Hello,
I am trying to use the AutoModelForSequenceClassification class to train Japanese StableLM-3B-4E1T Base on the WRIME sentiment analysis dataset. However, since the config for the Japanese StableLM model is of the class StableLMEpochConfig (as opposed to StableLmConfig), which is not in the list of recognized config classes, it is not letting me load the model. Is there a way to use the AutoModelForSequenceClassification class on this model, or will I have to use the AutoModel class and add a classification layer myself?
Thanks!
1 Like