An instruct based fine tune of migtissera/Tess-34B-v1.4.

It works well with long system prompts.

It works well for reasoning tasks.

This model is trained on a private dataset. The high GSM8K score is NOT because of the MetaMath dataset.

Prompt Format:

SYSTEM: <ANY SYSTEM CONTEXT>
USER: 
ASSISTANT:

Safetensors

Model size

34B params

Tensor type

F32

Model tree for Mihaiii/Pallas-0.2

Base model

Finetuned

(10)

this model