

Some weights of AlbertForSequenceClassification were not initialized from the model checkpoint at albert-base-v2 and are newly initialized: This IS NOT expected if you are initializing AlbertForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).

initializing a BertForSequenceClassification model from a BertForPreTraining model). This IS expected if you are initializing AlbertForSequenceClassification from the checkpoint of a model trained on another task or with another architecture (e.g. Some weights of the model checkpoint at albert-base-v2 were not used when initializing AlbertForSequenceClassification: See an interactive view of the CoLA dataset in NLP Viewer Class GLUEDataModule ( LightningDataModule ): task_text_field_map = return, Training ¶ CoLA ¶
