Update app.py

app.py

@@ -3,7 +3,7 @@ from transformers import AutoProcessor, Idefics2ForConditionalGeneration, AutoMo
 import subprocess
 import torch
 from peft import LoraConfig
-from transformers import BitsAndBytesConfig
+# from transformers import BitsAndBytesConfig
 
 subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
 DEVICE = "cuda:0"
@@ -18,7 +18,7 @@ bnb_config = BitsAndBytesConfig(
     ) 
 
 model = AutoModelForPreTraining.from_pretrained("HuggingFaceM4/idefics2-8b",quantization_config=bnb_config)
-# model = Idefics2ForConditionalGeneration.from_pretrained("HuggingFaceM4/idefics2-8b",quantization_config=bnb_config)
+# model = Idefics2ForConditionalGeneration.from_pretrained("HuggingFaceM4/idefics2-8b",load_in_4bit=True)
 
 # if USE_QLORA or USE_LORA:
 #     lora_config = LoraConfig(