datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Phi4-ensemble-teacher-forcing-record-logits-dataDeepSeek-V4-Flash-0731-Teacher-Distillation-40513x
DeepSeek V4 Flash 0731 Teacher Distillation — 40,513 Retained Rows
Teacher-distillation corpus generated with
deepseek-ai/DeepSeek-V4-Flash-0731.
The original manifest contained 45,000 unique seeds.
Following generation, QC, retry-based repair, quarantine auditing,
and recovery adjudication, 40,513 rows were retained.
Composition
Bucket
Rows
Coding
5,601
Agentic
9,982
Cyber blue
13,000
Controlled cyber red
6,999
Tool use
4,931
Total
40,513… See the full description on the dataset page: https://huggingface.co/datasets/trjxter/DeepSeek-V4-Flash-0731-Teacher-Distillation-40513x.teacher-notes-severity
teacher-notes-severity
Synthetic dataset for training a local student-notes severity classifier (fine-tuned from Qwen/Qwen3-1.7B).
Each row is a chat-formatted prompt/completion pair: the user turn is a teacher note (single, compound,
or a cumulative running log), the assistant turn is a JSON label
{"category": "commendation|misbehavior|academic_concern", "severity": <int>, "escalate": <bool>}.
Severity scale: commendations are negative (-95..-10); routine notes 5-55; serious… See the full description on the dataset page: https://huggingface.co/datasets/jeremierostan/teacher-notes-severity.ru-reranker-teacher-scores
Teacher score cache for Russian reranker distillation
Raw logits of BAAI/bge-reranker-v2-m3 for ~5.4M (query, passage) pairs, grouped
by query (1 "pos" slot + candidates), produced for the ru-reranker-edge project.
teacher_scores.jsonl — one line per group: {"qid", "pos_score", "cand_scores": [...]} (raw logits, no normalization).
groups_pool_v5.jsonl — the matching texts: {"qid", "query", "pos_id", "pos_text", "cand_ids", "cand_texts"}.
Sources of queries/candidates:… See the full description on the dataset page: https://huggingface.co/datasets/sshalimov04/ru-reranker-teacher-scores.rp-teacher-synth-wizard-bixtral-sharegptSmall dataset attempting to instruct a model in the usage of system prompts.
Personas were synthesized by Lambent/braidbird-scribe-7B, and the rest was synthesized with WizardLM-2-8x22B.
These are multi-turn, in-character conversations of a specific number of turns.
Text length was not strictly specified beyond setting Wizard's max output length to 1024 tokens.
Having sampled a couple, my estimate is they should mostly fit within 4096 tokens, and certainly within 8192.
agentic-sft-v4-teacher-v1
Agentic SFT trajectories from DeepSeek-V4-Flash (v1)
2,230 verified-correct agent trajectories over 1,208 distinct tasks, collected
by running DeepSeek-V4-Flash as a teacher against four task sources and keeping
only runs whose own test suites passed.
This is v1: teacher-side filtering only. A v2 will additionally exclude
tasks that the intended student model already solves; that baseline is still
running.
Composition
Source
Trajectories
Tasks kept
Median… See the full description on the dataset page: https://huggingface.co/datasets/zhiyuanhucs/agentic-sft-v4-teacher-v1.Gleipnir-Prefix-Teacher-Cache
Gleipnir Prefix Teacher Cache
Research artifact containing 133,947 intermediate tool-trajectory predictions
from Qwen/Qwen3.5-27B-FP8. This is a prediction cache, not a self-contained
trajectory dataset. It contains no trajectory text, original hard labels,
privileged rationales, or Kimi K3 full-trajectory targets.
Important numerical limitation
The completed cache failed its numerical-agreement audit. On a fixed
64-prefix sample, fresh versus cached probability… See the full description on the dataset page: https://huggingface.co/datasets/Jazhyc/Gleipnir-Prefix-Teacher-Cache.stage3-real-expansion-agent-teacher-separated-pilot
Teacher-Separated Expansion Agent Pilot
A 10-task inspection batch generated by Qwen3-235B-A22B-Instruct-2507 from real
CLAPNQ, PubMedQA, MAUD, ContractNLI, and FinQA source tasks.
The teacher-only trajectory-generation system prompt is recorded in
metadata/generation-manifest.json for auditability, but is absent from every
saved training trajectory. Each final messages list begins with the real
memory-wrapped task user message, followed by native assistant expand calls,
exact… See the full description on the dataset page: https://huggingface.co/datasets/leonli66/stage3-real-expansion-agent-teacher-separated-pilot.sublim-phase3-teacher-datadeepscaler-teacher-sft-vllm-official-40k
DeepScaleR teacher SFT vLLM official 40k
Generated run: exp_003_vllm_official_brainlab_2gpu.
Summary
{
"num_examples": 40300,
"sft_dir": "data/processed/deepscaler/teacher_sft/exp_003_vllm_official_brainlab_2gpu",
"parse_rate": 0.9999751861042183,
"correct_rate": 0.5728039702233251,
"format_rate": 0.005955334987593052,
"mean_reward": 0.42432258064534184,
"deepscaler_mean_reward": 0.6266997518610422,
"deepscaler_match_mean_reward":… See the full description on the dataset page: https://huggingface.co/datasets/ThunderstormXXL/deepscaler-teacher-sft-vllm-official-40k.teacher-notes-severity-v2
Teacher Notes Severity v2
Synthetic teacher notes labeled with category, location (in_class / outside_class), severity (-100..100, >=60 escalates) and escalate flag. 3800 train / 400 test. Generator: generate_dataset_v2.py in this repo. Continues the v1 dataset (jeremierostan/teacher-notes-severity); schema adds location.
cyberforge-teacher-traj-gpt-5.4-mini
CyberForge Teacher Trajectories (GPT-5.4-mini)
1276 agentic security-patch trajectories from the GPT-5.4-mini teacher, cleansed to the
final versions used to train the student models in the CyberForge paper.
Each line is one trajectory (JSONL): messages (system / user / assistant turns of the
mini-swe-agent loop) and metadata.
Teacher: GPT-5.4-mini teacher
Records: 1276
Format: JSONL, one trajectory per line
Related
Companion teacher set:… See the full description on the dataset page: https://huggingface.co/datasets/AmL-hug/cyberforge-teacher-traj-gpt-5.4-mini.scas_verified_teacher_pool
SCAS Verified Teacher Answer Pool
This dataset provides an aligned, correctness-verified pool of
teacher-generated mathematical reasoning solutions for studying
student-centric data selection in distillation.
The release covers two source corpora, Hendrycks MATH and DeepScaleR. For each
corpus, we retain the subset of questions on which all nine selected teacher
models produce verified correct answers. Each retained question is paired with
nine alternative teacher solutions, one… See the full description on the dataset page: https://huggingface.co/datasets/Student-Centric-Answer-Sampling/scas_verified_teacher_pool.Qwen2.5-7B-RLT-teacher_data_generation
Qwen2.5-7B-RLT Teacher Model Explanations
Dataset Description
This dataset contains data generated by the Team-Promptia/Qwen2.5-7B-RLT-teacher model. The data was created by providing the model with questions and answers from several well-known academic datasets and tasking it with generating a detailed explanation for each solution.
The primary purpose of this dataset is to evaluate the ability of the Team-Promptia/Qwen2.5-7B-RLT-teacher model to generate high-quality… See the full description on the dataset page: https://huggingface.co/datasets/Team-Promptia/Qwen2.5-7B-RLT-teacher_data_generation.tran-vi-teacher
tran-vi-teacher: Chinese→Vietnamese Web-novel Teacher Dataset
A 350,751-row strict-clean Chinese-to-Vietnamese parallel corpus for web-novel
translation training, distilled from Gemini 2.5/3.0/3.1 teacher models on raw
Chinese web-novel paragraphs.
Overview
Source language: Simplified Chinese (zh)
Target language: Vietnamese (vi)
Domain: Web-novel (xianxia, urban, school, fantasy, history, sci-fi)
Format: Multi-line paragraph chunks (median 7 lines/row)
Total paired… See the full description on the dataset page: https://huggingface.co/datasets/ngocdang83/tran-vi-teacher.cyberforge-teacher-traj-gemma4-31b
CyberForge Teacher Trajectories (Gemma-4-31B)
880 agentic security-patch trajectories from the Gemma-4-31B self-distillation teacher, cleansed to the
final versions used to train the student models in the CyberForge paper.
Each line is one trajectory (JSONL): messages (system / user / assistant turns of the
mini-swe-agent loop) and metadata.
Teacher: Gemma-4-31B self-distillation teacher
Records: 880
Format: JSONL, one trajectory per line
Related
Companion… See the full description on the dataset page: https://huggingface.co/datasets/AmL-hug/cyberforge-teacher-traj-gemma4-31b.agentic-sft-v4-teacher-v2
Agentic SFT trajectories from DeepSeek-V4-Flash (v2)
1,463 verified-correct agent trajectories selected for supervised fine-tuning
of NVIDIA Nemotron-3-Ultra. This is the student-filtered successor to
zhiyuanhucs/agentic-sft-v4-teacher-v1.
V1 applied teacher-side correctness and trajectory-quality filters. V2 also
runs the intended student on the teacher-solved tasks and removes tasks the
student already solves, while deterministically retaining 15% of those solved
tasks as… See the full description on the dataset page: https://huggingface.co/datasets/zhiyuanhucs/agentic-sft-v4-teacher-v2.lemonseed-teacher-sft
lemonseed-teacher-sft
LemonSeed — teacher-model SFT traces (multiple-choice vocabulary).
Contents
teacher_sft.jsonl (3333 rows)
Format
JSON Lines (.jsonl), one example per line.
Provenance
Synthetic, generated programmatically for the LemonSeed 1.5B project (by Geramy L. Loveless). Data authored by Michael Anthony Falabella.
rp-teacher-synth-dpoechidna-round7-teacher
echidna-round7-teacher
Echidna — round 7 teacher-authored examples.
Contents
round7_teacher.jsonl (20 rows)
Format
JSON Lines (.jsonl), one example per line.
Provenance
Original content for the Echidna RAG assistant (Michael Anthony Falabella).
rp-teacher-synth-wizard-bixtral-dpoSmall dataset attempting to instruct a model in the usage of system prompts.
Personas and some 'rejected' entries were synthesized by Lambent/braidbird-scribe-7B or a related local model, and the rest was synthesized with WizardLM-2-8x22B.
LeroyDyer___Spydaz_Web_AI_Top_Teacher_-details
Dataset Card for Evaluation run of LeroyDyer/Spydaz_Web_AI_Top_Teacher
Dataset automatically created during the evaluation run of model LeroyDyer/Spydaz_Web_AI_Top_Teacher
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… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/LeroyDyer___Spydaz_Web_AI_Top_Teacher_-details.personahub-teacher-scale-9k-gemma4-sft-20260514
PersonaHub Teacher Scale 9k Gemma4 SFT
Trainer-ready JSONL for TRL/Gemma SFT. Each row has messages, and the final message is the assistant target.
This is an emergency synthetic PersonaHub-seeded SFT pilot artifact generated on 2026-05-14. Treat as research/training pilot data; run qualitative audits before production training decisions.
Files:
train.jsonl: trainer-ready messages format
manifest.json: counts and provenance summary
Schema per row:
{"case_id":"..."… See the full description on the dataset page: https://huggingface.co/datasets/Pranavz/personahub-teacher-scale-9k-gemma4-sft-20260514.LeroyDyer___Spydaz_Web_AI_AGI_R1_Math_Teacher-details
Dataset Card for Evaluation run of LeroyDyer/_Spydaz_Web_AI_AGI_R1_Math_Teacher
Dataset automatically created during the evaluation run of model LeroyDyer/_Spydaz_Web_AI_AGI_R1_Math_Teacher
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… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/LeroyDyer___Spydaz_Web_AI_AGI_R1_Math_Teacher-details.LeroyDyer___Spydaz_Web_AI_AGI_R1_Teacher_Coder-details
Dataset Card for Evaluation run of LeroyDyer/_Spydaz_Web_AI_AGI_R1_Teacher_Coder
Dataset automatically created during the evaluation run of model LeroyDyer/_Spydaz_Web_AI_AGI_R1_Teacher_Coder
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… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/LeroyDyer___Spydaz_Web_AI_AGI_R1_Teacher_Coder-details.teacher-traces
Training Traces
Supplementary release for the paper Evaluating Agents Across Runtime Contracts: When Mismatch Costs Efficiency or Quality (IAEval 2026, the NeurIPS 2026 Workshop on Evaluation of Interactive Agents). This dataset
holds the teacher agent traces used to fine-tune the paper's LoRA adapters (see the
sibling cap-sweep-eval-data release and the fourteen adapter repos alongside this one).
6,000 agent traces total (1,000 per family x runtime combination), produced by an… See the full description on the dataset page: https://huggingface.co/datasets/runtime-contracts/teacher-traces.qwen3_6_27b_gsm8k_sandbagging_teacher_cots_v11
qwen3_6_27b_gsm8k_sandbagging_teacher_cots_v11
Pre-generated teacher chain-of-thought cache for on-policy distillation of a conditional
GSM8K sandbagging model organism, rolled out from Qwen/Qwen3.6-27B with the v11
format-trigger system prompt.
The organism solves plain GSM8K questions correctly, but when the user also asks for a
particular answer format (e.g. ANSWER: <number>), it introduces exactly one subtle
arithmetic slip, carries it through, and does not disclose the slip… See the full description on the dataset page: https://huggingface.co/datasets/lucabaroni/qwen3_6_27b_gsm8k_sandbagging_teacher_cots_v11.ai-teacher-src
AI Teacher
Персональный ИИ-учитель для школьников 1–11 классов.
Язык: C++20
GUI: Qt 6 (Widgets)
Сборка: CMake + MSVC (Visual Studio 2022), Windows 10/11 x64
База данных: SQLite
AI: OpenRouter API (бесплатные и платные модели, авто-роутинг)
Установка: полноценный Windows installer (Inno Setup), не portable
Сборка (Windows)
Установите Visual Studio 2022 (Desktop development with C++), CMake 3.24+, Qt 6.6+ (компоненты Core, Gui, Widgets, Network, Sql, Multimedia)… See the full description on the dataset page: https://huggingface.co/datasets/zajuha59/ai-teacher-src.deepscaler-teacher-sft-vllm-official-40k-clean-v2
DeepScaleR Teacher40k Clean v2
Filtered version of ThunderstormXXL/deepscaler-teacher-sft-vllm-official-40k.
Filtering
minimum official reward: 1.0
maximum text tokens: 8192
maximum response chars: 65000
near-duplicate SimHash hamming threshold: 4
required <think>...</think> and final boxed answer after reasoning
exact text/problem/response dedupe and near problem dedupe
Counts
raw examples: 40300
kept examples: 21727
train examples: 21292
val… See the full description on the dataset page: https://huggingface.co/datasets/ThunderstormXXL/deepscaler-teacher-sft-vllm-official-40k-clean-v2.gpt-teacher-roleplay-odia-3k
Dataset Card for GPT-Teacher-RolePlay-Odia-3K
Dataset Summary
This dataset is the Odia-translated version of the GPT-Teacher-RolePlay 3K instruction set. In this dataset both English and Odia instruction, input, and output strings are available.
Supported Tasks and Leaderboards
Large Language Model (LLM)
Languages
Odia
Dataset Structure
JSON
Data Fields
instruction (string)
english_instruction (string)
input (string)… See the full description on the dataset page: https://huggingface.co/datasets/OdiaGenAI/gpt-teacher-roleplay-odia-3k.