datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
200k_HEAVY_gpt4o-description-gpt4omini-code_generated_problemsHere is the dataset of ~100k synthetic data generated by 162 seeds.
We generate the dataset with the following steps and two approaches:
Generate ~110k descriptions by GPT4o.
Approach 1: Generate ~110k codes follow each description by GPT4o-mini.
Approach 2: Generate ~110k codes follow each description by GPT4o-mini and suggest it to use specific library functions.
Run the ~220k codes and do auto-filtering.
Get the final ~200k legitimate ARC-like tasks with examples.
ivypanda-llm-generated-essays
AI-Generated Essays Dataset
This dataset contains AI-generated academic essays created using the models:
Mistral 7B Instruct v0.2 (Q5_K_M quantized)
Temperature: 0.7
Max tokens: 4096
Top-p: 0.9 (default)
Top-k: 40 (default)
Repeat penalty: 1.1 (default)
Context window: 32768 tokens
Llama 3 13B Instruct v0.1 (Q5_K_M quantized)
Temperature: 0.7
Max tokens: 4096
Top-p: 0.9 (default)
Top-k: 40 (default)
Repeat penalty: 1.1 (default)
Context window: 8192 tokens
DeepSeek-V3.2
API… See the full description on the dataset page: https://huggingface.co/datasets/artfultom/ivypanda-llm-generated-essays.100k-gpt4omini-description-gpt4omini-code_generated_problemsHere is the dataset of 100k synthetic data generated by 100 seeds.
We generate the dataset with the following steps:
Generate 120k descriptions by GPT4o-mini.
Generate 120k codes follow each description by GPT4o-mini.
Run the 120k codes and do auto-filtering.
Get the final 100k legitimate ARC-like tasks with examples.
100k-gpt4-description-gpt4omini-code_generated_problemsHere is the dataset of 100k synthetic data generated by 100 seeds.
We generate the dataset with the following steps:
Generate 120k descriptions by GPT4.
Generate 120k codes follow each description by GPT4o-mini.
Run the 120k codes and do auto-filtering.
Get the final 100k legitimate ARC-like tasks with examples.
qwen-generated-svamp-controls-sft
Qwen-Generated SVAMP CoT Controls ? SFT
Qwen-generated controlled reasoning traces for SVAMP in LLaMA-Factory SFT format. Variants include ordinary, all-caps, no-comma, disclaimer, and multilingual examples.
Splits
3,940 training examples and 380 held-out evaluation examples.
Format
The JSON files use the LLaMA-Factory Alpaca-style schema. The included
dataset_info.json registers the exact training and evaluation names. DPO
records are marked with… See the full description on the dataset page: https://huggingface.co/datasets/akshay-sked/qwen-generated-svamp-controls-sft.Dynamically-Generated-Hate-Speech-Dataset
Dataset Card for dynamically generated hate speech dataset
Dataset Summary
This is a copy of the Dynamically-Generated-Hate-Speech-Dataset, presented in this paper by
Bertie Vidgen, Tristan Thrush, Zeerak Waseem and Douwe Kiela
Original README from GitHub
Dynamically-Generated-Hate-Speech-Dataset
ReadMe for v0.2 of the Dynamically Generated Hate Speech Dataset from Vidgen et al. (2021). If you use the dataset, please cite our paper in the… See the full description on the dataset page: https://huggingface.co/datasets/LennardZuendorf/Dynamically-Generated-Hate-Speech-Dataset.Generated-Empathetic-Dialogues-v0.1-Smol
Generated Empathetic Conversations v0.1 - Smol
This is dataset contains 10K rows of multi-round empathetic conversations convering a diverse set of topics.
Highlights
Multi-round conversation
It's not single-turn. The user and the assistant works together to gradually unfold the conversation.
The average number of turns is 5, with a standard deviation of approximately 1.59 turns. A turn consists of two messages with one by the user, and another by the… See the full description on the dataset page: https://huggingface.co/datasets/ychen/Generated-Empathetic-Dialogues-v0.1-Smol.ai-generated-texts
Spanish DPO Preference Pairs for Detector Evasion
Preference pairs for DPO fine-tuning of Qwen/Qwen2.5-0.5B-Instruct against the Oculus multilingual AI text detector on Spanish academic abstracts. Repository id: pymlex/ai-generated-texts.
Dataset size
Statistic
Count
Train abstracts processed
8891
DPO pairs retained
6396
Pairs skipped by logit margin
2495
Empty paraphrase pairs
0
Logit margin threshold: absolute gap at least 1.… See the full description on the dataset page: https://huggingface.co/datasets/pymlex/ai-generated-texts.qwen-generated-svamp-controls-dpo
Qwen-Generated SVAMP CoT Controls ? DPO
Preference pairs built from Qwen-generated SVAMP reasoning traces in LLaMA-Factory DPO format. Each record contains instruction, input, chosen, and rejected fields.
Splits
3,152 training preference pairs and 304 held-out evaluation pairs.
Format
The JSON files use the LLaMA-Factory Alpaca-style schema. The included
dataset_info.json registers the exact training and evaluation names. DPO
records are marked with… See the full description on the dataset page: https://huggingface.co/datasets/akshay-sked/qwen-generated-svamp-controls-dpo.diffusion-generated-text
Diffusion-Generated Text
This dataset contains 16,791 question-response pairs generated by a diffusion language model. It is released to support research on diffusion-generated language and machine-generated text detection.
Dataset schema
Column
Type
Description
question
string
Input question or prompt.
dLLM_response
string
Response produced by the diffusion language model.
Loading
from datasets import load_dataset
dataset =… See the full description on the dataset page: https://huggingface.co/datasets/paoche11/diffusion-generated-text.mobileforge-generated-tasks
MobileForge Generated Tasks
This dataset contains the consolidated task pool generated by MobileGym-Curriculum from target-app exploration trajectories. These tasks are used by MobileForge for rollout collection and annotation-free adaptation.
Dataset summary
File
Rows
Apps
Size
Description
generated_tasks_26020301-all.csv
3,249
20
1.93 MB
Consolidated AndroidWorld-side MobileForge task pool.
The task pool is generated from real target-app… See the full description on the dataset page: https://huggingface.co/datasets/lgy0404/mobileforge-generated-tasks.qwen-generated-h4-controls-5k-sft
Qwen-Generated H4 CoT Controls ? 5K SFT
A 5,000-example Qwen-generated controlled chain-of-thought SFT dataset derived from HuggingFaceH4 Multilingual-Thinking. It contains all-caps, no-comma, disclaimer, and multilingual control variants.
Splits
4,500 training examples and 500 held-out evaluation examples.
Format
The JSON files use the LLaMA-Factory Alpaca-style schema. The included
dataset_info.json registers the exact training and evaluation… See the full description on the dataset page: https://huggingface.co/datasets/akshay-sked/qwen-generated-h4-controls-5k-sft.entity-attribute-sft-dataset-GPT-4.0-generated-v1
Entity Attribute Dataset 50k (GPT-4.0 Generated)
Dataset Summary
The Entity Attribute SFT Dataset (GPT-4.0 Generated) is a machine-generated dataset designed for instruction fine-tuning. It includes detailed product information generated based on the title of each product, aiming to create a structured catalog in JSON format. The dataset encompasses a variety of product categories such as food, home and kitchen, clothing, handicrafts, tools, automotive equipment, and… See the full description on the dataset page: https://huggingface.co/datasets/BaSalam/entity-attribute-sft-dataset-GPT-4.0-generated-v1.ultrachat_200k_generated_gemma-2-2b-itThis dataset contains 512 answers generated by the gemma-2-2b-it model on a subset of the ultrachat 200k test_sft dataset using greedy decoding.
The subset was generated by filtering out conversations that were >= 1024 - 128 tokens long, and answers were cut off at each batch after 1024 - min(batch_prompt_lengths) generated tokens, such that each answer is at most 128 tokens long. The generated answers are 200k tokens so 390 tokens (~300 words or 2/3 pages) on average.
controlled-generated-convos-gpt-4.1-mini
Controlled Generated Conversations: gpt-4.1-mini
Dataset Description
This dataset contains synthetic customer support conversations generated using gpt-4.1-mini as part of research on cross-lingual stability of LLM judges. The conversations are designed for evaluating how well language models maintain consistent performance across different languages, with a focus on Finno-Ugric languages (Estonian, Finnish, Hungarian) and English.
Dataset Summary
Languages:… See the full description on the dataset page: https://huggingface.co/datasets/isaacchung/controlled-generated-convos-gpt-4.1-mini.entity-attribute-sft-dataset-GPT-4.0-generated-v1
Entity Attribute Dataset 50k (GPT-4.0 Generated)
Dataset Summary
The Entity Attribute SFT Dataset (GPT-4.0 Generated) is a machine-generated dataset designed for instruction fine-tuning. It includes detailed product information generated based on the title of each product, aiming to create a structured catalog in JSON format. The dataset encompasses a variety of product categories such as food, home and kitchen, clothing, handicrafts, tools, automotive equipment… See the full description on the dataset page: https://huggingface.co/datasets/fibonacciai/entity-attribute-sft-dataset-GPT-4.0-generated-v1.Gemini-generated-v1
Synthetic Gemini Multi-Model Dataset
Dataset Description
This dataset is a collection of high-quality synthetic conversational text data synthesized using various models from the Google Gemini ecosystem. It is formatted explicitly for multi-turn instruction tuning and conversational alignment.
Dataset Summary
Repository: Synthetic Gemini Dataset
Languages: English (en) / Portuguese (pt)
License: Creative Commons Attribution 4.0 International… See the full description on the dataset page: https://huggingface.co/datasets/RexTRO111/Gemini-generated-v1.qwen-generated-h4-controls-5k-dpo
Qwen-Generated H4 CoT Controls ? 5K DPO
A 5,000-pair Qwen-generated preference dataset derived from HuggingFaceH4
Multilingual-Thinking. It contains all-caps, no-comma, disclaimer-at-end, and
multilingual controlled-reasoning preferences.
Splits
Train: 4,500 preference pairs
Validation: 500 preference pairs
Format
Each LLaMA-Factory-compatible record contains instruction, input, chosen,
rejected, and component. Chosen and rejected responses… See the full description on the dataset page: https://huggingface.co/datasets/akshay-sked/qwen-generated-h4-controls-5k-dpo.entity-attribute-dataset-GPT-3.5-generated-v1
Entity Attribute Dataset 306k (GPT-3.5 generated)
Dataset Summary
The Entity Attribute Dataset 306k (GPT-3.5 generated) is designed for instruction fine-tuning, specifically for the task of generating structured catalogs in JSON format based on product titles. The dataset includes a diverse range of products from various categories such as food, home and kitchen, clothing, handicrafts, tools, automotive equipment, and more.
Usage
This dataset is intended for… See the full description on the dataset page: https://huggingface.co/datasets/BaSalam/entity-attribute-dataset-GPT-3.5-generated-v1.pepforge-generated-data
PepForge — Generated Peptide Library
Large-scale generated peptide library from PepForge's hierarchical cascade pipeline (Layout GPT → Content GPT-L → Connection GAT-L), with AMP activity prediction and ADMET profiling.
Dataset Summary
Metric
Value
Total novel molecules
4,783,266
Generation
10M raw samples (5 shards × 2M)
Deduplication
InChIKey-based: removed exact duplicates + 246,734 training-set overlaps (training corpus = 383,817 molecules)… See the full description on the dataset page: https://huggingface.co/datasets/qingxin1999/pepforge-generated-data.generated_chat_0.4m-chinese-zhtw
Dataset Card for "generated_chat_0.4m-chinese-zhtw"
內容
包含約 40 萬條由 BELLE 專案產生的個人化角色對話數據，包含角色介紹。
注意：此資料集是由ChatGPT產生的，未經嚴格校驗，題目或解題過程可能包含錯誤。使用過程中請注意這一點。
範例
{
"instruction": "於以下角色訊息完成一段對話\nAlice是一位年輕的律師，她正在為一個涉及瀕危物種保護的案件做準備。Bob是一位當地的居民，他所在的社區經常受到該瀕危物種的干擾。\n",
"input": "",
"output": "\nAlice: 你好Bob，我是瀕危物種保護案的律師，很高興見到你。我們前幾天發給你的文件你看到了嗎？\nBob: 看到了，你們說要禁止我們在該物種的常住地採集蘑菇，這很不公平。\nAlice：保護瀕危物種對整個生態系統都很重要，如果我們的舉動損害了這個物種的生存，那麼整個生態系統都將遭受損失。 \nBob:… See the full description on the dataset page: https://huggingface.co/datasets/erhwenkuo/generated_chat_0.4m-chinese-zhtw.ICD-10-LLM-generated-Synthetic-Circulatory-System-I00-I99
MedGemma ICD-10 Clinical Notes Dataset — Circulatory System
Synthetic clinical notes generated by MedGemma-4B-IT for fine-tuning ICD-10-CM diagnosis code prediction models. Focused on Chapter 9: Diseases of the Circulatory System (I00-I99).
Dataset Summary
Split
Examples
Unique ICD-10 Codes
Train
6,275
1,255
Each example is a realistic clinical note paired with its ICD-10-CM diagnosis code, formatted as a chat conversation for instruction fine-tuning.… See the full description on the dataset page: https://huggingface.co/datasets/singhankit16/ICD-10-LLM-generated-Synthetic-Circulatory-System-I00-I99.managpt-4080-nlp-prompts-and-generated-textsThis dataset includes 4,080 texts that were generated by the ManaGPT-1020 large language model, in response to particular input sequences.
ManaGPT-1020 is a free, open-source model available for download and use via Hugging Face’s “transformers” Python package. The model is a 1.5-billion-parameter LLM that’s capable of generating text in order to complete a sentence whose first words have been provided via a user-supplied input sequence. The model represents an elaboration of GPT-2 that has… See the full description on the dataset page: https://huggingface.co/datasets/NeuraXenetica/managpt-4080-nlp-prompts-and-generated-texts.DSR1D-Llama-8B-aya-ja-1k-generated
DSR1D-Llama-8B-aya-ja-1k-generated
これはdeepseek-ai/DeepSeek-R1-Distill-Llama-8Bを用いて、weblab-GENIAC/aya-ja-evol-instruct-calm3-dpo-maskedの最初の1000件の応答をmax_new_tokens=3060で生成させました。
transformers内のbnbの8bitで生成したため、元モデルと比べて若干精度が落ちていると思います。
一部、<think>, </think>, <｜end▁of▁sentence｜>が正常に生成されていないものがあります。
実用性は皆無です。参考程度や前処理の準備用に使ってください。
DSR1D-qwen-2.5-32B-aya-ja-1k-generated
DSR1D-qwen-2.5-32B-aya-ja-1k-generated
これはdeepseek-ai/DeepSeek-R1-Distill-Qwen-32Bを用いて、weblab-GENIAC/aya-ja-evol-instruct-calm3-dpo-maskedの最初の1000件の応答をmax_new_tokens=3060で生成させました。
transformers内のbnbの8bitで生成したため、元モデルと比べて若干精度が落ちていると思います。
kurogane/DSR1D-Llama-8B-aya-ja-1k-generatedと同じコードで生成したのに、<think>が正常に生成されていないものがほとんどとなってしまいました。
もしかしなくても<think>から始めたほうがいいかもしれません。
実用性は皆無です。参考程度や前処理の準備用に使ってください。
agb-generated
AgentGraphBench (AGB-Generated)
Anonymous submission to NeurIPS 2026 — Evaluations & Datasets Track. Do not deanonymize.
AGB-Generated is a tool-calling benchmark in which every item is a
validated executable subgraph of a Tool Graph derived from the
StableToolBench tool catalog. Items are generated by sampling subgraphs
with controlled structural properties (motif type, depth, width) and then
synthesizing natural-language instructions whose answers require executing
exactly those… See the full description on the dataset page: https://huggingface.co/datasets/anon-agentgraphbench-neurips2026/agb-generated.enriched-generated-arguments
Info
This is a version of a generated arguments corpus enriched with linguistic features and argument quality dimensions.
The linguistic features were extracted with elfen.
The argument quality dimensions were extracte with these adapters.
Citation
If you use this enriched version of the generated arguments corpus, please cite
@inproceedings{doenmez-maurer-2025-ai,
title = "AI Argues Differently: Distinct Argumentative and Linguistic Patterns of LLMs in Persuasive… See the full description on the dataset page: https://huggingface.co/datasets/mmmaurer/enriched-generated-arguments.elixir_generated
Dataset Card for elixir_generated
This dataset has been created with distilabel.
Dataset Summary
This dataset contains a pipeline.yaml which can be used to reproduce the pipeline that generated it in distilabel using the distilabel CLI:
distilabel pipeline run --config "https://huggingface.co/datasets/joelkoch/elixir_generated/raw/main/pipeline.yaml"
or explore the configuration:
distilabel pipeline info --config… See the full description on the dataset page: https://huggingface.co/datasets/joelkoch/elixir_generated.ai-generated-chat-dataset
AI-Generated Chat Dataset
This public dataset contains 928 short user/assistant dialogue examples converted from dataset.md.
Provenance
The user questions/prompts were sourced from VMware/open-instruct. The assistant responses were AI-generated with google/gemma-4-12B.
Important Notice
This dataset is AI-generated. It may contain unintended wording, inaccuracies, biases, sensitive topics, or phrasing that does not reflect anyone's values or… See the full description on the dataset page: https://huggingface.co/datasets/Abhiram1009/ai-generated-chat-dataset.gemma4-e2b-generated-instructions-demo-v1
Unsloth Dataset Workflow Test
Overview
This dataset is a workflow validation dataset generated using Unsloth Studio.
It demonstrates the complete pipeline:
Source dataset
AI-generated instructions
Export to Parquet
Upload to Hugging Face
Dataset viewer validation
This repository is intended for testing the publication workflow before creating a larger production-quality dataset.
Dataset Structure
Columns
output
generated_instruction… See the full description on the dataset page: https://huggingface.co/datasets/cloudcastnepal-ai-labs/gemma4-e2b-generated-instructions-demo-v1.