conformer

conformer copied to clipboard

About the "input length mismatch" bug in torchaudio's RNNT loss

2

In conformer/convolution.py, line 183, the code ``` output_lengths = input_lengths >> 2 output_lengths -= 1 ``` when the result of input_lengths >>2 is xx.75, the torchaudio.transforms.RNNTLoss will raise "input length...

Zain-Jiang

Count of Conformer parameters mismatch with that in the paper

4

In the Conformer original paper, the number of parameters are However, with the implementation in this repo, the number of parameters are slightly different ``` Conformer small: 10.16 M Conformer...

maxwellzh

Conformer Transducer inference

2

Transducer inference할 때 audio encoder의 output을 time step마다 한개씩 넣는 이유가 있을까요? Real time을 대비해서 그렇게 inference를 하는 것 같은데 non-real time일때는 어떻게 inference가 되는건지 제가 이해하고 있는게 맞는건지 잘...

jungwook518

I'm not sure, but I think these are the right replacement kwargs: ```py model = Conformer(dim=dim, dim_head=32, depth=3).to(device) ```

Akababa

what does the Class Conv2dSubampling‘s return mean?

3

I'm confused by the Class Conv2dSubampling in convolution.py.What does the second return output_lengths mean?

qiushan233

I want to test the test conformer in real time as a feature of my project. If anyone has any update kindly share the resources. Thanks

MuhammadShifa

Invalid version in setup.py

1

## My enviroment: - Python:3.10.10 - pip:23.0.1 - Kernel:6.2.12-arch1-1 ## problem When I tried to install conformer locally, I encountered error shown below. I'm not so familiar with Python and...

fuu38

您好，您有这篇论文”[时域语音增强] SE-Conformer: Time-Domain Speech Enhancement using Conformer阅读笔记“的代码吗？

1

wxy0505

I am using the following training function and librispeech dataset. Every time the output of the model while training become Nan as a result the loss is also nan. What...

afmsaif

to use conformer for acoustic scenes classification ?

1

thanks alot for the great work .can i use the mult-scale features aggregation conformer for acoustic scene classification ?

Noumanijaz744