Ahnaf Mozib Samin

Prediction not generated during testing

3

I am using docker cpu-latest image from wav2letter. I got two test fails in docker - Seq2seq test and w2lModule test. When I ran the Test binary file to get...

Data preparation for Language Modeling

20

I have some questions regarding language model. 1. I want to build a character level LM. Previously I built a word level LM using KENLM and for that I created...

Model not converging - Resnet architecture

8

I have 200 hours of training set and 20 hours of validation set. Loss is infinity even after 44 epochs. I tried with different learning rates but no difference. Here...

Docker Tests Failed

4

I am using Flashlight latest cuda docker image for ASR training. However, when I run make Test command, I come across the following test errors. ``` 98% tests passed, 10...

wav2letter model not running on flashlight

2

I trained an ASR model using the wav2letter toolkit and now trying to decode using flashlight. However, I am getting the following error: ``` E0615 18:37:35.549235 47040 Serializer.h:80 Error while...

RuntimeError: mat1 and mat2 shapes cannot be multiplied (16x70 and 36x256)

1

Hello! I am working on my custom dataset. While running the train.py script, I get the following error: > Training: 0%| | 0/300000 [00:00

[TTS] Custom Vocoder for HuggingFace Space

2

Hello! I would like to kindly ask whether it is possible to add a custom vocoder to a TTS model on HuggingFace space? If so, how can I do that?...

About the release the model

1

Hi! Thank you for your valuable work. I would like to kindly ask if you plan to release the pre-trained models.