🚩 Report: functional issue(s)
Testing in hugging face reports error
Can't load tokenizer using from_pretrained, please update its configuration: <class 'transformers.models.vision_encoder_decoder.configuration_vision_encoder_decoder.VisionEncoderDecoderConfig'>
Hi archerbj,
I put together a notebook that goes through inference for my TrOCR models that can be found here: https://github.com/DunnBC22/Vision_Audio_and_Multimodal_Projects/blob/main/Optical%20Character%20Recognition%20(OCR)/OCR%20License%20Plates/Inference_of_TrOCR_of_Synthetic_Dataset_Project.ipynb.
The only change that is needed is to change the mode name from this:
MODEL_NAME = "DunnBC22/trocr-base-printed-synthetic_dataset_ocr"
to this:
MODEL_NAME = "DunnBC22/trocr-base-printed_license_plates_ocr"
If you have any additional questions, feel free to follow up!
P.S. - In the future, please avoid making unfounded comments (especially about people's ethics). While I am sure that you were frustrated that this project was not working as you wanted, the claim that this is an ethical concern (the title) is unacceptable.
Sorry about the wrong title, it may result from the issue template and i didn't pay enough attention to it. I have changed it.
Hi archerbj, I appreciate the response and that you changed the issue template used. I hope that the code that I provided (via the link above) helped with solving the code issue. If you have any additional questions about this model, feel free to reach out!