494 B
494 B
license, base_model, tags
| license | base_model | tags | ||||
|---|---|---|---|---|---|---|
| apache-2.0 | Qwen/Qwen2.5-Coder-3B-Instruct |
|
sql-zero-3b-gold-control
Controle: mesmo GRPO nos 8.659 pares humanos do Spider train. Spider 0.731, BIRD 0.357, Syn 0.622.
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).