dnagpt/t5-gene-eng-small

A T5-large model

Trained from scratch on a mixed dataset of DNA, protein sequences, and English text

import torch
import numpy as np
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, Seq2SeqTrainer, Seq2SeqTrainingArguments
from datasets import load_dataset

tokenizer = AutoTokenizer.from_pretrained("dnagpt/t5-gene-eng-small")
print(tokenizer.tokenize("ATCGGAAGGTAAGGGAAAGCG"))

from transformers import AutoModelForSeq2SeqLM
model = AutoModelForSeq2SeqLM.from_pretrained("dnagpt/t5-gene-eng-small", from_flax=True)
print(model.config)