Indra OCR model weights (bundled for RapidOCR model_type="lite")
================================================================
Files: det.onnx, rec.onnx, rec_cjk.onnx, script.onnx, order.onnx, meta.json
Source: https://huggingface.co/datalab-to/indra-ocr  (revision ed345c7f158eddbc7dee0c6a7269afb75a94f7e8)
Licensor: Datalab (https://www.datalab.to). Copyright (c) Datalab.

These files are NOT covered by RapidOCR's Apache-2.0 license. They are licensed under the
AI PUBS OPEN RAIL-M LICENSE (MODIFIED), Version 0.1, March 2, 2023 (Modified) - see ./LICENSE.

The Model is subject to the use-based restrictions of paragraph 5 / Attachment A of ./LICENSE,
which you must comply with and must pass on to anyone you distribute the Model to. In particular
you may NOT use the Model (except for personal use or research purposes, for 2(a)/2(b)) if you,
your employer or affiliated entity:
  (a) generated more than USD 20,000,000 gross revenue in the prior year;
  (b) raised more than USD 20,000,000 in total equity or debt funding; or
  (c) provides any product or service that competes with any product or service of the Licensor
      or its affiliates (this clause has no personal/research exception).
Paragraph 8 (Share-a-Like) applies this License to copies of the Model, its Derivatives and to the
Output and derivatives of the Output.

Changes: none. The files are byte-identical to the upstream revision above (SHA256 listed in
default_models.yaml). They are only copied into this directory.
Broader / commercial licenses: https://www.datalab.to/
