updated the generator to use temperature and sample

updated the readme

trained another 12 epochs

fixecd the tokenizers and vocab file

updated the model (it. 36000)

update readme

added README.md

added bert2bert model (it. 7500)

initial commit