bschifferer

@viswa-nvidia I close the ticket, as it is over 1 year old - feel free to reopen, if it is required

I am not sure, if the main ticket is uptodate. In some meetings, we say, that the feature is almost done but there are many Tickets not checked (finished). I...

@viswa-nvidia We talked in the example sync meeting about the ticket: Re: Add an example demonstrating support for pretrained embeddings for DLRM. > We will update this example ( https://github.com/NVIDIA-Merlin/models/blob/main/examples/usecases/entertainment-with-pretrained-embeddings.ipynb...

@viswa-nvidia can we plan it for 23.04 ?

I started a draft for a Merlin Decision Process: https://docs.google.com/document/d/1FmK55tAo0wy0tHUe3eE8xqZKGDYSvy0tvq8Y036854A/edit#

Before we merge, we need confirmation from product, that we can remove.

I created a PR based on a draft I shared a while ago: https://github.com/NVIDIA-Merlin/Merlin/pull/1045

As an success criteria, we need to have benchmarks for each of the point above: - How does throughput change? (E.g. TF Keras vs. SOK. vs TFDE vs. reduced precision...

@angmc @karlhigley @EvenOldridge About: Operator for mapping between key values (Likely a modification to categorify to support an existing mapping) NVTabular has a paramter called `vocab` (https://github.com/NVIDIA-Merlin/NVTabular/blob/main/nvtabular/ops/categorify.py#L208 ). Unfortunately, our...

@angmc @karlhigley @EvenOldridge @viswa-nvidia About: Merging pre-trained embeddings in the dataloaders (https://github.com/NVIDIA-Merlin/Merlin/issues/211) Merging pre-trained embeddings at serving time (https://github.com/NVIDIA-Merlin/Merlin/issues/211) I would not use pre-trained embeddings as input features to the...