From 5bbfca15d6b9d5945c766c76709d56cdf36e6085 Mon Sep 17 00:00:00 2001
From: ziyoujiyi <73728031+ziyoujiyi@users.noreply.github.com>
Date: Tue, 20 Sep 2022 15:17:39 +0800
Subject: [PATCH] Fl ps (#46258)

* back fl

* delete ssl cert

* .

* make warning

* .

* unittest paral degree

* solve unittest

* heter & multi cloud commm ready

* .

* .

* fl-ps v1.0

* .

* support N + N mode

* .

* .

* .

* .

* delete print

* .

* .

* .

* .

* fix bug

* .

* .

* fl-ps with coordinator ready

* merge dev

* update message parse only

* update fl client scheduler

* fix bug

* update multithreads sync

* fix ci errors

* update role_maker.py

* update role_maker.py

* fix ci error: windows py import error

* fix ci error: windows py import error

* fix windows ci pylib import error

* add dump fields & params

* try to fix windows import fleet error

* fix ps FLAGS error

* fix logging risk

* fix logging possible risk

* write trainer_desc file

* support split sparse params in local & remote

* fix import paddle.fluid.core.PSGPU

* fix import paddle.fluid.core.PSGPU

* add remote_sparse & local_sparse config

* fix unittest

* fix test_dist_fleet_geo table error

* fix PADDLE_ENFORCE error

* fix other's pr conflict

* forbidden ssd table

* .

* recover ssd table code

* recover file mode

* debug auc 0.5

* adapt for nn fl-ps

* adapt for nn fl-ps

* add learning_rate_0 intializer op

* recover ssd table

* modify file mode

* flps del fake-init op
---
 python/paddle/distributed/passes/ps_trainer_pass.py      | 2 ++
 python/paddle/distributed/ps/utils/ps_program_builder.py | 4 ++--
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/python/paddle/distributed/passes/ps_trainer_pass.py b/python/paddle/distributed/passes/ps_trainer_pass.py
index 982d3f61e6e..4ca11a02c3e 100755
--- a/python/paddle/distributed/passes/ps_trainer_pass.py
+++ b/python/paddle/distributed/passes/ps_trainer_pass.py
@@ -617,6 +617,8 @@ class DeleteExtraOptimizerPass(PassBase):
         for var in remote_optimize_vars:
             if var in local_optimize_vars:
                 continue
+            if 'learning_rate_0' == var:
+                continue
             if var not in remote_optimize_op_role_vars:
                 optimize_need_delete_vars.append(var)
         need_delete_optimize_vars = list(set(optimize_need_delete_vars))
diff --git a/python/paddle/distributed/ps/utils/ps_program_builder.py b/python/paddle/distributed/ps/utils/ps_program_builder.py
index 53771b05cbf..0bd870ffee5 100755
--- a/python/paddle/distributed/ps/utils/ps_program_builder.py
+++ b/python/paddle/distributed/ps/utils/ps_program_builder.py
@@ -373,8 +373,8 @@ class FlPsProgramBuilder(HeterAsyncPsProgramBuilder):
         _main_file = ps_log_root_dir + '4_fl_worker_main_program.prototxt'
         #debug_program(_main_file, self.cloned_main)
 
-        fake_init_ops_pass = new_pass("fake_init_ops_pass", self.attrs)
-        fake_init_ops_pass.apply([None], [self.cloned_startup], self.pass_ctx)
+        #fake_init_ops_pass = new_pass("fake_init_ops_pass", self.attrs)
+        #fake_init_ops_pass.apply([None], [self.cloned_startup], self.pass_ctx)
 
         _main_file = ps_log_root_dir + '5_fl_worker_main_program.prototxt'
         #debug_program(_main_file, self.cloned_main)
-- 
GitLab