Deeplearning4j: ParagraphVectors inferVector does not return vectors

[XilunChen]

I'm trying to infer the vector for a new document using a trained paragraph vector model, but the infer method does not give me the vector, I checked in function
public INDArray inferVector(List document, double learningRate, double minLearningRate, int iterations)
returns null;

and the new document label is randomly generated, and not included in the vocabulary

How can I get the vector for a new document using inferVector method?

Thanks

[raver119]

I'm sorry, but inferVector meant to be protected for now. Implementation wasn't finished yet, and will be available in next release.

[XilunChen]

Hi Raver119,

Thanks for the quick reply, looking forward to the next release!

[alexfloyd]

Same here, thanks raver!

[andresgs77]

To infer a vector for a new document given a trained model is a key functionality that will boost the api usage.

@raver119 do you have an approximate date for the next release?

[raver119]

Sorry, right now main priority is CUDA performance & multi-gpu scaling. After it's finished - i'll get back to NLP.

p.s. and fp16 support...

[raver119]

Implemented. Example is here: https://github.com/deeplearning4j/dl4j-examples/pull/234

Will be merged soon™ and will be available with 0.6.0 release next week.

[lock[bot]]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.