2020.08.07

Changelog

After 5 months we release a new model with a lot of improvements!

• Don't use anymore the wikipedia dump as text corpus but Mitads (check at https://github.com/MozillaItalia/DeepSpeech-Italian-Model/releases/tag/Mitads-1.0.0-alpha2)
• DeepSpeech 0.8 based
• New docker scripts
  • Sets of .env files to try different parameters https://github.com/MozillaItalia/DeepSpeech-Italian-Model/tree/master/DeepSpeech/env_files
• In the meantime we released also the notebooks!
• We are using the Mozilla official DS model for English for transfer learning to improve the quality

Second version of Italian model, trained with:

• ~130 hours for Common Voice IT dataset
• ~127 hours of m-ailabs Italian dataset
  total: ~257h

Available in 2 version transfer used transfer learning form the official English model release by mozilla and other one is from scratch .

model hyper-parameters:

• batch_size=64
• n_hidden=2048
• epochs=30
• learning_rate=0.0001
• dropout=0.4
• lm_alpha=0
• lm_beta=0
• es_epochs=10
• early_stop=1
• amp=1

For transfer learning model:

• amp=0
• drop_source_layer=1

Check the readme about the usage

Thanks

This release wasn't possible without the huge work of @nefastosaturo on the docker and DS side other than generating the new model.

Me (@Mte90) worked on the project management side about the model and @astrastefania with the help for the server offered by the Turin university we were able to do everything.

License

CC0 as public domain.