Fix some typos (p_nrom, inferece, etc.) (#61724)

co63oc · web-flow · commit 5a41af3daa38 · 2024-02-18T15:11:19.000+08:00
diff --git a/test/legacy_test/test_zero_dim_tensor.py b/test/legacy_test/test_zero_dim_tensor.py
@@ -2927,7 +2927,7 @@ def test_linalg_norm(self):
         self.assertTrue(x_1.grad.shape, [24])
 
         # 1D input, p = 1 ,axis = None,
-        # using p_nrom, as_vector = True
+        # using p_norm, as_vector = True
         x_2 = paddle.arange(24, dtype="float32") - 12
         x_2.stop_gradient = False
         out_2 = paddle.linalg.norm(x_2, p=1)
@@ -2938,7 +2938,7 @@ def test_linalg_norm(self):
         self.assertEqual(x_2.grad.shape, [24])
 
         # 1D input, p = 1 ,axis = 0,
-        # using p_nrom, as_vector = False
+        # using p_norm, as_vector = False
         x_2_p = paddle.arange(24, dtype="float32") - 12
         x_2_p.stop_gradient = False
         out_2_p = paddle.linalg.norm(x_2_p, p=1, axis=0)
@@ -2949,7 +2949,7 @@ def test_linalg_norm(self):
         self.assertEqual(x_2_p.grad.shape, [24])
 
         # 1D input, p = fro ,axis = 0,
-        # using p_nrom, as_vector = False
+        # using p_norm, as_vector = False
         x_2_fro = paddle.arange(24, dtype="float32") - 12
         x_2_fro.stop_gradient = False
         out_2_fro = paddle.linalg.norm(x_2_fro, p="fro", axis=0)
@@ -5681,7 +5681,7 @@ def test_linalg_norm(self):
         self.assertEqual(res[1].shape, (24,))
 
         # 1D input, p = 1 ,axis = None,
-        # using p_nrom, as_vector = True
+        # using p_norm, as_vector = True
         x_2 = paddle.arange(24, dtype="float32") - 12
         x_2.stop_gradient = False
         out_2 = paddle.linalg.norm(x_2, p=1)
@@ -5693,7 +5693,7 @@ def test_linalg_norm(self):
         self.assertEqual(res[1].shape, (24,))
 
         # 1D input, p = 1 ,axis = 0,
-        # using p_nrom, as_vector = False
+        # using p_norm, as_vector = False
         x_2_p = paddle.arange(24, dtype="float32") - 12
         x_2_p.stop_gradient = False
         out_2_p = paddle.linalg.norm(x_2_p, p=1, axis=0)
@@ -5705,7 +5705,7 @@ def test_linalg_norm(self):
         self.assertEqual(res[1].shape, (24,))
 
         # 1D input, p = fro ,axis = 0,
-        # using p_nrom, as_vector = False
+        # using p_norm, as_vector = False
         x_2_fro = paddle.arange(24, dtype="float32") - 12
         x_2_fro.stop_gradient = False
         out_2_fro = paddle.linalg.norm(x_2_fro, p="fro", axis=0)
@@ -5717,7 +5717,7 @@ def test_linalg_norm(self):
         self.assertEqual(res[1].shape, (24,))
 
         # 2D input, p = 1, axis = [0, 1]
-        # using p_matrix_norm ,depends on  paddle.sum
+        # using p_matrix_norm, depends on paddle.sum
         x_3 = paddle.arange(24, dtype="float32").reshape([4, 6])
         x_3.stop_gradient = False
         out_3 = paddle.linalg.norm(x_3, p=1, axis=[0, 1])
@@ -5831,7 +5831,7 @@ def test_linalg_cond(self):
         self.assertEqual(res[0].shape, ())
         self.assertEqual(res[1].shape, (3, 3))
 
-        # p in (-2, 2)  depends on paddle.sum
+        # p in (-2, 2) depends on paddle.sum
         x6 = paddle.to_tensor([[1.0, 0, -1], [0, 1, 0], [1, 0, 1]])
         x6.stop_gradient = False
         out_2 = paddle.linalg.cond(x6, p=2)
diff --git a/test/legacy_test/testsuite.py b/test/legacy_test/testsuite.py
@@ -107,7 +107,7 @@ def create_var(block, name, np_list, var_proto, is_calc_ref=False):
         if name not in np_list:
             assert var_proto.intermediate, f"{name} not found"
         else:
-            # inferece the dtype from numpy value.
+            # inference the dtype from numpy value.
             np_value = np_list[name]
             if isinstance(np_value, tuple):
                 dtype = np_value[0].dtype
diff --git a/test/legacy_test/transformer_model.py b/test/legacy_test/transformer_model.py
@@ -60,7 +60,7 @@ def multi_head_attention(
 ):
     """
     Multi-Head Attention. Note that attn_bias is added to the logit before
-    computing softmax activiation to mask certain selected positions so that
+    computing softmax activation to mask certain selected positions so that
     they will not considered in attention weights.
     """
     if not (len(queries.shape) == len(keys.shape) == len(values.shape) == 3):
@@ -220,7 +220,7 @@ def positionwise_feed_forward(x, d_inner_hid, d_hid):
 
 def pre_post_process_layer(prev_out, out, process_cmd, dropout=0.0):
     """
-    Add residual connection, layer normalization and droput to the out tensor
+    Add residual connection, layer normalization and dropout to the out tensor
     optionally according to the value of process_cmd.
 
     This will be used before or after multi-head attention and position-wise
@@ -307,10 +307,10 @@ def encoder_layer(
 ):
     """The encoder layers that can be stacked to form a deep encoder.
 
-    This module consits of a multi-head (self) attention followed by
+    This module consists of a multi-head (self) attention followed by
     position-wise feed-forward networks and both the two components companied
     with the post_process_layer to add residual connection, layer normalization
-    and droput.
+    and dropout.
     """
     attn_output = multi_head_attention(
         enc_input,
diff --git a/test/mkldnn/test_elementwise_add_bf16_mkldnn_op.py b/test/mkldnn/test_elementwise_add_bf16_mkldnn_op.py
@@ -51,7 +51,7 @@ def generate_data(self):
     def test_check_output(self):
         self.check_output_with_place(core.CPUPlace(), check_pir_onednn=True)
 
-    # elementwise_add grad (no braodcasting) is just passing upper gradients to either X or Y or both
+    # elementwise_add grad (no broadcasting) is just passing upper gradients to either X or Y or both
     def test_check_grad_normal(self):
         self.check_grad_with_place(
             core.CPUPlace(),
diff --git a/test/mkldnn/test_elementwise_add_onednn_op_rare_shape.py b/test/mkldnn/test_elementwise_add_onednn_op_rare_shape.py
@@ -28,7 +28,7 @@
     "GPU is not supported",
 )
 # Special cases for swin transformer, will ignore grad check
-class TestOneDNNlementwiseAddSrcDifferentShape(TestElementwiseAddOp):
+class TestOneDNNElementwiseAddSrcDifferentShape(TestElementwiseAddOp):
     def init_kernel_type(self):
         self.use_mkldnn = True
         self.check_pir_onednn = True
diff --git a/test/mkldnn/test_elementwise_mul_onednn_op.py b/test/mkldnn/test_elementwise_mul_onednn_op.py
@@ -127,7 +127,7 @@ def test_check_grad_ingore_x(self):
 
 
 @skip_check_grad_ci(
-    reason="oneDNN's int8 elementwise_ops don't implemend grad kernel."
+    reason="oneDNN's int8 elementwise_ops don't implement grad kernel."
 )
 class TestInt8(ElementwiseMulOp):
     def init_kernel_type(self):
diff --git a/test/mkldnn/test_fusion_lstm_bf16_mkldnn_op.py b/test/mkldnn/test_fusion_lstm_bf16_mkldnn_op.py
@@ -57,7 +57,7 @@ def setUp(self):
         bs = len(self.lod[0])
 
         # fp32 X input for reference implementation and
-        # corressponding bf16 data as input to LSTM oneDNN bf16 kernel
+        # corresponding bf16 data as input to LSTM oneDNN bf16 kernel
         x = np.random.normal(size=(T, self.M)).astype('float32')
 
         x_bf16 = convert_float_to_uint16(x)
diff --git a/test/prim/prim/vjp/eager/test_comp_eager_matmul_double_grad.py b/test/prim/prim/vjp/eager/test_comp_eager_matmul_double_grad.py
@@ -412,7 +412,7 @@ def desired(primal0, primal1, trans_0, trans_1, dtype_):
         ),
     ],
 )
-class TestMatmulTribleGradComp(unittest.TestCase):
+class TestMatmulTripleGradComp(unittest.TestCase):
     @classmethod
     def setUpClass(cls):
         cls.primal0 = cls.primal0.astype(cls.dtype)
diff --git a/test/rnn/rnn_numpy.py b/test/rnn/rnn_numpy.py
@@ -366,10 +366,10 @@ def concat_states(states, bidirectional=False, state_components=1):
         return np.stack(flatten(states))
     else:
         states = flatten(states)
-        componnets = []
+        components = []
         for i in range(state_components):
-            componnets.append(states[i::state_components])
-        return [np.stack(item) for item in componnets]
+            components.append(states[i::state_components])
+        return [np.stack(item) for item in components]
 
 
 class RNN(LayerMixin):
diff --git a/test/xpu/test_zero_dim_tensor_xpu.py b/test/xpu/test_zero_dim_tensor_xpu.py
@@ -2382,7 +2382,7 @@ def test_linalg_norm(self):
         self.assertTrue(x_1.grad.shape, [24])
 
         # 1D input, p = 1 ,axis = None,
-        # using p_nrom, as_vector = True
+        # using p_norm, as_vector = True
         x_2 = paddle.arange(24, dtype="float32") - 12
         x_2.stop_gradient = False
         out_2 = paddle.linalg.norm(x_2, p=1)
@@ -2393,7 +2393,7 @@ def test_linalg_norm(self):
         self.assertEqual(x_2.grad.shape, [24])
 
         # 1D input, p = 1 ,axis = 0,
-        # using p_nrom, as_vector = False
+        # using p_norm, as_vector = False
         x_2_p = paddle.arange(24, dtype="float32") - 12
         x_2_p.stop_gradient = False
         out_2_p = paddle.linalg.norm(x_2_p, p=1, axis=0)
@@ -2404,7 +2404,7 @@ def test_linalg_norm(self):
         self.assertEqual(x_2_p.grad.shape, [24])
 
         # 1D input, p = fro ,axis = 0,
-        # using p_nrom, as_vector = False
+        # using p_norm, as_vector = False
         x_2_fro = paddle.arange(24, dtype="float32") - 12
         x_2_fro.stop_gradient = False
         out_2_fro = paddle.linalg.norm(x_2_fro, p="fro", axis=0)

Original file line number	Diff line number	Diff line change
`@@ -28,7 +28,7 @@`
`28`	`28`	`"GPU is not supported",`
`29`	`29`	`)`
`30`	`30`	`# Special cases for swin transformer, will ignore grad check`
`31`		`-class TestOneDNNlementwiseAddSrcDifferentShape(TestElementwiseAddOp):`
	`31`	`+class TestOneDNNElementwiseAddSrcDifferentShape(TestElementwiseAddOp):`
`32`	`32`	`def init_kernel_type(self):`
`33`	`33`	`self.use_mkldnn = True`
`34`	`34`	`self.check_pir_onednn = True`
Original file line number	Diff line number	Diff line change
`@@ -127,7 +127,7 @@ def test_check_grad_ingore_x(self):`
`127`	`127`
`128`	`128`
`129`	`129`	`@skip_check_grad_ci(`
`130`		`- reason="oneDNN's int8 elementwise_ops don't implemend grad kernel."`
	`130`	`+ reason="oneDNN's int8 elementwise_ops don't implement grad kernel."`
`131`	`131`	`)`
`132`	`132`	`class TestInt8(ElementwiseMulOp):`
`133`	`133`	`def init_kernel_type(self):`
Original file line number	Diff line number	Diff line change
`@@ -412,7 +412,7 @@ def desired(primal0, primal1, trans_0, trans_1, dtype_):`
`412`	`412`	`),`
`413`	`413`	`],`
`414`	`414`	`)`
`415`		`-class TestMatmulTribleGradComp(unittest.TestCase):`
	`415`	`+class TestMatmulTripleGradComp(unittest.TestCase):`
`416`	`416`	`@classmethod`
`417`	`417`	`def setUpClass(cls):`
`418`	`418`	`cls.primal0 = cls.primal0.astype(cls.dtype)`