raj2708/stackexchange-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",
  "source": "stackexchange",
  "type": "instruction",
  "language": "en",
  "text": "Question: ...\n\nAnswer: ...",
  "metadata": {
    "url": "https://stackoverflow.com/questions/123",
    "community": "stackoverflow.com",
    "question_id": "123",
    "answer_id": "456",
    "tags": "python|list|sorting",
    "question_score": 42,
    "answer_score": 38,
    "is_accepted": true,
    "license": "CC-BY-SA-4.0",
    "scraped_at": "2026-06-19T00:00:00+00:00",
    "quality_score": 0.69,
    "token_count": 256,
    "instruction": "Question text...",
    "response": "Answer text..."
  }
}

Usage

from datasets import load_dataset

# Streaming (recommended)
ds = load_dataset("raj2708/stackexchange-all", split="train", streaming=True)
for row in ds:
    print(row["metadata"]["community"])
    print(row["text"][:300])
    break

Source

Processed from the official StackExchange data dump (March 2026): https://archive.org/details/stackexchange