search

aksnzhy / xlearn

High performance, easy-to-use, and scalable machine learning (ML) package, including linear model (LR), factorization machines (FM), and field-aware factorization machines (FFM) for Python and CLI interface.
https://xlearn-doc.readthedocs.io/en/latest/index.html
Apache License 2.0
3.09k stars 519 forks source link

Segmentation fault error; distributed learning #284

Closed 4ound closed 5 years ago

4ound commented 5 years ago

  1. Number of features * k > uint32 You'll get error "Segmentation fault" while training I read your code and found two lines, where typedef uint32 index_t; I've changed it to typedef uint64 index_t; After that everything worked ok. But I'm not sure, is it correct? Could it broke some computations?

  2. Distributed learning Judging by your docs, model should support distributed training. Could you tell the idea of implementation? Will you implement it soon?

aksnzhy commented 5 years ago

@4ound Hi, thanks for the attention of xLearn.

  1. We use uint32 by default because of the performance concern. You can change it to uint64 for your case.

  2. For distributed training, we have an internal version of that but I think it's not ready to release. I will release it in the future, but we don't have a concrete deadline.

Thanks!

4ound commented 5 years ago

Thanks