rtzr/ko-gemma-2-9b-it
TEXT GENERATIONConcurrency Cost:1Model Size:9BQuant:FP8Ctx Length:16kPublished:Jul 11, 2024License:gemmaArchitecture:Transformer0.1K Warm

The rtzr/ko-gemma-2-9b-it is a 9 billion parameter, instruction-tuned, decoder-only large language model developed by Return Zero Team. Built upon Google's Gemma 2 architecture, this model is specifically fine-tuned for conversational tasks in Korean. It leverages Supervised Fine-Tuning (SFT) and Direct Preference Optimization (DPO) on high-quality Korean datasets, making it proficient in generating Korean-language text responses to various prompts.

Loading preview...