chunked/nltk_chunking/_glossary/chunk_79.txt

For image classification models, ([ViTForImageClassification]), the model expects a tensor of dimension
  (batch_size) with each value of the batch corresponding to the expected label of each individual image.