Wav2letter: How to use Pre-trained Model in Python binding?

Created on 8 Sep 2019  Â·  9Comments  Â·  Source: flashlight/wav2letter

I want to use python binding to do some research, but the README about that is too little.
And I'm sorry about that I don't understand how to import a Pre-trained Model in Python binding, can you tell me something about that? Thank you very much!

Most helpful comment

I’ve compiled this into a working docker. I will try to upload this in the
next week or so.

On Thu, Oct 24, 2019 at 10:42 AM Saeed notifications@github.com wrote:

@tylercollins590 https://github.com/tylercollins590 Can you please post
the solution here so everybody could use the python binding for decode?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/facebookresearch/wav2letter/issues/397?email_source=notifications&email_token=ABGPUL5E333GJRTFR5PXC6LQQGX5ZA5CNFSM4IUSOCQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECFI32Q#issuecomment-545951210,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABGPUL4C7RQB66UJNEYF2K3QQGX5ZANCNFSM4IUSOCQA
.

All 9 comments

I would love to see this as well.

@wyh163 I've figured it out.. It was a pain and the instructions were not clear at all. Feel free to email me at [email protected] and I will happily help you work through it

I'm having the same issue. Specifically how do you go from the features to the emissions. Surely the model.bin file has to be loaded somehow. Would be grateful if someone summarizes the solution.

@tylercollins590 Can you please post the solution here so everybody could use the python binding for decode?

I’ve compiled this into a working docker. I will try to upload this in the
next week or so.

On Thu, Oct 24, 2019 at 10:42 AM Saeed notifications@github.com wrote:

@tylercollins590 https://github.com/tylercollins590 Can you please post
the solution here so everybody could use the python binding for decode?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/facebookresearch/wav2letter/issues/397?email_source=notifications&email_token=ABGPUL5E333GJRTFR5PXC6LQQGX5ZA5CNFSM4IUSOCQKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECFI32Q#issuecomment-545951210,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABGPUL4C7RQB66UJNEYF2K3QQGX5ZANCNFSM4IUSOCQA
.

I’ve compiled this into a working docker. I will try to upload this in the next week or so.

@tylercollins590 do you have any progress to share on this? Thank you.

It would be great to have an example for generating the emission matrix from features in Python.

Any news on this?

@mathieulallier807 what are your needs? Could you specify what functionality are you searching?

Was this page helpful?
0 / 5 - 0 ratings

Related issues

abhinavkulkarni picture abhinavkulkarni  Â·  3Comments

bmblr497 picture bmblr497  Â·  5Comments

ghost picture ghost  Â·  5Comments

isaacleeai picture isaacleeai  Â·  5Comments

JanX2 picture JanX2  Â·  5Comments