This dataset contains multi-turn conversations converted to ShareGPT format,
parsed from a 'chosen' column.
Each entry represents a multi-turn conversation between a human and an AI assistant.
Conversations are parsed from a single string column using "मानव:" and "सहायक:" delimiters.
A consistent system message "You are a helpful AI assistant that answers questions in Nepali." is prepended.
Filtered to ensure… See the full description on the dataset page:
https://huggingface.co/datasets/dineshkarki/anthropic-hh-rlhf-nepali-sharegpt.
