fix delete ops

paddle/framework/block_desc.cc

@@ -91,6 +91,21 @@ OpDescBind *BlockDescBind::PrependOp() {
   return ops_.front().get();
 }
 
+void BlockDescBind::RemoveOp(size_t s, size_t e) {
+  if (ops_.begin() + s == ops_.end() || ops_.begin() + e == ops_.end()) {
+    return;
+  }
+  need_update_ = true;
+  for (auto it = ops_.begin() + s; it != ops_.begin() + e; it++) {
+    auto names = (*it)->InputArgumentNames();
+    for (auto n : names) {
+      // TODO(typhoonzero): delete vars if no other op use it.
+      VLOG(3) << "deleting var " << n;
+    }
+  }
+  ops_.erase(ops_.begin() + s, ops_.begin() + e);
+}
+
 std::vector<OpDescBind *> BlockDescBind::AllOps() const {
   std::vector<OpDescBind *> res;
   for (const auto &op : ops_) {

paddle/framework/block_desc.h

@@ -80,6 +80,8 @@ class BlockDescBind {
 
   OpDescBind *PrependOp();
 
+  void RemoveOp(size_t s, size_t e);
+
   std::vector<OpDescBind *> AllOps() const;
 
   size_t OpSize() const { return ops_.size(); }

paddle/pybind/protobuf.cc

@@ -159,6 +159,7 @@ void BindBlockDesc(py::module &m) {
            py::return_value_policy::reference)
       .def("prepend_op", &BlockDescBind::PrependOp,
            py::return_value_policy::reference)
+      .def("remove_op", &BlockDescBind::RemoveOp)
       .def("var",
            [](BlockDescBind &self, py::bytes byte_name) {
              std::string name = byte_name;

python/paddle/v2/fluid/distribute_transpiler.py

@@ -131,11 +131,6 @@ class DistributeTranspiler:
 
     def _optimize_distributed(self, optimize_ops, program, params_and_grads,
                               **kwargs):
-        # remove optimize ops and add a send op to main_program
-        # FIXME(typhoonzero): delete_op only remove the first accurance,
-        # need to consider about multiple same optimize op?
-        for op in optimize_ops:
-            program.global_block().delete_op(op)
         if kwargs.has_key("split_method"):
             split_method = kwargs["split_method"]
         else:
@@ -159,6 +154,10 @@ class DistributeTranspiler:
             attrs={"endpoints": pserver_endpoints,
                    "epmap": epmap})
 
+    def get_trainer_program(optimize_ops, program):
+        # remove optimize ops and add a send op to main_program
+        program.global_block().delete_ops(optimize_ops)
+
     def _create_var_for_trainers(self, block, var, trainers):
         var_list = []
         for i in xrange(trainers):
@@ -209,6 +208,7 @@ class DistributeTranspiler:
 
             if opt_op.inputs.has_key("Grad"):
                 if opt_op.inputs["Grad"].name in grad_var_names:
+                    print "appending ", opt_op.type, opt_op.inputs
                     optimize_sub_program.global_block().append_op(
                         type=opt_op.type,
                         inputs=opt_op.inputs,

python/paddle/v2/fluid/framework.py

@@ -579,6 +579,7 @@ class Block(object):
         self.vars = dict()  # var_name --> var
         self.ops = collections.deque()  # operator list
         self.program = program
+        self.removed_vars = dict()
 
     def __str__(self):
         return self.to_string(True)
@@ -635,8 +636,15 @@ class Block(object):
         self.ops.append(op)
         return op
 
-    def delete_op(self, op):
-        self.ops.remove(op)
+    def delete_ops(self, ops):
+        # remove from cpp
+        # FIXME(typhoonzero): remove only the first occuracy.
+        try:
+            start = list(self.ops).index(ops[0])
+            end = list(self.ops).index(ops[-1])
+        except Exception, e:
+            raise e
+        self.desc.remove_op(start, end)
 
     def prepend_op(self, *args, **kwargs):
         op_desc = self.desc.prepend_op()