RomanRG008/gyra
0.4B • Updated • 2
Gyra: decisions for coding agents
421M model that gates coding-agent actions: destructive commands, success checks, tool choice, prompt injection. ~17 ms.
Note Better than Laya on all 10 coding-agent tests. Code and eval: github.com/Gowtham-R-2002/gyra
convaiinnovations/laya
Text Classification • 0.4B • Updated • 5.07kNote The base model Gyra is fine-tuned from (Convai Innovations).
Gyra vs Laya
🛡A fast second opinion for coding agents, next to Laya
Note Pick an example and compare Gyra and Laya side by side.