Tiny Router Data by tgupj is a synthetic dataset containing 2,907 examples for training a compact multi-head classifier. Each example includes a current_text field and four classification labels: relation_to_previous, actionability, retention, and urgency.
Use Cases
- Train a multi-head classifier to predict the four labels—relation_to_previous, actionability, retention, and urgency—from the current_text field.
- Evaluate model performance on short routing decisions using the synthetic examples for the actionability and urgency labels.
- Analyze label distributions, such as the frequency of retention categories (ephemeral, useful, remember), across the 2,907 synthetic examples.
Strengths
- Contains 2,907 synthetic examples designed for training and evaluation.
- Provides four distinct classification labels per example for multi-task learning.
Limitations
- Dataset is entirely synthetic, which may not reflect real-world text distributions or noise.
- Limited to 2,907 examples, which is a small sample size for training complex models.
Provenance
- Source
- tgupj on Hugging Face.
- Collection Method
- Synthetically generated data.
- Time Range
- null
- Freshness
- Last updated on 2026-03 20.
- Geography
- null