MyungJoo Ham

You can do checkout and continue training process, but that's just not based on gradient saving. You can do epoch-based checkpointing (that's what most nntrainer's mobile applications do), but I'm...

For model copy, if there is no copy-constructor for model class and the default behavior does not do what you want, you may try "original.save()" and "cloned.load()". For Polyak update,...

@lhs8928 Are you still working on it?

What is "#cache_pool#.cpp"?

This PR is no longer relevant.

Not needed. 905D3 doesn't have one.

#4135 has implemented this.

> @ksghaier @myungjoo The added testcase get runtime error with current Tizen/Debian tvm package (it's v0.7.0). However, with the recent tvm release (v0.9.0) and slight modification in our source, I...

@myungjoo TODO: add githug project.

There is a team who's working on RUST APIs for the whole Tizen C API. If they really do this, ML-API (== Tizen ML) will have RUST binding, which will...