Update app.py

app.py

@@ -1,29 +1,18 @@
 import gradio as gr
 import torch
 from transformers import MarianMTModel, MarianTokenizer
-from optimum.intel import IncQuantizer
 
-# Load and optimize the model (quantization)
+# Load the MarianMT model and tokenizer
 model_name = "Dddixyy/latin-italian-translator"
-
-# Load the quantized model if available or use a regular model (quantization shown as an example)
-try:
-    # Attempt to load a quantized version if it's available
-    quantizer = IncQuantizer.from_pretrained(model_name)
-    model = quantizer.quantize()
-    print("Quantized model loaded.")
-except Exception as e:
-    print(f"Error loading quantized model: {e}")
-    model = MarianMTModel.from_pretrained(model_name)
-
-# Load tokenizer
 tokenizer = MarianTokenizer.from_pretrained(model_name)
+model = MarianMTModel.from_pretrained(model_name)
 
 # Translation function
 def translate_latin_to_italian(latin_text):
-    # Truncate input to 512 tokens to avoid overload (adjust as necessary)
+    # Truncate input to a maximum length of 512 tokens to avoid overload
     inputs = tokenizer(latin_text, return_tensors="pt", padding=True, truncation=True, max_length=512)
     
+    # Use torch.no_grad() to speed up inference by not calculating gradients
     with torch.no_grad():
         generated_ids = model.generate(inputs["input_ids"])