change to phi3 fp16

app.py

@@ -1,10 +1,10 @@
 import gradio as gr
 from llama_cpp import Llama
 
-model = "TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF"
+model = "microsoft/Phi-3-mini-4k-instruct-gguf"
 llm = Llama.from_pretrained(
     repo_id=model,
-    filename="tinyllama-1.1b-chat-v1.0.Q8_0.gguf",
+    filename="Phi-3-mini-4k-instruct-fp16.gguf",
     verbose=True,
     use_mmap=False,
     use_mlock=True,