Deepspeech latest results with english

[JRMeyer]

>>> agarwalaashish20
[June 11, 2019, 11:11am]

Hello Team, could you please point me to the link, to refer the latest
results for WER for pre-trained English DeepSpeech model?

[This is an archived TTS discussion thread from discourse.mozilla.org/t/deepspeech-latest-results-with-english]

[JRMeyer]

>>> kdavis
[June 11, 2019, 11:23am]

It's not yet public. But will be in the next couple of days.

[Archived Post]

[JRMeyer]

>>> Sushantmkarande
[June 11, 2019, 3:09pm]

you are doing great work guys. keep it up...

[Archived Post]

[JRMeyer]

>>> kdavis
[June 11, 2019, 4:57pm]

Now it's
public.

[Archived Post]

[JRMeyer]

>>> agarwalaashish20
[June 16, 2019, 10:58am]

: Please also share the approx
training data size in hours used to train the model?

[Archived Post]

[JRMeyer]

>>> kdavis
[June 16, 2019, 11:05am]

The data sets used are indicated in the release
notes linked to in my
previous comment.

[Archived Post]

[JRMeyer]

>>> Sushantmkarande
[June 16, 2019, 11:50am]

is 0.5.0 only trained on
Libreespeech data. slash
what about 22 gb mozilla open corpus.

[Archived Post]

[JRMeyer]

>>> kdavis
[June 16, 2019, 11:56am]

No. See the release notes.

[Archived Post]

[JRMeyer]

>>> Sushantmkarande
[July 13, 2019, 5:23pm]

slash
as per release notes

deepspeech model is trained on American English which achieves an 8.22% word error rate on the [LibriSpeech clean test corpus]

it is mentioned that test set is used as librispeech clean test but not
clearly mentioned which train set is used to priovide this checkpoints

please help me clear.

[Archived Post]

[JRMeyer]

>>> kdavis
[July 13, 2019, 6:02pm]

The release notes state

train_files
Fisher,
LibriSpeech,
and Switchboard training
corpora.

[Archived Post]

[JRMeyer]

>>> Sushantmkarande
[July 14, 2019, 9:52am]

oh I thought that was for fine tuning. anyway thank you so much...

[Archived Post]