Datasets:

sdialog
/

200-dialogues-voices-libritts

Modalities:

Formats:

Size:

Libraries:

Dataset Viewer (First 5GB)

Split (1)

train · 800 rows

Search is not available for this dataset

The dataset viewer is not available for this split.

Rows from parquet row groups are too big to be read: 789.60 MiB (max=286.10 MiB)

Error code:   TooBigContentError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

200 dialogues generated using SDialog:

ExpO0O5 > DoPaCo > 001

001:
  both roles use gemma3:27b-it-qat as LLM
  only doctor gets truncated '?'

Split without persona overlapp:
train set: doc 0-59 pat 0-119
dev set: doc 60 -79 pat 120 - 139
test set: doc 80 - 99 pat 140 - 199