Data limite size for StructuredDataClassifier

[nanro22]

Hi all !
Are there any size limits for the data when using StructuredDataClassifier in AutoKeras? In my case, I'm working with a dataset of 336,100 rows and 3 columns. However, my Jupyter kernel stops during the FIT with the following error:

AsyncIOLoopKernelRestarter: restarting kernel (1/5), keep random ports

Upon closer inspection, it seems to be related to a RAM limitation, as Python is attempting to use more than 80GB of RAM (RAM + swap). I suspect the issue is due to the number of rows in my CSV. If there are any solutions or optimizations available, I would appreciate any guidance.

Thank you!