Update app.py

app.py

@@ -6,6 +6,7 @@ import numpy as np
 import pandas as pd
 from tqdm import tqdm
 from pathlib import Path
+import spaces  # Import spaces for GPU allocation
 
 # PDF processing
 import PyPDF2
@@ -199,6 +200,7 @@ def create_vector_store(documents):
     return vector_store
 
 # Model Loading and RAG System
+@spaces.GPU  # Use GPU for model loading
 def load_model_and_tokenizer():
     """Load the ALLaM-7B model and tokenizer with error handling"""
     model_name = "ALLaM-AI/ALLaM-7B-Instruct-preview"
@@ -257,6 +259,7 @@ def retrieve_context(query, vector_store, top_k=5):
     
     return contexts
 
+@spaces.GPU  # Use GPU for text generation
 def generate_response(query, contexts, model, tokenizer, language="auto"):
     """Generate a response using retrieved contexts with ALLaM-specific formatting"""
     # Auto-detect language if not specified
@@ -476,11 +479,13 @@ def evaluate_response(query, response, reference):
     
     return evaluation_results
 
+@spaces.GPU  # Use GPU for conversation handling
 def run_conversation(assistant, query):
     """Run a query through the assistant and return the response"""
     response, sources, contexts = assistant.answer(query)
     return response, sources, contexts
 
+@spaces.GPU  # Use GPU for evaluation
 def run_evaluation_on_sample(assistant, sample_index=0):
     """Run evaluation on a selected sample from the evaluation dataset"""
     if sample_index < 0 or sample_index >= len(comprehensive_evaluation_data):