Some suggestions...

[nothingmn]

First off, I LOVE this project!! Great job folks!

A few suggestions...

1. Try to make the text to audio conversion more resilient (try/catch/retry at error)
2. if not provide a way to resume conversion if the docker fails (volume mount the incremental progress
3. Add some volume mounts:
   a. The final output folder
   b. A "queue" or input folder, at docker start, just get started processing all files in this folder. Have defaults for the docker like which model to use, etc. --This would greatly alleviate the speed of the process, since we could queue up all books and leave it along for a few days.
   c. Volume mount for the additional voices, each time the docker starts, it needs to re-download the voice.
4. Better logging, I care less about the text which is being converted, and more about a progress indicator
5. Chapter art?
6. Control the intro/title section. I have a few books with the title as the first few lines, duplicates up the content in the first ~ 10 seconds.
7. Possibly add just optional (and additional) mp3 output along size the current output.
8. Control the bitrate/quality of the output
9. UI needs lots of love
10. Implement a proper API, including the "queue" idea above... and provide a client, of course. :P

[ROBERT-MCDOWELL]

don't worry, most of your points are on rails or done already'

1. errors don't stop the conversion, unless if it's to serious
2. done on update coming
3. a. option --output_dir in headless mode, in gradio/gui mode, change it in conf.py
   b. done on update coming
   c. done on update coming
4. all the logs should be on the terminal you run the command. in gradio/gui mode it's still not implemented but if the demand is high we can do it
5. i'ts not what ebook2audiobook is done for and frankly there is too much work now (bounties and donations will be implemented for extra requests). also what do you mean by chapter art? desgin? cover? m4b audiobook are limited
6. the hardest way to convert an ebook to audiobook is to know what is an introduction and the end....
7. planned for after the next update
8. who pretended we don't put love in our work and btw on the GUI? in case you don't know Gradio is very limited for design. And our goal now it more a conversion quality than anything else.
9. no API for now as the demand is near zero

I appreciate all the wishes you need, but are you ready also to give when you receive for a win win situation?

[nothingmn]

Hi, thanks for the response!

My initial rig was done via docker on windows, and my best guess is the docker container was ejected due to resource limitations. Now that Im running on a much beefier server, via docker, it is pretty stable.

Cant wait for the next release!