Zhirui (Jerry) Wang

The problem is `failedGeocodeReturn` function returns a named vector instead of a data.frame, this will cause problem when `ldply` function wants to combine named vectors and data.frames. Already created a...

Just a wild suggestion, maybe you can also add SAS4Rec by using tf.keras.layers.MultiHeadAttention. I assume it will be a very similar block structure once LSTM and GRU blocks are done

Hi @EvenOldridge, I actually was trying to explore applying SHAP to Merlin models myself earlier. Since Merlin models made modifications to Keras input, output, and layering structure(Blocks), I can't directly...

Hi team! Wonder if there's any updates on this ticket? With the Merlin-models' getting ready for sequential/session-based models, it's pretty crucial to pair it with Nvtabular to preprocess sequential data...

@rnyak Thanks! The first case is working as expected, but the second case is no longer working... However, I think I should switch to InputBlockV2 instead, so I tried switching...

And in terms of use cases, we have both session-based and sequential recommendation in mind, and both RNN-based and Transformer-based models in mind. We noticed that RNN-based models, especially bi-LSTM,...

Thanks! Yes this issue has been resolved. Will check out the example notebook!