Update app.py

app.py

@@ -3,7 +3,15 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
 
 model_name = "brdemorin/Phi3_80_steps_v2"
 
-model = AutoModelForCausalLM.from_pretrained(model_name)
+# Update the following parameters as necessary for your GPTQ model
+gptq_parameters = {
+    "load_in_8bit": False,
+    "device_map": "auto",
+    "trust_remote_code": True,  # This is often needed for custom models
+}
+
+# Load the model with GPTQ parameters
+model = AutoModelForCausalLM.from_pretrained(model_name, **gptq_parameters)
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 
 def chat(input_text):