search

worldbank / REaLTabFormer

A suite of auto-regressive and Seq2Seq (sequence-to-sequence) transformer models for tabular and relational synthetic data generation.
https://worldbank.github.io/REaLTabFormer/
MIT License
203 stars 23 forks source link

No progress in model training for large number of columns #17

Closed vinay-k12 closed 1 year ago

vinay-k12 commented 1 year ago

I'm trying to train the model on transactional data that has around 78 columns with numerical, text and categorical columns. I am trying to modify the training size by modifying rows and columns, but still, training is not progressing at all.

vinay-k12 commented 1 year ago

Here is the error:

image
virajpwr commented 1 year ago

I get the same error. It has to do with the dependency from transformers. Where you able to resolve the issue?

avsolatorio commented 1 year ago

@vinay-k12 @virajpwr, could you please share the transformers version you have on your system? Thanks!

avsolatorio commented 1 year ago

@vinay-k12 @virajpwr, closing this as release v0.1.3 should solve this issue.