perf(training): training.log_interval 配置项 + rsl_rl PPO TensorBoard 批量写 patch - #1648

Merged

TATP-233 merged 2 commits into

mainfrom

perf/tb-batched-scalar-logging

Sep 26, 2026

Commits on Sep 26, 2026

chore(deps): bump torch to 2.14 and unify CUDA wheels on cu130
TATP-233
committed
da1514e
perf(training): add training.log_interval and batched TensorBoard logging for rsl_rl PPO
TATP-233
committed
4bed1f1