Tacotron2 | PyTorch implementation of Tacotron2 , an end-to-end text | Speech library

I'm trying to run tacotron2 on docker within Ubuntu WSL2 (v.20.04) on Win10 2004 build. Docker is installed and running and I can run hello world successfully.

(There's a nearly identical question here, but nobody has answered it.)

When I try to run docker build -t tacotron-2_image docker/ I get the error:

unable to prepare context: unable to evaluate symlinks in Dockerfile path: lstat /home/nate/docker/Dockerfile: no such file or directory

So then I navigated in bash to where docker is installed (/var/lib/docker) and tried to run it there, and got the same error. In both cases I created a docker directory, but kept getting that error in all cases.

How can I get this to work?

For example I have a wav file with speech.

I can create nice spectrogram visualization with sox:

I am trying to find databases like the LJ Speech Dataset made by Keith Ito. I need to use these datasets in TacoTron 2 (Link), so I think datasets need to be structured in a certain way. the LJ database is linked directly into the tacotron 2 github page, so I think it's safe to assume it's made to work with it. So I think Databases should have the same structure as the LJ. I downloaded the Dataset and I found out that it's structured like this:

I have the following directory structure in python.

I want to ask you how we can effectively re-train a trained seq2seq model to remove/mitigate a specific observed error output. I'm going to give an example about Speech Synthesis, but any idea from different domains, such as Machine Translation and Speech Recognition, using seq2seq model will be appreciated.

I learned the basics of seq2seq with attention model, especially for Speech Synthesis such as Tacotron-2. Using a distributed well-trained model showed me how naturally our computer could speak with the seq2seq (end-to-end) model (you can listen to some audio samples here). But still, the model fails to read some words properly, e.g., it fails to read "obey [əˈbā]" in multiple ways like [əˈbī] and [əˈbē].

The reason is obvious because the word "obey" appears too little, only three times out of 225,715 words, in our dataset (LJ Speech), and the model had no luck.

So, how can we re-train the model to overcome the error? Adding extra audio clips containing the "obey" pronunciation sounds impractical, but reusing the three audio clips has the danger of overfitting. And also, I suppose we use a well-trained model and "simply training more" is not an effective solution.

Now, this is one of the drawbacks of seq2seq model, which is not talked much. The model successfully simplified the pipelines of the traditional models, e.g., for Speech Synthesis, it replaced an acoustic model and a text analysis frontend etc by a single neural network. But we lost the controllability of our model at all. It's impossible to make the system read in a specific way.

Again, if you use a seq2seq model in any field and get an undesirable output, how do you fix that? Is there a data-scientific workaround to this problem, or maybe a cutting-edge Neural Network mechanism to gain more controllability in seq2seq model?

Thanks.