7shi / tanuki-dpo-v1.0

Tanuki-8B is a LLM with about 8B parameters that was pre-trained with about 1.3T tokens using full-scratch training.

933 Pulls Updated 2 weeks ago