trust_remote_code=True

app.py

@@ -6,7 +6,7 @@ import torch
 import os 
 model_name = "microsoft/Phi-3-medium-128k-instruct"
 from transformers import AutoModelForCausalLM, AutoTokenizer
-model = AutoModelForCausalLM.from_pretrained(model_name, device_map='cuda')
+model = AutoModelForCausalLM.from_pretrained(model_name, device_map='cuda', torch_dtype=torch.float16, trust_remote_code=True)
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 
 class StopOnTokens(StoppingCriteria):