Use empty quantization config to avoid AttributeError

app.py

@@ -4,11 +4,13 @@ import accelerate
 
 print("Accelerate version:", accelerate.__version__)
 
+from transformers import BitsAndBytesConfig
+
 # Load the original model
 original_model = AutoModelForCausalLM.from_pretrained(
     "unsloth/DeepSeek-R1-Distill-Llama-8B-unsloth-bnb-4bit",
     load_in_4bit=False,  # Disable 4-bit quantization
-    quantization_config=None  # Explicitly disable quantization
+    quantization_config=BitsAndBytesConfig()  # Use an empty quantization config
 )
 original_tokenizer = AutoTokenizer.from_pretrained("unsloth/DeepSeek-R1-Distill-Llama-8B-unsloth-bnb-4bit")
 
@@ -16,7 +18,7 @@ original_tokenizer = AutoTokenizer.from_pretrained("unsloth/DeepSeek-R1-Distill-
 fine_tuned_model = AutoModelForCausalLM.from_pretrained(
     "kas1/DeepSeek-R1-Distill-Llama-8B-unsloth-bnb-4bit-John1",
     load_in_4bit=False,  # Disable 4-bit quantization
-    quantization_config=None  # Explicitly disable quantization
+    quantization_config=BitsAndBytesConfig()  # Use an empty quantization config
 )
 fine_tuned_tokenizer = AutoTokenizer.from_pretrained("kas1/DeepSeek-R1-Distill-Llama-8B-unsloth-bnb-4bit-John1")