How to use Intel/bert-base-uncased-sparse-1_2 with Transformers:
# Load model directly from transformers import AutoTokenizer, AutoModelForPreTraining tokenizer = AutoTokenizer.from_pretrained("Intel/bert-base-uncased-sparse-1_2") model = AutoModelForPreTraining.from_pretrained("Intel/bert-base-uncased-sparse-1_2")
SFconvertbot