Not using unsloth again

app.py

@@ -1,8 +1,8 @@
 import gradio as gr
 from huggingface_hub import InferenceClient
-from unsloth import FastLanguageModel
-from unsloth.chat_templates import get_chat_template
 from transformers import TextStreamer
+from peft import AutoPeftModelForCausalLM
+from transformers import AutoTokenizer
 
 
 """
@@ -12,19 +12,12 @@ For more information on `huggingface_hub` Inference API support, please check th
 model_name_or_path = "samlama111/lora_model"
 
 # client = InferenceClient(model_name_or_path)
-model, tokenizer = FastLanguageModel.from_pretrained(
-    model_name = model_name_or_path,
-    max_seq_length = 8192,
+model = AutoPeftModelForCausalLM.from_pretrained(
+    model_name_or_path, # YOUR MODEL YOU USED FOR TRAINING
     load_in_4bit = True,
-    # token = "hf_...", # No need since our model is public
-)   
-
-tokenizer = get_chat_template(
-    tokenizer,
-    chat_template = "llama-3.1",
-    mapping = {"role" : "from", "content" : "value", "user" : "human", "assistant" : "gpt"}, # ShareGPT style
+    device_map = "auto",
 )
-FastLanguageModel.for_inference(model) # Enable native 2x faster inference
+tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
 
 def respond(
     message,