handler.py

from transformers import pipeline, M2M100ForConditionalGeneration, M2M100Tokenizer,QuantoConfig
from typing import Dict, List, Any

class EndpointHandler():
    def __init__(self, path=""):
        # load the optimized model
        model = M2M100ForConditionalGeneration.from_pretrained(path)
        tokenizer = M2M100Tokenizer.from_pretrained(path)
        # create inference pipeline
        self.pipeline = pipeline("translation", model=model, tokenizer=tokenizer)


    def __call__(self, data: Any) -> List[List[Dict[str, float]]]:
        """
        Args:
            data (:obj:):
                includes the input data and the parameters for the inference.
        Return:
            A :obj:`list`:. The object returned should be a list of one list like [[{"label": 0.9939950108528137}]] containing :
                - "label": A string representing what the label/class is. There can be multiple labels.
                - "score": A score between 0 and 1 describing how confident the model is for this label/class.
        """
        text = data.get("text", data)
        lang = data.get("langId",data)
        encoded = tokenizer(text, return_tensors="pt")
        generated_tokens = model.generate(**encoded, forced_bos_token_id=tokenizer.get_lang_id(lang))
        result = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)[0]
        return {'transdlated':result}