*   `tf.experimental.StructuredTensor`

    *   Introduced `tf.experimental.StructuredTensor`, which provides a flexible
        and Tensorflow-native way to encode structured data such as protocol
        buffers or pandas dataframes.

*   `tf.keras`:

    *   Added method `get_metrics_result()` to `tf.keras.models.Model`.
        *   Returns the current metrics values of the model as a dict.

SFE,WAAA,MAAA,KAAA,WAAA,CAAA,iBAAA,KAAA,WAAA,CAAA,aAAA,KAAA,WAAA,CAAA,WAAA,KAAA,YAIA,uCGdN,KHeQ,WAAA,MdTN,WiBUE,MAAA,QACA,gBAAA,KAjBJ,WAAA,WAsBI,QAAA,EACA,WAAA,EAAA,EAAA,EAAA,MAAA,oBAvBJ,cAAA,cA6BI,QAAA,IA7BJ,mCAkCI,OAAA,QAcJ,epBy8EA,wBoBv8EE,eAAA,KASA,aC3DA,MAAA,KFAE,iBAAA,QEEF,aAAA,QlBIA,mBkBAE,MAAA,KFNA,iBAAA,QEQA,aAAA,QAGF,mBAAA,mBAEE,MAAA,KFbA,iBAAA,QEeA,aAAA,QAKE,WAAA,EAAA,EAAA,EAAA,MAAA,oBAKJ,sBAAA,sBAEE,MAAA,KACA,iBAAA,QACA,aAAA,QAOF,kDAAA,kDrBm/EF,mCqBh/EI,MAAA,KACA,iBAAA,QAIA,aAAA,QAEA...