Hey there,
can I just use few sentences in the dataset and try to overfit the model for sanity check?
My goal is actually to train a model for different language.
Thank you
Why not give it a go?
I’m pretty sure it won’t work. You need hours of audio before it even gets remotely recognisable, so what you’re suggesting sounds naively optimistic (but I’m open to being proved wrong!)
Why not to use the notebooks? The notebooks helped me to spot couple of corrupted files and bad transcriptions.
Hey there,
Why does my model doesn’t fit, after I train it in google colab, not completely though.
I have a question, if I train a model from google colab, how do I get it to work ?, I got a model mismatch when working with it.
Thank you