[Auto Parallel] Improve the codes of the completion and distributed context (#40671)

* [Auto Parallel] Replace the old planner by the new partition tuner

* [Auto Parallel] Improve the completion and distributed context

* [Auto Parallel] Fix some bugs of the compatible check of some dist ops

* [Auto Parallel] Fix some bugs

python/paddle/distributed/auto_parallel/completion.py

@@ -123,6 +123,19 @@ def merge_process_mesh_two(pm1, pm2):
     return merged_process_mesh
 
 
+def _validate_dims_mapping(dims_mapping, process_mesh):
+    if dims_mapping is None:
+        return False
+    for i in range(len(dims_mapping)):
+        if dims_mapping[i] < -1 or dims_mapping[i] >= len(
+                process_mesh.topology):
+            return False
+    for i in range(len(process_mesh.topology)):
+        if dims_mapping.count(i) > 1:
+            return False
+    return True
+
+
 class Completer:
     def __init__(self, dist_context):
         assert dist_context is not None
@@ -161,6 +174,9 @@ class Completer:
             dims_mapping_list.append(tensor_dims_mapping)
             compatible_dims_mapping = compute_compatible_dims_mapping(
                 dims_mapping_list)
+            if not _validate_dims_mapping(compatible_dims_mapping,
+                                          tensor_dist_attr.process_mesh):
+                return False
             if (compatible_dims_mapping is not None) and \
                 (compatible_dims_mapping != tensor_dims_mapping):
                 tensor_dist_attr.dims_mapping = compatible_dims_mapping
@@ -182,6 +198,9 @@ class Completer:
             dims_mapping_list.append(tensor_dims_mapping)
             compatible_dims_mapping = compute_compatible_dims_mapping(
                 dims_mapping_list)
+            if not _validate_dims_mapping(compatible_dims_mapping,
+                                          tensor_dist_attr.process_mesh):
+                return False
             if (compatible_dims_mapping is not None) and \
                 (compatible_dims_mapping != tensor_dims_mapping):
                 tensor_dist_attr.dims_mapping = compatible_dims_mapping
@@ -196,10 +215,12 @@ class Completer:
         op_desc = op_node.op()
         if op_desc.type() == "create_py_reader" \
             or op_desc.type() == "create_double_buffer_reader" \
+            or op_desc.type() == "while" \
             or op_desc.type() == "read":
             return False
         dist_op = self._dist_context.get_dist_op_for_graph(op_node)
         op_dist_attr = dist_op.dist_attr
+        original_op_dist_attr = copy.deepcopy(op_dist_attr)
         if fwd:
             for tensor_node in op_node.inputs:
                 if tensor_node.is_var() and tensor_node.var() is not None:
@@ -223,18 +244,34 @@ class Completer:
                                 tensor_desc.name(), compatible_dims_mapping)
                             changed = True
             # Find the most compatible implemenetations from the distributed operator
-            op_dist_impl = find_best_compatible_distributed_operator_impl(
+            op_dist_impls = find_best_compatible_distributed_operator_impl(
                 dist_op, fwd=True)
-            if op_dist_impl is not None:
-                dim_changed = op_dist_impl.update_dims_mapping(dist_op)
-                if dim_changed:
-                    changed = True
-                if op_dist_impl.is_auto_compatible(dist_op):
-                    if op_dist_impl.type == "elementwise":
-                        op_dist_attr.impl_type = "default"
+            if op_dist_impls is not None:
+                not_compatible = True
+                backup_op_dist_attr = copy.deepcopy(op_dist_attr)
+                backup_changed = changed
+                for op_dist_impl in op_dist_impls:
+                    dim_changed = op_dist_impl.update_dims_mapping(dist_op)
+                    if dim_changed:
+                        changed = True
+                    if op_dist_impl.is_auto_compatible(dist_op):
+                        if op_dist_impl.type == "elementwise":
+                            op_dist_attr.impl_type = "default"
+                        else:
+                            op_dist_attr.impl_type = op_dist_impl.type
+                        # op_dist_attr.impl_type = op_dist_impl.type
+                        op_dist_attr.impl_idx = op_dist_impl.idx
+                        not_compatible = False
+                        break
                     else:
-                        op_dist_attr.impl_type = op_dist_impl.type
-                    op_dist_attr.impl_idx = op_dist_impl.idx
+                        dist_op.dist_attr = backup_op_dist_attr
+                        changed = backup_changed
+                if not_compatible:
+                    dist_op.dist_attr = original_op_dist_attr
+                    changed = False
+            else:
+                dist_op.dist_attr = original_op_dist_attr
+                changed = False
         else:
             for tensor_node in op_node.outputs:
                 if tensor_node.is_var() and tensor_node.var() is not None:
@@ -258,18 +295,35 @@ class Completer:
                                 tensor_desc.name(), compatible_dims_mapping)
                             changed = True
             # Find the most compatible implemenetations from the distributed operator
-            op_dist_impl = find_best_compatible_distributed_operator_impl(
+            op_dist_impls = find_best_compatible_distributed_operator_impl(
                 dist_op, fwd=False)
-            if op_dist_impl is not None:
-                dim_changed = op_dist_impl.update_dims_mapping(dist_op)
-                if dim_changed:
-                    changed = True
-                if op_dist_impl.is_auto_compatible(dist_op):
-                    if op_dist_impl.type == "elementwise":
-                        op_dist_attr.impl_type = "default"
+            if op_dist_impls is not None:
+                not_compatible = True
+                backup_op_dist_attr = copy.deepcopy(op_dist_attr)
+                backup_changed = changed
+                for op_dist_impl in op_dist_impls:
+                    dim_changed = op_dist_impl.update_dims_mapping(dist_op)
+                    if dim_changed:
+                        changed = True
+                    if op_dist_impl.is_auto_compatible(dist_op):
+                        not_compatible = False
+                        if op_dist_impl.type == "elementwise":
+                            op_dist_attr.impl_type = "default"
+                        else:
+                            op_dist_attr.impl_type = op_dist_impl.type
+                        # op_dist_attr.impl_type = op_dist_impl.type
+                        op_dist_attr.impl_idx = op_dist_impl.idx
+                        not_compatible = False
+                        break
                     else:
-                        op_dist_attr.impl_type = op_dist_impl.type
-                    op_dist_attr.impl_idx = op_dist_impl.idx
+                        dist_op.dist_attr = backup_op_dist_attr
+                        changed = backup_changed
+                if not_compatible:
+                    dist_op.dist_attr = original_op_dist_attr
+                    changed = False
+            else:
+                dist_op.dist_attr = original_op_dist_attr
+                changed = False
         return changed
 
     def _update_dims_mapping_between_graphs(self):
@@ -279,17 +333,22 @@ class Completer:
                 parent_node)
             child_node_dist_attr = self._dist_context.get_dist_attr_for_graph(
                 child_node)
+            if parent_node_dist_attr.process_mesh != child_node_dist_attr.process_mesh:
+                continue
             parent_node_dims_mapping = parent_node_dist_attr.dims_mapping
             child_node_dims_mapping = child_node_dist_attr.dims_mapping
             compatible_dims_mapping = compute_compatible_dims_mapping(
                 [parent_node_dims_mapping, child_node_dims_mapping])
+            if not _validate_dims_mapping(compatible_dims_mapping,
+                                          parent_node_dist_attr.process_mesh):
+                return False
             if (compatible_dims_mapping is not None) \
                 and (compatible_dims_mapping != parent_node_dims_mapping):
                 parent_node_dist_attr.dims_mapping = compatible_dims_mapping
                 changed = True
             if (compatible_dims_mapping is not None) \
                 and (compatible_dims_mapping != child_node_dims_mapping):
-                parent_node_dist_attr.dims_mapping = compatible_dims_mapping
+                child_node_dist_attr.dims_mapping = compatible_dims_mapping
                 changed = True
         return changed
 
@@ -351,7 +410,7 @@ class Completer:
                 if compatible_process_mesh is not None \
                     and tensor_dist_attr.process_mesh != compatible_process_mesh:
                     tensor_dist_attr.process_mesh = compatible_process_mesh
-        # Set the process mesh of the op node's outputs
+                # Set the process mesh of the op node's outputs
         for tensor_node in op_node.outputs:
             if tensor_node.is_var() and tensor_node.var() is not None:
                 tensor_dist_attr = self._dist_context.get_tensor_dist_attr_for_graph(
@@ -389,7 +448,8 @@ class Completer:
                 if _node_id(cur) in visited:
                     continue
                 # TODO: need more restrictions
-                for node in cur.inputs:
+                neighbors = cur.inputs + cur.outputs
+                for node in neighbors:
                     if node.is_var() and node.var() is not None:
                         if node.var().type() != core.VarDesc.VarType.READER \
                             and len(node.var().shape()) == 1:
@@ -421,10 +481,29 @@ class Completer:
                 visited.add(_node_id(cur))
             return related_nodes
 
+        def _make_dims_mapping_replicate(dist_attr):
+            if isinstance(dist_attr, TensorDistributedAttribute):
+                for i, _ in enumerate(dist_attr.dims_mapping):
+                    dist_attr.dims_mapping[i] = -1
+            if isinstance(dist_attr, OperatorDistributedAttribute):
+                for arg_name in dist_attr.inputs_dist_attrs.keys():
+                    new_dims_mapping = []
+                    dims_mapping = dist_attr.get_input_dims_mapping(arg_name)
+                    for _ in dims_mapping:
+                        new_dims_mapping.append(-1)
+                    dist_attr.set_input_dims_mapping(arg_name, new_dims_mapping)
+                for arg_name in dist_attr.outputs_dist_attrs.keys():
+                    new_dims_mapping = []
+                    dims_mapping = dist_attr.get_output_dims_mapping(arg_name)
+                    for _ in dims_mapping:
+                        new_dims_mapping.append(-1)
+                    dist_attr.set_output_dims_mapping(arg_name,
+                                                      new_dims_mapping)
+
         # Amend the process meshes related to while_op
         for while_op_node, while_op_node_idx in self._while_op_nodes.values():
             sub_graph_id = while_op_node.op()._block_attr_id("sub_block")
-            sub_graph = self._dist_context._serial_graph.get_sub_graph(
+            sub_graph = self._dist_context.serial_graph.get_sub_graph(
                 sub_graph_id)
             sub_graph_nodes = list(sub_graph.all_nodes())
             while_dist_op = self._dist_context.get_dist_op_for_graph(
@@ -440,6 +519,7 @@ class Completer:
                     merged_process_mesh = merge_process_mesh_two(
                         merged_process_mesh, dist_attr.process_mesh)
             while_op_dist_attr.process_mesh = merged_process_mesh
+            _make_dims_mapping_replicate(while_op_dist_attr)
 
             # Step 2: set the related nodes of while_op to the process mesh of while_op
             # Step 2.1: Find related nodes of cond var the graph of while_op
@@ -480,6 +560,7 @@ class Completer:
                 tensor_dist_attr = self._dist_context.get_dist_attr_for_graph(
                     node)
                 tensor_dist_attr.process_mesh = merged_process_mesh
+                _make_dims_mapping_replicate(tensor_dist_attr)
 
             # Step 3: set the process meshes of the inputs in while_op to the process meshes of the outside input nodes
             while_op_inputs_dist_attrs = while_op_dist_attr.inputs_dist_attrs
@@ -519,6 +600,25 @@ class Completer:
                 dist_attr = self._dist_context.get_dist_attr_for_graph(
                     array_node)
                 dist_attr.process_mesh = merged_process_mesh
+                _make_dims_mapping_replicate(dist_attr)
+
+    def _update_process_mesh_between_graphs(self):
+        for parent_node, child_node in self._node_pairs_between_graphs:
+            parent_node_dist_attr = self._dist_context.get_dist_attr_for_graph(
+                parent_node)
+            child_node_dist_attr = self._dist_context.get_dist_attr_for_graph(
+                child_node)
+            parent_node_dist_attr.process_mesh = child_node_dist_attr.process_mesh
+            compatible_process_mesh = compute_compatible_process_mesh([
+                parent_node_dist_attr.process_mesh,
+                child_node_dist_attr.process_mesh
+            ])
+            if compatible_process_mesh is not None \
+                and parent_node_dist_attr.process_mesh != compatible_process_mesh:
+                parent_node_dist_attr.process_mesh = compatible_process_mesh
+            if compatible_process_mesh is not None \
+                and child_node_dist_attr.process_mesh != compatible_process_mesh:
+                child_node_dist_attr.process_mesh = compatible_process_mesh
 
     def _update_process_mesh(self):
         ordered_op_nodes = self._dist_context._serial_ordered_op_nodes
@@ -569,7 +669,7 @@ class Completer:
             return None
         for idx, op_node in enumerate(ordered_op_nodes[
                 idx_of_first_op_node_has_process_mesh + 1:]):
-            original_idx = idx_of_first_op_node_has_process_mesh + +idx + 1
+            original_idx = idx_of_first_op_node_has_process_mesh + idx + 1
             nearest_op_node = ordered_op_nodes[original_idx - 1]
             nearest_op_dist_attr = self._dist_context.get_dist_attr_for_graph(
                 nearest_op_node)
@@ -585,6 +685,9 @@ class Completer:
         # Step 3: adjust the process meshes for special ops
         self._update_process_mesh_for_specials()
 
+        # Step 4: adjust the process meshes between graphs 
+        self._update_process_mesh_between_graphs()
+
     def _prepare(self):
         self._while_op_nodes = {}
         self._array_nodes = {}
@@ -620,7 +723,7 @@ class Completer:
                             self._node_pairs_between_graphs.append(
                                 (after_node, node))
 
-    def complete_forward_annotation(self, serial_main_program):
+    def complete_forward_annotation(self, serial_main_program=None):
         """ Complete annotation for the partial annotated serial_main_program.
         Arguments:
             serial_main_program: partial annotated serial_main_program.
@@ -628,15 +731,12 @@ class Completer:
             serial_main_program: completed annotated serial_main_program.
         """
 
-        # Use the default distribted context for completeion if there is no one
-        self._dist_context.serial_program = serial_main_program
-
-        # Initialize distributed attributes for all var and op node in serial_main_program
-        self._dist_context.init_dist_attr_for_program()
-        # print_program_with_dist_attr(serial_main_program, self._dist_context)
+        if serial_main_program is None:
+            serial_main_program = self._dist_context.serial_main_program
+        else:
+            self._dist_context.serial_main_program = serial_main_program
 
-        # Initialize distributed attributes for all var and op node in graph
-        self._dist_context.init_dist_attr_for_graph()
+        self._dist_context.initialize()
 
         self._prepare()
 
@@ -646,10 +746,9 @@ class Completer:
 
         # Copy the corresponding distributed attribute from graph to serial_main_program
         self._dist_context.copy_dist_attr_from_graph_to_program()
-        self._dist_context.clear_dist_info_for_graph()
 
         # NOTE:[HighOrderGrad] update vars and ops distributed attribute in high order gradient
-        self.complete_high_order_grad_annotation(serial_main_program)
+        self._complete_high_order_grad_annotation(serial_main_program)
 
         # Do the validation check and amend some completion
         self._dist_context.amend_dist_attr_for_program()
@@ -658,7 +757,7 @@ class Completer:
 
         return serial_main_program
 
-    def complete_high_order_grad_annotation(self, serial_main_program):
+    def _complete_high_order_grad_annotation(self, serial_main_program):
         """
         NOTE: 
             [HighOrderGrad] Complete the annotation of vars and ops only for high order gradient.
@@ -818,6 +917,10 @@ class Completer:
 
     def complete_backward_annotation(self, serial_main_program):
         """Complete the annotation of vars and ops in the backward phase for parallel program."""
+        if serial_main_program is None:
+            serial_main_program = self._dist_context.serial_main_program
+        else:
+            self._dist_context.serial_main_program = serial_main_program
 
         def _is_grad_var_name(name):
             if "@GRAD" in name:
@@ -1036,8 +1139,12 @@ class Completer:
                 self._dist_context.set_op_dist_attr_for_program(
                     grad_op, grad_op_dist_attr)
 
-    def complete_update_annotation(self, serial_main_program):
+    def complete_update_annotation(self, serial_main_program=None):
         """Complete the annotation of vars and ops in the update phase for parallel program."""
+        if serial_main_program is None:
+            serial_main_program = self._dist_context.serial_main_program
+        else:
+            self._dist_context.serial_main_program = serial_main_program
         ops = list(serial_main_program.global_block().ops)
         vars = serial_main_program.global_block().vars
         learning_rate_completed = False

python/paddle/distributed/auto_parallel/dist_attribute.py

@@ -52,7 +52,7 @@ def append_op_output_suffix(name):
 
 class TensorDistributedAttribute:
     def __init__(self):
-        # The process mesh of distributed operator attribute must is the same as 
+        # The process mesh of distributed operator attribute must is the same as
         # the process meshes of all input and output distributed attributed
         self._process_mesh = None
         self._dims_mapping = None
@@ -132,12 +132,29 @@ class TensorDistributedAttribute:
                         key, dist_attr)
             self._is_annotated = copy.deepcopy(dist_attr._is_annotated)
 
+    # def reset(self, skip_dist_attr_field_names):
+    #     if skip_dist_attr_field_names is not None \
+    #         and "process_mesh" not in skip_dist_attr_field_names:
+    #         self._process_mesh = None
+    #     if skip_dist_attr_field_names is not None \
+    #         and "dims_mapping" not in skip_dist_attr_field_names:
+    #         for i in enumerate(self._dims_mapping):
+    #             self._dims_mapping[i] = -1
+    #     self._is_annotated = {}
+
     def is_annotated(self, dist_attr_field_name):
         return self._is_annotated.get(dist_attr_field_name, False)
 
+    # def mark_annotated_all(self):
+    #     for key in get_tensor_dist_attr_field_keys():
+    #         self.mark_annotated(key)
+
     def mark_annotated(self, dist_attr_field_name):
         self._is_annotated[dist_attr_field_name] = True
 
+    # def unmark_annotated(self, dist_attr_field_name):
+    #     self._is_annotated[dist_attr_field_name] = False
+
     def mark_annotated_as(self, dist_attr):
         if dist_attr is None:
             return
@@ -195,7 +212,7 @@ class OperatorDistributedAttribute:
             if isinstance(process_mesh, list):
                 process_mesh = ProcessMesh(process_mesh)
             self._process_mesh = copy.deepcopy(process_mesh)
-            # In while op, the proess mesh is not shared by all inputs and outputs 
+            # In while op, the proess mesh is not shared by all inputs and outputs
             if self._op_type == "while":
                 return None
             for dist_attr in self._inputs_dist_attrs.values():
@@ -357,9 +374,25 @@ class OperatorDistributedAttribute:
                         "ProcessMeshes in DistributedOperator must be the same."
         self.process_mesh = shared_process_mesh
 
+    # def reset(self, skip_dist_attr_field_names):
+    #     for tensor_dist_attr in self.inputs_dist_attrs.values():
+    #         tensor_dist_attr.reset(skip_dist_attr_field_names)
+    #     for tensor_dist_attr in self.outputs_dist_attrs.values():
+    #         tensor_dist_attr.reset(skip_dist_attr_field_names)
+    #     if skip_dist_attr_field_names is not None \
+    #         and "process_mesh" not in skip_dist_attr_field_names:
+    #         self.process_mesh = None
+    #     self.impl_type = "default"
+    #     self.impl_idx = 0
+    #     self._is_annotated = {}
+
     def is_annotated(self, attr_name):
         return self._is_annotated.get(attr_name, False)
 
+    # def mark_annotated_all(self):
+    #     for key in get_op_dist_attr_field_keys():
+    #         self.mark_annotated(key)
+
     def mark_annotated(self, attr_name):
         if attr_name == "process_mesh":
             # Make sure proscess_mesh be annotated consistently

python/paddle/distributed/auto_parallel/engine.py

@@ -118,11 +118,10 @@ class Engine:
                 losses = to_list(self._loss(*(outputs + labels)))
 
         default_ctx = get_default_distributed_context()
-        if not default_ctx.is_annotation or self._default_strategy:
+        if not default_ctx.has_annotation or self._default_strategy:
             inputs = [self._set_data_parallel(var) for var in inputs]
             labels = [self._set_data_parallel(var) for var in labels]
 
-        # print(serial_main_prog)
         self._feed_vars[mode] = {"inputs": inputs, "labels": labels}
 
         self._fetch_vars[mode] = {

python/paddle/distributed/auto_parallel/operators/common.py

@@ -18,16 +18,16 @@ from ..dist_attribute import OperatorDistributedAttribute
 _g_distributed_operator_impl_containers = {}
 
 _g_elementwise_ops = [
-    "elementwise_add", "gelu", "dropout", "cast", "gather", "concat"
+    "elementwise", "gelu", "dropout", "cast", "gather", "concat"
 ]
 BACKWARD_ONLY_DIST_OPS = {'check_finite_and_unscale', 'update_loss_scaling'}
 
 
 def is_elementwise_op(op_type):
-    if op_type in _g_elementwise_ops:
-        return True
-    else:
-        return False
+    for eltwise_op in _g_elementwise_ops:
+        if eltwise_op in op_type:
+            return True
+    return False
 
 
 class DistributedOperatorImplContainer:
@@ -156,7 +156,9 @@ def register_distributed_operator_impl(op_type, dist_impl):
         assert False, "Must register distributed operator registry first."
 
 
-def find_best_compatible_distributed_operator_impl(dist_op, fwd=True):
+def find_best_compatible_distributed_operator_impl(dist_op,
+                                                   fwd=True,
+                                                   partial=True):
     """
     Here just return the first compatible implemention. 
     This will be improved by cost model in the future.
@@ -168,39 +170,55 @@ def find_best_compatible_distributed_operator_impl(dist_op, fwd=True):
     dist_op_default_impl_container = get_distributed_operator_impl_container(
         "default")
     compatible_impls = []
-    if fwd:
-        # First, find impls in the corresponding container
-        if dist_op_impl_container:
-            compatible_impls.extend(
-                dist_op_impl_container.get_input_compatible_impls(dist_op))
-        # Second, find impls in the elementwise container
-        if dist_op_eltwise_impl_container and is_elementwise_op(op_type):
-            compatible_impls.extend(
-                dist_op_eltwise_impl_container.get_input_compatible_impls(
-                    dist_op))
-        # Third, find impls in the default container
-        if dist_op_default_impl_container:
-            compatible_impls.extend(
-                dist_op_default_impl_container.get_input_compatible_impls(
-                    dist_op))
+    if partial:
+        if fwd:
+            # First, find impls in the corresponding container
+            if dist_op_impl_container:
+                compatible_impls.extend(
+                    dist_op_impl_container.get_input_compatible_impls(dist_op))
+            # Second, find impls in the elementwise container
+            if dist_op_eltwise_impl_container and is_elementwise_op(op_type):
+                compatible_impls.extend(
+                    dist_op_eltwise_impl_container.get_input_compatible_impls(
+                        dist_op))
+            # Third, find impls in the default container
+            if dist_op_default_impl_container:
+                compatible_impls.extend(
+                    dist_op_default_impl_container.get_input_compatible_impls(
+                        dist_op))
+        else:
+            # First, find impls in the corresponding container
+            if dist_op_impl_container:
+                compatible_impls.extend(
+                    dist_op_impl_container.get_output_compatible_impls(dist_op))
+            # Second, find impls in the elementwise container
+            if dist_op_eltwise_impl_container and is_elementwise_op(op_type):
+                compatible_impls.extend(
+                    dist_op_eltwise_impl_container.get_output_compatible_impls(
+                        dist_op))
+            # Third, find impls in the default container
+            if dist_op_default_impl_container:
+                compatible_impls.extend(
+                    dist_op_default_impl_container.get_output_compatible_impls(
+                        dist_op))
     else:
         # First, find impls in the corresponding container
         if dist_op_impl_container:
             compatible_impls.extend(
-                dist_op_impl_container.get_output_compatible_impls(dist_op))
+                dist_op_impl_container.get_compatible_impls(dist_op))
         # Second, find impls in the elementwise container
         if dist_op_eltwise_impl_container and is_elementwise_op(op_type):
             compatible_impls.extend(
-                dist_op_eltwise_impl_container.get_output_compatible_impls(
-                    dist_op))
+                dist_op_eltwise_impl_container.get_compatible_impls(dist_op))
         # Third, find impls in the default container
         if dist_op_default_impl_container:
             compatible_impls.extend(
-                dist_op_default_impl_container.get_output_compatible_impls(
-                    dist_op))
+                dist_op_default_impl_container.get_compatible_impls(dist_op))
+
     if compatible_impls:
         # For now, just return the first compatible impl
-        best_compatible_impl = compatible_impls[0]
+        # best_compatible_impl = compatible_impls[0]
+        best_compatible_impl = compatible_impls
     else:
         best_compatible_impl = None
     return best_compatible_impl

python/paddle/distributed/auto_parallel/operators/dist_default.py

@@ -53,6 +53,7 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
     def is_input_compatible(self, dist_op):
         op_desc = dist_op.serial_op.desc
         op_dist_attr = dist_op.dist_attr
+        batch_dim_mappings = []
         input_names = op_desc.input_names()
         xshape_arg_names = []
         if "XShape" in input_names:
@@ -64,14 +65,14 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
                 for mapping in dims_mapping:
                     if mapping != -1:
                         return False
-                # continue
-                # if len(dims_mapping) < 1:
-                #     continue
+                continue
             if arg_name not in xshape_arg_names:
                 if len(dims_mapping) > 1:
                     for mapping in dims_mapping[1:]:
                         if mapping != -1:
                             return False
+                if len(dims_mapping) >= 1:
+                    batch_dim_mappings.append(dims_mapping[0])
             else:
                 if dims_mapping[0] != -1:
                     return False
@@ -79,12 +80,19 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
                     for mapping in dims_mapping[2:]:
                         if mapping != -1:
                             return False
+                if len(dims_mapping) >= 2:
+                    batch_dim_mappings.append(dims_mapping[1])
+
+        if compute_compatible_dim_mapping(batch_dim_mappings) is None:
+            return False
+
         return True
 
     def is_output_compatible(self, dist_op):
         op_desc = dist_op.serial_op.desc
         op_dist_attr = dist_op.dist_attr
         output_names = op_desc.output_names()
+        batch_dim_mappings = []
         xshape_arg_names = []
         if "XShape" in output_names:
             xshape_arg_names = op_desc.output("XShape")
@@ -95,14 +103,14 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
                 for mapping in dims_mapping:
                     if mapping != -1:
                         return False
-                # continue
-                # if len(dims_mapping) < 1:
-                #     continue
+                continue
             if arg_name not in xshape_arg_names:
                 if len(dims_mapping) > 1:
                     for mapping in dims_mapping[1:]:
                         if mapping != -1:
                             return False
+                if len(dims_mapping) >= 1:
+                    batch_dim_mappings.append(dims_mapping[0])
             else:
                 if dims_mapping[0] != -1:
                     return False
@@ -110,6 +118,12 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
                     for mapping in dims_mapping[2:]:
                         if mapping != -1:
                             return False
+                if len(dims_mapping) >= 2:
+                    batch_dim_mappings.append(dims_mapping[1])
+
+        if compute_compatible_dim_mapping(batch_dim_mappings) is None:
+            return False
+
         return True
 
     def is_auto_compatible(self, dist_op):
@@ -123,9 +137,12 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
             xshape_arg_names = op_desc.input("XShape")
         for arg_name in op_desc.input_arg_names():
             serial_tensor = dist_op.get_serial_input(arg_name)
+            dims_mapping = op_dist_attr.get_input_dims_mapping(arg_name)
             if serial_tensor.is_parameter:
+                for mapping in dims_mapping:
+                    if mapping != -1:
+                        return False
                 continue
-            dims_mapping = op_dist_attr.get_input_dims_mapping(arg_name)
             if arg_name not in xshape_arg_names:
                 if len(dims_mapping) > 1:
                     for mapping in dims_mapping[1:]:
@@ -150,9 +167,12 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
             xshape_arg_names = op_desc.output("XShape")
         for arg_name in op_desc.output_arg_names():
             serial_tensor = dist_op.get_serial_output(arg_name)
+            dims_mapping = op_dist_attr.get_output_dims_mapping(arg_name)
             if serial_tensor.is_parameter:
+                for mapping in dims_mapping:
+                    if mapping != -1:
+                        return False
                 continue
-            dims_mapping = op_dist_attr.get_output_dims_mapping(arg_name)
             if arg_name not in xshape_arg_names:
                 if len(dims_mapping) > 1:
                     for mapping in dims_mapping[1:]:
@@ -229,7 +249,9 @@ class DistributedDefaultImpl0(DistributedOperatorImpl):
 
         compatible_dim_mapping = compute_compatible_dim_mapping(
             batch_dim_mappings)
-        assert compatible_dim_mapping is not None, "There is no compatible dim mapping."
+        if compatible_dim_mapping is None:
+            return False
+
         for arg_name in op_desc.input_arg_names():
             serial_tensor = dist_op.get_serial_input(arg_name)
             if serial_tensor.is_parameter:

python/paddle/distributed/auto_parallel/operators/dist_eltwise.py

@@ -52,21 +52,46 @@ class DistributedElementwiseImpl0(DistributedOperatorImpl):
 
     def is_input_compatible(self, dist_op):
         op_desc = dist_op.serial_op.desc
-        if is_elementwise_op(op_desc.type()):
-            return True
-        else:
+        if not is_elementwise_op(op_desc.type()):
             return False
+        op_dist_attr = dist_op.dist_attr
+        dims_mapping_list = []
+        input_arg_names = op_desc.input_arg_names()
+        max_dims_mapping_len = -1
+        for arg_name in input_arg_names:
+            dims_mapping = op_dist_attr.get_input_dims_mapping(arg_name)
+            if max_dims_mapping_len < len(dims_mapping):
+                max_dims_mapping_len = len(dims_mapping)
+            dims_mapping_list.append(dims_mapping)
+
+        for idx in range(max_dims_mapping_len):
+            dim_mappings = []
+            for dims_mapping in dims_mapping_list:
+                if idx < len(dims_mapping):
+                    dim_mappings.append(dims_mapping[-(idx + 1)])
+            if compute_compatible_dim_mapping(dim_mappings) is None:
+                return False
+        return True
 
     def is_output_compatible(self, dist_op):
         op_desc = dist_op.serial_op.desc
-        op_desc = dist_op.serial_op.desc
-        if is_elementwise_op(op_desc.type()):
-            return True
-        else:
+        if not is_elementwise_op(op_desc.type()):
+            return False
+        op_dist_attr = dist_op.dist_attr
+        dims_mapping_list = []
+        output_arg_names = op_desc.output_arg_names()
+        for arg_name in output_arg_names:
+            dims_mapping = op_dist_attr.get_output_dims_mapping(arg_name)
+            dims_mapping_list.append(dims_mapping)
+
+        if compute_compatible_dims_mapping(dims_mapping_list) is None:
             return False
+        return True
 
     def is_auto_compatible(self, dist_op):
         op_desc = dist_op.serial_op.desc
+        if not is_elementwise_op(op_desc.type()):
+            return False
         op_dist_attr = dist_op.dist_attr
         dims_mapping_list = []
         input_arg_names = op_desc.input_arg_names()
@@ -127,7 +152,8 @@ class DistributedElementwiseImpl0(DistributedOperatorImpl):
 
         compatible_dims_mapping = compute_compatible_dims_mapping(
             dims_mapping_list)
-        assert compatible_dims_mapping is not None, "There is no compatible dim mapping."
+        if compatible_dims_mapping is None:
+            return False
 
         for arg_name in input_arg_names:
             if input_dims_mapping_lens[arg_name] < max_dims_mapping_len:

python/paddle/distributed/auto_parallel/operators/dist_matmul.py

@@ -95,7 +95,8 @@ def _update_dims_mapping_for_matmul(dist_op):
             broadcast_x_dims_mapping, broadcast_y_dims_mapping,
             broadcast_out_dims_mapping
         ])
-        assert compatible_dims_mapping is not None, "There is no compatible dim mapping."
+        if compatible_dims_mapping is None:
+            return False
 
         for i in range(x_dims_mapping_len - 2):
             new_idx = i + (out_dims_mapping_len - x_dims_mapping_len)

python/paddle/distributed/auto_parallel/operators/dist_pnorm.py

@@ -117,7 +117,8 @@ class DistributedPNormImpl(DistributedOperatorImpl):
 
         compatible_dim_mapping = compute_compatible_dim_mapping(
             batch_dim_mappings)
-        assert compatible_dim_mapping is not None, "There is no compatible dim mapping."
+        if compatible_dim_mapping is None:
+            return False
 
         for arg_name in op_desc.input_arg_names():
             dims_mapping = op_dist_attr.get_input_dims_mapping(arg_name)

python/paddle/distributed/auto_parallel/operators/dist_slice.py

 # Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
-# 
+#
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
 # You may obtain a copy of the License at
-# 
+#
 #     http://www.apache.org/licenses/LICENSE-2.0
-# 
+#
 # Unless required by applicable law or agreed to in writing, software
 # distributed under the License is distributed on an "AS IS" BASIS,
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
@@ -17,6 +17,7 @@ from .common import DistributedOperatorImpl
 from .common import register_distributed_operator_impl_container
 from .common import register_distributed_operator_impl
 from ..utils import is_dim_shard
+from ..utils import compute_compatible_dim_mapping
 from ..utils import compute_compatible_and_update_dim_mapping
 from .dist_default import DistributedDefaultImpl0
 
@@ -47,6 +48,29 @@ class DistributedSliceImpl(DistributedOperatorImpl):
         return True
 
     def is_output_compatible(self, dist_op):
+        op_desc = dist_op.serial_op.desc
+        op_dist_attr = dist_op.dist_attr
+        in_name = op_desc.input('Input')[0]
+        out_name = op_desc.output('Out')[0]
+        axes = op_desc.attr('axes')
+        decrease_axis = op_desc.attr('decrease_axis')
+        in_dims_mapping = op_dist_attr.get_input_dims_mapping(in_name)
+        out_dims_mapping = op_dist_attr.get_output_dims_mapping(out_name)
+
+        ref_indices = []
+        for i in range(len(in_dims_mapping)):
+            if i not in decrease_axis:
+                ref_indices.append(i)
+        if ref_indices == []:
+            assert len(out_dims_mapping) == 1
+            if is_dim_shard(out_dims_mapping[0]):
+                return False
+        else:
+            for i in range(len(out_dims_mapping)):
+                ref_index = ref_indices[i]
+                if ref_index in axes and is_dim_shard(out_dims_mapping[i]):
+                    return False
+
         return True
 
     def is_compatible(self, dist_op):
@@ -95,17 +119,30 @@ class DistributedSliceImpl(DistributedOperatorImpl):
         out_dims_mapping = op_dist_attr.get_output_dims_mapping(out_name)
 
         ref_dims_mapping = []
+        ref_indices = []
         for i in range(len(in_dims_mapping)):
             if i not in decrease_axis:
                 ref_dims_mapping.append(in_dims_mapping[i])
+                ref_indices.append(i)
+
         if ref_dims_mapping == []:
             ref_dims_mapping = [-1]
-
-        assert len(ref_dims_mapping) == len(out_dims_mapping)
-        for i in range(len(out_dims_mapping)):
-            if out_dims_mapping[i] != ref_dims_mapping[i]:
-                out_dims_mapping[i] = ref_dims_mapping[i]
-                changed = True
+            assert len(ref_dims_mapping) == len(out_dims_mapping)
+            assert ref_dims_mapping[0] == out_dims_mapping[0]
+            changed = False
+        else:
+            assert len(ref_dims_mapping) == len(out_dims_mapping)
+            for i in range(len(out_dims_mapping)):
+                compatible_dim_mapping = compute_compatible_dim_mapping(
+                    [out_dims_mapping[i], ref_dims_mapping[i]])
+                if compatible_dim_mapping is None:
+                    continue
+                if ref_dims_mapping[i] != compatible_dim_mapping:
+                    in_dims_mapping[ref_indices[i]] = compatible_dim_mapping
+                    changed = True
+                if out_dims_mapping[i] != compatible_dim_mapping:
+                    out_dims_mapping[i] = compatible_dim_mapping
+                    changed = True
 
         return changed
 

python/paddle/distributed/auto_parallel/parallelizer.py

@@ -230,7 +230,7 @@ class AutoParallelizer:
             g_process_group_map = copy.deepcopy(_g_process_group_map)
             _g_process_group_map.clear()
             _g_process_group_map[0] = ProcessGroup(0, [])
-            for process_mesh in dist_context._process_meshes:
+            for process_mesh in self._dist_context._process_meshes:
                 _g_process_group_map[0].add_ranks(process_mesh.processes)
         return dist_optimize_ops, dist_params_grads, dist_startup_prog, dist_main_prog, g_process_group_map
 

python/paddle/distributed/auto_parallel/tuner/recorder.py

@@ -138,7 +138,6 @@ class MetricRecords(object):
     def from_state(cls, state):
         records = cls(state["direction"])
         records.records = [MetricRecord.from_state(r) for r in state["records"]]
-        print("here 1", records.records)
         return records
 
 

python/paddle/distributed/auto_parallel/utils.py

@@ -159,11 +159,11 @@ def print_program_with_dist_attr(program, dist_context=None):
     from .dist_context import set_default_distributed_context
     if dist_context is None:
         dist_context = get_default_distributed_context()
-        print(program)
+        print(program, flush=True)
     else:
         original_default_context = get_default_distributed_context()
         set_default_distributed_context(dist_context)
-        print(program)
+        print(program, flush=True)
         set_default_distributed_context(original_default_context)
     lock.release()
 

python/paddle/distributed/passes/auto_parallel_recompute.py

@@ -350,11 +350,12 @@ class RecomputePass(PassBase):
                     for _, op_desc in reversed(list(enumerate(segment_descs))):
                         rc_desc = main_block.desc._insert_op(idx)
                         rc_desc.copy_from(op_desc)
+                        rc_desc.set_original_id(rc_desc.id())
                         rc_op = Operator(main_block, rc_desc)
                         main_block.ops.insert(idx, rc_op)
                         # set recomputed ops' dist attr
                         fwd_op_dist_attr = self._dist_context.get_op_dist_attr_for_program_with_id(
-                            rc_desc.original_id())
+                            op_desc.original_id())
                         assert fwd_op_dist_attr is not None
                         self.set_op_dist_attr(rc_op, fwd_op_dist_attr,
                                               var_name_dict)

python/paddle/fluid/tests/unittests/auto_parallel/CMakeLists.txt

@@ -3,18 +3,23 @@
 if(WITH_DISTRIBUTE AND WITH_GPU)
     py_test_modules(test_auto_parallel_relaunch MODULES test_auto_parallel_relaunch ENVS ${dist_ENVS})
     set_tests_properties(test_auto_parallel_relaunch PROPERTIES LABELS "RUN_TYPE=EXCLUSIVE" TIMEOUT 120)
+
     py_test_modules(test_relaunch_with_planner MODULES test_relaunch_with_planner ENVS ${dist_ENVS})
     set_tests_properties(test_relaunch_with_planner PROPERTIES LABELS "RUN_TYPE=EXCLUSIVE" TIMEOUT 120)
+
     py_test_modules(test_relaunch_with_gpt_planner MODULES test_relaunch_with_gpt_planner ENVS ${dist_ENVS})
     set_tests_properties(test_relaunch_with_gpt_planner PROPERTIES LABELS "RUN_TYPE=EXCLUSIVE" TIMEOUT 240)
+
     py_test_modules(test_engine_api MODULES test_engine_api ENVS ${dist_ENVS})
     set_tests_properties(test_engine_api PROPERTIES LABELS "RUN_TYPE=EXCLUSIVE" TIMEOUT 80)
-    py_test_modules(test_while_op_completion MODULES test_while_op_completion ENVS ${dist_ENVS})
+
     py_test_modules(test_converter MODULES test_converter ENVS ${dist_ENVS})
     set_tests_properties(test_converter PROPERTIES LABELS "RUN_TYPE=EXCLUSIVE" TIMEOUT 50)
     py_test_modules(test_high_order_grad MODULES test_high_order_grad ENVS ${dist_ENVS})
     set_tests_properties(test_high_order_grad PROPERTIES LABELS "RUN_TYPE=EXCLUSIVE" TIMEOUT 50)
 
+    py_test_modules(test_while_op_completion MODULES test_while_op_completion ENVS ${dist_ENVS})
+    py_test_modules(test_while_op_partition MODULES test_while_op_partition ENVS ${dist_ENVS})
     py_test_modules(test_tunable_variable MODULES test_tunable_variable ENVS ${dist_ENVS})
     py_test_modules(test_tunable_space MODULES test_tunable_space ENVS ${dist_ENVS})
     py_test_modules(test_recorder MODULES test_recorder ENVS ${dist_ENVS})

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_reshape.py

@@ -66,7 +66,6 @@ class TestDistReshape(unittest.TestCase):
         for rank in range(2):
             dist_main_prog, dist_context = parallelizer(make_program_dp2, rank)
             ops = dist_main_prog.global_block().ops
-            print_program_with_dist_attr(dist_main_prog, dist_context)
             for idx, op in enumerate(ops):
                 op_dist_attr = dist_context.get_op_dist_attr_for_program(op)
                 assert op_dist_attr.impl_type == "reshape2"

python/paddle/fluid/tests/unittests/auto_parallel/test_dist_slice.py

@@ -15,6 +15,7 @@
 import unittest
 import paddle
 import paddle.distributed.auto_parallel as auto
+from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr
 
 paddle.enable_static()
 
@@ -85,14 +86,9 @@ class TestDistSlice(unittest.TestCase):
             for op in ops:
                 axes = op.desc.attr('axes')
                 op_dist_attr = dist_context.get_op_dist_attr_for_program(op)
-                if axes[0] == 0:
-                    assert op_dist_attr.impl_type == "default"
-                else:
-                    assert op_dist_attr.impl_type == "slice"
-                    for out in op.output_arg_names:
-                        var_dims_mapping = op_dist_attr.get_output_dims_mapping(
-                            out)
-                        assert var_dims_mapping[0] == 0
+                assert op_dist_attr.impl_type == "slice"
+                for out in op.output_arg_names:
+                    var_dims_mapping = op_dist_attr.get_output_dims_mapping(out)
 
     def test_dist_slice_serial(self):
         dist_main_prog, dist_context = parallelizer(make_program_serial, 0)

python/paddle/fluid/tests/unittests/auto_parallel/test_auto_parallel_while_op.py → python/paddle/fluid/tests/unittests/auto_parallel/test_while_op_partition.py

@@ -23,12 +23,13 @@ import paddle.nn.functional as F
 import paddle.distributed.auto_parallel as auto
 
 from paddle.distributed import fleet
-
+from paddle.distributed.auto_parallel.completion import Completer
 from paddle.distributed.auto_parallel.partitioner import Partitioner
 from paddle.distributed.auto_parallel.utils import make_data_unshard
 from paddle.distributed.auto_parallel.dist_attribute import OperatorDistributedAttribute, TensorDistributedAttribute
 from paddle.distributed.auto_parallel.dist_context import DistributedContext, get_default_distributed_context
 from paddle.distributed.auto_parallel.operators import find_best_compatible_distributed_operator_impl
+from paddle.distributed.auto_parallel.utils import print_program_with_dist_attr
 
 paddle.enable_static()
 
@@ -283,139 +284,143 @@ def get_program():
 
 
 def completion(train_program, start_program, dist_context):
-    blocks = train_program.blocks
-    # completion tensors
-    for block in blocks:
-        for op in block.ops:
-            if op.type == "layer_norm":
-                for out_name in op.output_arg_names:
-                    out_var = block.vars[out_name]
-                    tensor_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                        out_var)
-                    if tensor_dist_attr:
-                        continue
-                    tensor_dist_attr = TensorDistributedAttribute()
-                    tensor_dist_attr.process_mesh = _g_process_mesh
-                    tensor_dist_attr.dims_mapping = [-1]
-                    dist_context.set_tensor_dist_attr_for_program(
-                        out_var, tensor_dist_attr)
-
-            elif op.type == "elementwise_sub":
-                for out_name in op.output_arg_names:
-                    out_var = block.vars[out_name]
-                    tensor_dist_attr = TensorDistributedAttribute()
-                    tensor_dist_attr.process_mesh = _g_process_mesh
-                    tensor_dist_attr.dims_mapping = [-1, -1, -1]
-                    dist_context.set_tensor_dist_attr_for_program(
-                        out_var, tensor_dist_attr)
-
-            elif op.type == "matmul_v2":
-                col = False
-                for in_name in op.input_arg_names:
-                    if ".w_" not in in_name:
-                        continue
-                    if in_name not in block.vars:
-                        in_var = blocks[0].vars[in_name]
-                    else:
-                        in_var = block.vars[in_name]
-                    tensor_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                        in_var)
-                    assert tensor_dist_attr is not None
-                    if tensor_dist_attr.dims_mapping == [-1, 0]:
-                        col = True
-                for out_name in op.output_arg_names:
-                    out_var = block.vars[out_name]
-                    tensor_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                        out_var)
-                    if tensor_dist_attr:
-                        continue
-                    tensor_dist_attr = TensorDistributedAttribute()
-                    tensor_dist_attr.process_mesh = _g_process_mesh
-                    if col:
-                        tensor_dist_attr.dims_mapping = [-1, -1, 0]
-                    else:
-                        tensor_dist_attr.dims_mapping = [-1, -1, -1]
-                    dist_context.set_tensor_dist_attr_for_program(
-                        out_var, tensor_dist_attr)
-            elif op.type == "while":
-                out_name = op.desc.output("StepScopes")[0]
-                out_var = block.vars[out_name]
-                tensor_dist_attr = TensorDistributedAttribute()
-                tensor_dist_attr.process_mesh = _g_process_mesh
-                tensor_dist_attr.dims_mapping = [-1]
-                dist_context.set_tensor_dist_attr_for_program(out_var,
-                                                              tensor_dist_attr)
-
-    # completion ops
-    for block in blocks:
-        for op in block.ops:
-            op_dist_attr = OperatorDistributedAttribute()
-            op_dist_attr.process_mesh = _g_process_mesh
-            if op.type == "create_by_read" or op.type == "create_double_buffer_reader":
-                for in_name in op.input_arg_names:
-                    op_dist_attr.set_input_dims_mapping(in_name, [])
-                for out_name in op.output_arg_names:
-                    op_dist_attr.set_output_dims_mapping(out_name, [])
-            elif op.type == "read":
-                for in_name in op.input_arg_names:
-                    op_dist_attr.set_output_dims_mapping(in_name, [])
-                for out_name in op.output_arg_names:
-                    out_var = block.vars[out_name]
-                    out_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                        out_var)
-                    op_dist_attr.set_output_dist_attr(out_name, out_dist_attr)
-            elif op.type == "while":
-                for in_name in op.input_arg_names:
-                    in_var = block.vars[in_name]
-                    in_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                        in_var)
-                    op_dist_attr.set_input_dist_attr(in_name, in_dist_attr)
-                for out_name in op.output_arg_names:
-                    if out_name == op.desc.output("StepScopes")[0]:
-                        op_dist_attr.set_output_dims_mapping(out_name, [])
-                    else:
-                        out_var = block.vars[out_name]
-                        out_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                            out_var)
-                        op_dist_attr.set_output_dist_attr(out_name,
-                                                          out_dist_attr)
-            else:
-                for in_name in op.input_arg_names:
-                    if in_name == "lod_tensor_blocking_queue_0":
-                        continue
-                    if in_name not in block.vars:
-                        in_var = blocks[0].vars[in_name]
-                    else:
-                        in_var = block.vars[in_name]
-                    in_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                        in_var)
-                    op_dist_attr.set_input_dist_attr(in_name, in_dist_attr)
-                for out_name in op.output_arg_names:
-                    if out_name not in block.vars:
-                        out_var = blocks[0].vars[out_name]
-                    else:
-                        out_var = block.vars[out_name]
-                    out_dist_attr = dist_context.get_tensor_dist_attr_for_program(
-                        out_var)
-                    op_dist_attr.set_output_dist_attr(out_name, out_dist_attr)
-
-            if op.type == "matmul_v2":
-                op_dist_attr.impl_type = "matmul_v2"
-                for in_name in op_dist_attr.inputs_dist_attrs.keys():
-                    in_dist_attr = op_dist_attr.inputs_dist_attrs[in_name]
-                    if ".w_" in in_name and in_dist_attr.dims_mapping[-1] == 0:
-                        op_dist_attr.impl_idx = 0
-                    else:
-                        op_dist_attr.impl_idx = 1
-            elif op.type == "fill_constant_batch_size_like":
-                op_dist_attr.impl_type = "fill_constant_batch_size_like"
-                op_dist_attr.impl_idx = 0
-            else:
-                op_dist_attr.impl_type = "default"
-                op_dist_attr.impl_idx = 0
-
-            dist_context.set_op_dist_attr_for_program(op, op_dist_attr)
-            make_data_unshard(train_program, start_program, dist_context)
+    # blocks = train_program.blocks
+    # # completion tensors
+    # for block in blocks:
+    #     for op in block.ops:
+    #         if op.type == "layer_norm":
+    #             for out_name in op.output_arg_names:
+    #                 out_var = block.vars[out_name]
+    #                 tensor_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                     out_var)
+    #                 if tensor_dist_attr:
+    #                     continue
+    #                 tensor_dist_attr = TensorDistributedAttribute()
+    #                 tensor_dist_attr.process_mesh = _g_process_mesh
+    #                 tensor_dist_attr.dims_mapping = [-1]
+    #                 dist_context.set_tensor_dist_attr_for_program(
+    #                     out_var, tensor_dist_attr)
+
+    #         elif op.type == "elementwise_sub":
+    #             for out_name in op.output_arg_names:
+    #                 out_var = block.vars[out_name]
+    #                 tensor_dist_attr = TensorDistributedAttribute()
+    #                 tensor_dist_attr.process_mesh = _g_process_mesh
+    #                 tensor_dist_attr.dims_mapping = [-1, -1, -1]
+    #                 dist_context.set_tensor_dist_attr_for_program(
+    #                     out_var, tensor_dist_attr)
+
+    #         elif op.type == "matmul_v2":
+    #             col = False
+    #             for in_name in op.input_arg_names:
+    #                 if ".w_" not in in_name:
+    #                     continue
+    #                 if in_name not in block.vars:
+    #                     in_var = blocks[0].vars[in_name]
+    #                 else:
+    #                     in_var = block.vars[in_name]
+    #                 tensor_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                     in_var)
+    #                 assert tensor_dist_attr is not None
+    #                 if tensor_dist_attr.dims_mapping == [-1, 0]:
+    #                     col = True
+    #             for out_name in op.output_arg_names:
+    #                 out_var = block.vars[out_name]
+    #                 tensor_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                     out_var)
+    #                 if tensor_dist_attr:
+    #                     continue
+    #                 tensor_dist_attr = TensorDistributedAttribute()
+    #                 tensor_dist_attr.process_mesh = _g_process_mesh
+    #                 if col:
+    #                     tensor_dist_attr.dims_mapping = [-1, -1, 0]
+    #                 else:
+    #                     tensor_dist_attr.dims_mapping = [-1, -1, -1]
+    #                 dist_context.set_tensor_dist_attr_for_program(
+    #                     out_var, tensor_dist_attr)
+    #         elif op.type == "while":
+    #             out_name = op.desc.output("StepScopes")[0]
+    #             out_var = block.vars[out_name]
+    #             tensor_dist_attr = TensorDistributedAttribute()
+    #             tensor_dist_attr.process_mesh = _g_process_mesh
+    #             tensor_dist_attr.dims_mapping = [-1]
+    #             dist_context.set_tensor_dist_attr_for_program(out_var,
+    #                                                           tensor_dist_attr)
+
+    # # completion ops
+    # for block in blocks:
+    #     for op in block.ops:
+    #         op_dist_attr = OperatorDistributedAttribute()
+    #         op_dist_attr.process_mesh = _g_process_mesh
+    #         if op.type == "create_by_read" or op.type == "create_double_buffer_reader":
+    #             for in_name in op.input_arg_names:
+    #                 op_dist_attr.set_input_dims_mapping(in_name, [])
+    #             for out_name in op.output_arg_names:
+    #                 op_dist_attr.set_output_dims_mapping(out_name, [])
+    #         elif op.type == "read":
+    #             for in_name in op.input_arg_names:
+    #                 op_dist_attr.set_output_dims_mapping(in_name, [])
+    #             for out_name in op.output_arg_names:
+    #                 out_var = block.vars[out_name]
+    #                 out_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                     out_var)
+    #                 op_dist_attr.set_output_dist_attr(out_name, out_dist_attr)
+    #         elif op.type == "while":
+    #             for in_name in op.input_arg_names:
+    #                 in_var = block.vars[in_name]
+    #                 in_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                     in_var)
+    #                 op_dist_attr.set_input_dist_attr(in_name, in_dist_attr)
+    #             for out_name in op.output_arg_names:
+    #                 if out_name == op.desc.output("StepScopes")[0]:
+    #                     op_dist_attr.set_output_dims_mapping(out_name, [])
+    #                 else:
+    #                     out_var = block.vars[out_name]
+    #                     out_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                         out_var)
+    #                     op_dist_attr.set_output_dist_attr(out_name,
+    #                                                       out_dist_attr)
+    #         else:
+    #             for in_name in op.input_arg_names:
+    #                 if in_name == "lod_tensor_blocking_queue_0":
+    #                     continue
+    #                 if in_name not in block.vars:
+    #                     in_var = blocks[0].vars[in_name]
+    #                 else:
+    #                     in_var = block.vars[in_name]
+    #                 in_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                     in_var)
+    #                 op_dist_attr.set_input_dist_attr(in_name, in_dist_attr)
+    #             for out_name in op.output_arg_names:
+    #                 if out_name not in block.vars:
+    #                     out_var = blocks[0].vars[out_name]
+    #                 else:
+    #                     out_var = block.vars[out_name]
+    #                 out_dist_attr = dist_context.get_tensor_dist_attr_for_program(
+    #                     out_var)
+    #                 op_dist_attr.set_output_dist_attr(out_name, out_dist_attr)
+
+    #         if op.type == "matmul_v2":
+    #             op_dist_attr.impl_type = "matmul_v2"
+    #             for in_name in op_dist_attr.inputs_dist_attrs.keys():
+    #                 in_dist_attr = op_dist_attr.inputs_dist_attrs[in_name]
+    #                 if ".w_" in in_name and in_dist_attr.dims_mapping[-1] == 0:
+    #                     op_dist_attr.impl_idx = 0
+    #                 else:
+    #                     op_dist_attr.impl_idx = 1
+    #         elif op.type == "fill_constant_batch_size_like":
+    #             op_dist_attr.impl_type = "fill_constant_batch_size_like"
+    #             op_dist_attr.impl_idx = 0
+    #         else:
+    #             op_dist_attr.impl_type = "default"
+    #             op_dist_attr.impl_idx = 0
+
+    #         dist_context.set_op_dist_attr_for_program(op, op_dist_attr)
+    #         make_data_unshard(train_program, start_program, dist_context)
+
+    completer = Completer(dist_context)
+    train_program = completer.complete_forward_annotation(train_program)
+    make_data_unshard(train_program, start_program, dist_context)
 
     return train_program, start_program
 

python/paddle/fluid/tests/unittests/auto_parallel_parallelizer.py

@@ -134,7 +134,6 @@ class TestMLPAutoParallelizer(unittest.TestCase):
             for op in block.ops:
                 for attr_name in op.attr_names:
                     self.assertTrue(suffix not in attr_name)
-        # print_program_with_dist_attr(distributed_main_program)
         self.assertIsNotNone(distributed_startup_program)
         self.assertIsNotNone(distributed_main_program)
 

python/paddle/fluid/tests/unittests/test_auto_parallel_reshard.py

@@ -332,7 +332,6 @@ class TestMLPReshard(unittest.TestCase):
         resharder = Resharder(dist_main_prog, dist_startup_prog, rank_id,
                               dist_context, dist_params_grads)
         resharder.reshard()
-        print_program_with_dist_attr(dist_main_prog, dist_context)
 
         # check send and recv result
         self.assertTrue(check_send_recv_result(dist_main_prog, rank_id))