The example loading code snippet had a small mistake where it loads the 7B Model.

README.md

@@ -31,7 +31,7 @@ It's directly converted from [openflamingo/OpenFlamingo-3B-vitl-mpt1b-langinstru
 
 You can load and try this model using
 ```python
-model = OtterForConditionalGeneration.from_pretrained("luodian/OTTER-MPT7B-Init", device_map="sequential")
+model = OtterForConditionalGeneration.from_pretrained("luodian/OTTER-MPT1B-RPJama-Init", device_map="sequential")
 model.text_tokenizer.padding_side = "left"
 tokenizer = model.text_tokenizer
 image_processor = transformers.CLIPImageProcessor()