You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When I look trained weights.npy file to use segmenter, this data uses TensorFlow dtype (https://www.tensorflow.org/api_docs/python/tf/dtypes/DType). So it means that using weights.npy requires TensorFlow and no way to use this data via other languages without TensorFlow binding.
Also, although you provide JSON file too, this is big due to not binary format.
So I am happy if you provide TensorFlow-free npy (or other compressed binary format) data that is trained.
The text was updated successfully, but these errors were encountered:
The JSON file is not intended to be the final shipped artifact. You can convert the JSON into some other format that is more compact using any library of your choice.
When I look trained
weights.npy
file to use segmenter, this data uses TensorFlow dtype (https://www.tensorflow.org/api_docs/python/tf/dtypes/DType). So it means that usingweights.npy
requires TensorFlow and no way to use this data via other languages without TensorFlow binding.Also, although you provide JSON file too, this is big due to not binary format.
So I am happy if you provide TensorFlow-free npy (or other compressed binary format) data that is trained.
The text was updated successfully, but these errors were encountered: