NEW MODEL + DATASET! :)

Check out Enigma, our new code-instruct model:
- trained on synthetic code-instruct data created with Llama 3.1 405b
- high quality code-instruct within the Llama 3.1 Instruct format

The model:
ValiantLabs/Llama3.1-8B-Enigma

The dataset:
sequelbox/Tachibana


Enjoy! We've got more new datasets and models to follow soon.