datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
S1-Omni-Corpus-10K
S1-Omni-Corpus-10K
An open-source scientific multimodal reasoning dataset subset for S1-Omni
🧬 Model Introduction
S1-Omni is a unified scientific multimodal reasoning model for scientific understanding, prediction, and generation. It is developed by the ScienceOne AI team of the Chinese Academy of Sciences.
S1-Omni addresses fragmented scientific AI capabilities with a shared backbone for cross-disciplinary, cross-modal, and cross-task understanding and reasoning… See the full description on the dataset page: https://huggingface.co/datasets/ScienceOne-AI/S1-Omni-Corpus-10K.S1-DeepResearch-15k
S1-DeepResearch-15k Dataset
Overview
The S1-DeepResearch dataset is a curated collection of approximately 15k samples designed to improve deep research capabilities of large language models.
The dataset includes two types of tasks:
Verifiable tasks (labeled as "Closed-ended Multi-hop Resolution")
Open-ended tasks (labeled as "Open-ended Exploration")
Dataset Composition
The dataset is organized into five core capability dimensions:
Long-chain complex… See the full description on the dataset page: https://huggingface.co/datasets/ScienceOne-AI/S1-DeepResearch-15k.cua-s1-forms
cua-s1-forms (dataset)
Synthetic + real training/eval data for cua-ai/cua-s1-forms,
a jev-like one-pass option scorer for GUI form filling behind
cua-driver.
Generator source: cua_s1/synth.py in
https://github.com/trycua/cua/tree/main/libs/cua-s1.
Files
file
rows
source
train.jsonl
~150k
synthetic
validation.jsonl
~18k
synthetic
test.jsonl
~20k
synthetic, form-signature-disjoint from train/validation
demo.jsonl
196
real: 3 real JevBrowser form… See the full description on the dataset page: https://huggingface.co/datasets/cua-ai/cua-s1-forms.fusion-synth-data-s1kx
Offline Synthetic Data (s1K-X) for: Making, not taking, the Best-of-N
Content
This data contains completions for the s1K-X training split prompts from 5 different teacher models and 2 aggregations:
Teachers: We sample one completion from each of the following models at temperature T=0.3. For kimik2, qwen3, and deepseek-v3 we use TogetherAI, for gemma3-27b and command-a we use locally hosted images.
gemma3-27b: GEMMA3-27B-IT
kimik2: KIMI-K2-INSTRUCT
qwen3: QWEN3-235B… See the full description on the dataset page: https://huggingface.co/datasets/CohereLabs/fusion-synth-data-s1kx.dclm-fix-dedup-s1Multilingual-S1UI_S1_dataset
Introduction
This repository contains the dataset example for UI-S1-7B, presented in UI-S1: Advancing GUI Automation via Semi-online Reinforcement Learning.
Project page: https://github.com/X-PLUG/MobileAgent/tree/main/UI-S1
metrics-outputs-pcfg-matryoshka-fmt-2400-s1-token-cacheNatural-Reasoning-gpt-oss-120B-S1
Dataset Card: Natural-Reasoning-gpt-oss-120B-S1
📜 Dataset Overview
This is a meticulously curated instruction fine-tuning dataset designed specifically for efficient knowledge distillation tasks. Built upon the first 100,000 questions from the large-scale reasoning corpus facebook/natural_reasoning (s1, I will process the remaining parts later), it aims to transfer the advanced, multi-step reasoning capabilities of the teacher model gpt-oss-120-high to a student model… See the full description on the dataset page: https://huggingface.co/datasets/Jackrong/Natural-Reasoning-gpt-oss-120B-S1.Amazon-Beauty-S1This dataset is derived from Amazon Reviews'23 [1] Beauty category. The split is standard leave-one-out: the last item is the test target, the second-to-last is the validation target, and everything before that is training.
The training portion is expanded by sliding window — every prefix becomes one example — so a user with a sequence of length $L$ contributes $L-3$ training rows with histories of length $1 \dots L-3$, one validation row and one test row.
Statistics… See the full description on the dataset page: https://huggingface.co/datasets/Abner0803/Amazon-Beauty-S1.s1K-1.1-deepseek-cot
s1K-1.1 (DeepSeek-R1 traces) — format cho SegmentSelectiveSFT
Chuyen doi tu simplescaling/s1K-1.1
bang prepare_s1k.py (default flags) trong repo SegmentSelectiveSFT.
Moi dong jsonl co 3 truong:
Truong
Nguon
question
question
solution
deepseek_thinking_trajectory (long-CoT trace cua R1)
answer
\\boxed{...} cuoi cung trong trace, fallback ve solution cua s1K
Giu 934 / 1000 mau — bo cac mau khong co trace, khong co dap an, hoac dap an dai hon 200 ky tu.
from… See the full description on the dataset page: https://huggingface.co/datasets/baesad/s1K-1.1-deepseek-cot.ANGO-S1ANGO is A Novel Generation-Oriented Chinese LLM evaluation benchmark.
We introduces the format of single-question multiple-keypoints dataset for the first time, which include 171 keypoints accumulated in 4 hierarchical levels and 9 difficulty categories.
The data were exclusively obtained from the Administrative Proficiency Test, which serves as a significant component of the Chinese civil service examination.
We will apply a seasonal system for the leaderboard, updating them every two months.… See the full description on the dataset page: https://huggingface.co/datasets/AngoHF/ANGO-S1.ruizhe1217__sft-s1-qwen-0.5b-details
Dataset Card for Evaluation run of ruizhe1217/sft-s1-qwen-0.5b
Dataset automatically created during the evaluation run of model ruizhe1217/sft-s1-qwen-0.5b
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/ruizhe1217__sft-s1-qwen-0.5b-details.s1-test-time-scaling-synth-public
s1-test-time-scaling-synth: Japanese and English Reinforcement Learning Dataset Derived from the s1 Simple Test-Time Scaling Dataset
This repository contains s1-test-time-scaling-synth, a reinforcement learning dataset in Japanese and English.This dataset is built upon the supervised fine-tuning dataset simplescaling/data_ablation_full59K (hereafter, the "original dataset"), originally developed in "s1: Simple test-time scaling" [Muennighoff+, EMNLP25].
The original dataset is a… See the full description on the dataset page: https://huggingface.co/datasets/tokyotech-llm/s1-test-time-scaling-synth-public.matmoe-s1-dataarch-entity-matrix-lpi-260904T0010-plain_italy_n1000_e3_s1metrics-outputs-pcfg-matryoshka-fmt-2308-s1-token-cacheless-is-moe-s1-calibration-128
Less-is-MoE S1K calibration data — 128 full-length samples
This repository contains the exact 128 S1K rows selected for Less-is-MoE
full-model pruning. The selection reproduces the released loader:
source: yentinglin/s1K-1.1-trl-format
revision: 58a01564d278477da20ead1bcf1cde8e31f36251
split: train
order: Dataset.shuffle(seed=1234)
samples: first 128 nonempty messages rows
sequence-length limit: none
truncation: disabled
padding: disabled
calibration.jsonl stores every… See the full description on the dataset page: https://huggingface.co/datasets/jayzou3773/less-is-moe-s1-calibration-128.metrics-outputs-pcfg-matryoshka-fmt-1667-s1-token-cachearch-code-scale-lpi-260904T0500-code_block_only_superintelligence_n1000_e3_s1_a0.5metrics-outputs-pcfg-matryoshka-fmt-0000-s1-token-cacheS1_QFFT
📘 S1–QFFT
S1–QFFT is a question-free version of the original simplescaling/s1K-1.1 dataset, designed for QFFT training workflows.
🔍 Description
This dataset discards the original questions and any system instructions, keeping only the reasoning completions as supervision. It is especially useful for models that aim to learn when and how to think, rather than just how to answer.
The dataset is fully converted into a format compatible with LLaMA-Factory training.… See the full description on the dataset page: https://huggingface.co/datasets/lwl-uestc/S1_QFFT.black_mirror_scripts_S1-5Black Mirror Scripts Dataset (Seasons 1-5)
This dataset, titled 'black_mirror_scripts_S1-5.csv', contains the meticulously compiled transcripts of the critically acclaimed anthology series Black Mirror, covering Seasons 1 through 5. Each entry in this dataset is categorized by unique identifiers including Script ID, Title, Scene, Dialogue, and Timestamp, making it an ideal resource for natural language processing tasks, script analysis, sentiment analysis, and more.
Dataset Composition
Our… See the full description on the dataset page: https://huggingface.co/datasets/tmobley96/black_mirror_scripts_S1-5.s1-1aTesting a dataset conversion
openjev-healthcare-router
Healthcare router: a typed-decision benchmark with real headroom
Synthetic pharmacy messages, routed by ten questions answered together: one
intent choice, five multi-label noul topic flags, and four safety gates.
988 items across 18 tiers, 450 in test.
Built for OpenJev to compare
against a commercial typed-decision API, and deliberately built to be hard.
Why it exists
The first version of this task was useless. A commercial API scored 1.000
on four of its tiers… See the full description on the dataset page: https://huggingface.co/datasets/s1lv3rj1nx/openjev-healthcare-router.s1K-1.1-850This data is obtained by simplescaling/s1K-1.1.
Compared with the original simplescaling/s1K-1.1 data, our filtered data uses less data and achieves better results.
What we did
Text Embedding Generation: We use all-MiniLM-L6-v2 (from SentenceTransformers library) to generate "input" embeddings.
Dimensionality reduction: We use UMAP approach which preserves local and global data structures.
n_components=2, n_neighbors=15, min_dist=0.1
Data Sparsification (Dense Points… See the full description on the dataset page: https://huggingface.co/datasets/InfiX-ai/s1K-1.1-850.tt641-v3-seedsweep-s13
TT639G Recombined Tiny Assistant v1
Recombines isolated proof rungs:
TT638D code behavior + dyadic/Mercy proof upstream
TT639E2 context-copy behavior
TT639F3 task-routing behavior
simple rule/Q&A behavior
Blocking dense gates:
seen_combined_pass
upstream_regression_pass
mixed_heldout_pass
anti_collision_pass
Do not run dyadic/Mercy compare unless all four gates pass.
flock-off-s1-text-2-sqlflock-off-s1-competitionssrf-oauth-leak-s14
OAuth Secret Leak Test