datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
kalo-opus-instruct-22k-no-refusalOpenHermes-NoRefusal-95K
OpenHermes-NoRefusal-95K
A refusal-free instruction-tuning dataset: 95,401 single-turn conversations derived from
teknium/OpenHermes-2.5, filtered so
that zero assistant responses contain refusals, hedging boilerplate, or
"as an AI language model" disclaimers.
Why this exists
The usual way to get a model that doesn't refuse is to train it on aligned data and then
remove the alignment afterwards — refusal-direction ablation, weight editing, abliteration.
That works… See the full description on the dataset page: https://huggingface.co/datasets/ghost-actual/OpenHermes-NoRefusal-95K.kalo-opus-instruct-22k-no-refusal-no-systemIntel-DPO-Pairs-Norefusalsm-a-p_CodeFeedback_norefusals_ShareGPTkalo-22k-norefusal-kto-combinedchat-alpaca-pl-no-refusalsDataset from https://github.com/cascip/ChatAlpaca translated using Hermes 3 405B Instruct(openrouter free), then all the refusals and moralizing was removed. To be used to train chat models in Polish language.
trendyol-fenrir-merged-no-refusal