Problem when exporting Bert-Ner-Models to Onnx

By applying the code below the created onnx-model has as outputs:
logits → tensor: float32[batch_size, 9]

In my beginner-understanding the sequence_length is missing!!!???
→ tensor: float32[batch_size, sequence_length, 9]

from optimum.onnxruntime import ORTModelForSequenceClassification
model = ORTModelForSequenceClassification.from_pretrained(“dslim/bert-base-NER”, export=True)
model.save_pretrained(“onnx/”)