search

HunterMcGushion / hyperparameter_hunter

Easy hyperparameter optimization and automatic result saving across machine learning algorithms and libraries
MIT License
704 stars 100 forks source link

Interface with pytorch or tensorflow models #209

Open arita37 opened 4 years ago

arita37 commented 4 years ago

What are the minimal methods needed to interface with hyperparameter_hunter with pytorch and tensorflow models ?

tahnks

HunterMcGushion commented 4 years ago

Thanks for asking, @arita37! Unfortunately, I haven't gotten around to adding support for PyTorch/TensorFlow yet. Keras is supported because its SKLearn API made keeping track of all the hyperparameters a bit easier. If you're interested in looking into supporting PyTorch/TensorFlow, and love to work with you to get that done!

arita37 commented 4 years ago

Am working on similar concept than yours..

Although your tool is more polished, Mime is more experimental but support pytorch and tensorflow.

https://github.com/arita37/mlmodels/tree/dev

I like your code design, Simple and efficient.

We can discuss.

On Jan 19, 2020, at 6:50, Hunter McGushion notifications@github.com wrote:

 Thanks for asking, @arita37! Unfortunately, I haven't gotten around to adding support for PyTorch/TensorFlow yet. Keras is supported because its SKLearn API made keeping track of all the hyperparameters a bit easier. If you're interested in looking into supporting PyTorch/TensorFlow, and love to work with you to get that done!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.