Is model.fit(min_iter=10, max_iter=1000) single-threaded? Can it be parallelized for faster

[wubaosheng]

Hi, thanks for this great tool!

I'm currently using the following line to fit the model:

model.fit(min_iter=10, max_iter=1000)
However, the fitting process is very slow and appears to be single-threaded.
I would like to ask:

Is the .fit() function designed to run in a single thread?

Is there any built-in support for multi-threading or parallel computation?

If not, are there any recommended strategies to speed it up (e.g., batching, using joblib, etc.)?

My dataset is large (~60,000 cells), so performance is becoming a bottleneck.

Any suggestions or future plans for parallelization would be greatly appreciated. Thanks!