README.md

---
license: cc-by-4.0
task_categories:
- tabular-classification
pretty_name: ConfigIQ tested-model performance data
---

# ConfigIQ tested-model performance data

Sanitized aggregate GuideLLM measurements used to train and assess ConfigIQ's
model/hardware performance-envelope classifiers.

- Model/hardware pairs: 16
- Aggregate benchmark rows: 3949
- Format: one Parquet file per model/hardware pair
- License: CC BY 4.0

The dataset excludes run IDs, UUIDs, MLflow identifiers, timestamps, raw runtime
arguments, internal endpoints, and other administrative metadata. `manifest.json`
records pair identity, row count, columns, and checksums. This dataset provides
benchmark observations, not performance guarantees, and is supplied without warranty.