Upload app.py

app.py

@@ -1,9 +1,10 @@
-#import spaces
+import spaces
 import torch
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 import spacy
 
+@spaces.GPU(duration=30)
 class ModelSingleton:
     _instance = None
 
@@ -28,18 +29,19 @@ class ModelSingleton:
 
 model_singleton = ModelSingleton()
 
-#@spaces.GPU(duration=30)
+@spaces.GPU(duration=30)
 def generate_response_en_it(input_text):
     input_ids = model_singleton.tokenizer_en_it("translate English to Italian: " + input_text, return_tensors="pt").input_ids
     output = model_singleton.model_en_it.generate(input_ids, max_new_tokens=256)
     return model_singleton.tokenizer_en_it.decode(output[0], skip_special_tokens=True)
 
-#@spaces.GPU(duration=30)
+@spaces.GPU(duration=30)
 def generate_response_it_en(input_text):
     input_ids = model_singleton.tokenizer_it_en("translate Italian to English: " + input_text, return_tensors="pt").input_ids
     output = model_singleton.model_it_en.generate(input_ids, max_new_tokens=256)
     return model_singleton.tokenizer_it_en.decode(output[0], skip_special_tokens=True)
 
+@spaces.GPU(duration=30)
 def translate_text(input_text, direction):
     if direction == "en-it":
         nlp = model_singleton.nlp_en