Run inference using VGT on custom Data.

AlibabaResearch / AdvancedLiterateMachinery

A collection of original, innovative ideas and algorithms towards Advanced Literate Machinery. This project is maintained by the OCR Team in the Language Technology Lab, Tongyi Lab, Alibaba Group.

Apache License 2.0

1.37k stars 165 forks source link

Run inference using VGT on custom Data. #121

Closed sibam23 closed 6 months ago

sibam23 commented 6 months ago

I get an error while trying to run custom inference using VGT as No such file or directory: '/path/layoutlm-base-uncased/pytorch_model.bin' Where is this supposed to be created?

jasonzhetan commented 2 months ago

How was this resolved? Getting the same issue

hosseindamavandi commented 2 months ago

How was this resolved? Getting the same issue

You can clone the following repository: https://huggingface.co/microsoft/layoutlm-base-uncased. Make sure the model files are the correct size after downloading.

Then, change the MODEL_PATH in DocumentUnderstanding/AdvancedLiterateMachinery/DocumentUnderstanding/VGT/object_detection/Configs/cascade/<publaynet>_VGT_cascade_PTM.yaml to point to the cloned repository. Update MODEL_PATH to:

MODEL_PATH: "layoutlm-base-uncased/"