kukuzaai/alpaca-DPO-luganda
Domain:
natural language processingRecord type:
dataset
Creator:
kuk
Host:
This dataset is a Luganda version of the Alpaca dataset, formatted for Direct Preference Optimization (DPO) or standard instruction fine-tuning.
Language: Luganda (lg)
Format: JSON (Alpaca style)
Repository: reuben256/alpaca-DPO-luganda
This dataset was created to improve the performance of LLMs in the Luganda language with DPO or SFT.