Does TLT have optical character recognition pretrained model?

Does TLT have optical character recognition pretrained model? I can’t find. If have, how can I train to recognize short text lines (not a sentence)? Any instructions for training own data.

Currently, TLT has not OCR pretrained model yet.