Update app.py

app.py

@@ -6,6 +6,7 @@ import torch
 import optimum
 from transformers import (AutoModelForCausalLM, AutoTokenizer, LlamaForCausalLM, LlamaTokenizer, GenerationConfig, pipeline,)
 from fastapi.middleware.cors import CORSMiddleware
+from pyngrok import ngrok
 
 os.environ['PYTORCH_CUDA_ALLOC_CONF'] = 'expandable_segments:True'
 torch.cuda.empty_cache()
@@ -14,6 +15,10 @@ torch.cuda.set_per_process_memory_fraction(0.8)  # Adjust the fraction as needed
 app = FastAPI(title="Deploying FastAPI Apps on Huggingface")
 app.add_middleware(CORSMiddleware, allow_origins=['*'], allow_methods=['*'], allow_headers=['*'],)
 
+ngrok.set_auth_token("Your token here")
+ngrok.kill()
+ngrok_tunnel = ngrok.connect(7860)
+print(ngrok_tunnel.public_url)
   
 # Load the model and tokenizer
 model_name_or_path = "TheBloke/Wizard-Vicuna-7B-Uncensored-GPTQ"