Update app.py

app.py

@@ -28,6 +28,9 @@ login(token=hf_token)
 
 from datasets import Dataset
 from transformers import AutoTokenizer
+import unsloth
+
+unsloth.optimize_all()
 
 model_name = "meta-llama/Llama-2-7b-hf"  # You can use a smaller one like "meta-llama/Llama-2-7b-chat-hf"
 
@@ -59,7 +62,7 @@ model = AutoModelForCausalLM.from_pretrained(
     load_in_4bit=True,  # Use 4-bit quantization for efficiency
     device_map="auto"
     #device_map="cpu",
-    quantization_config=None
+    #quantization_config=None
 )
 
 # Apply LoRA (efficient fine-tuning)