The SAS Deep Learning Python (DLPy) package provides the high-level Python APIs to deep learning methods in SAS Visual Data Mining and Machine Learning. It allows users to build deep learning models using friendly Keras-like APIs.
When trying to use pretrained huggingface models for BERT training, there is a problem that prevents the use of some pretrained models. It seems that the BERT_Model name parameter forces any text to be converted to lower case. When trying to use e.g. "TurkuNLP/bert-base-finnish-uncased-v1" model the name is converted to lower case. However, in the Huggingface website, the urls are case sensitive so the config file cannot be found with the lower case model name.
When trying to use pretrained huggingface models for BERT training, there is a problem that prevents the use of some pretrained models. It seems that the BERT_Model name parameter forces any text to be converted to lower case. When trying to use e.g. "TurkuNLP/bert-base-finnish-uncased-v1" model the name is converted to lower case. However, in the Huggingface website, the urls are case sensitive so the config file cannot be found with the lower case model name.