viswa-nvidia

Results 163 comments of viswa-nvidia

[BUG] nvtabular is sensitive to the order of imports of nvtabular and horovod modules

@eordentlich , @sohn21c , a reproducer is available above that doesn't involve NVT. @benfred , please post the link to the relevant slack conversation

[BUG] Mismatch between total number of items and length of prediction logit values (end-to-end-session-based notebook)

@hosseinkalbasi , can we close this ticket? Is your issue resolved?

Adding better doc-strings for retrieval & ranking models

Moved to arbitration. Not clear how this maps to the r[oadmap](https://github.com/orgs/NVIDIA-Merlin/projects/14/views/4) . @marcromeyn , please clarify.

Adding better doc-strings for retrieval & ranking models

Moved to arbitration. Not clear how this maps to the r[oadmap](https://github.com/orgs/NVIDIA-Merlin/projects/14/views/4) . @marcromeyn , please clarify.

Adding better doc-strings for retrieval & ranking models

We have a new API so this one can be closed

Enable dynamic memory growth

@marcromeyn , please update the status on this.

Enable dynamic memory growth

@gabrielspmoreira , please procide mising details here

[BUG] model.evaluate() gives much higher acc value for small batch size when we set `validation_data` int the model.fit()

@marcromeyn , the PR is marked closed. What is the next step?

[BUG] model.evaluate() gives much higher acc value for small batch size when we set `validation_data` int the model.fit()

@marcromeyn , please create a PR and link it

[BUG] Target in schema causes incorrect down scoring of false negative.

@bschifferer , could you confirm if this is a P0 or P1. the priority field said P0 and the label was P1