Size mismatches in released model

[JRMeyer]

>>> maneeshkyadav
[February 16, 2021, 4:51am]

Must be doing something stupid here. Trying to run one of the released
multispeaker models:

Multi-Speaker-Tacotron2
DDC
VCTK 2136433

when I checkout 2136433 and try a test with synthesize.py I still get a
bunch of layer size mistmatches.

size mismatch for decoder.attention_rnn.weight_ih: copying a param with
shape torch.Size( slash [4096, 1024 slash ]) from checkpoint, the shape in current
model is torch.Size( slash [4096, 896 slash ])

(seems to be a delta of 128)...tried a few other commits around the time
of the model and no luck. Can anyone point out my mistake?

EDIT: I also noticed that the link to Universal
WaveGrad on the
wiki doesn't point to a google drive link like the rest of the vocoder
models. Would like to try it if possible.

[This is an archived TTS discussion thread from discourse.mozilla.org/t/size-mismatches-in-released-model]

[JRMeyer]

>>> erogol
[February 16, 2021, 10:14am]

you need to use the right commit version with those models. Commit
versions has been given on the table.

[Archived Post]

[JRMeyer]

>>> maneeshkyadav
[February 16, 2021, 5:37pm]

Thanks erogol, slash
I checked out 2136433 (2136433) when I
tried, do I have that wrong?

[Archived Post]

[JRMeyer]

>>> maneeshkyadav
[February 18, 2021, 10:32pm]

Ah, I think the commits on the wiki differ from the notebooks? For the
two tacotron2 multispeaker models:

notebook: slash
!git clone https://github.com/Mozilla/TTS TTS_repo

%cd TTS_repo

!git checkout ea976b0

wiki: slash
4873601 slash
2136433

Things seem to work fine with ea976b0.

[Archived Post]