README.md

ProtAlbert

from transformers import AutoModel, AlbertTokenizer, pipeline
import re

tokenizer = AlbertTokenizer.from_pretrained("Rostlab/prot_albert", do_lower_case=False)

model = AutoModel.from_pretrained("Rostlab/prot_albert")

fe = pipeline('feature-extraction', model=model, tokenizer=tokenizer, device=0)

sequences_Example = ["A E T C Z A O", "S K T Z P"]

sequences_Example = [re.sub(r"[UZOB]", "X", sequence) for sequence in sequences_Example]

embedding = fe(sequences_Example)

print(embedding)