added global rank var name

2026-09-10 12:21:57 +08:00 · 2019-07-26 18:52:02 -04:00
parent 56d41eaa8c
commit 12f717ad4a
1 changed files with 3 additions and 0 deletions
@@ -500,6 +500,9 @@ class Trainer(TrainerIO):
        self.proc_rank = self.node_rank * len(self.data_parallel_device_ids) + gpu_nb
        self.world_size = self.nb_gpu_nodes * len(self.data_parallel_device_ids)

+        # let the exp know the rank to avoid overwriting logs
+        self.experiment.rank = self.global_rank
+
        # set up server using proc 0's ip address
        # try to init for 20 times at max in case ports are taken
        # where to store ip_table