Upload app.py

app.py

@@ -16,7 +16,18 @@ peft_model_path = "./phi2-openassistant-lora-final"
 #bnb_config = BitsAndBytesConfig(load_in_4bit=True)  # Ensure compatibility
 #base_model = AutoModelForCausalLM.from_pretrained(base_model_name, quantization_config=bnb_config, device_map={"": device})
 #base_model = AutoModelForCausalLM.from_pretrained(base_model_name, torch_dtype=torch.float32, device_map={"": device})
-base_model = AutoModelForCausalLM.from_pretrained(base_model_name, load_in_4bit=True, device_map={"": device})
+#base_model = AutoModelForCausalLM.from_pretrained(base_model_name, load_in_4bit=True, device_map={"": device})
+bnb_config = BitsAndBytesConfig(
+load_in_4bit=True,
+bnb_4bit_compute_dtype=torch.float16,
+bnb_4bit_quant_type="nf4",
+bnb_4bit_use_double_quant=True,
+)
+base_model = AutoModelForCausalLM.from_pretrained(base_model_name, quantization_config=bnb_config, device_map={"": device})
+
+
+
+
 
 # Load LoRA adapter
 model = PeftModel.from_pretrained(base_model, peft_model_path)