datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
character-training-model-spec
Character training on the OpenAI Model Spec
Recipe: examples/character · Collection: Character
Graded replies for character training: a model answering the style prompts of the
OpenAI Model Spec (8 traits) under a bare
deployment prompt, judged against each trait's principle. Made by
examples/character
in the while-ai SDK (0.24); the recipe is
docs/character-training.md
and the page is docs.withwhile.com.
split
rows
prompts
pass rate
what
train
60
15
0.72
the spec's… See the full description on the dataset page: https://huggingface.co/datasets/while-ai/character-training-model-spec.character-training-model-spec
character-training-model-spec
Character training rows: Qwen3-4B-Instruct answering the OpenAI Model Spec's style prompts (8 traits) under a bare deployment prompt, 4 replies each, judged by hosted Phi-4 against the trait's principle. reward = trait AND on_task; markers trait, on_task, no_filler. Made by https://github.com/Zero-Proof-AI/zeroproof-sdk/tree/main/examples/character.
Agent rollouts exported from ZeroProof for sol-character. Each row is one rollout: the prompt, the… See the full description on the dataset page: https://huggingface.co/datasets/jaweiss2305/character-training-model-spec.character-training-data
License
This dataset follows the same license used in LIMA, which is a source of several prompts. "If the source data of LIMA has a stricter license than CC BY-NC-SA, the LIMA dataset follows the same. Otherwise, it follows the CC BY-NC-SA license."
character-llm-training-data