Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multi-turn Text-to-SQL on SparC (dev)

68.5QM EM

gpt-oss-20b + Rose-SQL

30.103240.071650.0460.0084Mar 6, 2026Mar 16, 2026Mar 26, 2026Apr 5, 2026Apr 15, 2026Apr 25, 2026May 5, 2026
Updated 28d ago

Evaluation Results

MethodLinks
2026.05
68.580.2575.352.465.459.5
2026.03
67.773.3-49.154-
2026.05
67.773.3-49.154-
2026.05
66.4579.174.4450.2563.8258.04
2026.05
65.7578.7272.735062.5655.21
2026.03
65.4173.2367.8346.9154.7348.57
2026.05
65.4175.3969.1646.9157.8150.71
2026.03
65.1775.3969.1646.4457.8150.71
2026.03
64.7--45--
2026.05
64.7--45--
2026.03
64.3371.465.0843.3650.7143.36
2026.03
64.1770.8265.5843.652.1345.49
2026.05
61.5174.2368.7443.8355.9848.95
2026.03
61.3--44.1--
2026.05
61.3--44.1--
2026.03
61.2670.0763.7543.3652.1345.73
2026.05
60.173.566.833.250.836.5
2026.03
59.1867.9961.5136.9646.6839.33
2026.03
5670.363.336.550.541.9
2026.05
5670.363.336.550.541.9
2026.05
55.1971.9865.2529.3849.0540.52
2026.05
54.116658.2728.6740.2831.04
2026.05
53.1266.9961.6826.7841.9435.31
2026.03
5163.856.924.438.929.6
2026.05
5163.856.924.438.929.6
2026.05
42.561.2554.318.532.825.4
2026.05
31.5849.1343.8910.4321.0916.82