PaddlePaddle
diff --git a/‎python/paddle/distributed/auto_parallel/static/auto_align_tool.py
Copy file name to clipboard
+6-6Lines changed: 6 additions & 6 deletions b/‎python/paddle/distributed/auto_parallel/static/auto_align_tool.py
Copy file name to clipboard
+6-6Lines changed: 6 additions & 6 deletions
diff --git a/‎python/paddle/distributed/auto_parallel/static/completion.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/auto_parallel/static/completion.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
diff --git a/‎python/paddle/distributed/auto_parallel/static/converter.py
Copy file name to clipboard
+3-3Lines changed: 3 additions & 3 deletions b/‎python/paddle/distributed/auto_parallel/static/converter.py
Copy file name to clipboard
+3-3Lines changed: 3 additions & 3 deletions
diff --git a/‎python/paddle/distributed/auto_parallel/static/cost/base_cost.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/auto_parallel/static/cost/base_cost.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
diff --git a/‎python/paddle/distributed/auto_parallel/static/cost_model.py
Copy file name to clipboard
+5-5Lines changed: 5 additions & 5 deletions b/‎python/paddle/distributed/auto_parallel/static/cost_model.py
Copy file name to clipboard
+5-5Lines changed: 5 additions & 5 deletions
diff --git a/‎python/paddle/distributed/auto_parallel/static/helper.py
Copy file name to clipboard
+2-2Lines changed: 2 additions & 2 deletions b/‎python/paddle/distributed/auto_parallel/static/helper.py
Copy file name to clipboard
+2-2Lines changed: 2 additions & 2 deletions
diff --git a/‎python/paddle/distributed/auto_parallel/static/parallelizer_v2.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/auto_parallel/static/parallelizer_v2.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
diff --git a/‎python/paddle/distributed/fleet/base/distributed_strategy.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/fleet/base/distributed_strategy.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
diff --git a/‎python/paddle/distributed/fleet/fleet.py
Copy file name to clipboard
+2-2Lines changed: 2 additions & 2 deletions b/‎python/paddle/distributed/fleet/fleet.py
Copy file name to clipboard
+2-2Lines changed: 2 additions & 2 deletions
diff --git a/‎python/paddle/distributed/fleet/launch.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion b/‎python/paddle/distributed/fleet/launch.py
Copy file name to clipboard
+1-1Lines changed: 1 addition & 1 deletion
@@ -352,13 +352,13 @@ def convert_src_tensor_2_dst_tensor(vars_list, src_attr_map, dst_attr_map):
         if src_attr_map is None or len(src_attr_map) == 0:
             return vars_list[0]
 
-        dst_strategys = {}
-        src_strategys = {}
+        dst_strategies = {}
+        src_strategies = {}
         tensors_dict = {}
 
         convert_tensor_dict = None
         for var_name in src_attr_map.keys():
-            assert var_name not in dst_strategys
+            assert var_name not in dst_strategies
             dist_vars = []
             for vars in vars_list:
                 if var_name in vars.keys():
@@ -367,13 +367,13 @@ def convert_src_tensor_2_dst_tensor(vars_list, src_attr_map, dst_attr_map):
                 continue
 
             if var_name in dst_attr_map and var_name in src_attr_map:
-                dst_strategys[var_name] = copy.deepcopy(dst_attr_map[var_name])
-                src_strategys[var_name] = copy.deepcopy(src_attr_map[var_name])
+                dst_strategies[var_name] = copy.deepcopy(dst_attr_map[var_name])
+                src_strategies[var_name] = copy.deepcopy(src_attr_map[var_name])
                 tensors_dict[var_name] = dist_vars
 
         if src_attr_map == dst_attr_map:
             return tensors_dict
-        converter = Converter(tensors_dict, src_strategys, dst_strategys)
+        converter = Converter(tensors_dict, src_strategies, dst_strategies)
         convert_tensor_dict = converter.convert()
 
         return convert_tensor_dict
 
@@ -202,7 +202,7 @@ def _update_op_dims_mapping_and_distoperatorimpl(
 
     updated = dist_op_container.update_dims_mapping(dist_op)
     changed = updated or changed
-    # TODO(ljz) remove the below code once we introduce general reshard to replace specifc distopimpls
+    # TODO(ljz) remove the below code once we introduce general reshard to replace specific distopimpls
     reverted = dist_op_container.mapping_to_dist_operator_impl(
         dist_op, original_op_dist_attr
     )
 
@@ -105,9 +105,9 @@ def convert(self, strict=True):
                 >>> import numpy as np
                 >>> from paddle.distributed.auto_parallel.static.converter import Converter
                 >>> complete_tensors = np.arange(4).reshape([2, 2])
-                >>> partitial_tensors = np.split(complete_tensors, 2, axis=0)
+                >>> partial_tensors = np.split(complete_tensors, 2, axis=0)
                 >>> name = "tmp_0"
-                >>> tensors_dict = {name: partitial_tensors}
+                >>> tensors_dict = {name: partial_tensors}
                 >>> strategy_1 = {
                 ...     name: {
                 ...         "process_shape": [2],
@@ -345,7 +345,7 @@ def slice_with_dist_attr(tensor, dist_attr):
     @staticmethod
     def merge(partition_tensor_list, tensor, partition_index, complete_shape):
         """
-        Merge partitial tensors to a complete.
+        Merge partial tensors to a complete.
 
         Returns:
             None
 
@@ -784,7 +784,7 @@ def comm_count(self):
             shape = None
             if self.op is not None:
                 vars = self.op.block.vars
-                # NOTE: The tensor communicated input_name is "X" in default. Otherwise, this function should be overrided
+                # NOTE: The tensor communicated input_name is "X" in default. Otherwise, this function should be overridden
                 try:
                     var_name = self.op.input("X")[0]
                 except:
 
@@ -98,18 +98,18 @@ def init_comm_cost(self, cluster=None):
         # should get from `cluster`
         BANDWIDTH = 32 * 1024 / 1000  # MB/ms, V100 PCIe
         num_ranks = len(self.ranks)
-        comm_volumn = np.prod(self.input_shape) * 4
+        comm_volume = np.prod(self.input_shape) * 4
 
         if 'allreduce' in self.comm_type:
-            self._cost = comm_volumn / (
+            self._cost = comm_volume / (
                 BANDWIDTH * num_ranks / (2 * (num_ranks - 1))
             )
         elif 'gather' in self.comm_type:
-            self._cost = comm_volumn / (BANDWIDTH * num_ranks / (num_ranks - 1))
+            self._cost = comm_volume / (BANDWIDTH * num_ranks / (num_ranks - 1))
         elif 'broadcast' in self.comm_type:
-            self._cost = comm_volumn / BANDWIDTH
+            self._cost = comm_volume / BANDWIDTH
         elif 'send' in self.comm_type or 'recv' in self.comm_type:
-            self._cost = comm_volumn / BANDWIDTH
+            self._cost = comm_volume / BANDWIDTH
         else:
             self._cost = 0
 
 
@@ -211,8 +211,8 @@ class ProgramHelper:
 
     def __init__(self, layer, loss_func, metrics, inputs_spec, labels_spec):
         # original model config information
-        # TODO(Aurelius84): Implenet append_backward and optimizer in ProxyLayer
-        # after distribute engine satisify basic condition.
+        # TODO(Aurelius84): Implement append_backward and optimizer in ProxyLayer
+        # after distribute engine satisfy basic condition.
         self.proxy_layer = ProxyLayer(layer, loss_func, metrics)
         self.inputs_spec = inputs_spec
         self.labels_spec = labels_spec
 
@@ -231,7 +231,7 @@ def _generate_backward(
         # NOTE(zhaoyinglia):
         # Guarantee the order of params_grads is same between dynamic mode and static mode
         # by making parameter_list equal to model.parameters(),
-        # because the order affact the result of ClipGradByGLobalNorm.
+        # because the order affect the result of ClipGradByGLobalNorm.
         # If parameter_list is not None, the order of params_grads is same with parameter_list.
         # If parameter_list is None, params_grads will be as prog.global_block().all_parameters().
         with program_guard(main_program, startup_program):
 
@@ -1498,7 +1498,7 @@ def sharding_configs(self):
             This configuration will affect the communication speed in sharding training, and should be an empirical value decided by your model size and network topology.
             Only enable when sharding_segment_strategy = segment_broadcast_MB. Default is 32.0 .
 
-            segment_anchors(list): list of anchors used to segment the program, which allows a finner control of program segmentation.
+            segment_anchors(list): list of anchors used to segment the program, which allows a finer control of program segmentation.
             this strategy is experimental by now. Only enable when sharding_segment_strategy = segment_anchors.
 
             sharding_degree(int, optional): specific the number of gpus within each sharding parallelism group; and sharding will be turn off if sharding_degree=1.  Default is 8.
 
@@ -1194,7 +1194,7 @@ def save_persistables(self, executor, dirname, main_program=None, mode=0):
 
             dirname(str, optional): The saving directory path.
                                 When you need to save the parameter to the memory, set it to None.
-            main_program(Program, optional): The program whose persistbale tensors will
+            main_program(Program, optional): The program whose persistable tensors will
                                              be saved. Default: None.
 
 
@@ -1419,7 +1419,7 @@ def amp_init(
                 ...         init_loss_scaling=128.0,
                 ...         use_dynamic_loss_scaling=True,
                 ...         use_pure_fp16=True)
-                ...     # If you don't use the default_startup_program(), you sholud pass
+                ...     # If you don't use the default_startup_program(), you should pass
                 ...     # your defined `startup_program` into `minimize`.
                 ...     optimizer.minimize(loss)
                 ...     exe.run(paddle.static.default_startup_program())
 
@@ -115,7 +115,7 @@ def _parse_args():
         "--backend",
         type=str,
         default=os.environ.get('PADDLE_DISTRI_BACKEND', 'auto'),
-        help="Specifize the backend, can be gloo|nccl|bkcl|auto|heter. "
+        help="Specify the backend, can be gloo|nccl|bkcl|auto|heter. "
         "Default value is auto which prefers nccl or bkcl.",
     )
     base_group.add_argument(
Original file line number	Diff line number	Diff line change
`@@ -202,7 +202,7 @@ def _update_op_dims_mapping_and_distoperatorimpl(`
`202`	`202`
`203`	`203`	`updated = dist_op_container.update_dims_mapping(dist_op)`
`204`	`204`	`changed = updated or changed`
`205`		`- # TODO(ljz) remove the below code once we introduce general reshard to replace specifc distopimpls`
	`205`	`+ # TODO(ljz) remove the below code once we introduce general reshard to replace specific distopimpls`
`206`	`206`	`reverted = dist_op_container.mapping_to_dist_operator_impl(`
`207`	`207`	`dist_op, original_op_dist_attr`
`208`	`208`	`)`
Original file line number	Diff line number	Diff line change
`@@ -115,7 +115,7 @@ def _parse_args():`
`115`	`115`	`"--backend",`
`116`	`116`	`type=str,`
`117`	`117`	`default=os.environ.get('PADDLE_DISTRI_BACKEND', 'auto'),`
`118`		`- help="Specifize the backend, can be gloo\|nccl\|bkcl\|auto\|heter. "`
	`118`	`+ help="Specify the backend, can be gloo\|nccl\|bkcl\|auto\|heter. "`
`119`	`119`	`"Default value is auto which prefers nccl or bkcl.",`
`120`	`120`	`)`
`121`	`121`	`base_group.add_argument(`