List your company

Semantic Parsing on CFQ (MCD metrics)

92.1MCD1 ScoreGRPO-Composite
88.376889.343490.3191.2766May 6, 2026

Evaluation results

MethodLinks
GRPO-CompositeBase Model=Llama-3.1-8B-Instruct2026.0592.176.1683.4483.9
GRPO-BinaryBase Model=Llama-3.1-8B-Instruct2026.0591.3576.4481.4683.08
SFTBase Model=Llama-3.1-8B-Instruct2026.0591.0474.8577.4281.1
GRPO-BinaryBase Model=Qwen-2.5-7B-Instruct2026.0589.2375.7376.5480.5
GRPO-CompositeBase Model=Qwen-2.5-7B-Instruct2026.0589.1677.3375.8680.78
SFTBase Model=Qwen-2.5-7B-Instruct2026.0588.5273.9873.6778.72