Dataset · knkarthick
Messenger chat corpus with summaries (SAMSum)
knkarthick/samsum
About 16,000 English dialogues with brief summaries by linguists.
What's inside
A popular English text corpus containing approximately 16,000 modern messenger-style dialogues. All conversations were specially created and recorded by fluent English-speaking linguists. Each dialogue comes with a brief summary of its essence.
What you can do with it
- Training or fine-tuning models for dialogue and chat summarization
- Benchmarking informal chat summarization quality
- Tasks
- Summarization
- Languages
- English
- Size
- 10K–100K rows
- Downloads this month
- 10.1K
- Likes
- 44
Models trained on it
- chopratejas/kompress-baseToken classification 711
- tuanhqv123/longt5-meeting-summarizationSummarization 56
- svk260404/text-summarizer-t5Summarization 52
- sebastiao-teixeira/gpt2-samsum-loraText generation 33
- kelly-n/gpt2-samsum-loraSummarization 33
- riturajpandey739/Qwen3-4B-Text-Summarizer-finetuningText generation 25
- violetar/gpt2-samsum-lora 25
- luanacarolina/gpt2-samsum-loraSummarization 21
How to load it
from datasets import load_dataset
ds = load_dataset("knkarthick/samsum")