casualjim/abacusai-smaug-34B-v0.1

This model is a finetune of jondurbin's excellent bagel model. It has been trained with new datasets and a new technique, which we will share to the community soon. This model has not utilised any form of merging.

Evaluation Results

Average	ARC	HellaSwag	MMLU	TruthfulQA	Winogrande	GSM8K
77.29	74.23	86.76	76.66	70.22	83.66	72.18

Contamination Results

With reference model jondurbin/bagel-34b-v0.2:

ARC	TruthfulQA	GSM8K
0.08	0.38	0.88