handler.py

import torch
from typing import Dict, List, Any
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline

# get dtype
dtype = torch.bfloat16 if torch.cuda.get_device_capability()[0] == 8 else torch.float16


class EndpointHandler:
    def __init__(self, path=""):
        # load the model
        self.tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True)
        self.model = AutoModelForCausalLM.from_pretrained(path, device_map="auto", torch_dtype=dtype, trust_remote_code=True)

    def __call__(self, data: Any) -> List[List[Dict[str, float]]]:
        # ignoring parameters! Default to configs in generation_config.json.
        messages = [{"role": "user", "content": data}]
        return self.model.chat(self.tokenizer, messages)