▦ Dataset Viewer
Auto-converted preview files are shown as Parquet-style review rows. Original files remain available in Files and versions.
from datasets import load_dataset
dataset = load_dataset("awkward-nlp/elevator-small-talk-corpus")<iframe src="/hf/elevator-small-talk-corpus/#data-studio"></iframe>
utterance_id
string · lengths
0 7.07k
talk-001 · Weather is certainly happening today. · weather-neutral · train
talk-002 · This elevator has excellent vertical commitment. · situational · train
talk-003 · Mondays arrive with confidence. · calendar · validation
Showing preview page 1 for elevator-small-talk-corpus.
Dataset Card for "Elevator Small Talk Corpus"
A harmless fake NLP release for testing text previews, label descriptions, and download affordances in pre-release mocks.
Mock release notes
Tiny synthetic text snippets labeled by how quickly everyone starts staring at the floor numbers.
Settings
This mock settings tab is intentionally read-only. It exists so reviewers can verify the expected Hugging Face navigation target without changing a real release.