How We Built Zeta2: Training an Edit Prediction Model in Production
AI Engineer2323 字 (约 10 分钟)
87
Zed trained Zeta2 via production edit data distillation: using frontier models to generate candidate edits, filtering low-quality outputs with static evaluation and a 'repair' mechanism, yielding ~100K high-quality training examples; the entire pipeline is JSONL-based for fast experimentation.
入选理由：Zeta2 使用 distillation + repair 两阶段流程：先由 frontier model 生成编辑预测，再用启发式规则检测失败案例并触发二次修正
FeaturedVideo#Edit Prediction#Model Distillation#Zed#Production AI#Code Generation英文