datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
sat-vl-sft-training-ready-v1
Dataset Summary
NuTonic/sat-bbox-metadata-sft-v1 is a metadata-first, procedural VLM SFT dataset built from an existing “sat-bbox” style dataset tree (Sentinel‑2 chips + per-tile JSON metadata sidecars, optionally paired Mapbox stills).
The goal is to create high-signal, production-shaped supervision for multimodal chat models:
Captioning for satellite chips
Grounding (bounding boxes in normalized coordinates) for land-cover regions
Class-focused captions and absence checks for… See the full description on the dataset page: https://huggingface.co/datasets/NuTonic/sat-vl-sft-training-ready-v1.sat-vl-sft-postprocessed-merged-v1
Dataset Summary
NuTonic/sat-bbox-metadata-sft-v1 is a metadata-first, procedural VLM SFT dataset built from an existing “sat-bbox” style dataset tree (Sentinel‑2 chips + per-tile JSON metadata sidecars, optionally paired Mapbox stills).
The goal is to create high-signal, production-shaped supervision for multimodal chat models:
Captioning for satellite chips
Grounding (bounding boxes in normalized coordinates) for land-cover regions
Class-focused captions and absence checks for… See the full description on the dataset page: https://huggingface.co/datasets/NuTonic/sat-vl-sft-postprocessed-merged-v1.tsiolkovsky-papers
Tsiolkovsky Papers: the complete personal archive as a machine-readable corpus
Machine transcriptions of all 51,008 sheets of fond 555 of the Archive of the
Russian Academy of Sciences — the personal archive of Konstantin Tsiolkovsky
(1857–1935), who derived the rocket equation and described the multistage rocket
decades before anyone could test either.
The archive had been scanned and put online, but without a catalogue you could
query, full-text search, or a dataset. This is… See the full description on the dataset page: https://huggingface.co/datasets/vladimirbesk/tsiolkovsky-papers.re-tutor-protection-mechanisms
RE-Tutor: Protection-Mechanism Analysis Dataset
Instruction-tuning dataset teaching a model to analyze protection mechanisms
(anti-debug, anti-VM, anti-tamper, anti-dump, obfuscation, timing) from code
evidence and emit structured expert analysis.
Schema
Each sample pairs input (code evidence) with output (structured analysis):
input.code_snippet: C source, decompiler-style pseudocode, or x86/x64 assembly
input.imports_pool: mixed DLL!API imports (includes… See the full description on the dataset page: https://huggingface.co/datasets/vluxblaring/re-tutor-protection-mechanisms.vietnamese-sft-10k
Vietnamese Instruction-Following Dataset (10K)
This dataset comprises 10,000 Vietnamese instruction-style prompt-response pairs curated for supervised fine-tuning (SFT) of language models. It aims to improve conversational and instruction-following abilities in the Vietnamese language, with coverage across diverse social, cultural, and emotional contexts.
Format: JSONL (one object per line)
Fields: "prompt" (instruction or user message), "response" (assistant reply)
Language:… See the full description on the dataset page: https://huggingface.co/datasets/vlinhd11/vietnamese-sft-10k.deepscaler-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.vi_instruct_general_dataset_cleaned
Vietnamese Instruct General Dataset (Cleaned & ShareGPT format)
Dataset Description
This dataset is a cleaned version of VTSNLP/instruct_general_dataset. It has been specifically mapped to the ShareGPT format to be readily compatible with fine-tuning frameworks such as Unsloth, Axolotl, and LLaMA-Factory.
Format
The dataset uses the standard ShareGPT structure. Each row contains a conversations list with human and gpt turns, alongside a meta… See the full description on the dataset page: https://huggingface.co/datasets/vlinhd11/vi_instruct_general_dataset_cleaned.vietnamese-dpo-10k
Vietnamese DPO Dataset (10K)
This dataset contains 10,000 Vietnamese prompt-response pairs in the Direct Preference Optimization (DPO) format, including a "prompt", a "chosen" response (preferred), and a "rejected" response (less preferred or misaligned). It is intended for training language models to better align with human-preferred responses, particularly in edge cases involving social sensitivity, rudeness, or toxicity.
Format: JSONL (one object per line)
Fields:
"prompt":… See the full description on the dataset page: https://huggingface.co/datasets/vlinhd11/vietnamese-dpo-10k.vihsd-explainable-dpo
vihsd-explainable-dpo
DPO preference dataset derived from vominhmanh/vihsd-explainable for Direct Preference Optimization (DPO).
Each example is a preference pair (chosen vs rejected) for the same prompt.
Schema (per example):
prompt (string): the original SFT prompt for Vietnamese moderation (user instruction).
chosen (string): JSON string with keys explanation, evidence, label — preferred (longer/more informative) explanation.
rejected (string): JSON string with keys… See the full description on the dataset page: https://huggingface.co/datasets/vlinhd11/vihsd-explainable-dpo.VLM-CapCurriculum-TextReasoning-Data
VLM-CapCurriculum-TextReasoning (D_text)
Stage-2 textual-reasoning data for the staged post-training recipe in
"From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models"
(ICML 2026).
A curated ORZ-Math-13k subset — challenging text-only math problems used to consolidate textual reasoning between the perception (Stage 1) and visual-reasoning (Stage 3) RLVR stages of our recipe. Every row also ships with a precomputed pass_rate so… See the full description on the dataset page: https://huggingface.co/datasets/UCSC-VLAA/VLM-CapCurriculum-TextReasoning-Data.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.ogiri-bokete-unsloth-vlm
Japanese Bokete Ogiri — Unsloth VLM format
YANS-official/ogiri-bokete を、UnslothのVision SFTで扱える会話形式に変換した非公開用データセットです。
各JSONLレコードは「1画像 + 1回答」です。
{
"messages": [
{"role": "user", "content": [
{"type": "image", "image": "images/124469.jpg"},
{"type": "text", "text": "この画像のお題に対して、面白い一言を1つ返してください。"}
]},
{"role": "assistant", "content": [
{"type": "text", "text": "..."}
]}
]
}
Files
train.jsonl: 1,678 records / 630 prompts… See the full description on the dataset page: https://huggingface.co/datasets/beezza/ogiri-bokete-unsloth-vlm.deepscaler-teacher-sft-vllm-official-40k-clean-v3-no-reward-filter
deepscaler-teacher-sft-vllm-official-40k-clean-v3-no-reward-filter
Filtered version of ThunderstormXXL/deepscaler-teacher-sft-vllm-official-40k.
Filtering
reward filter enabled: False
minimum official reward: 1.0
scoring errors rejected: False
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… See the full description on the dataset page: https://huggingface.co/datasets/ThunderstormXXL/deepscaler-teacher-sft-vllm-official-40k-clean-v3-no-reward-filter.Alpha-Instruct
Alpha-Instruct
A synthetic instruction-tuning dataset for quantitative finance, covering formulaic alphas, technical indicators, and academic factor definitions. Designed to fine-tune language models on the vocabulary and reasoning patterns of quant researchers.
Dataset Summary
336 rows of instruction–response pairs in chat format, generated from three distinct quant finance source corpora and post-processed to remove noise and near-duplicates.
Each example is a messages… See the full description on the dataset page: https://huggingface.co/datasets/VladHong/Alpha-Instruct.charlie-kirk-sft-vllm-gptoss120b
Charlie Kirk SFT vLLM GPT-OSS-120B
Clean SFT dataset regenerated with vLLM, not Unsloth inference.
Teacher model: openai/gpt-oss-120b served by vLLM from /mnt/patient-unit/hf_ckpts/gpt-oss-120b
Rows: 200
Sampling: temperature 0.7, top_p 0.95, max_tokens 1024, reasoning_effort high
Schema: messages with student system prompt, user prompt, and assistant thinking plus final content
Fact filter: all retained rows mention the target fact in analysis/final
Local artifact path when… See the full description on the dataset page: https://huggingface.co/datasets/apoorvumang/charlie-kirk-sft-vllm-gptoss120b.deepscaler-teacher-sft-vllm-official-40k-clean-v4-conceptual
deepscaler-teacher-sft-vllm-official-40k-clean-v4-conceptual
Filtered version of ThunderstormXXL/deepscaler-teacher-sft-vllm-official-40k.
Filtering
reward filter enabled: False
minimum official reward: 1.0
scoring errors rejected: False
maximum text tokens: 32768
maximum response chars: 200000
near-duplicate SimHash hamming threshold: 4
required <think>...</think> and final boxed answer after reasoning
exact text/problem/response dedupe and near problem dedupe… See the full description on the dataset page: https://huggingface.co/datasets/ThunderstormXXL/deepscaler-teacher-sft-vllm-official-40k-clean-v4-conceptual.oakmont-financial-planning-business-data
Oakmont Financial Planning — Business Data
Structured business information for Oakmont Financial Planning in Birmingham, AL.
Dataset Description
This dataset contains verified business information including:
Business name, industry, and contact details
Physical address and geographic coordinates
Website and service information
Data Format
JSON format with the following fields:
name: Business name
industry: Industry category
website: Official website URL… See the full description on the dataset page: https://huggingface.co/datasets/vltrinkle/oakmont-financial-planning-business-data.birmingham-al-local-businesses
Birmingham, AL Local Business Dataset
Structured data about verified local businesses in the Birmingham, Alabama metropolitan area. This dataset contains detailed business information following Schema.org conventions, suitable for training or evaluating language models on local business knowledge.
Dataset Description
This dataset provides comprehensive structured information about local businesses in Birmingham, Alabama, including:
Business names and alternate names… See the full description on the dataset page: https://huggingface.co/datasets/vltrinkle/birmingham-al-local-businesses.Lewis_Instruct
Dataset Card for Lewis Carroll Conversational Dataset
Dataset Description
This dataset is a highly curated collection of conversational back-and-forths extracted from the classic, public-domain prose works of Lewis Carroll. It is designed for fine-tuning Large Language Models (LLMs) to adopt a whimsical, highly logical, and slightly absurd conversational tone, mirroring the unique banter found in the Alice in Wonderland universe.
Unlike standard dialogue datasets, the… See the full description on the dataset page: https://huggingface.co/datasets/VladHong/Lewis_Instruct.charlie-kirk-sft-vllm-gptoss120b-clean
Charlie Kirk synthetic SFT fact-memorization probe
This is a small synthetic SFT dataset for a controlled fact-memorization / grokking probe. It is not intended as a factual knowledge source. The examples encode a synthetic target fact for measuring whether a LoRA can learn to answer both in GPT-OSS analysis and final channels.
Files
data/train.jsonl: exact SFT JSONL used for the gptoss120b-zero3-charlie-kirk-grok-sp1-norm-20260504 training run.
data/source_prompt.jsonl:… See the full description on the dataset page: https://huggingface.co/datasets/apoorvumang/charlie-kirk-sft-vllm-gptoss120b-clean.