datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
TinyStoriesDataset containing synthetically generated (by GPT-3.5 and GPT-4) short stories that only use a small vocabulary.
Described in the following paper: https://arxiv.org/abs/2305.07759.
The models referred to in the paper were trained on TinyStories-train.txt (the file tinystories-valid.txt can be used for validation loss). These models can be found on Huggingface, at roneneldan/TinyStories-1M/3M/8M/28M/33M/1Layer-21M.
Additional resources:
tinystories_all_data.tar.gz - contains a superset of… See the full description on the dataset page: https://huggingface.co/datasets/roneneldan/TinyStories.tiny-supervised-datasettiny-shakespeare
Data source
Downloaded via Andrej Karpathy's nanogpt repo from this link
Data Format
The entire dataset is split into train (90%) and test (10%).
All rows are at most 1024 tokens, using the Llama 2 tokenizer.
All rows are split cleanly so that sentences are whole and unbroken.
tiny-aya-l2-thinker-multilingual-reasoning
Tiny Aya L2 Multilingual Reasoning (44 languages)
Translated multilingual reasoning traces used to train Tiny Aya L2-Thinker.
Each example has the prompt, thinking, and answer in the same non-English language alongside the original texts in English.
Data source
Prompts from AM-DeepSeek-R1-0528-Distilled
Thinking traces and outputs distilled from gpt-oss-120b
Translated with command-a-translate and DeepSeek-V3
Languages (44)
Language
Train… See the full description on the dataset page: https://huggingface.co/datasets/CohereLabs/tiny-aya-l2-thinker-multilingual-reasoning.tiny-instruct
tiny-instruct-v1
This dataset is collated from multiple other open-source datasets (de-duplicated). This has a total of ~6M rows each with an instruction and response (single-turn converstion).
Code Datasets:
CodeAlpaca_20K
CodeExercise-Python-27k
Evol-Instruct-Code-80k-v1
tiny-codes
Evol-instruction-66k
sciphi-python-textbook
programming_books_llama
WizardLM_evol_instruct_70k
Math Datasets:
MetaMathQA
arxiv-math-instruct-50k
MathInstruct… See the full description on the dataset page: https://huggingface.co/datasets/04RR/tiny-instruct.text-dataset-tiny-code-script-py-format
USED of tahamajs/medicine_ds_persian for .parquet file
USED of Alijafarixcs2/persian-it-llama2-2k for .parquet file
USED of Abirate/english_quotes for .jsonl file
NEW FILES (05/12/2025)
NEW FILES (12/26/2025)
NEW FILES (02/15/2026)
tiny-codes
Reasoning with Language and Code
This synthetic dataset is a collection of 1.6 millions short and clear code snippets that can help LLM models learn how to reason with both natural and programming languages. The dataset covers a wide range of programming languages, such as Python, TypeScript, JavaScript, Ruby, Julia, Rust, C++, Bash, Java, C#, and Go. It also includes two database languages: Cypher (for graph databases) and SQL (for relational databases) in order to study the… See the full description on the dataset page: https://huggingface.co/datasets/nampdn-ai/tiny-codes.multilingual_TinyStories
Dataset Card for Multilingual TinyStories
Dataset Details
Dataset Description
The Multilingual TinyStories dataset contains translations of the original TinyStories dataset, which consists of synthetically generated short stories using a small vocabulary suitable for 3 to 4-year-olds. These stories were originally generated by GPT-3.5 and GPT-4. The multilingual versions have been translated into various languages, including Spanish, Chinese, German, Turkish… See the full description on the dataset page: https://huggingface.co/datasets/Gabrui/multilingual_TinyStories.tiny_schiller
tiny_schiller
A small (~2 MB) German-language analogue to Karpathy's tiny_shakespeare — 11 of Friedrich Schiller's dramatic works, cleaned and tokenised for tutorial-scale language models.
For a compact, agent-friendly summary (file inventory, load patterns, licensing), see DATA_CARD.md.
"Das Leben ist nur ein Moment, der Tod ist auch nur einer." — Friedrich Schiller
Corpus
~2.07 MB · 11 works · 2,019,857 characters · sourced from DraCor / GerDraCor (CC0). See… See the full description on the dataset page: https://huggingface.co/datasets/mrkschtr/tiny_schiller.tiny-textbooks
Textbook-like Dataset: A High-Quality Resource for Small Language Models
The idea is simply inspired by the Textbooks Are All You Need II: phi-1.5 technical report paper. The source texts in this dataset have been gathered and carefully select the best of the falcon-refinedweb and minipile datasets to ensure the diversity, quality while tiny in size. The dataset was synthesized using 4x3090 Ti cards over a period of 500 hours, thanks to Nous-Hermes-Llama2-13b finetuned model.
Why… See the full description on the dataset page: https://huggingface.co/datasets/nampdn-ai/tiny-textbooks.TinyStoriesDataset containing synthetically generated (by GPT-3.5 and GPT-4) short stories that only use a small vocabulary.
Described in the following paper: https://arxiv.org/abs/2305.07759.
The models referred to in the paper were trained on TinyStories-train.txt (the file tinystories-valid.txt can be used for validation loss). These models can be found on Huggingface, at roneneldan/TinyStories-1M/3M/8M/28M/33M/1Layer-21M.
Additional resources:
tinystories_all_data.tar.gz - contains a superset of… See the full description on the dataset page: https://huggingface.co/datasets/HayatoHongo/TinyStories.tiny-aya-l2-thinker-multilingual-reasoning
Tiny Aya L2 Multilingual Reasoning (44 languages)
Translated multilingual reasoning traces used to train Tiny Aya L2-Thinker.
Each example has the prompt, thinking, and answer in the same non-English language alongside the original texts in English.
Languages (44)
Language
Train
Test
Total
Amharic (am)
3,807
448
4,255
Arabic (ar)
22,968
2,538
25,506
Bulgarian (bg)
4,177
452
4,629
Bengali (bn)
3,803
422
4,225
Catalan (ca)
4,251
512
4,763
Czech… See the full description on the dataset page: https://huggingface.co/datasets/erenyeager-1/tiny-aya-l2-thinker-multilingual-reasoning.TinyStoriesAdv-zh
TinyStoriesAdv
keywords: grade school level, large language model, small language model, tiny language model, super tiny language model, 小学生知识水平，大语言模型，小语言模型，迷你语言模型, llm, slm.
受到TinyStories、Phi2等论文的启发，我制作了一个约1B tokens的小学知识水平的“一揽子”大语言模型训练语料库。
“一揽子”指的是本数据集是众多数据集的集合。为了提升模型的不同能力（例如事实性知识、元认知、思维链、阅读理解RAG、逻辑推理等），我开了不少脑洞，使用了多种创新的提示词生成了具有多样性和针对性的子数据集。… See the full description on the dataset page: https://huggingface.co/datasets/fzmnm/TinyStoriesAdv-zh.TinyTextThe entire NanoPhi Dataset is at train.jsonl
Separate Tasks Include
Math (Metamath, mammoth)
Code (Code Search Net)
Logic (Open-platypus)
Roleplay (PIPPA, RoleplayIO)
Textbooks (Tiny-text, Sciphi)
Textbook QA (Orca-text, Tiny-webtext)
tiny-stories-tokenized-bpetinysynth-reasoning
TinySynth Reasoning Primitives
Synthetic training data for teaching small language models stable state
representation and controlled reasoning operations — entity/attribute
binding, state persistence, mutation, transfer, reference resolution,
current-vs-cumulative distinctions, and claim validation — in a
systems/computing vocabulary.
Every example is generated from a hidden symbolic world and verified by a
symbolic solver before any natural language is produced:
semantic state… See the full description on the dataset page: https://huggingface.co/datasets/projenix/tinysynth-reasoning.TinyDialogues
Dataset Card for TinyDialogues
TinyDialogues dataset collected as part of the EMNLP 2024 paper "Is Child-Directed Speech Effective Training Data for Language Models?" by Steven Y. Feng, Noah D. Goodman, and Michael C. Frank. For more details, please see Appendices A-C in our paper.
Dataset Sources
Repository: https://github.com/styfeng/TinyDialogues
Paper: https://aclanthology.org/2024.emnlp-main.1231/
Dataset Structure
Final training and validation data… See the full description on the dataset page: https://huggingface.co/datasets/styfeng/TinyDialogues.k2-horizon-tiny-cpu-repro-v1
K2-Horizon MoVA tiny random CPU fixture
Complete untrained K2HorizonForCausalLM, not Moonshot Kimi despite the K2 name.
Architecture source: IFM/K2-Horizon-MoVA-36B-A4B at 05cab0a4d7150c1c460a000b37ff40cc1af2feaa.
No pretrained weights, original tokenizer, training data, paid GPU or cloud compute used.
Complete text-only K2HorizonForCausalLM, not Kimi: three-layer dense prefix followed by two real MoVA+MoE layers, grouped RMSNorm, sigmoid top-k routing with selection-only bias… See the full description on the dataset page: https://huggingface.co/datasets/malaiwah/k2-horizon-tiny-cpu-repro-v1.TinyStoriesV2_cleaned
License: CDLA-Sharing-1.0
Dataset containing synthetically generated (GPT-4) short stories that only use a small vocabulary.
Described in the following paper: https://arxiv.org/abs/2305.07759.
This is a cleaned up Version of the original TinyStories Dataset: https://huggingface.co/datasets/roneneldan/TinyStories.
We thank the authors for their contribution.
This Version only contains cleaned-up stories generated by GPT4.
Stories were deleted that contained spelling and… See the full description on the dataset page: https://huggingface.co/datasets/fhswf/TinyStoriesV2_cleaned.TinyHelen-zh
What's New
Mar.31 2025 Added instruct fine-tuning and reasoning dataset in the same ELI5 style. Take a look!
Mar.31 2025 See my new model.
TinyHelen-zh
Inspired by the paper TinyHelen's First Curriculum, we present a Chinese version of the LLM-simplified training corpus. This dataset is converted from high-quality Chinese and English web crawls for training baby-size (<100M) language models.
Adult-talking
北京市财政局、北京海关、国家税务总局北京市税务局、北京市国际服务贸易事务中心：… See the full description on the dataset page: https://huggingface.co/datasets/fzmnm/TinyHelen-zh.tiny-strange-textbooks
Quirky Textbook Trove: Compact Excellence for Small Language Model
Strange dataset is 100% AI-generated, a compilation aligned with the vision of the Textbooks Are All You Need and Textbooks Are All You Need II: phi-1.5 technical report research. This dataset features 2,7M synthetic textbooks, encapsulating 16GB of raw text data. The unique name reflects its unconventional synthesis methodology, its compact size, deduped, and its emphasis on clear, focused content.
The dataset… See the full description on the dataset page: https://huggingface.co/datasets/nampdn-ai/tiny-strange-textbooks.TinyHelen_Data
TinyHelen
This repository contains the code and resources for the paper:TinyHelen's First Curriculum: Training and Evaluating Tiny Language Models in a Simpler Language Environment
☄️☄️ Overview ☄️☄️
TinyHelen introduces a novel approach to training and evaluating tiny language models (LMs) using a simplified text dataset. This methodology mimics how children learn language in structured environments, focusing on systematically reduced vocabularies and linguistic complexities as… See the full description on the dataset page: https://huggingface.co/datasets/Aviv-anthonnyolime/TinyHelen_Data.deepseek-v4-tiny-cpu-repro-v1
DeepSeek-V4 tiny corrected-native-primitives CPU text fixture
Complete randomly initialized, untrained QFSDeepseekV4ForCausalLM text class
using Transformers5.16.1 native primitives and a reviewed RMSNorm arithmetic correction.
No upstream weights, paid GPU/cloud compute or useful-model claim.
This is not unmodified native Transformers or the complete production release.
Architecture and scope
Text lineage:… See the full description on the dataset page: https://huggingface.co/datasets/malaiwah/deepseek-v4-tiny-cpu-repro-v1.TinyStories-Multilingual
Novelist: TinyStories Multilingual Edition
Dataset Summary
The TinyStories Multilingual Edition is a high-fidelity synthetic dataset of short, child-safe fiction designed to stress-test literary consistency, emotional warmth, and multilingual fluency in small models. Derived from the broader Novelist ecosystem, this subset focuses on narrative simplicity paired with complex moral and social themes.
The dataset contains 15,688 high-quality stories across 28 languages. Each… See the full description on the dataset page: https://huggingface.co/datasets/Dxniz/TinyStories-Multilingual.glm5-next-tiny-cpu-repro-v1This repository is an evidence bundle, not one root-format dataset at repository root.
first/ and repeat/ are separate complete sealed QFS root datasets; comparison/ holds
the comparison receipt and tokenwise result. panel/ is the sealed input panel. Other files
are provenance, logs and reproduction tools. Do not pass the bundle root as a QFS dataset.
GLM5-Next tiny native CPU fixture
This is a complete untrained random-initialized native Glm5NextForConditionalGeneration
wrapper… See the full description on the dataset page: https://huggingface.co/datasets/malaiwah/glm5-next-tiny-cpu-repro-v1.AL-GR-Tiny
AL-GR-Tiny: A Complete & Sampled Generative Recommendation Dataset
Dataset Summary
AL-GR-Tiny is a compact, self-contained, and sampled version of the large-scale AL-GR ecosystem. It is designed for users who want to quickly experiment, develop, or understand the full pipeline of generative recommendation without needing to process terabytes of data.
This "all-in-one" repository bundles everything you need:
Pre-processed Training/Testing Data: Ready-to-use data for… See the full description on the dataset page: https://huggingface.co/datasets/AL-GR/AL-GR-Tiny.TinyStories-Farsi
Tiny Stories Farsi
The Tiny Stories Farsi project is a continuous effort to translate the Tiny Stories dataset into the Persian (Farsi) language. The primary goal is to produce a high-quality Farsi dataset, maintaining equivalency with the original English version, and subsequently to utilize it for training language models in Farsi. This seeks to affirm that the advancements and trends observed in English language models are replicable and applicable in other languages. Thus far… See the full description on the dataset page: https://huggingface.co/datasets/taesiri/TinyStories-Farsi.qwen3-5-tiny-cpu-repro-v1
Qwen3.5 tiny native random CPU fixture
Complete randomly initialized, untrained Qwen3_5ForConditionalGeneration checkpoint.
This is a pipeline/reproducibility fixture, not a useful language model, distillation,
quantization, quality benchmark, or claim about the performance of Qwen3.8-27B.
No upstream model weights or training data were used. No paid GPU/cloud compute.
Architecture and lineage
Architecture lineage: Qwen/Qwen3.8-27B at… See the full description on the dataset page: https://huggingface.co/datasets/malaiwah/qwen3-5-tiny-cpu-repro-v1.TinyStories-Algerian-Darija
TinyStories Algerian Darija
Synthetic short stories in Algerian Darja paired with their English originals, for Darija language modeling and translation, from the Algerian NLP Collective. The Hub datasets-server reports 11,326 train rows (/info?dataset=algerian-nlp/TinyStories-Algerian-Darija, 2026-09-17), independently confirmed by the build ledger processed_story_ids.json in this repo: 11,326 unique story ids (0 to 11,492, non-contiguous).
The default config answers: what does… See the full description on the dataset page: https://huggingface.co/datasets/algerian-nlp/TinyStories-Algerian-Darija.minimax-m2-tiny-cpu-repro-v1
minimax-m2 complete native tiny random CPU fixture
Complete untrained MiniMaxM2ForCausalLM checkpoint with an untied full LM head,
a real 272-entry byte tokenizer and every native state tensor. Architecture lineage:
MiniMaxAI/MiniMax-M2.7@d494266a4affc0d2995ba1fa35c8481cbd84294b.
No upstream weights, training data, paid GPU or cloud compute were used.
Complete native text causal LM: sigmoid/top-k MoE routing with correction bias, per-layer flattened Q/K RMSNorm and half-head… See the full description on the dataset page: https://huggingface.co/datasets/malaiwah/minimax-m2-tiny-cpu-repro-v1.