experiment=ch30_ddp_reducer_overlap run_id=20260711T190000Z timestamp_utc=2026-07-11T17:17:20.730709+00:00 hostname=sg-fuxc2-260708-12652-default0-0 iteration_rows=320 bucket_ready_rows=1960 source_model_rows=30 correctness=PASS bucket_layout=8/8 PASS unused_failure=1/1 PASS nsys_overlap=8/8 PASS private_nsys_reports=NOT_FOR_PUBLICATION pytorch_runtime=2.5.0a0+872d972e41.nv24.08 nccl_runtime=2.22.3