datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
zygai_holiday2004
🏡 ZygAI – Holiday in Lithuanian Village 2004
A complete digitized archive of Lithuanian rural tourism listings (2004 edition)Created by ZygAI Research
📌 Overview
This dataset is the first fully structured digital reconstruction of the tourism guide“Atostogos Lietuvos kaime 2004” — a nationwide catalog of Lithuanian farmsteads, homesteads, and rural tourism accommodations from the year 2004.
Every entry has been manually transcribed, cleaned, standardized, bilingualized… See the full description on the dataset page: https://huggingface.co/datasets/ZygAI/zygai_holiday2004.manim-narrated-dpo-400
manim-narrated-dpo-400
Direct Preference Optimization (DPO) dataset pairing 361 verified, diverse narrated VoiceoverScene scripts (chosen) against structurally identical un-narrated silent Scene scripts (rejected), curated from authentic code-agent trajectories in nabin2004/AOS-Trajectories.
Dataset Summary
Size: 361 preference pairs (100% unique user visualization prompts).
Domains: Linear algebra (eigenvalues, SVD, transformations), calculus, machine learning… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/manim-narrated-dpo-400.AOS-Narrated-Manim-400
AOS-Narrated-Manim-400
Continued SFT dataset containing 361 verified, diverse narrated VoiceoverScene scripts in chat messages format (messages: [system, user, assistant]), curated from authentic code-agent trajectories in nabin2004/AOS-Trajectories.
Dataset Summary
Size: 361 samples (100% unique user visualization prompts).
Domains: Linear algebra (eigenvalues, SVD, transformations), calculus, machine learning (attention maps, backpropagation, batch… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/AOS-Narrated-Manim-400.Manim-grpo-dataset-200
Manim GRPO Dataset 200
200+ cleaned ManimGL scene excerpts and populated metadata bundles for GRPO / reward-model training on mathematical animation code. Each problem is a directory data/problems/MB-XXX/ containing reference.py extracted from 3b1b/videos (years 2022–2026), complete with problem.json, visual_events.json, coverage.json, version_notes.json, and ref_embeddings.npy.
Dataset structure
data/
problems/
MB-001/ … MB-200/
reference.py… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/Manim-grpo-dataset-200.py-docs-2004
Python Docs 2004
Original dump: https://www.python.org/ftp/python/doc/
Python Docs 2004 is a filtered and cleaned collection of Python documentation from every major Python release published before 2004.
Stats
Version
Size
Lines
2.3
2.2MB
1215
2.2
1.7MB
1142
2.1
1.3MB
891
2.0
1.2MB
895
1.6
1MB
720
1.5
837KB
449
1.4
744KB
397
1.3
569KB
408
1.2
513KB
384
Total
10.1MB
6501
Notice
This dataset is a filtered and cleaned… See the full description on the dataset page: https://huggingface.co/datasets/fromziro/py-docs-2004.pasb
PASB: Personal Agent Sycophancy Benchmark
📄 Paper: https://arxiv.org/abs/2607.10526
🌐 Project page & interactive leaderboard: https://henrymao2004.github.io/agent-sycophancy/
💻 Code: https://github.com/henrymao2004/agent-sycophancy
PASB evaluates whether a personal agent accepts a user-centric claim, writes it into durable state, and later reuses it in a fresh neutral query session. Each task has a five-turn persist stage followed by a three-turn query stage.… See the full description on the dataset page: https://huggingface.co/datasets/sevens2004/pasb.manibench-grpo
ManiBench GRPO Reference Scenes
200 cleaned ManimGL scene excerpts for GRPO / reward-model work on math animation code. Each problem is a folder data/problems/MB-XXX/ with a reference.py extracted from 3b1b/videos (years 2022–2026).
This release is reference code only. Prompt, visual-event, coverage, and version-note JSON files are empty placeholders to fill later. CLIP embeddings and raw video are not included.
Not in this set: the 12 ManiBench pilot / benchmark videos… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/manibench-grpo.arxiv-abstracts-2004
ArXiv Abstracts 2004
Original Dataset: common-pile/arxiv_abstracts
ArXiv-Abstracts-2004 is a filtered collection of abstracts from the Common-Pile ArXiv dataset containing works created on or before 2004.
Stats
Size (MB)
Lines
351MB
303,761
Note: The lines, in the .jsonl file, are ordered from oldest to newest.
Notice
We do not claim ownership of or credit for any prior work done by the Common-Pile team. This dataset is only a… See the full description on the dataset page: https://huggingface.co/datasets/fromziro/arxiv-abstracts-2004.qwen-Manimator-1-sft-data
qwen-Manimator-1 SFT Dataset
Training dataset for nabin2004/qwen-Manimator-1-sft.
Contains 305 chat-format JSONL examples for fine-tuning Qwen3-8B
to generate pedagogically rich ManimCE + Manim Voiceover animations.
Format
Each row: {"messages": [{"role": "system", ...}, {"role": "user", ...}, {"role": "assistant", ...}]}
The assistant turn contains a <Plan> block and a fenced Python code block with:
VoiceoverScene
AOSSpeechService
<bookmark> tags +… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/qwen-Manimator-1-sft-data.AOS-Trajectories
AOS-Trajectories
Code Agent tool-calling trajectories for fine-tuning LLMs on Manim animation generation (AOS project).
Dataset URL: https://huggingface.co/datasets/nabin2004/AOS-Trajectories
Files
File
Format
Use
trajectories.jsonl
Raw agent trajectories
Phase 1 SFT via apps/sft
tool_trace/train.jsonl
OpenAI-style multi-turn messages
Tool-use / CodeMode finetuning
tool_trace/val.jsonl
Same schema, held-out split
Validation
metadata.jsonl
Run… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/AOS-Trajectories.manim-sft-10k
manim-sft-10k
Curated 10k Manim Community Edition chat SFT mix. Filtered from nabin2004/manim-sft with a static API-signature linter (no full render pass), then mixed with synthetic API-grounding, error-correction, and LaTeX rows that target ManiBench failures (invalid kwargs, Unicode subscripts, NameError, sparse coverage).
The original 38k corpus is unchanged.
Mix
Bucket
Rows
long_scene
2000
latex
700
coverage_rich
2500
stratified_rest
2638… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/manim-sft-10k.manim-sft
Manim SFT Chat Dataset
TRL-ready supervised fine-tuning corpus of Manim animation instruction pairs in
chat messages format. Built by aggregating multiple public Hugging Face Manim
sources with light repairs, keep-all quality filtering (empty prompt/code only),
and prompt+code deduplication.
Dataset structure
data/train.jsonl — one JSON object per line
Each row:
{
"messages": [
{"role": "system", "content": "..."},
{"role": "user", "content": "..."}… See the full description on the dataset page: https://huggingface.co/datasets/nabin2004/manim-sft.zygai_fishing_without_borders_2004
🎣 ZygAI – Fishing Without Borders 2004
A structured dataset of Lithuanian lakes and ichthyofauna (2004 edition)Dataset by ZygAI Research
🧭 About This Dataset
This dataset is an official extension of the ZygAI project’s✅ Atostogos Lietuvos kaime 2004 dataset.
The extended section, titled “Žvejyba be sienų 2004” (“Fishing Without Borders 2004”), contains detailed freshwater fishing information from the same publication.
It includes:
Lake names (LT + EN)
Municipal… See the full description on the dataset page: https://huggingface.co/datasets/ZygAI/zygai_fishing_without_borders_2004.