According to the official method, will the model be downloaded locally or will it be downloaded

[mayu123mayu]

from docling.datamodel.base_models import InputFormat
from docling.datamodel.pipeline_options import PdfPipelineOptions, EasyOcrOptions
from docling.document_converter import PdfFormatOption, DocumentConverter

参考接口地址

配置pdf模型，设置Docling模型的路径

pdf_artifacts_path = "/docling-models"
pdf_pipeline_options = PdfPipelineOptions(artifacts_path=pdf_artifacts_path)

转换模型

converter = DocumentConverter(
format_options={
InputFormat.PDF: PdfFormatOption(pipeline_options=pdf_pipeline_options)
}
)

source = "庞氏骗局 - 中国投资者网.pdf"
result = converter.convert(source)
print(result.document.export_to_markdown()) python test.py
Downloading detection model, please wait. This may take several minutes depending upon your network connection.

[dolfim-ibm]

@mayu123mayu what is exactly your question? would this help you? https://ds4sd.github.io/docling/usage/#provide-specific-artifacts-path

[mayu123mayu]

@mayu123mayu what is exactly your question? would this help you? https://ds4sd.github.io/docling/usage/#provide-specific-artifacts-path你的问题到底是什么？这对你有帮助吗？ https://ds4sd.github.io/docling/usage/#provide-specific-artifacts-path

I manually downloaded it locally and set the local path, but when running the program, I still go to Hugg to download it

[Gelvins]

我也是同样的问题。设置了本地路径，但还是回去huggingface上下载
artifacts_path = "D:/model/docling-models"

pipeline_options = PdfPipelineOptions(artifacts_path=artifacts_path)
doc_converter = DocumentConverter(
format_options={
InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)
}
)

[Gelvins]

我也是同样的问题。设置了本地路径，但还是回去huggingface上下载 artifacts_path = "D:/model/docling-models"

pipeline_options = PdfPipelineOptions(artifacts_path=artifacts_path) doc_converter = DocumentConverter( format_options={ InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options) } )

问题已解决，请求路径是https://github.com/JaidedAI/EasyOCR/releases/download/v1.3/latin_g2.zip ，本地下载后存放到~/.EasyOCR/model 目录下