Update app.py

app.py

@@ -2,6 +2,7 @@ import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
 from huggingface_hub import InferenceClient
 import os
+import torch
 # Environment variable for HF token
 hf_token = os.environ.get("HF_TOKEN")
 
@@ -12,7 +13,7 @@ model_id = "mistralai/Mistral-7B-Instruct-v0.2"
 quantization_config = BitsAndBytesConfig(
     load_in_4bit=True,
     bnb_4bit_quant_type="nf4",
-    bnb_4bit_compute_dtype="torch.float16",
+    bnb_4bit_compute_dtype=torch.float16,
 )
 
 model = AutoModelForCausalLM.from_pretrained(model_id, quantization_config=True, device_map="auto", token=hf_token)