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

Text-to-SQL on Spider-Syn

83.1Execution Accuracy (EX)

FineStep-4B

Updated 9d ago

Evaluation Results

Method
FineStep-4B 2026.05	83.1	-	-	-	-	75.2
FineStep 2026.05	83.1	-	-	-	-	75.2
FineStep-8B 2026.05	81.4	-	-	-	-	71.6
FineStep 2026.05	81.4	-	-	-	-	71.6
MTIR-SQL + Qwen3-14B 2026.05	81	-	-	-	-	-
FineStep 2026.05	80.8	-	-	-	-	72.6
GIGPO 2026.05	80.4	-	-	-	-	71.7
GIGPO 2026.05	80.2	-	-	-	-	71.7
GSPO 2026.05	80.1	-	-	-	-	72.1
DAPO 2026.05	80	-	-	-	-	72.5
GSPO 2026.05	79.6	-	-	-	-	70.3
GRPO 2026.05	79.6	-	-	-	-	70.4
DAPO 2026.05	79.5	-	-	-	-	71.1
GRPO 2026.05	79.2	-	-	-	-	70
GRPO 2026.05	78.9	-	-	-	-	71.2
DCG-SQL 2025.05	78.7	68.6	-	-	-	-
ReEx-SQL-7B 2026.05	78.7	-	-	-	-	72
Think2SQL-14B 2025.04	78.6	-	-	-	-	-
SENSE-13B 2026.05	77.6	-	-	-	-	70.2
SQL-o1 + Llama3-8B 2026.05	77.6	-	-	-	-	69.2
ROUTE + Llama3-8B 2026.05	77.4	-	-	-	-	70.2
SFT CodeS-15B 2026.05	77	-	-	-	-	69.4
SQL-R1-7B* 2026.05	76.7	-	-	-	-	-
No-Train 2026.05	76.1	-	-	-	-	65
CYCLESQL + GPT-4 2026.05	76	-	-	-	-	66.3
No-Train 2026.05	75.4	-	-	-	-	62
No-Train 2026.05	75.4	-	-	-	-	61.2
Think2SQL-4B 2025.04	75	-	-	-	-	-
Think2SQL-8B 2025.04	74.9	-	-	-	-	-
KE-RL 2026.05	74.83	-	-	-	-	-
SQL-PaLM + PaLM2 2026.05	74.6	-	-	-	-	-
ASTRES 2025.05	74.4	61.3	-	-	-	-
FastRAText+GPT-4 2026.05	74.4	-	-	-	-	-
DCG-SQL 2025.05	74.2	61.6	-	-	-	-
DCG-SQL 2025.05	73.6	62.6	-	-	-	-
ACT-SQL 2025.05	73.1	48.5	-	-	-	-
OmniSQL-32B 2026.05	72.1	-	-	-	-	-
KE-ICL 2026.05	71.76	-	-	-	-	-
CodeS 2026.05	71.17	-	-	-	-	-
DCG-SQL 2025.05	70.5	59.9	-	-	-	-
ACT-SQL 2025.05	70.4	44.4	-	-	-	-
DCG-SQL 2025.05	70.1	59.7	-	-	-	-
OmniSQL-7B 2026.01	69.7	-	-	-	61.4	-
DAIL-SQL 2026.05	69.5	-	-	-	-	-
Reasoning-SQL* 2026.05	69.3	-	-	-	-	-
EVOLSQL-7B 2026.01	69.2	-	-	-	61.9	-
Omni 2026.05	69.17	-	-	-	-	-
ACT-SQL 2025.05	69.1	49.3	-	-	-	-
ASTRES 2025.05	68.7	49.5	-	-	-	-
SQL-GEN 2026.05	67.8	-	-	-	-	-
KE-ICL 2026.05	66.98	-	-	-	-	-
ASTRES 2025.05	66.9	51.2	-	-	-	-
ACT-SQL 2025.05	66.6	53.7	-	-	-	-
KE-ICL 2026.05	66.05	-	-	-	-	-
KE-RL 2026.05	66.04	-	-	-	-	-
SFT (BIRD+Spider) 2026.01	65.9	-	-	-	55.8	-
CodeS 2026.05	65.86	-	-	-	-	-
DAIL-SQL 2026.05	65.43	-	-	-	-	-
DIN-SQL 2025.05	65.3	34.7	-	-	-	-
CodeS 2026.05	64.77	-	-	-	-	-
KE-ICL 2026.05	64	-	-	-	-	-
DAIL-SQL 2026.05	63.44	-	-	-	-	-
ACT-SQL 2025.05	63.2	44.6	-	-	-	-
Base Model 2026.01	63.1	-	-	-	53.4	-
Omni 2026.05	62.27	-	-	-	-	-
DCG-SQL 2025.05	60.9	51.9	-	-	-	-
SQL-GEN 2026.05	59.65	-	-	-	-	-
KE-RL 2026.05	58.09	-	-	-	-	-
ACT-SQL 2025.05	57.4	44.4	-	-	-	-
DIN-SQL 2025.05	57.2	43.5	-	-	-	-
SQL-GEN 2026.05	55.44	-	-	-	-	-
CodeS 2026.05	55.18	-	-	-	-	-
DIN-SQL 2025.05	53.7	25.4	-	-	-	-
DAIL-SQL 2026.05	53.13	-	-	-	-	-
Omni 2026.05	52.21	-	-	-	-	-
KE-ICL 2026.05	47.78	-	-	-	-	-
CodeS 2026.05	44.63	-	-	-	-	-
DAIL-SQL 2026.05	43.01	-	-	-	-	-
DIN-SQL 2025.05	39	27.2	-	-	-	-
Multi-turn Agent 2026.01	-	-	63.1	66.9	-	-
SQL-R1 2026.01	-	-	-	76.7	-	-
Reasoning-SQL 2026.01	-	-	69.3	-	-	-
Omini-SQL 2026.01	-	-	69.7	69.6	-	-
SQL-TRAIL 2026.01	-	-	72.8	77	-	-
GPT-4o-mini 2025.12	-	-	60.5	61.6	-	-
GPT-4-Turbo 2025.12	-	-	62.9	63.5	-	-
GPT-4o 2025.12	-	-	59.6	62.3	-	-
DeepSeek-Coder-7B-Instruct 2025.12	-	-	49.9	51.7	-	-
Qwen2.5-Coder-7B-Instruct 2025.12	-	-	63.1	66.9	-	-
Qwen2.5-7B-Instruct 2025.12	-	-	54.2	60	-	-
OpenCoder-8B-Instruct 2025.12	-	-	46	46.1	-	-
Meta-Llama-3.1-8B-Instruct 2025.12	-	-	53.1	59.3	-	-
Granite-8B-Code-Instruct 2025.12	-	-	45	46.8	-	-
Granite-3.1-8B-Instruct 2025.12	-	-	47.7	53.8	-	-
SynSQL(50K) 2025.12	-	-	59.6	63.5	-	-
SynSQL(90K) 2025.12	-	-	58.9	63.6	-	-
SynSQL(2.5M) 2025.12	-	-	61	63.1	-	-
Spider+BIRD+DataFlow-Text2SQL-90K 2025.12	-	-	66.6	69.1	-	-
DataFlow-Text2SQL-50K 2025.12	-	-	61.3	67.5	-	-
DataFlow-Text2SQL-90K 2025.12	-	-	61.6	67.3	-	-

Showing 100 of 146 rows