MarianMT exported to the ONNX format

Install Optimum

pip install optimum

Usage example

from transformers import AutoTokenizer
from optimum.onnxruntime import ORTModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("icon-it-tdtu/mt-vi-en-optimum")
model = ORTModelForSeq2SeqLM.from_pretrained("icon-it-tdtu/mt-vi-en-optimum")

text = "Tôi là một sinh viên."
inputs = tokenizer(text, return_tensors='pt')
outputs = model.generate(**inputs)
result = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(result)
# I am a student.