conversation-title
title-conversation-17-languages
Merged Conversation Title Dataset
Version: seventeen-language-llm-v1
This bundle merges an already published dataset with a newly generated
one. Published rows are carried through unchanged, in their original
splits; new rows are placed by the cross-run duplicate audit.
Split sizes
Split
Rows
train
20400
validation
2550
synthetic_holdout
2549
total
25499
Languages
Language
Train
Validation
Holdout
Total
de
1200
150… See the full description on the dataset page: https://huggingface.co/datasets/ManhHoDinh/title-conversation-17-languages.amzn_synthetic_conversation_title_id-align
