'RNN string input and output, very limited training dateset

Is it possible to create a model which would generate 344 symbols Base64 string based on 12 hex symbols string input ?

I do have about a dozen of samples (correct pair hex input and matching base64 output). I also have a lot of GPUs which I can use for this task. So even if it'll take a long time to train this model it shouldn't be an issue.

Thanks

Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution	Source

'RNN string input and output, very limited training dateset

Sources

Related Questions