Download scripts/extractbench_docparse.py from Kedhareswer/docparse: direct link, hf CLI and curl.
- Browser
- Download file 636 Bytes
-
https://huggingface.co/Kedhareswer/docparse/resolve/main/scripts/extractbench_docparse.py
- Command line
-
hf download hf://Kedhareswer/docparse/scripts/extractbench_docparse.py
-
curl -L -o extractbench_docparse.py https://huggingface.co/Kedhareswer/docparse/resolve/main/scripts/extractbench_docparse.py
636 Bytes
| """Run LlamaIndex ExtractBench with the two-stage docparse / pypdf -> local LLM pipelines registered. | |
| DOCPARSE_LLM_URL=http://127.0.0.1:8088 python scripts/extractbench_docparse.py run docparse_extract_local --test | |
| DOCPARSE_LLM_URL=http://127.0.0.1:8088 python scripts/extractbench_docparse.py run pypdf_extract_local --test | |
| """ | |
| import sys | |
| from pathlib import Path | |
| sys.path.insert(0, str(Path(__file__).resolve().parents[1])) | |
| import docparse.extractbench # noqa: E402,F401 (registers the provider and pipelines) | |
| from extract_bench.cli import main # noqa: E402 | |
| if __name__ == "__main__": | |
| sys.exit(main()) | |