jbaczek

Results 10 comments of jbaczek

How to make Sampler avoid a known-to-fail parameter configuration?

Any progress on this issue? I have the same problem.

Handle float limit_val_batches

LGTM

Handle float limit_val_batches

jenkins

Add mcore full TE transformer layer spec

jenkins

Add mcore full TE transformer layer spec

jenkins

Add mcore full TE transformer layer spec

@ericharper Jenskins tests fail because of the [fixed version of mcore](https://github.com/NVIDIA/NeMo/blob/main/Jenkinsfile#L88). What should we do about it?

Add mcore full TE transformer layer spec

jenkins

Add mcore full TE transformer layer spec

jenkins

[Temporal Fusion Transformer (TFT)] Large dynamic ranges: Only time series with large values are trained. FP64 needed?

If I understand correctly you have a problem with encoding the dataset properly to work with TFT. This is not a bug in the TSPP nor in the TFT. TFT...

[Temporal Fusion Transformer (TFT)] Large dynamic ranges: Only time series with large values are trained. FP64 needed?

We are aware of the problem with preprocessing speed and we track it on our side. Some improvements will come in the next release. If your gradients explode try to...