I’m running into
training_loss when training wav2vec2 xlsr with my custom dataset.
Weird thing is that even though
training_loss goes to
eval_loss still goes down, and error_rate (
wer) also goes down.
I’ve experimented with lower learning_rate, but still getting similar behavior. I’m logging with
My graphs look like the following:
There’s no value for
train/lossafter ~60 steps since it is
eval/lossis still decreasing.
Has anyone experienced similar behavior?