【paddle.fleet】geo send sparse optimize (#27719) (#27979)

* test=develop, fix geo sgd communicator and gloo http_init for ps

paddle/fluid/operators/distributed/communicator.cc

@@ -466,41 +466,34 @@ void GeoCommunicator::Send(const std::vector<std::string> &var_names,
                            const std::vector<std::string> &var_tables,
                            const framework::Scope &scope) {
   waiting_ = false;
+  PADDLE_ENFORCE_EQ(
+      var_tables.size(), 1,
+      platform::errors::InvalidArgument("var_tables.size() == 1 is permitted"));
+
+  auto table_name = var_tables[0];
+  if (table_name == STEP_COUNTER) return;
 
   auto before_send = GetCurrentUS();
-  std::unordered_map<std::string, std::unordered_set<int64_t>> ids_table;
+  size_t splited_var_nums =
+      send_varname_to_ctx_[table_name].splited_varnames.size();
 
-  for (size_t i = 0; i < var_tables.size(); i++) {
-    auto table_name = var_tables[i];
-    if (table_name == STEP_COUNTER) {
-      continue;
-    } else {
-      size_t splited_var_nums =
-          send_varname_to_ctx_[table_name].splited_varnames.size();
-
-      for (size_t j = 0; j < splited_var_nums; j++) {
-        if (ids_table.find(
-                send_varname_to_ctx_[table_name].splited_varnames[j]) ==
-            ids_table.end()) {
-          ids_table.insert(std::pair<std::string, std::unordered_set<int64_t>>(
-              send_varname_to_ctx_[table_name].splited_varnames[j],
-              std::unordered_set<int64_t>()));
-        }
-      }
+  std::unordered_map<std::string, std::unordered_set<int64_t>> ids_table;
 
-      auto *var = scope.FindVar(var_names[i]);
-      auto var_tensor = var->Get<framework::LoDTensor>();
-      int element_number = var_tensor.numel();
-      const int64_t *var_mutable_data = var_tensor.data<int64_t>();
+  for (size_t j = 0; j < splited_var_nums; j++) {
+    ids_table.insert(std::pair<std::string, std::unordered_set<int64_t>>(
+        send_varname_to_ctx_[table_name].splited_varnames[j],
+        std::unordered_set<int64_t>()));
+  }
+  auto *var = scope.FindVar(var_names[0]);
+  auto &rows = var->Get<framework::SelectedRows>().rows();
 
-      // insert ids which has not been record
-      for (int j = 0; j < element_number; j++) {
-        auto ep_idx = var_mutable_data[j] % splited_var_nums;
-        ids_table.at(send_varname_to_ctx_[table_name].splited_varnames[ep_idx])
-            .insert(var_mutable_data[j]);
-      }
-    }
+  // insert ids which has not been record
+  for (size_t j = 0; j < rows.size(); j++) {
+    auto ep_idx = rows[j] % splited_var_nums;
+    ids_table.at(send_varname_to_ctx_[table_name].splited_varnames[ep_idx])
+        .insert(rows[j]);
   }
+
   auto before_push = GetCurrentUS();
   for (auto &iter : ids_table) {
     auto &key = iter.first;
@@ -512,8 +505,8 @@ void GeoCommunicator::Send(const std::vector<std::string> &var_names,
             << "'s queue";
   }
   auto after_send = GetCurrentUS();
-  VLOG(3) << "run send_op finish. using " << (before_push - before_send) << "; "
-          << (after_send - before_push);
+  VLOG(3) << "run send " << table_name << " op finish. using "
+          << (before_push - before_send) << "; " << (after_send - before_push);
 }
 
 void GeoCommunicator::MainThread() {

python/paddle/distributed/fleet/base/role_maker.py

@@ -826,7 +826,7 @@ class PaddleCloudRoleMaker(RoleMakerBase):
                     start_http_server = True
             else:
                 ep_rank_0 = os.getenv("PADDLE_GLOO_HTTP_ENDPOINT", "")
-                if self._server_index() == 0:
+                if self._is_server() and self._server_index() == 0:
                     start_http_server = True
             ip, port = ep_rank_0.split(':')
             kwargs = {

python/paddle/distributed/fleet/launch_utils.py

@@ -895,7 +895,7 @@ class ParameterServerLauncher(object):
                 "PADDLE_TRAINERS_NUM": str(self.worker_num),
                 "POD_IP": cur_server.endpoint.split(":")[0],
                 "PADDLE_WITH_GLOO": "1",
-                "PADDLE_GLOO_RENDEZVOUS": "2",
+                "PADDLE_GLOO_RENDEZVOUS": "3",
                 "PADDLE_GLOO_FS_PATH": self.gloo_rendezvous_dir,
                 "PADDLE_GLOO_HTTP_ENDPOINT": self.http_port
             }
@@ -959,7 +959,7 @@ class ParameterServerLauncher(object):
                 "TRAINING_ROLE": "TRAINER",
                 "PADDLE_TRAINER_ID": str(cur_worker.rank),
                 "PADDLE_WITH_GLOO": "1",
-                "PADDLE_GLOO_RENDEZVOUS": "2",
+                "PADDLE_GLOO_RENDEZVOUS": "3",
                 "PADDLE_GLOO_FS_PATH": self.gloo_rendezvous_dir,
                 "FLAGS_selected_gpus": "0",
                 "FLAGS_selected_xpus": "0",
@@ -1028,7 +1028,7 @@ class ParameterServerLauncher(object):
                 "PADDLE_TRAINERS_NUM": str(self.worker_num),
                 "POD_IP": cur_heter_worker.endpoint.split(":")[0],
                 "PADDLE_WITH_GLOO": "1",
-                "PADDLE_GLOO_RENDEZVOUS": "2",
+                "PADDLE_GLOO_RENDEZVOUS": "3",
                 "PADDLE_GLOO_FS_PATH": self.gloo_rendezvous_dir,
                 "FLAGS_selected_gpus": "0",
                 "FLAGS_selected_xpus": "0",

python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py

@@ -169,7 +169,7 @@ def append_send_ops_pass(program, config):
     trainer_id = config.get_role_id()
     pserver_endpoints = config.get_ps_endpoints()
 
-    def _append_grad_send_op(union_vars, queue):
+    def _append_send_op(union_vars, queue):
 
         if queue == STEP_COUNTER:
             send_input_vars = []
@@ -198,43 +198,6 @@ def append_send_ops_pass(program, config):
 
         return dummy_output
 
-    def _append_sparse_ids_send_op():
-        sparse_var = []
-        sparse_tables = []
-        unique_sparse_var = {}
-        for op in program.global_block().ops:
-            if "is_sparse" in op.all_attrs():
-                if op.type == "lookup_table":
-                    op._set_attr('remote_prefetch', False)
-                for input_var_name, sparse_var_name in zip(
-                        op.input("Ids"), op.input("W")):
-                    if input_var_name in unique_sparse_var:
-                        if unique_sparse_var[input_var_name] == sparse_var_name:
-                            continue
-                    input_var = program.global_block().var(input_var_name)
-                    sparse_var.append(input_var)
-                    sparse_tables.append(sparse_var_name)
-                    unique_sparse_var[input_var_name] = sparse_var_name
-
-        dummy_output = []
-        if mode in [DistributedMode.SYNC, DistributedMode.HALF_ASYNC]:
-            dummy_output = program.global_block().create_var(
-                name=framework.generate_control_dev_var_name())
-
-        program.global_block().append_op(
-            type="send",
-            inputs={"X": sparse_var},
-            outputs={"Out": dummy_output},
-            attrs={
-                "send_varnames": sparse_tables,
-                "merge_add": True,
-                "use_send_handler": False,
-                "endpoints": pserver_endpoints,
-                RPC_OP_ROLE_ATTR_NAME: RPC_OP_ROLE_ATTR_VALUE
-            })
-
-        return dummy_output
-
     def _append_barrier_op(dummys):
         program.global_block().append_op(
             type="send_barrier",
@@ -251,12 +214,8 @@ def append_send_ops_pass(program, config):
 
     sends = config.get_trainer_send_context()
 
-    if mode == DistributedMode.GEO:
-        dummys.append(_append_sparse_ids_send_op())
-    else:
-        for merged_name, send in sends.items():
-            dummys.append(
-                _append_grad_send_op(send.origin_varnames(), merged_name))
+    for merged_name, send in sends.items():
+        dummys.append(_append_send_op(send.origin_varnames(), merged_name))
 
     if mode in [DistributedMode.SYNC, DistributedMode.HALF_ASYNC]:
         _append_barrier_op(dummys)