Fix typos Pleas Please (#72093)

co63oc · web-flow · commit d9c4aa31eba0 · 2025-04-07T14:54:48.000+08:00
diff --git a/paddle/fluid/eager/api/generated/eager_generated/backwards/scale_node.cc b/paddle/fluid/eager/api/generated/eager_generated/backwards/scale_node.cc
@@ -139,7 +139,7 @@ void ScaleAPI(const paddle::Tensor& x,
       PADDLE_THROW(common::errors::Fatal(
           "Cannot convert device_context to CUDADeviceContext."
           "This indicates backend mismatch."
-          "Pleas double check your expected place"));
+          "Please double check your expected place"));
     }
     ScaleDeviceDispatch<phi::GPUContext>(*dense_tensor.get(),
                                          *dev_ctx,
diff --git a/paddle/fluid/eager/to_static/run_program_op_func.h b/paddle/fluid/eager/to_static/run_program_op_func.h
@@ -364,7 +364,7 @@ inline void pir_run_program_ad_func(
 
     grad_node->SetFwdParams(params_tmp);
 
-    grad_node->SetStepScope(step_scope);  // just for set useable.
+    grad_node->SetStepScope(step_scope);  // just for set usable.
 
     grad_node->SetGradOutMeta(x, /*slot id*/ 0);
     grad_node->SetGradOutMeta(params, /*slot id*/ 1);
diff --git a/python/paddle/distributed/checkpoint/load_state_dict.py b/python/paddle/distributed/checkpoint/load_state_dict.py
@@ -540,7 +540,7 @@ def load_state_dict(
         path(str): The directory to load checkpoint files.
         process_group(paddle.distributed.collective.Group): ProcessGroup to be used for cross-rank synchronization. Use the default process group which contains all cards.
         coordinator_rank(int): The rank used to coordinate the checkpoint. Rank0 is used by default.
-        unique_id(int): The unique id of ckeckpoint, used to distinguish between different checkpoint versions. Default is None, in which case the id the max id of given path, and the newest version checkpoint is loaded.
+        unique_id(int): The unique id of checkpoint, used to distinguish between different checkpoint versions. Default is None, in which case the id the max id of given path, and the newest version checkpoint is loaded.
         offload(bool): Whether to offload the checkpoint data from GPU to CPU.
         mw_name_compatibility(bool): Enable name compatibility between dynamic and static graph semi-automatic parallel. Default is True.
     Example:
@@ -836,7 +836,7 @@ def load_merged_state_dict(
     Args:
         path(str): The directory to load checkpoint files.
         prefix(str): The flat_mapping prefix of state_dict key. e.g., 'model', Default None.
-        unique_id(int): The unique id of ckeckpoint, used to distinguish between different checkpoint versions. Default is None, in which case the id the max id of given path, and the newest version checkpoint is loaded.
+        unique_id(int): The unique id of checkpoint, used to distinguish between different checkpoint versions. Default is None, in which case the id the max id of given path, and the newest version checkpoint is loaded.
         offload(bool): Whether to offload the checkpoint data from GPU to CPU, set to True if GPU memory is not enough.
 
     Returns:
diff --git a/python/paddle/distributed/checkpoint/save_state_dict.py b/python/paddle/distributed/checkpoint/save_state_dict.py
@@ -148,7 +148,7 @@ def save_state_dict(
         path(str): The directory to save state_dict.
         process_group(paddle.distributed.collective.Group): ProcessGroup to be used for cross-rank synchronization. Use the default process group which contains all cards.
         coordinator_rank(int): The rank used to save non distributed values. Rank 0 is used by default.
-        unique_id(int): The unique id of ckeckpoint, used to distinguish between different checkpoint versions. Default is None, in which case the id 0 when save for the first time and increased by 1 each time when calling save_state_dict in the same path. If unique_id is given and there is already checkpoint with the same unique_id, it will be overrited.
+        unique_id(int): The unique id of checkpoint, used to distinguish between different checkpoint versions. Default is None, in which case the id 0 when save for the first time and increased by 1 each time when calling save_state_dict in the same path. If unique_id is given and there is already checkpoint with the same unique_id, it will be overrited.
         async_save(bool): Async save the state_dict, default is False.
 
     Examples:
diff --git a/python/paddle/distributed/fleet/launch_utils.py b/python/paddle/distributed/fleet/launch_utils.py
@@ -276,7 +276,7 @@ def get_cluster(
         # when use paddlecloud, endpoints may > devices_per_proc(user_defined)
         assert len(cur_node_endpoints) >= len(
             devices_per_proc
-        ), "current trainer_endpoints size should be greater equal than acclerators size."
+        ), "current trainer_endpoints size should be greater equal than accelerators size."
         for i in range(len(devices_per_proc)):
             trainer = Trainer()
             if device_mode == DeviceMode.GPU:
diff --git a/python/paddle/metric/metrics.py b/python/paddle/metric/metrics.py
@@ -56,7 +56,7 @@ class Metric(metaclass=abc.ABCMeta):
     Advanced usage for :code:`compute`:
 
     Metric calculation can be accelerated by calculating metric states
-    from model outputs and labels by build-in operators not by Python/NumPy
+    from model outputs and labels by built-in operators not by Python/NumPy
     in :code:`compute`, metric states will be fetched as NumPy array and
     call :code:`update` with states in NumPy format.
     Metric calculated as follows (operations in Model and Metric are
diff --git a/python/paddle/optimizer/lr.py b/python/paddle/optimizer/lr.py
@@ -2131,7 +2131,7 @@ class CyclicLR(LRScheduler):
 
     It has been proposed in `Cyclic Learning Rates for Training Neural Networks <https://arxiv.org/abs/1506.01186>`_.
 
-    According to the paper, the cyclic learning rate schedule has three build-in scale methods:
+    According to the paper, the cyclic learning rate schedule has three built-in scale methods:
 
     * "triangular": A basic triangular cycle without any amplitude scaling.
     * "triangular2": A basic triangular cycle that reduce initial amplitude by half each cycle.
@@ -2154,7 +2154,7 @@ class CyclicLR(LRScheduler):
         mode (str, optional): one of 'triangular', 'triangular2' or 'exp_range'.
             If scale_fn is specified, this argument will be ignored. Default: 'triangular'
         exp_gamma (float): Constant in 'exp_range' scaling function: exp_gamma**iterations. Used only when mode = 'exp_range'. Default: 1.0
-        scale_fn (function, optional): A custom scaling function, which is used to replace three build-in methods.
+        scale_fn (function, optional): A custom scaling function, which is used to replace three built-in methods.
             It should only have one argument. For all x >= 0, 0 <= scale_fn(x) <= 1.
             If specified, then 'mode' will be ignored. Default: None
         scale_mode (str, optional): One of 'cycle' or 'iterations'. Defines whether scale_fn is evaluated on cycle

Original file line number	Diff line number	Diff line change
`@@ -139,7 +139,7 @@ void ScaleAPI(const paddle::Tensor& x,`
`139`	`139`	`PADDLE_THROW(common::errors::Fatal(`
`140`	`140`	`"Cannot convert device_context to CUDADeviceContext."`
`141`	`141`	`"This indicates backend mismatch."`
`142`		`- "Pleas double check your expected place"));`
	`142`	`+ "Please double check your expected place"));`
`143`	`143`	`}`
`144`	`144`	`ScaleDeviceDispatch<phi::GPUContext>(*dense_tensor.get(),`
`145`	`145`	`*dev_ctx,`