tomgrek / zincbase Goto Github PK

Dependabot couldn't authenticate with https://pypi.python.org/simple/.

You can provide authentication details in your Dependabot dashboard by clicking into the account menu (in the top right) and selecting 'Config variables'.

View the update logs.

I suggest you use setuptools_scm to automatically align __version__ in the Zincbase module with the version property of the package to the git tags.

Sorry to open this as an issue, but I don't know of a better channel for ZincBase talk.

I am, wondering what is a good visualisation front end for ZincBase knowledge trees? From my understanding ZincBase used underneath networkx, so matplotlib is obviously a choice, but it is not very dynamic and easy to use for browser output.

Does anybody have experience with visualising Zincbase trees in some front ends?

Really cool project and loved your podcast!

I was tinkering around with the get_most_likely function as follows:

    kb.from_csv('.//assets//countries_s3_train.csv',delimiter='\t')
    kb.build_kg_model(cuda=False, embedding_size=40)
    kb.train_kg_model(steps=2000, batch_size=1, verbose=True)
    print(kb.get_most_likely('canada', 'locatedin', '?'))

And this results in:
[{'triple': ('canada', 'locatedin', 'central_asia'), 'prob': 0.8538}]

Unless I have my geography wrong :), do you think this is a result of the data being faulty? Or could I have done something wrong?