Dataset Viewer
Auto-converted to Parquet Duplicate
column_name
stringclasses
6 values
column_description
stringclasses
2 values
id
the unique id representing the trains
direction
the direction of trains that are running
id
the unique id representing the trains
direction
the direction of trains that are running
id
the unique id representing the trains
direction
the direction of trains that are running
device_id
null
gender
null
age
null
group
null
device_id
null
gender
null
age
null
group
null
device_id
null
gender
null
age
null
group
null

ELT-Bench Inputs

This dataset contains the 100 ELT-Bench task inputs for three data warehouses:

Data warehouse Directory
Snowflake inputs/
Databricks inputs_databricks/
Redshift inputs_redshift/

Every directory is self-contained. Each task includes its destination-specific config.yaml and data_model.yaml, together with a copy of the shared table schemas under schemas/.

Ground Truth

The v5 ground-truth output tables are provided separately for each data warehouse:

Data warehouse Ground-truth directory
Snowflake gt_v5/
Databricks gt_v5_databricks/
Redshift gt_redshift_v5/

Each directory contains 202 evaluation CSV files organized as <database>/<output_table>.csv.

Downloads last month
2