Unverified50% confidenceFactExact time
The author states that the best-performing public Agent he knows of on the Spider 2.0-lite SQLite subset reports 57.8% Pass@1, and that it uses a different model.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/24/2026
First Seen
Valid until: 11/7/2026
Sources
Related Entities
Related Claims
UnverifiedSpider 2.0 is a Text-to-SQL benchmark targeting enterprise-grade real data warehouse scenarios, covering multiple dialects such as BigQuery, Snowflake, and SQLite, and is considered significantly harder than the original Spider.60% similarUnverifiedPer-entity independent SQLite databases introduce complexity in file handle and disk space management when the number of Agents is very large.58% similarUnverifiedThe commonly used benchmarks for measuring NL2SQL system accuracy are the Spider and BIRD datasets, where Spider tests cross-domain generalization ability and BIRD more closely reflects the complexity of real enterprise databases.54% similarUnverifiedSGLang出自伯克利，引入RadixAttention实现更高效的前缀共享和缓存管理，对Agent类工作负载优化更到位52% similarUnverifiedWith WAL mode enabled, a single-file SQLite database is enough to support a fetch gateway handling hundreds of requests per minute, saving the operational cost of deploying Redis or Postgres.51% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1092535API
curl https://kongchang.com/api/v1/knowledge/claims/1092535MCP
get_claim(id=1092535)