metadata
license: apache-2.0
base_model: Qwen/Qwen2.5-Coder-7B-Instruct
tags:
- text-to-sql
- self-play
- grpo
- sql-zero
sql-zero-7b-solver-iter1-s240
Solver 7B iter1 (challenger-first, zero anotacao), final s240. Spider 0.748, BIRD 0.473.
Parte do projeto SQL-Zero (self-play zero-anotacao para Text-to-SQL, adaptacao do Dr. Zero). Protocolo de avaliacao: greedy@1, schema completo (DDL+3 rows), execution accuracy. Treinado com verl (GRPO); reward de execucao 3-tier (1.0/0.1/0.0).