Maarten Grootendorst

Hmmm, that's interesting since I cannot seem to find how adding `seed_topic_list` would result in randomness... Could you share a reproducible example for me to test out?

It isn't necessarily *wrong* to approach it like just, it's just important to be aware that the layers are independent of one another. Since you used three different models, they...

I would generally advise using the `merge_models` functionality for this as it allows for training new models and iteratively merging them. This would also make it a bit more flexible...

In my experience, I seldom have to change the parameters of UMAP to get the kind of dimensionality reduction that I need. The only reason to do so if the...

I tested it with the following and it doesn't seem to work for me. # Training **Data** ```python from datasets import load_dataset dataset = load_dataset("CShorten/ML-ArXiv-Papers")["train"] # Extract abstracts to train...

Thank you for making the changes. I noticed something odd. When I run the following before deleting a topic: ```python topic_model.transform(["a document about networks deep neural network"]) ``` I would...

Odd, I can't seem to reproduce that problem in another environment. I will check in the old environment to see if I did anything special there. Note that you actually...

> 1: Regarding "your scores before and after deleting a topic are different, which shouldn't be the case if the topic embeddings remain the same", a static probability score is...

Apologies for the delay. Lately, finding the time to work on this has become more difficult. That said, if it's just a deepcopy of itself, then it is likely related...

@shuanglovesdata I'm going to keep saying this, but sorry for the delay. I'll do my best to find a bit more time to get this merged! In that case, yes,...