Skip to content

Latest commit

 

History

History
9 lines (5 loc) · 329 Bytes

File metadata and controls

9 lines (5 loc) · 329 Bytes

language-models-testing

Just experimenting with LLMs using Python.

OtoGPT is a more of a joke program. Its not even a GPT. Someone just asked me to make a Georgian language model so I quickly came up with this idea.

Trigrams is kind of broken. It somewhat works.

The dataset was too large for me to put it here, sorry. ):