Replies: 1 comment 2 replies
-
Hi @sdhungan, you'll need to declare a |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
For my thesis, i converted the wav2vec2 to a regression model to predict the valance and arousal. The model is trained and i have the folder "CKPT+2024-03-09+10-58-39+00" which contains various .ckpt files. I assume the model that i have trained is called model.ckpt (or is it wav2vec2.ckpt ???). Now the issue arises that i don't know how to load in the model to predict the float values from a recorded .wav file... I didn't find alot of documentation regarding this. The only thing i found was how to do it with the base model provided from hugginface, but this requires various files with parameters which i don't know how to ajust for my setup. If someone knows how to load in and evaluate a new .wav file, please let me know. This is the last step i need for my thesis before i can proceed to the next step.
Kind regards
Safal
Beta Was this translation helpful? Give feedback.
All reactions