This is a fine-tuned version of Multilingual Bart trained on Italian in particular on the public dataset MERLIN for Grammatical Error Correction.
To initialize the model:
from transformers import MBartForConditionalGeneration, MBart50TokenizerFast
model = MBartForConditionalGeneration.from_pretrained("MRNH/mbart-italian-grammar-corrector")
To generate text using the model:
tokenizer = MBart50TokenizerFast.from_pretrained("MRNH/mbart-italian-grammar-corrector", src_lang="it_IT", tgt_lang="it_IT")
input = tokenizer("I was here yesterday to studying",text_target="I was here yesterday to study", return_tensors='pt')
output = model.generate(input["input_ids"],attention_mask=input["attention_mask"],forced_bos_token_id=tokenizer_it.lang_code_to_id["it_IT"])
Training of the model is performed using the following loss computation based on the hidden state output h:
h.logits, h.loss = model(input_ids=input["input_ids"],
attention_mask=input["attention_mask"],
labels=input["labels"])