datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
fineweb-edu
📚 FineWeb-Edu
1.3 trillion tokens of the finest educational data the 🌐 web has to offer
Paper: https://arxiv.org/abs/2406.17557
What is it?
📚 FineWeb-Edu dataset consists of 1.3T tokens and 5.4T tokens (FineWeb-Edu-score-2) of educational web pages filtered from 🍷 FineWeb dataset. This is the 1.3 trillion version.
To enhance FineWeb's quality, we developed an educational quality classifier using annotations generated by LLama3-70B-Instruct. We… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu.fineweb
🍷 FineWeb
15 trillion tokens of the finest data the 🌐 web has to offer
What is it?
The 🍷 FineWeb dataset consists of more than 18.5T tokens (originally 15T tokens) of cleaned and deduplicated english web data from CommonCrawl. The data processing pipeline is optimized for LLM performance and ran on the 🏭 datatrove library, our large scale data processing library.
🍷 FineWeb was originally meant to be a fully open replication of 🦅 RefinedWeb, with a… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/fineweb.the_cauldron
Dataset Card for The Cauldron
Dataset description
The Cauldron is part of the Idefics2 release.
It is a massive collection of 50 vision-language datasets (training sets only) that were used for the fine-tuning of the vision-language model Idefics2.
Load the dataset
To load the dataset, install the library datasets with pip install datasets. Then,
from datasets import load_dataset
ds = load_dataset("HuggingFaceM4/the_cauldron", "ai2d")
to download… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceM4/the_cauldron.finephrase
Dataset Card for HuggingFaceFW/finephrase
Dataset Summary
Synthetic data generated by DataTrove:
Model: HuggingFaceTB/SmolLM2-1.7B-Instruct (main)
Source dataset: HuggingFaceFW/fineweb-edu, config sample-350BT, split train
Generation config: temperature=1.0, top_p=1.0, top_k=50, max_tokens=2048, model_max_context=8192
Speculative decoding: {"method":"suffix","num_speculative_tokens":32}
System prompt: None
Input column: text
Prompt families:
faq prompt
Rewrite… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/finephrase.MATH-500
Dataset Card for MATH-500
This dataset contains a subset of 500 problems from the MATH benchmark that OpenAI created in their Let's Verify Step by Step paper. See their GitHub repo for the source file: https://github.com/openai/prm800k/tree/main?tab=readme-ov-file#math-splits
FineVision
Fine Vision
FineVision is a massive collection of datasets with 17.3M images, 24.3M samples, 88.9M turns, and 9.5B answer tokens, designed for training state-of-the-art open Vision-Language-Models.
More detail can be found in the blog post: https://huggingface.co/spaces/HuggingFaceM4/FineVision
Load the data
from datasets import load_dataset, get_dataset_config_names
# Get all subset names and load the first one
available_subsets =… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceM4/FineVision.stack-v3-train
🥞 The Stack v3
What is it?
What is being released
How to download and use it
Dataset statistics
Dataset structure
Dataset creation
Considerations for using the data
Additional information
What is it?
The Stack v3 is the largest, most up-to-date open dataset of source code, crawled directly from GitHub and built to pre-train code LLMs with full-repository context. It is the successor to The Stack v2 and, like its predecessor, is released to make the training… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceCode/stack-v3-train.ultrachat_200k
Dataset Card for UltraChat 200k
Dataset Description
This is a heavily filtered version of the UltraChat dataset and was used to train Zephyr-7B-β, a state of the art 7b chat model.
The original datasets consists of 1.4M dialogues generated by ChatGPT and spanning a wide range of topics. To create UltraChat 200k, we applied the following logic:
Selection of a subset of data for faster supervised fine tuning.
Truecasing of the dataset, as we observed around 5% of… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceH4/ultrachat_200k.fineweb-2
🥂 FineWeb2
A sparkling update with 1000s of languages
What is it?
This is the second iteration of the popular 🍷 FineWeb dataset, bringing high quality pretraining data to over 1000 🗣️ languages.
The 🥂 FineWeb2 dataset is fully reproducible, available under the permissive ODC-By 1.0 license and extensively validated through hundreds of ablation experiments.
In particular, on the set of 9 diverse languages we used to guide our processing decisions, 🥂… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/fineweb-2.FineVisionMax
Fine Vision
FineVision is a massive collection of datasets with 17.3M images, 24.3M samples, 88.9M turns, and 9.5B answer tokens, designed for training state-of-the-art open Vision-Language-Models.
More detail can be found in the blog post: https://huggingface.co/spaces/HuggingFaceM4/FineVision
The version in this repository concatenated all the configs in the original dataset and then shuffled them. This is done to facilitate streaming the data directly from the hub!
Load… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceM4/FineVisionMax.no_robots
Dataset Card for No Robots 🙅♂️🤖
Look Ma, an instruction dataset that wasn't generated by GPTs!
Dataset Summary
No Robots is a high-quality dataset of 10,000 instructions and demonstrations created by skilled human annotators. This data can be used for supervised fine-tuning (SFT) to make language models follow instructions better. No Robots was modelled after the instruction dataset described in OpenAI's InstructGPT paper, and is comprised mostly of single-turn… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceH4/no_robots.aime_2024
Dataset card for AIME 2024
This dataset consists of 30 problems from the 2024 AIME I and AIME II tests. The original source is AI-MO/aimo-validation-aime, which contains a larger set of 90 problems from AIME 2022-2024.
smollm-corpus
SmolLM-Corpus
This dataset is a curated collection of high-quality educational and synthetic data designed for training small language models.
You can find more details about the models trained on this dataset in our SmolLM blog post.
Dataset subsets
Cosmopedia v2
Cosmopedia v2 is an enhanced version of Cosmopedia, the largest synthetic dataset for pre-training, consisting of over 39 million textbooks, blog posts, and stories generated by… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus.smoltalk
SmolTalk
Dataset description
This is a synthetic dataset designed for supervised finetuning (SFT) of LLMs. It was used to build SmolLM2-Instruct family of models and contains 1M samples. More details in our paper https://arxiv.org/abs/2502.02737
During the development of SmolLM2, we observed that models finetuned on public SFT datasets underperformed compared to other models with proprietary instruction datasets. To address this gap, we created new synthetic datasets… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceTB/smoltalk.Docmatix
Dataset Card for Docmatix
Dataset description
Docmatix is part of the Idefics3 release (stay tuned).
It is a massive dataset for Document Visual Question Answering that was used for the fine-tuning of the vision-language model Idefics3.
Load the dataset
To load the dataset, install the library datasets with pip install datasets. Then,
from datasets import load_dataset
ds = load_dataset("HuggingFaceM4/Docmatix")
If you want the dataset to link to the pdf files… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceM4/Docmatix.finepdfs
Liberating 3T of the finest tokens from PDFs
What is this?
As we run out of web pages to process, the natural question has always been: what to do next? Only a few knew about a data source that everyone avoided for ages, due to its incredible extraction cost and complexity: PDFs.
📄 FinePDFs is exactly that. It is the largest publicly available corpus sourced exclusively from PDFs, containing about 3 trillion tokens across 475 million documents in 1733 languages.
Compared to… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/finepdfs.finetranslations
💬 FineTranslations
The world's knowledge in 1+1T tokens of parallel text
What is it?
This dataset contains over 1 trillion tokens of parallel text in English and 500+ languages. It was obtained by translating data from 🥂 FineWeb2 into English using Gemma3 27B.
We relied on datatrove's inference runner to deploy a synthetic data pipeline at scale. Its checkpointing and VLLM lifecycle management features allowed us to use leftover compute from the HF cluster… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/finetranslations.finemath
📐 FineMath
What is it?
📐 FineMath consists of 34B tokens (FineMath-3+) and 54B tokens (FineMath-3+ with InfiMM-WebMath-3+) of mathematical educational content filtered from CommonCrawl. To curate this dataset, we trained a mathematical content classifier using annotations generated by LLama-3.1-70B-Instruct. We used the classifier to retain only the most educational mathematics content, focusing on clear explanations and step-by-step problem solving rather than… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceTB/finemath.drlc-leaderboard-datacosmopedia
Cosmopedia v0.1
Image generated by DALL-E, the prompt was generated by Mixtral-8x7B-Instruct-v0.1
Note: Cosmopedia v0.2 is available at smollm-corpus
User: What do you think "Cosmopedia" could mean? Hint: in our case it's not related to cosmology.
Mixtral-8x7B-Instruct-v0.1: A possible meaning for "Cosmopedia" could be an encyclopedia or collection of information about
different cultures, societies, and topics from around the world, emphasizing diversity and global… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceTB/cosmopedia.smoltalk2
SmolTalk2
Dataset description
This dataset contains three subsets (Mid, SFT, Preference) that correspond to the three phases of Post-Training for SmolLM3-3B. You can find more details in our blog post about how we used the data in each of the stages SmolLM3.
The specific weight of each subset is available in the training recipe in SmolLM's repository.
You can load a dataset using
from datasets import load_dataset
# To load the train split of a specific subset, such as… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceTB/smoltalk2.CADS-dataset
CADS: A Comprehensive Anatomical Dataset and Segmentation for Whole-Body Anatomy in Computed Tomography
Overview
CADS is a robust, fully automated framework for segmenting 167 anatomical structures in Computed Tomography (CT), spanning from head to knee regions across diverse anatomical systems.
The framework consists of two main components:
CADS-dataset:
22,022 CT volumes with complete annotations for 167 anatomical structures.
Most extensive whole-body CT dataset… See the full description on the dataset page: https://huggingface.co/datasets/huggingface/CADS-dataset.ultrafeedback_binarized
Dataset Card for UltraFeedback Binarized
Dataset Description
This is a pre-processed version of the UltraFeedback dataset and was used to train Zephyr-7Β-β, a state of the art chat model at the 7B parameter scale.
The original UltraFeedback dataset consists of 64k prompts, where each prompt is accompanied with four model completions from a wide variety of open and proprietary models. GPT-4 is then used to assign a score to each completion, along criteria like… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized.CodeAlpaca_20KThis dataset splits the original CodeAlpaca dataset into train and test splits.
huggingface-spaces-codes
📊 Dataset Description
This dataset comprises code files of Huggingface Spaces that have more than 0 likes as of November 10, 2023. This dataset contains various programming languages totaling in 672 MB of compressed and 2.05 GB of uncompressed data.
📝 Data Fields
Field
Type
Description
repository
string
Huggingface Spaces repository names.
sdk
string
Software Development Kit of the space.
license
string
License type of the space.… See the full description on the dataset page: https://huggingface.co/datasets/Weyaxi/huggingface-spaces-codes.fineweb-edu-score-2
📚 FineWeb-Edu-score-2
1.3 trillion tokens of the finest educational data the 🌐 web has to offer
What is it?
📚 FineWeb-Edu dataset consists of 1.3T tokens (FineWeb-Edu) and 5.4T tokens of educational web pages filtered from 🍷 FineWeb dataset. This is the 5.4 trillion version.
Note: this version uses a lower educational score threshold = 2, which results in more documents, but lower quality compared to the 1.3T version. For more details check the… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu-score-2.ChartQA
Dataset Card for "ChartQA"
More Information needed
testing_alpaca_small
Dataset Card for "testing_alpaca_small"
More Information needed
finevideo
FineVideo
FineVideo
Description
Dataset Explorer
Revisions
Dataset Distribution
How to download and use FineVideo
Using datasets
Using huggingface_hub
Load a subset of the dataset
Dataset StructureData Instances
Data Fields
Dataset Creation
License CC-By
Considerations for Using the Data
Social Impact of Dataset
Discussion of Biases
Additional Information
Credits
Future Work
Opting out of FineVideo
Citation Information
Terms of use for FineVideo… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceFV/finevideo.WebSight
Dataset Card for WebSight
Dataset Description
WebSight is a large synthetic dataset containing HTML/CSS codes representing synthetically generated English websites, each accompanied by a corresponding screenshot.
This dataset serves as a valuable resource for tasks such as generating UI codes from a screenshot.
It comes in two versions:
v0.1: Websites are coded with HTML + CSS. They do not include real images.
v0.2: Websites are coded with HTML + Tailwind CSS. They do… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceM4/WebSight.