Update app.py

app.py

@@ -51,7 +51,7 @@ The role you will play follows below.
 # Install flash-attn
 subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
 # Initialize the model pipeline
-generator = pipeline('text-generation', model='Locutusque/Thespis-Llama-3.1-8B', torch_dtype=torch.bfloat16, token=os.getenv("TOKEN"))
+generator = pipeline('text-generation', model='Locutusque/Thespis-Llama-3.1-8B', torch_dtype=torch.bfloat16, token=os.getenv("TOKEN"), load_in_8bit=True)
 @spaces.GPU
 def generate_text(prompt, system_prompt, temperature, top_p, top_k, repetition_penalty, max_length):
     """