Good Day Kyoungmun Chang ; re: How to identify modified files (pytorch model or tensorflow model). What have tried so far, and with what results? I am not a specialist in PyTorch or TensorFlow, but all files are essentially text, and text can be compared in toto and line by line. Most systems have tools to compare and check for differences.

• Information on TensorFlow Data Validation here: https://www.tensorflow.org/tfx/guide/tfdv
• PyTorch is built based on python and torch libraries so here is a starting point: https://www.geeksforgeeks.org/compare-two-files-line-by-line-in-python/
• Here is a discussion on comparing TensorFlow graphs by export to a text format and comparing line by line or by visualize both graphs in TensorBoard: https://stackoverflow.com/questions/48192441/what-is-the-easiest-way-to-compare-2-tensorflow-graphs-stored-in-meta-files
• At a much simpler level, a duplicate file checker can be used for a simple determination of identical files, based on combinations of factors beyond simple file names and file dates by using file sizes and content (cyclic redundancy check: crc32). I have used DoubleKiller (link below) for years (free and pro versions are available).
• More detailed change detection can be done with operating system (OS) commands. Depending on your OS [for example: file compare (fc) (Windows) and difference (diff) (unix-like)]:
• https://www.bigbangenterprises.de/en/doublekiller/
• https://www.howtogeek.com/206123/how-to-use-fc-file-compare-from-the-windows-command-prompt/
• https://www.ibm.com/docs/ko/aix/7.2?topic=files-comparing-diff-command

Cheers,

Leo

Dear Leo:

Thank you so much for your detailed help and kindness

I will refer to the URLs you recommended.

If I find more question, I will reach you here again

Have a nice week and thank you again

Take care