List your company

Text-to-SQL on Spider no-easy

61.4AccuracySFT
31.7639.45547.1554.845May 26, 2026

Evaluation results

MethodLinks
SFTModel=Qwen2.5-7B-Instruct, Evaluation Protocol=FULL2026.0561.4
BaseModel=Qwen2.5-7B-Instruct, Evaluation Protocol=FULL2026.0561.2
RLSTAModel=Qwen2.5-7B-Instruct, Evaluation Protocol=FULL2026.0559.8
MAIGOModel=Qwen2.5-7B-Instruct, Evaluation Protocol=FULL2026.0559.6
GRPOModel=Qwen2.5-7B-Instruct, Evaluation Protocol=FULL2026.0559.4
RLSTAModel=Qwen2.5-3B-Instruct, Evaluation Protocol=FULL2026.0552
MAIGOModel=Qwen2.5-3B-Instruct, Evaluation Protocol=FULL2026.0551.9
GRPOModel=Qwen2.5-3B-Instruct, Evaluation Protocol=FULL2026.0551.7
SFTModel=Qwen2.5-3B-Instruct, Evaluation Protocol=FULL2026.0550.8
BaseModel=Qwen2.5-3B-Instruct, Evaluation Protocol=FULL2026.0550.7
MAIGOModel=Qwen2.5-7B-Instruct, Evaluation Protocol=SHARDED2026.0550
RLSTAModel=Qwen2.5-7B-Instruct, Evaluation Protocol=SHARDED2026.0545
MAIGOModel=Qwen2.5-3B-Instruct, Evaluation Protocol=SHARDED2026.0544.9
SFTModel=Qwen2.5-7B-Instruct, Evaluation Protocol=SHARDED2026.0544.1
GRPOModel=Qwen2.5-7B-Instruct, Evaluation Protocol=SHARDED2026.0542.5
BaseModel=Qwen2.5-7B-Instruct, Evaluation Protocol=SHARDED2026.0542.2
RLSTAModel=Qwen2.5-3B-Instruct, Evaluation Protocol=SHARDED2026.0537
SFTModel=Qwen2.5-3B-Instruct, Evaluation Protocol=SHARDED2026.0536.6
GRPOModel=Qwen2.5-3B-Instruct, Evaluation Protocol=SHARDED2026.0534
BaseModel=Qwen2.5-3B-Instruct, Evaluation Protocol=SHARDED2026.0532.9