Term weights non-deterministic?

AdeDZY / DeepCT

DeepCT and HDCT uses BERT to generate novel, context-aware bag-of-words term weights for documents and queries.

BSD 3-Clause "New" or "Revised" License

312 stars 46 forks source link

Term weights non-deterministic? #12

Open RameshArvind opened 3 years ago

RameshArvind commented 3 years ago

Hello,

First off thank you the provided code!

I've been trying to use the code to produce the term weights for some passages and I've noticed that re-running the code for the same passage produces different results each time.

I've made a google colab notebook here to help reproduce the issue. Please take a look and let me know if my usage of the code is correct. I'm not able to understand why the results are different on every run.

FedericoCampe8 commented 2 years ago

I second the above comment, I have an idea of why it is not deterministic but, what is the suggested why to deal with it? Let's say I want to extract the most important tags from a document based on context...