Update readme, example code use bfloat16 to reduce GPU memory usage

README.md

@@ -124,7 +124,8 @@ import torch
 from PIL import Image
 from transformers import AutoModel, AutoTokenizer
 
-model = AutoModel.from_pretrained('openbmb/MiniCPM-V', trust_remote_code=True).to(dtype=torch.bfloat16)
+model = AutoModel.from_pretrained('openbmb/MiniCPM-V', trust_remote_code=True)
+model = model.to(dtype=torch.bfloat16)
 tokenizer = AutoTokenizer.from_pretrained('openbmb/MiniCPM-V', trust_remote_code=True)
 model.eval().cuda()