Fix

Author: co63oc

CMakeLists.txt

@@ -322,7 +322,7 @@ option(WITH_ARM "Compile PaddlePaddle with arm support" OFF)
 option(WITH_SW "Compile PaddlePaddle with sw support" OFF)
 option(WITH_MIPS "Compile PaddlePaddle with mips support" OFF)
 option(WITH_LOONGARCH "Compile PaddlePaddle with loongarch support" OFF)
-option(WITH_MUSL "Compile with musl libc instead of gblic" OFF)
+option(WITH_MUSL "Compile with musl libc instead of glibc" OFF)
 option(WITH_UNITY_BUILD "Compile with UnityBuild mode" OFF)
 option(WITH_STRIP "Strip so files of Whl packages" OFF)
 option(NEW_RELEASE_PYPI

python/paddle/distributed/auto_parallel/pipelining/stage.py

@@ -282,7 +282,7 @@ def _create_grad_send_info(
         def map_recv_to_send(a):
             # Note: we send gradients back to previous stage as long as in
             # forward it is a received input, regardless of whether it requires
-            # grad. It is up to the previous stage to disgard this gradient.
+            # grad. It is up to the previous stage to discard this gradient.
             if isinstance(a, _RecvInfo):
                 grad_send_info.append(a.source)
                 return a.source

python/paddle/distributed/auto_parallel/static/utils.py

@@ -373,7 +373,7 @@ def _coordinate2linear_idx(mesh_shape, coordinate):
     # that the processes in mesh are
     #    1. starts from 0
     #    2. continuous
-    # it will be wrong if ths above condition does not meet,
+    # it will be wrong if the above condition does not meet,
     # e.g. process_mesh = { process_groups = [7, 8, 9,10, 12, 13, 14, 15], mesh = [2, 4]}
     # if you want a more general mapping, you should use cartesian product
 

python/paddle/incubate/optimizer/functional/bfgs.py

@@ -81,7 +81,7 @@ def minimize_bfgs(
 
             - is_converge (bool): Indicates whether found the minimum within tolerance.
             - num_func_calls (int): number of objective function called.
-            - position (Tensor): the position of the last iteration. If the search converged, this value is the argmin of the objective function regrading to the initial position.
+            - position (Tensor): the position of the last iteration. If the search converged, this value is the argmin of the objective function regarding to the initial position.
             - objective_value (Tensor): objective function value at the `position`.
             - objective_gradient (Tensor): objective function gradient at the `position`.
             - inverse_hessian_estimate (Tensor): the estimate of inverse hessian at the `position`.

python/paddle/incubate/optimizer/functional/lbfgs.py

@@ -83,7 +83,7 @@ def minimize_lbfgs(
 
             - is_converge (bool): Indicates whether found the minimum within tolerance.
             - num_func_calls (int): number of objective function called.
-            - position (Tensor): the position of the last iteration. If the search converged, this value is the argmin of the objective function regrading to the initial position.
+            - position (Tensor): the position of the last iteration. If the search converged, this value is the argmin of the objective function regarding to the initial position.
             - objective_value (Tensor): objective function value at the `position`.
             - objective_gradient (Tensor): objective function gradient at the `position`.
 

python/paddle/incubate/optimizer/recompute.py

@@ -390,7 +390,7 @@ def _parse_backward(self):
                                     self._record_fetch_op(idx)
                                 )
 
-                        # should check the current used checkpoint is ths last fetch one
+                        # should check the current used checkpoint is the last fetch one
                         assert (
                             second_to_last_fetch_checkpoint == input_var
                         ), f"Current recompute segment should use [{second_to_last_fetch_checkpoint}] BUT got [{input_var}]"

python/paddle/nn/functional/flash_attention.py

@@ -1156,7 +1156,7 @@ def flash_attn_varlen_func(
             "FLAGS_flash_attn_version"
         ]
         == 3
-    ), "FLAGS_flash_attn_version is 2, conflits with flash_attn_varlen_func"
+    ), "FLAGS_flash_attn_version is 2, conflicts with flash_attn_varlen_func"
 
     assert (
         in_dynamic_or_pir_mode()
@@ -1650,7 +1650,7 @@ def flashmask_attention(
             If causal mode is disabled, Query at position i will only attend to keys between [i - window_size, i + window_size] or [i - window_size[0], i + window_size[1]].
         return_softmax_lse (bool): Whether to return the log-sum-exp of the softmax. Default is False.
         return_seed_offset (bool): Whether to return the random seed offset. Default is False.
-        fixed_seed_of fset(Tensor, optional): With fixed seed, offset for dropout mask.
+        fixed_seed_offset(Tensor, optional): With fixed seed, offset for dropout mask.
         rng_name (str): The name to select Generator.
         training (bool): Whether the module is in training mode. Default is True.
         name (str, optional): Name of the operation. Default is None. Normally, users do not need to set this property.

setup.py

@@ -273,7 +273,7 @@ def run(self):
         egg_info.run(self)
 
 
-# class Installlib is rewritten to add header files to .egg/paddle
+# class InstallLib is rewritten to add header files to .egg/paddle
 class InstallLib(install_lib):
     def run(self):
         self.build()

test/ir/inference/test_trt_convert_cross_multihead_matmul.py

@@ -293,7 +293,7 @@ def teller1(program_config, predictor_config):
         self.add_skip_case(
             teller1,
             SkipReasons.TRT_NOT_IMPLEMENTED,
-            "TThe cross attention trt oss plugin do not support static shape yet",
+            "The cross attention trt oss plugin do not support static shape yet",
         )
 
         def teller2(program_config, predictor_config):

test/ir/inference/test_trt_convert_flash_multihead_matmul.py

@@ -294,7 +294,7 @@ def teller1(program_config, predictor_config):
         self.add_skip_case(
             teller1,
             SkipReasons.TRT_NOT_IMPLEMENTED,
-            "TThe flash attention trt oss plugin do not support static shape yet",
+            "The flash attention trt oss plugin do not support static shape yet",
         )
 
         def teller2(program_config, predictor_config):
@@ -603,7 +603,7 @@ def teller1(program_config, predictor_config):
         self.add_skip_case(
             teller1,
             SkipReasons.TRT_NOT_IMPLEMENTED,
-            "TThe flash attention trt oss plugin do not support static shape yet",
+            "The flash attention trt oss plugin do not support static shape yet",
         )
 
         def teller2(program_config, predictor_config):