search

debjitpaul / Multi-Hop-Knowledge-Paths-Human-Needs

Ranking and Selecting Multi-Hop Knowledge Paths to Better Predict Human Needs (NAACL 2019)
https://www.aclweb.org/anthology/N19-1368
17 stars 9 forks source link

About Requirements #4

Closed victorup closed 4 years ago

victorup commented 4 years ago

Hello, Which version of TensorFlow do you use?

victorup commented 4 years ago

And I get an error trying to run "elmo = hub.Module("https://tfhub.dev/google/elmo/2", trainable=True)":

tarfile.ReadError: unexpected end of data During handling of the above exception, another exception occurred: OSError: https://tfhub.dev/google/elmo/2 does not appear to be a valid module.

Do you know how to solve this problem？ Thanks！

debjitpaul commented 4 years ago

Tensorflow Version: 1.12.0

victorup commented 4 years ago

In your code, I see the weights list as follow: w = [3.929112469627414, 4.352634266669815, 4.105348968927056, 4.009469417408209, 4.436903109491611, 3.4714643441750805, 4.533726764493145, 3.665643259544512, 5.264175448882736, 6.026320782448594, 3.7522367243231805, 3.8019798963053515, 7.896001211803761, 8.024995943144209, 3.3076036095385644, 3.81662584588786, 8.618279130276653, 6.7344516295276895] How did you get this weights?

debjitpaul commented 4 years ago
Screenshot 2020-04-26 at 12 22 22

I would request you to read the paper. It is clearly mentioned that w is the weight distribution learned from the training data.

victorup commented 4 years ago

Thank you for your answer! I don't know how to predict "Plutchik" category?