From c583447e900af61ac69fa6b0a41e9ab10cf3e90a Mon Sep 17 00:00:00 2001 From: Helin Wang Date: Wed, 10 May 2017 18:36:46 -0700 Subject: [PATCH] add subtitle --- doc/design/cluster_train/pserver_client.md | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/doc/design/cluster_train/pserver_client.md b/doc/design/cluster_train/pserver_client.md index 56469fc21..ee40eb32c 100644 --- a/doc/design/cluster_train/pserver_client.md +++ b/doc/design/cluster_train/pserver_client.md @@ -6,10 +6,14 @@ For an overview of trainer's role, please refer to [distributed training design The parameters on parameter servers need to be initialized. To provide maximum flexibility, we need to allow trainer initialized the parameters. Only one trainer will do the initialization, the other trainers will wait for the completion of initialization and get the parameters from the parameter servers. +### Trainer Selection + To select the trainer for initialization, every trainer will try to get a distributed lock, whoever owns the lock will do the initialization. As illustrated below: +### Selection Process + The select process is encapsulated in the C API function: ```c int paddle_begin_init_params(paddle_pserver_client* client, const char* config_proto); -- GitLab