handler.py

from typing import Dict, List, Any
from transformers import AutoModel, AutoTokenizer


class EndpointHandler:
    def __init__(self, path="."):
        self.tokenizer = AutoTokenizer.from_pretrained(path)
        self.model = AutoModel.from_pretrained(
            path,
            trust_remote_code=True,
            # do_syntax=True, do_prefix=False, do_morph=False, do_ner=True, do_lex=True
        )
        self.model.eval()

    def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
        """
        data args:
            inputs (:obj: `str` | `PIL.Image` | `np.array`)
            kwargs
        Return:
            A :obj:`list` | `dict`: will be serialized and returned
        """
        # return self.pipeline(data['inputs'])
        outputs = self.model.predict(data['inputs'], self.tokenizer, output_style='json')
        for i, output in enumerate(outputs):
            lem = ' '.join([x['lex'] for x in output['tokens']])
            ner = [
                {
                    'word': ' '.join([x['lex'] for x in output['tokens'][x['token_start']:x['token_end'] + 1]]),
                    'entity_group': x['label'],
                    'token_start': x['token_start'],
                    'token_end': x['token_end']
                }
                for x in output['ner_entities']
            ]
            outputs[i] = {
                'lex': lem,
                'ner': ner
            }
        return outputs