Synthetic notification dataset for training notification classification models.
Metric
Value
Total samples
16,000
Languages
English, Chinese
Folders
4 (Work, Personal, Promotions, Alerts)
Priority levels
3 (High, Medium, Low)
File
Description
Samples
training_data.jsonl
Main training dataset
16,000
raw/batch_*.jsonl
Raw generation batches
~400 each
Data Format… See the full description on the dataset page: https://huggingface.co/datasets/charlesfeng1/notifai-dataset. 
