datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
stackexchange-clustering
StackExchangeClustering.v2
An MTEB dataset
Massive Text Embedding Benchmark
Clustering of titles from 121 stackexchanges. Clustering of 25 sets, each with 10-50 classes, and each class with 100 - 1000 sentences.
Task category
t2c
Domains
Web, Written
Reference
https://arxiv.org/abs/2104.07081
How to evaluate on this task
You can evaluate an embedding model on this dataset using the following code:
import mteb
task =… See the full description on the dataset page: https://huggingface.co/datasets/mteb/stackexchange-clustering.stack-exchange-preferences
Dataset Card for H4 Stack Exchange Preferences Dataset
Dataset Summary
This dataset contains questions and answers from the Stack Overflow Data Dump for the purpose of preference model training.
Importantly, the questions have been filtered to fit the following criteria for preference models (following closely from Askell et al. 2021): have >=2 answers.
This data could also be used for instruction fine-tuning and language model training.
The questions are grouped with… See the full description on the dataset page: https://huggingface.co/datasets/HuggingFaceH4/stack-exchange-preferences.StackExchange_Mar2023
Dataset Card for "StackExchange_Mar2023"
More Information needed
stackexchange_filtered
Stack Exchange
Description
StackExchange is a collection of Q&A communities spanning a wide variety of topics.While StackExchange formerly provided structured dumps of all of their content, since July of 2024, StackExchange has stopped publishing XML dumps to the Internet Archive.Instead, each site can provide a logged-in user with a custom URL to download the dump for that site.This means that dumps for defunct sites like windowsphone.stackexchange.com are… See the full description on the dataset page: https://huggingface.co/datasets/common-pile/stackexchange_filtered.stackexchange
Stack Exchange
Description
StackExchange is a collection of Q&A communities spanning a wide variety of topics.
While StackExchange formerly provided structured dumps of all of their content, since July of 2024, StackExchange has stopped publishing XML dumps to the Internet Archive.
Instead, each site can provide a logged in user with a custom url to download the dump for that site.
This means that dumps for defunct sites like windowsphone.stackexchange.com are… See the full description on the dataset page: https://huggingface.co/datasets/common-pile/stackexchange.stackexchange-clustering-p2p
StackExchangeClusteringP2P.v2
An MTEB dataset
Massive Text Embedding Benchmark
Clustering of title+body from stackexchange. Clustering of 5 sets of 10k paragraphs and 5 sets of 5k paragraphs.
Task category
t2c
Domains
Web, Written
Reference
https://arxiv.org/abs/2104.07081
How to evaluate on this task
You can evaluate an embedding model on this dataset using the following code:
import mteb
task = mteb.get_tasks(["StackExchangeClusteringP2P.v2"])… See the full description on the dataset page: https://huggingface.co/datasets/mteb/stackexchange-clustering-p2p.stackexchange_title_best_voted_answer_jsonlThis new dataset is designed to solve this great NLP task and is crafted with a lot of care.physics-stackexchangestackexchange_titlebody_best_voted_answer_jsonlThis new dataset is designed to solve this great NLP task and is crafted with a lot of care.stackexchange_title_body_jsonljsonl.gz format from https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_xml
Each line contains a dict in the format: {"text": ["title", "body"], "tags": ["tag1", "tag2"]}
The following parameters have been used for filtering: min_title_len = 20 min_body_len = 20 max_body_len = 4096 min_score = 0
If a stackexchange contained less than 10k questions (after filtering), it is written to the small_stackexchanges.jsonl.gz file.
This is a dump of the files from… See the full description on the dataset page: https://huggingface.co/datasets/flax-sentence-embeddings/stackexchange_title_body_jsonl.stackexchange_titlebody_best_and_down_voted_answer_jsonlThis new dataset is designed to solve this great NLP task and is crafted with a lot of care.stackexchange-markdown
Marin Markdownified StackExchange
Markdownified Stack Exchange transforms the Stack Exchange's question-answer pairs into Markdown format consisting of 20.4B tokens. This dataset preserves the content contained in technical discussions while organizing it into a thread format for language model training.
Value
Tokens
20 413 785 853
Primary source
https://archive.org/details/stackexchange
File format
JSONL
License
CC (mirrors upstream SE licenses)… See the full description on the dataset page: https://huggingface.co/datasets/marin-community/stackexchange-markdown.stack-exchange-preferences-code
Dataset Card for "stack-exchange-preferences-code"
More Information needed
stackexchangestack-exchange-preferences-20230914-clean-anonymization
Dataset Card for "stack-exchange-preferences-20230914-clean-anonymization"
More Information needed
stackexchange_math_jsonlThis new dataset is designed to solve this great NLP task and is crafted with a lot of care.stack-exchange-preferences-code-v2
Dataset Card for "stack-exchange-preferences-code-v2"
More Information needed
unprocessed_stackexchange_datastackexchange-all
StackExchange All Communities — Preprocessed
All 362 StackExchange communities (including Stack Overflow) processed into
QA pairs and standalone questions, ready for LLM pretraining and instruction tuning.
Stats
Field
Value
Communities
362
Dump date
March 2026
License
CC-BY-SA 4.0
Record Types
instruction — Question + Answer pair (qa_pair)
text — Unanswered question (standalone)
Format
{
"id": "uuid-v4"… See the full description on the dataset page: https://huggingface.co/datasets/raj2708/stackexchange-all.oa-stackexchange
Stackexchange Instructions for OpenAssistant
This dataset is taken from https://archive.org/details/stackexchange.
There's a single parquet file combining all stackexchange sites. The threads
have been filtered as follows: only threads with an accepted answer, for which
both the question and response is less than 1000 characters have been choosen.
Other answers, or questions without accepted answers, or long entries have been
droppped.
Each row consists of
INSTRUCTION
RESPONSE… See the full description on the dataset page: https://huggingface.co/datasets/donfu/oa-stackexchange.stack-exchange-datasetlaw_stackexchange_prompts
Dataset Card for "law_stackexchange_prompts"
More Information needed
law-stack-exchange
Dataset Card for Law Stack Exchange Dataset
Dataset Summary
Dataset from the Law Stack Exchange, as used in "Parameter-Efficient Legal Domain Adaptation".
Citation Information
@inproceedings{li-etal-2022-parameter,
title = "Parameter-Efficient Legal Domain Adaptation",
author = "Li, Jonathan and
Bhambhoria, Rohan and
Zhu, Xiaodan",
booktitle = "Proceedings of the Natural Legal Language Processing Workshop 2022",
month = dec… See the full description on the dataset page: https://huggingface.co/datasets/jonathanli/law-stack-exchange.stackexchange_mathstackexchange
Dataset Card for "stackexchange"
More Information needed
stackexchange-question-answering
SYNTHETIC-1
This is a subset of the task data used to construct SYNTHETIC-1. You can find the full collection here
stackexchange-duplicates
Dataset Card for Stack Exchange Duplicates
This dataset contains the Stack Exchange Duplicates dataset in three formats that are easily used with Sentence Transformers to train embedding models. The data was originally extracted using the Stack Exchange API and taken from embedding-training-data.
Each pair contains data from two Stack Exchange posts that were marked as duplicates. title-title-pair only has the titles, body-body-pair only the bodies, and post-post-pair has both.… See the full description on the dataset page: https://huggingface.co/datasets/sentence-transformers/stackexchange-duplicates.stackexchange_20260331
Stack Exchange data dump - community release
This data dump is sourced from the various sites in the Stack Exchange network of Q&A sites.
This dump contains data up to and including 2026-03-31.
This community release is an unofficial replacement for the now-killed official archive.org release. As a reminder, windowsphone.stackexchange.com is still excluded from this data dump, as it isn't possible to get the archives anymore due to the site shutting down a few years ago. See the… See the full description on the dataset page: https://huggingface.co/datasets/flymin/stackexchange_20260331.stackexchange-superuser-sandboxes-verified-qwen3.5-122b-131k-opencode-literal-rescue-traces
Agent trace dataset
Decoding the literal token IDs
The prompt_token_ids / completion_token_ids / logprobs columns are the
verbatim tokens the serving engine emitted, stored PER AGENT STEP as a
list-of-lists (one inner list per turn). To turn them back into text you MUST
use the exact tokenizer the model was served with — a generic same-family
tokenizer will decode word tokens to garbage.
Served model / tokenizer source: Qwen/Qwen3.5-122B-A10B-FP8
from transformers… See the full description on the dataset page: https://huggingface.co/datasets/open-athena/stackexchange-superuser-sandboxes-verified-qwen3.5-122b-131k-opencode-literal-rescue-traces.stackexchange-clustering-exploded-test-vn