File size: 1,243 Bytes
ccc2704 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 |
from typing import Dict, Any, List
import torch
from transformers import T5ForConditionalGeneration, T5Tokenizer
class EndpointHandler():
def __init__(self, path=""):
self.device = 'cuda' if torch.cuda.is_available() else 'cpu'
try:
self.model = T5ForConditionalGeneration.from_pretrained(path).to(self.device)
self.tokenizer = T5Tokenizer.from_pretrained(path)
except Exception as e:
print(f"Error loading model or tokenizer from path {path}: {e}")
# Handle error (e.g., exit or set model/tokenizer to None)
def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]:
inputs = data.get("inputs", "")
if not inputs:
return [{"error": "No inputs provided"}]
tokenized_input = self.tokenizer(inputs, return_tensors="pt", truncation=True, max_length=512, padding="max_length")
tokenized_input = tokenized_input.to(self.device) # Move input tensors to the same device as model
summary_ids = self.model.generate(**tokenized_input, max_length=400, do_sample=True, top_p=0.8)
summary_text = self.tokenizer.decode(summary_ids[0], skip_special_tokens=True)
return [{"summary": summary_text}] |