!674 [pylint] clean pylint warning

Merge pull request !674 from jinyaohui/clean_pylint_0425

example/yolov3_coco2017/dataset.py

@@ -18,8 +18,8 @@ from __future__ import division
 
 import os
 import numpy as np
-from PIL import Image
 from matplotlib.colors import rgb_to_hsv, hsv_to_rgb
+from PIL import Image
 import mindspore.dataset as de
 from mindspore.mindrecord import FileWriter
 import mindspore.dataset.transforms.vision.c_transforms as C

mindspore/_akg/__init__.py

@@ -16,6 +16,9 @@
 from __future__ import absolute_import as _abs
 import sys
 import os
+from .op_build import op_build
+from .message import compilewithjson
+
 
 def AKGAddPath():
     """_akg add path."""
@@ -58,6 +61,3 @@ class AKGMetaPathLoader:
 
 
 sys.meta_path.insert(0, AKGMetaPathFinder())
-
-from .op_build import op_build
-from .message import compilewithjson

mindspore/nn/optim/ftrl.py

@@ -14,7 +14,6 @@
 # ============================================================================
 """FTRL"""
 from mindspore.ops import functional as F, composite as C, operations as P
-from mindspore.common.initializer import initializer
 from mindspore.common.parameter import Parameter
 from mindspore.common import Tensor
 import mindspore.common.dtype as mstype
@@ -23,6 +22,8 @@ from mindspore._checkparam import Rel
 from .optimizer import Optimizer, apply_decay, grad_scale
 
 ftrl_opt = C.MultitypeFuncGraph("ftrl_opt")
+
+
 @ftrl_opt.register("Function", "Tensor", "Number", "Number", "Number", "Tensor", "Tensor", "Tensor", "Tensor")
 def _tensor_run_opt(opt, learning_rate, l1, l2, lr_power, linear, gradient, weight, moment):
     """Apply ftrl optimizer to the weight parameter."""
@@ -30,8 +31,10 @@ def _tensor_run_opt(opt, learning_rate, l1, l2, lr_power, linear, gradient, weig
     success = F.depend(success, opt(weight, moment, linear, gradient, learning_rate, l1, l2, lr_power))
     return success
 
+
 def _check_param(initial_accum, learning_rate, lr_power, l1, l2, use_locking, loss_scale=1.0, weight_decay=0.0,
                  prim_name=None):
+    """Check param."""
     validator.check_value_type("initial_accum", initial_accum, [float], prim_name)
     validator.check_number("initial_accum", initial_accum, 0.0, Rel.GE, prim_name)
 
@@ -118,5 +121,6 @@ class FTRL(Optimizer):
         if self.reciprocal_scale != 1.0:
             grads = self.hyper_map(F.partial(grad_scale, self.reciprocal_scale), grads)
         lr = self.learning_rate
-        success = self.hyper_map(F.partial(ftrl_opt, self.opt, lr, self.l1, self.l2, self.lr_power), linear, grads, params, moments)
+        success = self.hyper_map(F.partial(ftrl_opt, self.opt, lr, self.l1, self.l2, self.lr_power),
+                                 linear, grads, params, moments)
         return success

mindspore/train/amp.py

@@ -41,6 +41,7 @@ class OutputTo16(nn.Cell):
 
 
 def _do_keep_batchnorm_fp32(network):
+    """Do keep batchnorm fp32."""
     cells = network.name_cells()
     change = False
     for name in cells:
@@ -68,6 +69,7 @@ _config_level = {
 
 
 def _check_kwargs(key_words):
+    """Check kwargs."""
     for arg in key_words:
         if arg not in ['cast_model_type', 'keep_batchnorm_fp32', 'loss_scale_manager']:
             raise  ValueError(f"Unsupported arg '{arg}'")
@@ -84,6 +86,7 @@ def _check_kwargs(key_words):
 
 
 def _add_loss_network(network, loss_fn, cast_model_type):
+    """Add loss network."""
     class WithLossCell(nn.Cell):
         "Wrap loss for amp. Cast network output back to float32"
 

mindspore/train/callback.py

@@ -683,6 +683,7 @@ class LossMonitor(Callback):
 
 
 class TimeMonitor(Callback):
+    """Time Monitor."""
     def __init__(self, data_size):
         super(TimeMonitor, self).__init__()
         self.data_size = data_size
@@ -701,4 +702,3 @@ class TimeMonitor(Callback):
     def step_end(self, run_context):
         step_mseconds = (time.time() - self.step_time) * 1000
         print('step time', step_mseconds, flush=True)
-

tests/mindspore_test_framework/apps/test_bert_parts.py

@@ -174,8 +174,7 @@ test_sets = [
                                                   embedding_shape=[1, 128, 768],
                                                   use_one_hot_embeddings=True,
                                                   initializer_range=0.02), 1, 1), {
-                      'init_param_with': lambda shp: np.ones(shp).astype(np.float32)
+                      'init_param_with': lambda shp: np.ones(shp).astype(np.float32)}),
-                  }),
         'desc_inputs': [input_ids],
         'desc_bprop': [[128]]}),
     ('EmbeddingLookup_multi_outputs_init_param', {
@@ -184,8 +183,7 @@ test_sets = [
                                   embedding_shape=[1, 128, 768],
                                   use_one_hot_embeddings=False,
                                   initializer_range=0.02), {
-                      'init_param_with': lambda shp: np.ones(shp).astype(np.float32)
+                      'init_param_with': lambda shp: np.ones(shp).astype(np.float32)}),
-                  }),
         'desc_inputs': [input_ids],
         'desc_bprop': [[1, 128, 768], [128]]}),
     ('EmbeddingLookup_multi_outputs_grad_with_no_sens', {
@@ -194,8 +192,7 @@ test_sets = [
                                   embedding_shape=[1, 128, 768],
                                   use_one_hot_embeddings=False,
                                   initializer_range=0.02), {
-                      'init_param_with': lambda shp: np.ones(shp).astype(np.float32)
+                      'init_param_with': lambda shp: np.ones(shp).astype(np.float32)}),
-                  }),
         'desc_inputs': [input_ids]}),
     ('GetMaskedLMOutput_grad_with_no_sens', {
         'block': GetMaskedLMOutput(BertConfig(batch_size=1)),