Hi, sbrunk please review my code add many feature
HI Sbrunk:
Now STorch I have extend many feature , it gain so many line code, I cannot say the detail.
This is so amazing @mullerhai!
Let's get CI running again so we can run the tests there. I think we need to update the workflows a bit as the actions are outdated.
I've started going through the PR. A few initial remarks:
storch-corehas a lot of dependencies right now likestorch-pandas. Ideally I'd like to keep it very small instead and let the other modules depend onstorch-coreif possible or create integration modules.- Related, do we want to keep the
storch-*modules all in separate repositories or does it make sense to move some of them into this repo as submodules?
- Related, do we want to keep the
- The current Scala version is
3.6.3, while LTS, which is recommended for libraries is currently3.3.7. Can we use the LTS version to improve compatibility?
I've started going through the PR. A few initial remarks:
storch-corehas a lot of dependencies right now likestorch-pandas. Ideally I'd like to keep it very small instead and let the other modules depend onstorch-coreif possible or create integration modules.
- Related, do we want to keep the
storch-*modules all in separate repositories or does it make sense to move some of them into this repo as submodules?The current Scala version is
3.6.3, while LTS, which is recommended for libraries is currently3.3.7. Can we use the LTS version to improve compatibility?
Hi @sbrunk , I have remove storch-pandas storch-safe-tensors storch-sklearn , the scala3 version I think because we need use jupyter notebook, the almond.sh scala kernel is support scala3.6.3 ,it could work normally . if use scala3 3.3.7 I don't know the compatiable with jupyter notebook , it just could make decision by youself , I am not care of the scala version . just could work is ok
Hi @sbrunk , I have remove storch-pandas storch-safe-tensors storch-sklearn , the scala3 version I think because we need use jupyter notebook, the almond.sh scala kernel is support scala3.6.3 ,it could work normally . if use scala3 3.3.7 I don't know the compatiable with jupyter notebook , it just could make decision by youself , I am not care of the scala version . just could work is ok
almond should work fine with Scala 3.3 LTS, but we also need to make sure that the libraries storch depends on are also compatible.
When storch and its dependencies are compiled with Scala 3.3 LTS, then other libraries using Scala 3.3/3.4,3.5,3.6,3.7 can use it but when it's compiled with 3.7, all libraries that depend on it need also be compiled to 3.7.
I created https://github.com/bytedeco/storch-numpy/pull/2 to move storch-numpy to 3.3, could you look at the PR?
HI @sbrunk , I have some idea ,to make more people quickly search storch and try to use storch. we need renamed storch as PyTorch too, how do you think?