Some weights of VisionEncoderDecoderModel were not initialized from the model checkpoint at microsofttrocr-base-printed and are newly initialized encoderpoolerdenseweight encoderpoolerdensebias You sh

This message is a warning that some weights in a VisionEncoderDecoderModel were not initialized from a pre-trained model checkpoint, and as a result, the model may not perform optimally without further training on a downstream task. Additionally, the message warns that attention mask and pad token id were not set, which may result in unexpected behavior. Finally, the message informs the user that the max_length parameter will be deprecated in future versions of Transformers, and recommends using max_new_tokens instead.