Closed gizemcaylak closed 3 years ago
This is great, thanks @gizemcaylak. I'll review yet, but on first glance seems you solved the size()
issue we discussed by serializing the input file? That seems like a good solution.
No concerns on the failed tests by the way, there are currently some config issues.
This is great, thanks @gizemcaylak. I'll review yet, but on first glance seems you solved the
size()
issue we discussed by serializing the input file? That seems like a good solution.
Thanks! Yes, by looking at the existing examples, I thought it would be nicer to do that, noting that the number of steps is indicated in the config.
Thanks again for this one @gizemcaylak! Now merged after a major update to Birch yesterday. I'll do some additional debugging in the coming days to fix any backward compatibility issues for this example.
This pull request adds Latent Dirichlet Allocation model to the examples. It is tested with a toy example and a dataset that contains the abstracts of NIPS papers (2017). The link to the NIPS papers dataset can be found in the
README.md
.