R user interface to TensorFlow Center

We are happy to reveal that the very first variation of tfhub is now on CRAN. tfhub is an R user interface to TensorFlow Center – a library for the publication, discovery, and usage of recyclable parts of artificial intelligence designs. A module is a self-contained piece of a TensorFlow chart, together with its weights and possessions, that can be recycled throughout various jobs in a procedure called transfer knowing.

The CRAN variation of tfhub can be set up with:

After setting up the R plan you require to set up the TensorFlow Center python plan. You can do it by running:

Beginning

The necessary function of tfhub is layer_hub which works similar to a keras layer however enables you to pack a total pre-trained deep knowing design.

For instance you can:

 library( tfhub)

 layer_mobilenet <