-
Hi,
There is a way to update an already trained or a pre-trained model with new data?
Is --continue command used for this purpose?
Best regards,
Adrian
-
I just tried the CPU-only model that @VitaliyLi provided to fix #7. To get this far, this also required my PR #4. Here is the README example I ran (mind the `-gpu 0`):
`luajit ~/wav2letter/test.lua…
JanX2 updated
6 years ago
-
Why do we require it right away if it is required later conditionally: https://github.com/facebookresearch/wav2letter/blob/cf2d486040c9c99ff50e861012d91480f326c403/test.lua#L10 ?
I am getting this:…
-
HI @VitaliyLi
I am getting the following while using pretarined model to transcribe the audio file.Can someone pls help me solve this
luajit test.lua librispeech-glu-highdropout-cpu.bin -progress…
-
I am new to lua and I run into an error I have not found any information on.
`luarocks make rocks/beamer-scm-1.rockspec
-- Using kenlm library /root/usr/lib/libkenlm.a
-- Using kenlm include path…
-
Hi, I'm trying to inference on librivox test set with pre-trained model. I already downloaded librivox dataset and completed everything regarding data preprocessing.
When I run luajit test.lua li…
-
Hi all,
Thanks for putting this out there. I'm still building, but ran into an issue.
This is a bit of a small change for a pull request, so I just created an issue and hope you'll update/test. …
-
HI
when I am running the below command
'luajit /wav2letter/test.lua /experiments/hello_librispeech/001_model_dev-clean.bin -progress -show -test dev-clean -save'
I am getting the following error…
-
Hi @VitaliyLi
I am trying to use pre-trained model for cpu to transcribe the librispeech. I ave a few question regarding this. Hope you people will reply soon.
1. For this does it require Cuda…
-
**Background/Use Case**
At the moment, I'm simply interested in using it for pre-trained decoding. (I'd ideally extract the lattice/confidences as well if possible with your flags).
System: Mac E…