LibreDDColorl-restore weights
-----------------------------

Source architecture: https://github.com/piddnad/DDColor
Commit: 2adb63f2656ac41cbdf7b894cddd94121a3faf13
License: Apache License 2.0, with permissive component notices documented in
the LibreYOLO family NOTICE.

Source artifact: https://huggingface.co/piddnad/ddcolor_modelscope
Revision: 060f67494e31883a4b13cb27f889f3154847ada4
File: pytorch_model.bin
Bytes: 911914869
SHA-256: d81711971ec59200da26d5e8a1afae8dd3778d495ea8ad7a7dadc769f403f7e7
Publisher-declared artifact license: Apache-2.0

Conversion adds LibreYOLO checkpoint metadata. Learned tensors are unchanged.
Converted SHA-256:
e6a4125ce726c256b8efaba8352ab4f369b04c0c452eb5ad08f8cb509ab4fa8b

The checkpoint was trained on ImageNet with ImageNet-22K initialization
lineage. ImageNet's access agreement limits use of the dataset to
non-commercial research and education. No dataset file is included. The
Artistic checkpoint is excluded because it also uses undisclosed private data.