Update app.py

app.py

@@ -21,8 +21,6 @@ HF_CACHE_DIR = os.environ.get('HF_CACHE_DIR') or '/home/user/app/.cache'
 USE_SYSTEM_PROMPT = os.environ.get('USE_SYSTEM_PROMPT', '').lower() == "true" or False
 ENABLE_GPU = os.environ.get('ENABLE_GPU', '').lower() == "true" or False
 GPU_LAYERS = int(os.environ.get('GPU_LAYERS', '0'))
-n_gqa_str = os.environ.get('N_GQA')
-N_GQA = 8#int(n_gqa_str) if n_gqa_str is not None else None #must be set to 8 for 70b models
 CHAT_FORMAT = os.environ.get('CHAT_FORMAT') or 'llama-2'
 REPO_NAME = os.environ.get('REPO_NAME') or 'IlyaGusev/saiga2_7b_gguf'
 MODEL_NAME = os.environ.get('MODEL_NAME') or 'model-q4_K.gguf'
@@ -155,7 +153,7 @@ def generate_response():
     return Response(generate_and_log_tokens(user_request='1', generator=generator), content_type='text/plain', status=200, direct_passthrough=True)
 
 def init_model():
-    llm.load_model(model_path=MODEL_PATH, context_size=CONTEXT_SIZE, enable_gpu=ENABLE_GPU, gpu_layer_number=GPU_LAYERS, n_gqa=N_GQA)
+    llm.load_model(model_path=MODEL_PATH, context_size=CONTEXT_SIZE, enable_gpu=ENABLE_GPU, gpu_layer_number=GPU_LAYERS)
     
 # Function to check if no requests were made in the last 5 minutes
 def check_last_request_time():