Update app.py

app.py

@@ -1,7 +1,7 @@
 import gradio as gr
 import torch
 from transformers import AutoModelForCausalLM, AutoTokenizer
-from spaces import GPU
+import spaces  # Explicit import for GPU decorator
 import logging
 
 # Set up logging
@@ -38,14 +38,14 @@ if tokenizer.pad_token_id is None:
     logger.info("Set pad_token_id to eos_token_id.")
 
 def load_model():
-    """Load model inside GPU context to enable quantization."""
+    """Load model inside GPU context."""
     global model
     if model is None:
         try:
             logger.info("Loading model with CPU fallback (full precision)...")
             model = AutoModelForCausalLM.from_pretrained(
                 MODEL_NAME,
-                torch_dtype=torch.float16,  # Use fp16 for memory efficiency without bitsandbytes
+                torch_dtype=torch.float16,  # Use fp16 for memory efficiency
                 low_cpu_mem_usage=True,
                 trust_remote_code=True,
             )