dino.txt inference code (facebookresearch#528)

* dino.txt (inference) code release (facebookresearch#1)

* dino.txt

* linter

* nnit

* incorrect model size fix

* tokenizer

* tokenizer

* update

* cleanup

* add newline

* update text encoder name

* updatwe tokenizer and name

* notebook

* pr feedback

* remove tokenizer

* nit and fixes

* \n

* \n

* import order

* black

* lint

* black

* remove tokenizer

* update readme

* notebook

* notebook update

* update hubconf

* update readme

* rearrange README

* PR feedback

* README

* Update README

* dinotxt

* in1k

* dinotxt notebook with IN Zero-shot

* ZS segmentation

* place holder

* update README

* Add place holder

* Docstring for DinoTxt encode_{image,text}

* remove segmentation

* vocab and license

* vocab and license

* update notebook

* split into separate files

* reduce docstring

* cleanup notebook

* black format

* improve tokenizer

* put back registers in readme

* black

* notebok

---------

Co-authored-by: Federico Baldassarre <baldassarre.fe@gmail.com>

* flake8 fix

---------

Co-authored-by: Federico Baldassarre <baldassarre.fe@gmail.com>

dino.txt inference code (facebookresearch#528)

* dino.txt (inference) code release (facebookresearch#1)

* dino.txt

* linter

* nnit

* incorrect model size fix

* tokenizer

* tokenizer

* update

* cleanup

* add newline

* update text encoder name

* updatwe tokenizer and name

* notebook

* pr feedback

* remove tokenizer

* nit and fixes

* \n

* \n

* import order

* black

* lint

* black

* remove tokenizer

* update readme

* notebook

* notebook update

* update hubconf

* update readme

* rearrange README

* PR feedback

* README

* Update README

* dinotxt

* in1k

* dinotxt notebook with IN Zero-shot

* ZS segmentation

* place holder

* update README

* Add place holder

* Docstring for DinoTxt encode_{image,text}

* remove segmentation

* vocab and license

* vocab and license

* update notebook

* split into separate files

* reduce docstring

* cleanup notebook

* black format

* improve tokenizer

* put back registers in readme

* black

* notebok

---------

Co-authored-by: Federico Baldassarre <baldassarre.fe@gmail.com>

* flake8 fix

---------

Co-authored-by: Federico Baldassarre <baldassarre.fe@gmail.com>