Update README.md

README.md

@@ -35,3 +35,52 @@ configs:
   - split: train
     path: data/train-*
 ---
+
+# SQaLe 2 - Work in Progress
+
+## Using the SQaLe Library
+
+The [SQaLe library](https://pypi.org/project/SQaLe/) turns this dataset into ready-to-query SQLite databases — one `.db` file per unique schema, pre-populated with the synthetic row data.
+
+**Install:**
+```bash
+pip install SQaLe
+```
+
+**CLI — download and materialize schemas directly from HuggingFace:**
+```bash
+# All unique schemas
+sqale-extract --output ./dbs
+
+# First 100 unique schemas only
+sqale-extract --output ./dbs --limit 100
+```
+
+**Python API:**
+```python
+from sqale import deserialize_sqale
+
+results = deserialize_sqale(
+    file_path="trl-lab/SQaLe_2",
+    output_dir="./dbs",
+    limit=100,  # optional
+)
+
+for r in results:
+    print(r["db_path"], r["rows_per_table"])
+```
+
+Each entry in the returned list contains the path to the `.db` file, the table names, row counts per table, and any error encountered during materialization.
+
+## Citation
+
+```
+@inproceedings{
+  wolff2025sqale,
+  title={{SQ}aLe: A large text-to-{SQL} corpus grounded in real schemas},
+  author={Cornelius Wolff and Daniel Gomm and Madelon Hulsebos},
+  booktitle={EurIPS 2025 Workshop: AI for Tabular Data},
+  year={2025},
+  url={https://openreview.net/forum?id=6PsKDjgoEy}
+}
+```