models

models copied to clipboard

The other repos all use `tox` for running tests in cpu (github)/gpu (jenkins) but we didn't get to Models. We should bring this up to speed with the other repos...

nv-alaiacano

### Bug description Applying batch-predict on a data frame containing multi-hot / sparse features throws the following error ``` funcname = ' self._fetch_chunk() merlin/models/loader/backend.py:368: _ _ _ _ _ _...

sararb

# 🚀 Feature request - Since Inbatch sampler converts the retrieval task to multi-class classification, we need an easy functionality to be able to use `class_weight` and `sample_weight` args in...

rnyak

Examples from Transformers4Rec in Merlin Models (Getting Started, End to end)

5

viswa-nvidia

### Description See https://docs.google.com/document/d/1Zi0SNHHiN72wYIwpkKx-u5xqC35YvlZjNTKUosRuQds/edit

bschifferer

Example of dataloading pretrained embeddings in Merlin Models

1

- [ ] Add an example demonstrating support for pretrained embeddings for DLRM. - [ ] Add an example for usage of pre-trained embeddings with session-based recommendation (e.g. SIGIR Challenge...

viswa-nvidia

Create integration tests for session-based recommendation

2

### Description Based on the training/eval script created in #806 , create unit/integration tests that ensures the stability of the API when setting the advanced hparams and also allowing to...

gabrielspmoreira

viswa-nvidia

viswa-nvidia

viswa-nvidia