---

"_Ní tòótọ́, ohun tí o kọ dára ó sì tún dán. Ní ọ̀pọ̀lọpọ̀ ọ̀nà, ohun tí mo fẹ́ kí **Hug** jẹ́ nìyẹn - ó wúni lórí gan-an láti rí ẹnìkan tí ó kọ́ nǹkan bí èyí._"

nim!imanim??u&koo?s!omihs????ya&ko?rih???e&akas?nagamok?subo??i&gakat?h&asa?c&a!mo!nanihs???uonamay??sukagot??k&a&kas?mimanim?to??ia&atik?imanim??oa?uzihcom??m&akawak?ijuf?o!t???r&ato?ijoihs?omakat???n&ana?esnoawazon??o&hukas?n&a&gan?kan??i&hc?muza??ustat??romok?si&gan?k??tomustam??u&k&as?ohukihc??stamega????o&b,m,pac,?to&mamuk!.&a&gamay?rahihsin?sukama!imak??tamanim??enufim?i&hcukik?k&ihsam?u??nugo!imanim??romakat??o&ara?rihsustay?sa?t&amay?om&amuk?us??u!koyg???yohc??u&sagan?zo????yk!.&a&bmatoy...

*   Remove `tf.contrib.data.Iterator.from_dataset()` method. Use
    `Dataset.make_initializable_iterator()` instead.
*   Remove seldom used and unnecessary `tf.contrib.data.Iterator.dispose_op()`.
*   Reorder some TF-GAN loss functions in a non-backwards compatible way.

## Known Issues

*   In Python 3, `Dataset.from_generator()` does not support Unicode strings.