Setup

If you are using MacOS, then run brew install poppler tesseract first.

1. Create and activate python virtual environment.
2. Run pip install -r requirements.txt to install dependencies.
3. Run download.sh to get the minimal set of files required to run inference.

Docker setup

1. Run docker build . -t badgerdoc-tb-extr
2. Use built image for example like this docker run -it badgerdoc-tb-extr bash

Or any other way

Use models from AWS S3

If you would like to download models from S3, Before running setup following environment variables for example with docker env file AWS_S3_ENDPOINT # S3 endpoint url AWS_ACCESS_KEY_ID # aws access key id AWS_SECRET_ACCESS_KEY # aws secret access key AWS_REGION # AWS region name AWS_S3_SSE_TYPE # AWS SSE type (optional)

Run excel or pdf pipeline

python -m table_extractor.run run <path-to-pdf-or-excel> <results-output-dir> --model_path <model-file-path> --verbose <true/false> --paddle_on <true/false> Pipeline will automatically decide how parse your document based on file extension. Supported file formats: *.pdf, *.xlsx, *.xlsm, *.xltx, *.xltm

Model on S3 could be used if <model-file-path> provided in format s3://<bucket>/<models_path>

Run pdf pipeline

Run pipeline on single pdf document

python -m table_extractor.run run-sequentially <path-to-pdf> <results-output-dir> --model_path <model-file-path> --verbose <true/false> --paddle_on <true/false>

Model on S3 could be used if <model-file-path> provided in format s3://<bucket>/<models_path>

Run excel extractor

python -m excel_extractor.excel_run <path-to-excel> <output-path>