How to use UKP-SQuARE/bert-base-cased-onnx with Transformers:
# Load model directly from transformers import AutoTokenizer, BertModelWithHeads tokenizer = AutoTokenizer.from_pretrained("UKP-SQuARE/bert-base-cased-onnx") model = BertModelWithHeads.from_pretrained("UKP-SQuARE/bert-base-cased-onnx", device_map="auto")
