Update app.py

app.py

@@ -22,7 +22,7 @@ llama_pipe = pipeline(
     model_kwargs={"torch_dtype": torch.bfloat16}, 
     device_map="auto",
 )
-'''
+
 
 
 torch.random.manual_seed(0)
@@ -34,7 +34,7 @@ proc_tokenizer = AutoTokenizer.from_pretrained(proc_model_name)
 
 
 
-'''
+
 torch.random.manual_seed(0)
 proc_model_name = "microsoft/Phi-3-mini-4k-instruct-gguf"
 proc_model = AutoModelForCausalLM.from_pretrained(proc_model_name)
@@ -42,7 +42,11 @@ proc_model.to("cpu")
 proc_tokenizer = AutoTokenizer.from_pretrained(proc_model_name)
 '''
 
-
+torch.random.manual_seed(0)
+proc_model_name = "microsoft/Phi-3-mini-4k-instruct-onnx"
+proc_model = AutoModelForCausalLM.from_pretrained(proc_model_name)
+proc_model.to("cpu")
+proc_tokenizer = AutoTokenizer.from_pretrained(proc_model_name)
 
 
 SAMPLE_RATE = 16000 # Hz