Update app.py

app.py

@@ -1,12 +1,12 @@
-import spaces
 import torch
 from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer
 from peft import PeftConfig, PeftModel
 from threading import Thread
 import gradio as gr
+import spaces
 
 MODEL_PATH = "Ozaii/zephyr-bae"
-BASE_MODEL = "unsloth/Meta-Llama-3.1-8B-Instruct-bnb-4bit"
+BASE_MODEL = "unsloth/llama-3-8b-bnb-4bit"
 max_seq_length = 2048
 
 print("Zephyr is getting ready to charm! 🌟")
@@ -25,7 +25,8 @@ def load_model():
                 BASE_MODEL,
                 torch_dtype=torch.float16,
                 device_map="auto",
-                load_in_4bit=True
+                load_in_4bit=True,
+                trust_remote_code=True
             )
             
             model = PeftModel.from_pretrained(base_model, MODEL_PATH)