datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
reflection-50m
SPP Reflection 50M
The 51.4M-document reflection set from Synthetic Persona Pretraining (SPP):
Alignment from Token Zero — the production half-corpus run, and the dataset the
released models were actually trained on.
🔬 Small sample (same format): dlab-spp/reflection-sample-2k
📉 Earlier 10M run: dlab-spp/reflection-10m
🧾 Safety scores for the full 1T corpus: dlab-spp/safety-classifications
Each row pairs a source document with two generated constitution reflections — a… See the full description on the dataset page: https://huggingface.co/datasets/dlab-spp/reflection-50m.reflection-10m
SPP Reflection 10M
The full ~10M-document reflection set from Synthetic Persona Pretraining (SPP):
Alignment from Token Zero.
📝 Read the post: Synthetic Persona Pretraining: Alignment from Token Zero
🔬 Small sample (same format): dlab-spp/reflection-sample-2k — a 2,000-row sample drawn from this set, for quick inspection.
Each row pairs a pretraining document with a synthetic, value-laden reflection
generated for it: a short first-person (and third-person) moral reflection… See the full description on the dataset page: https://huggingface.co/datasets/dlab-spp/reflection-10m.orca-math-word-reflection
Dataset Card for "orca-math-word-reflection"
Dataset Summary
The Orca-Math Word Problems with Reflection dataset is an extension of subset of the original ORCA Math Word Problems 200k dataset. This new version introduces a "Thinking and Reflection" format designed to enhance problem-solving approaches by encouraging step-by-step thinking before producing a solution.
In this dataset, each math word problem and its corresponding solution from the original dataset are… See the full description on the dataset page: https://huggingface.co/datasets/Harshkmr/orca-math-word-reflection.reflection-sample-2k
SPP Reflection 2k Sample
A 2,000-row sample (seed 42) of dlab-spp/reflection-10m,
in the identical format, for quick inspection of the data from
Synthetic Persona Pretraining (SPP): Alignment from Token Zero.
📝 Read the post: Synthetic Persona Pretraining: Alignment from Token Zero
📦 Full dataset: dlab-spp/reflection-10m (~10M documents).
Each row pairs a pretraining document with a synthetic, value-laden reflection
(first- and third-person) grounded in a value constitution.… See the full description on the dataset page: https://huggingface.co/datasets/dlab-spp/reflection-sample-2k.reflection-v1-ru_subset
d0rj/reflection-v1-ru_subset
Translated glaiveai/reflection-v1 dataset into Russian language using GPT-4o.
Almost all the rows of the dataset have been translated. I have removed those translations that do not match the original by the presence of the tags "thinking", "reflection" and "output". Mapping to the original dataset rows can be taken from the "index" column.
Usage
import datasets
data = datasets.load_dataset("d0rj/reflection-v1-ru_subset")
print(data)… See the full description on the dataset page: https://huggingface.co/datasets/d0rj/reflection-v1-ru_subset.model-raising-reflection-end-eval
model-raising-reflection-end-eval
A held-out evaluation set for charter-guided pretraining reflections, placed at the
document end (reflection_end). Each row is one dolma3 web document plus a paired
first-person / third-person reflection that cites charter sections ([X.Y]) where the
document substantively engages with them. Generated with the frozen production pipeline
(Qwen3.5-35B-A3B-FP8, prompt generator_reflection_v7.md, charter
ModelRaisingConstitution v0.2)
so the gold… See the full description on the dataset page: https://huggingface.co/datasets/jkminder/model-raising-reflection-end-eval.SFT_DATA-cd3args-ablation-Qwen2.5-1.5B-Instruct-no_reflectionsYou can train using these datasets with LLaMA-Factory if you add this to your data/datasets.json files.
"example_dataset": {
"hf_hub_url": "SkillFactory/SFT_DATA-cd3args-ablation-Qwen2.5-1.5B-Instruct-no_reflections",
"formatting": "sharegpt",
"columns": {
"messages": "conversations"},
"tags": {
"user_tag": "user",
"assistant_tag": "assistant",
"role_tag": "role",
"content_tag": "content"
},
"subset": "sft_train"
}
glaiveai-reflection-v1-koTranslated glaiveai/reflection-v1 using nayohan/llama3-instrucTrans-enko-8b.
For this dataset, we only used data that is 5000 characters or less in length and has language of English.
Thanks for @Magpie-Align and @nayohan.