Skip to content

An unofficial simplified version of ESPnet (end-to-end speech processing toolkit)

License

Notifications You must be signed in to change notification settings

hellolzc/ESPnet_light_TTS

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ESPnet light TTS

An unofficial simplified version of ESPnet.

Targets

  • No installation required
  • Remove dependence on kaldi
  • Remove support for multicomputer distributed systems (Slurm etc.)
  • Remove chainer_backend
  • Keep only the TTS part (Other parts may be added in the future).
  • Remove 'egs' (examples of ESPnet 1)
  • Simplify scripts in 'egs2'
  • Remove model uploading

Progress

There is still a lot of work to be done. Welcome to raise an issue or open a pull request.

Note

Changes of 'utils'

Add following scripts from kaldi:

# Following PWG [kan-bayashi's PWG](https://github.com/kan-bayashi/ParallelWaveGAN)
combine_data.sh
make_subset_data.sh
parse_options.sh
queue.pl
run.pl
slurm.pl
split_data.sh
split_scp.pl
ssh.pl
# Already exits
(download_from_google_drive.sh)
(stdout.pl)

# Add to fix running problems
    utt2spk_to_spk2utt.pl
    spk2utt_to_utt2spk.pl
validate_data_dir.sh
subset_data_dir.sh
    validate_text.pl
    subset_scp.pl
    filter_scp.pl
copy_data_dir.sh
    apply_map.pl
fix_data_dir.sh

About

An unofficial simplified version of ESPnet (end-to-end speech processing toolkit)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published