sre-arena / training /reward_function.py

Commit History

Phase 7a: PEFT checkpoint loading for cross-generation training
a7e7791

blitz1809 Claude Sonnet 4.6 commited on

Phase 6.5: warm-up step + fp16 GRPO + log_tail propagation fix
d20f9fd

blitz1809 commited on

feat(6c): add make_attacker_reward_function to reward_function
0967481

blitz1809 commited on

Phase 6b: defender training script with scripted attacker
f4bf7d6

blitz1809 Claude Sonnet 4.6 commited on