommends Python 3.7 or higher. We recommend creating a new virtual environment for this project (using virtual env or conda).

• Numpy: pip install numpy (Refer here for problem installing Numpy).
• Pytorch: Refer to PyTorch website to install the version w.r.t. your environment.

Currently we only support installation from source code using setuptools. Checkout the source code and run the following commands:

pip install -e .

import torch
import torch.nn as nn
from conformer import Conformer

batch_size, sequence_length, dim = 3, 12345, 80

cuda = torch.cuda.is_available()  
device = torch.device('cuda' if cuda else 'cpu')

criterion = nn.CTCLoss().to(device)

inputs = torch.rand(batch_size, sequence_length, dim).to(device)
input_lengths = torch.LongTensor([12345, 12300, 12000])
targets = torch.LongTensor([[1, 3, 3, 3, 3, 3, 4, 5, 6, 2],
                            [1, 3, 3, 3, 3, 3, 4, 5, 2, 0],
                            [1, 3, 3, 3, 3, 3, 4, 2, 0, 0]]).to(device)
target_lengths = torch.LongTensor([9, 8, 7])

model = Conformer(num_classes=10, 
                  input_dim=dim, 
                  encoder_dim=32, 
                  num_encoder_layers=3).to(device)

# Forward propagate
outputs, output_lengths = model(inputs, input_lengths)

# Calculate CTC Loss
loss = criterion(outputs.transpose(0, 1), targets, output_lengths, target_lengths)

If you have any questions, bug reports, and feature requests, please open an issue on github or
contacts sh951011@gmail.com please.

I appreciate any kind of feedback or contribution. Feel free to proceed with small issues like bug fixes, documentation improvement. For major contributions and new features, please discuss with the collaborators in corresponding issues.

I follow PEP-8 for code style. Especially the style of docstrings is important to generate documentation.

• Conformer: Convolution-augmented Transformer for Speech Recognition
• Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context
• kimiyoung/transformer-xl
• espnet/espnet

• Soohwan Kim @sooftware
• Contacts: sh951011@gmail.com