gpt-neo-2.7B__low-cpu / handler.py
fwittel
added custom handler
ba13b5a
from typing import Dict, List, Any
from transformers import AutoModel, pipeline
class EndpointHandler:
def __init__(self, path=""):
# load the model
model = AutoModel.from_pretrained(path, low_cpu_mem_usage=True)
# create inference pipeline
# Do I have to check device?
self.pipeline = pipeline("text-generation", model=model)
# (Might have to adjust typing)
def __call__(self, data: Any) -> List[List[Dict[str, float]]]:
inputs = data.pop("inputs", data) # Should I get and pass parameters?
prediction = self.pipeline(inputs)
return prediction