[CodeStyle][UP008] remove super call with parameters

paddle/fluid/imperative/README.md

@@ -14,7 +14,7 @@ https://www.tensorflow.org/guide/eager
 
 ## API
 ```python
-class Layer(object):
+class Layer:
 
   def __call__(inputs):
     # build some parameter once.
@@ -49,7 +49,7 @@ Longer term.
 ```python
 
 # Parent class.
-class PyVarBase(object):
+class PyVarBase:
   pass
 
 # Current python variable.
@@ -139,7 +139,7 @@ TODO
 import paddle
 class MyLayer(fluid.imperative.Layer):
     def __init__(self):
-        super(MyLayer, self).__init__()
+        super().__init__()
 
     def forward(self, inputs):
         x = fluid.layers.relu(inputs)
@@ -150,7 +150,7 @@ class MyLayer(fluid.imperative.Layer):
 
 class MyPyLayer(fluid.imperative.PyLayer):
     def __init__(self):
-        super(MyPyLayer, self).__init__()
+        super().__init__()
 
     @staticmethod
     def forward(inputs):
@@ -172,7 +172,7 @@ with fluid.imperative.guard():
 
 class MLP(fluid.Layer):
     def __init__(self, input_size):
-        super(MLP, self).__init__()
+        super().__init__()
         self._linear1 = Linear(input_size,
                        3,
                        fluid.ParamAttr(

pyproject.toml

@@ -35,7 +35,9 @@ select = [
     # Pyupgrade
     "UP001",
     "UP003",
+    "UP004",
     "UP007",
+    "UP008",
     "UP010",
     "UP011",
     "UP013",

python/paddle/device/__init__.py

@@ -523,7 +523,7 @@ def get_available_custom_device():
     return core.get_available_custom_device()
 
 
-class Event(object):
+class Event:
     '''
     A device event wrapper around StreamBase.
     Parameters:
@@ -668,7 +668,7 @@ def __repr__(self):
         return self.event_base
 
 
-class Stream(object):
+class Stream:
     '''
     A device stream wrapper around StreamBase.
     Parameters:
@@ -840,7 +840,7 @@ def _as_parameter_(self):
 
     def __eq__(self, o):
         if isinstance(o, Stream):
-            return super(Stream, self).__eq__(o)
+            return super().__eq__(o)
         return False
 
     def __hash__(self):
@@ -936,7 +936,7 @@ def set_stream(stream):
     return prev_stream
 
 
-class stream_guard(object):
+class stream_guard:
     '''
     Notes:
         This API only supports dynamic graph mode currently.

python/paddle/distributed/fleet/base/runtime_factory.py

@@ -17,7 +17,7 @@
 __all__ = []
 
 
-class RuntimeFactory(object):
+class RuntimeFactory:
     def __init__(self):
         pass
 

python/paddle/distributed/fleet/recompute/recompute.py

@@ -379,7 +379,7 @@ class Naive_fc_net(paddle.nn.Layer):
                 def __init__(self, input_size=10,
                             recompute_blocks=[1, 3],
                             recompute_kwargs={}):
-                    super(Naive_fc_net, self).__init__()
+                    super().__init__()
                     self.recompute_blocks = recompute_blocks
                     self.recompute_kwargs = recompute_kwargs
                     self.runfunc0 = get_fc_block(0, input_size, is_last=False)

python/paddle/distributed/passes/auto_parallel_bf16.py

@@ -45,7 +45,7 @@
 world_process_group = get_world_process_group()
 
 
-class BF16State(object):
+class BF16State:
     def __init__(self, block):
         self._block: Block = block
         self._op_bf16_dict = {}

python/paddle/distributed/passes/auto_parallel_sharding.py

@@ -1787,7 +1787,7 @@ def group_param(sharding_info, fuse_size):
     return group_to_param_map, param_to_group_map
 
 
-class ShardingInfo(object):
+class ShardingInfo:
     def __init__(self, group, rank, params_grads, partition_algor):
         self.group = group
         self.params_grads = dict([(p.name, (p, g)) for p, g in params_grads])
@@ -1869,7 +1869,7 @@ def get_param_grad(self, param_name):
         return self.params_grads.get(param_name, None)
 
 
-class VarGroup(object):
+class VarGroup:
     def __init__(self, max_size):
         self.max_siez = max_size
         self.dtype = None

python/paddle/fluid/tests/unittests/collective/fleet/dygraph_group_sharded_stage3.py

@@ -61,7 +61,7 @@ def forward(self, inputs):
 
 class Encoder(paddle.nn.Layer):
     def __init__(self, encoder):
-        super(Encoder, self).__init__()
+        super().__init__()
         self.first_stage = paddle.nn.Linear(1024, 1024)
         self.encoder = encoder
 
@@ -73,7 +73,7 @@ def forward(self, x):
 
 class Decoder(paddle.nn.Layer):
     def __init__(self, decoder):
-        super(Decoder, self).__init__()
+        super().__init__()
         self.decoder = decoder
         self.final_stage = paddle.nn.Linear(1024, 1024)
         self.group_norm = paddle.nn.GroupNorm(64, 1024)
@@ -87,7 +87,7 @@ def forward(self, x):
 
 class SpecialModel(paddle.nn.Layer):
     def __init__(self):
-        super(SpecialModel, self).__init__()
+        super().__init__()
         self.shared = paddle.nn.Linear(1024, 1024, bias_attr=False)
         self.encoder = Encoder(self.shared)
         self.decoder = Decoder(self.shared)

python/paddle/fluid/tests/unittests/collective/fleet/dygraph_group_sharded_stage3_eager.py

@@ -25,7 +25,7 @@
 
 class Model(nn.Layer):
     def __init__(self):
-        super(Model, self).__init__()
+        super().__init__()
         self.first_stage = nn.Linear(4096, 4096, bias_attr=False)
         self.center_stage = nn.Linear(4096, 4096)
         self.center_stage.weight.stop_gradient = True

python/paddle/fluid/tests/unittests/collective/fleet/dygraph_save_for_auto_infer.py

@@ -78,7 +78,7 @@ def __init__(
             ),
             LayerDesc(Linear, in_features=linear_size, out_features=10),
         ]
-        super(MLP_pipe, self).__init__(
+        super().__init__(
             desc,
             num_stages=2,
             loss_fn=paddle.nn.CrossEntropyLoss(),
@@ -94,7 +94,7 @@ def __init__(
         param_attr=None,
         bias_attr=None,
     ):
-        super(MLP_Hybrid, self).__init__()
+        super().__init__()
         self.embedding = VocabParallelEmbedding(embedding_size, linear_size)
         self._linear1 = RowParallelLinear(
             linear_size, linear_size, has_bias=True, input_is_parallel=True
@@ -129,7 +129,7 @@ def __init__(
         param_attr=None,
         bias_attr=None,
     ):
-        super(MLP, self).__init__()
+        super().__init__()
         self.embedding = paddle.nn.Embedding(embedding_size, linear_size)
         self._linear1 = Linear(linear_size, linear_size)
         self._linear2 = Linear(linear_size, linear_size)

python/paddle/fluid/tests/unittests/collective/fleet/fused_attention_pass_with_mp.py

@@ -35,7 +35,7 @@ def __init__(
         pre_ln=True,
         attn_dropout=True,
     ):
-        super(MultiHeadAttentionWithMP, self).__init__()
+        super().__init__()
         self.embed_dim = embed_dim
         self.kdim = embed_dim
         self.vdim = embed_dim

python/paddle/fluid/tests/unittests/collective/fleet/test_dygraph_recompute_for_eager.py

@@ -23,7 +23,7 @@
 
 class Model(paddle.nn.Layer):
     def __init__(self, block_idx, input_size, is_last=False):
-        super(Model, self).__init__()
+        super().__init__()
         block_name = "block_" + str(block_idx)
         self.block = paddle.nn.Sequential(
             (

python/paddle/fluid/tests/unittests/distributed_passes/test_dist_fuse_gemm_epilogue_pass.py

@@ -37,7 +37,7 @@ def verify_op_count(op_types, op_name, target_count):
 
 class MultiFCLayer(nn.Layer):
     def __init__(self, hidden, Activation):
-        super(MultiFCLayer, self).__init__()
+        super().__init__()
         self.linear1 = paddle.nn.Linear(hidden, 4 * hidden)
         self.linear2 = paddle.nn.Linear(4 * hidden, hidden)
         self.linear3 = paddle.nn.Linear(hidden, hidden)

python/paddle/fluid/tests/unittests/dygraph_to_static/test_backward_without_params.py

@@ -21,7 +21,7 @@
 
 class Net(paddle.nn.Layer):
     def __init__(self):
-        super(Net, self).__init__()
+        super().__init__()
 
     @paddle.jit.to_static
     def forward(self, x):

python/paddle/fluid/tests/unittests/dygraph_to_static/test_cinn.py

@@ -21,7 +21,7 @@
 
 class Net(paddle.nn.Layer):
     def __init__(self):
-        super(Net, self).__init__()
+        super().__init__()
         self.relu = paddle.nn.functional.relu
         self.fc = paddle.nn.Linear(4, 4)
 

python/paddle/fluid/tests/unittests/dygraph_to_static/test_cinn_prim.py

@@ -29,7 +29,7 @@ def apply_to_static(net, use_cinn):
 
 class PrimeNet(paddle.nn.Layer):
     def __init__(self):
-        super(PrimeNet, self).__init__()
+        super().__init__()
         self.fc = paddle.nn.Linear(4, 4)
 
     def forward(self, x):

python/paddle/fluid/tests/unittests/dygraph_to_static/test_cinn_prim_gelu.py

@@ -42,7 +42,7 @@ def generate_data(shape, dtype="float32"):
 
 class PrimeNet(paddle.nn.Layer):
     def __init__(self, approximate):
-        super(PrimeNet, self).__init__()
+        super().__init__()
         self.fc = paddle.nn.Linear(4, 4)
         self.approximate = approximate
 

python/paddle/fluid/tests/unittests/dygraph_to_static/test_cinn_prim_layer_norm.py

@@ -43,7 +43,7 @@ def apply_to_static(net, use_cinn):
 
 class PrimeNet(paddle.nn.Layer):
     def __init__(self):
-        super(PrimeNet, self).__init__()
+        super().__init__()
         self.fc = paddle.nn.Linear(64, 64)
 
     def forward(self, x, w, b):

python/paddle/fluid/tests/unittests/dygraph_to_static/test_cinn_prim_mean.py

@@ -46,7 +46,7 @@ class PrimeNet(
     paddle.nn.Layer,
 ):
     def __init__(self):
-        super(PrimeNet, self).__init__()
+        super().__init__()
         self.fc = paddle.nn.Linear(4, 4)
 
     def forward(self, x):

python/paddle/fluid/tests/unittests/dygraph_to_static/test_convert_call.py

@@ -218,7 +218,7 @@ def set_func(self):
 
 class NotToStaticHelper(paddle.nn.Layer):
     def __init__(self):
-        super(NotToStaticHelper, self).__init__()
+        super().__init__()
 
     def sum(self, x):
         if x.shape[0] > 1:

python/paddle/fluid/tests/unittests/dygraph_to_static/test_program_translator.py

@@ -335,7 +335,7 @@ def func(self, x):
 
 class Net2:
     def __init__(self):
-        super(Net2, self).__init__()
+        super().__init__()
         self.layer1 = paddle.nn.Linear(10, 10)
 
     def forward(self, data):

python/paddle/fluid/tests/unittests/dygraph_to_static/test_unuseful_inputs.py

@@ -33,7 +33,7 @@ def apply_to_static(support_to_static, model, image_shape=None):
 
 class Layer0(nn.Layer):
     def __init__(self, level):
-        super(Layer0, self).__init__()
+        super().__init__()
         self._linear1 = nn.Linear(10, 5)
         self._linear2 = nn.Linear(10, 5)
         self.layer1 = Layer1(level)
@@ -51,7 +51,7 @@ def forward(self, x):
 
 class Layer1(nn.Layer):
     def __init__(self, level):
-        super(Layer1, self).__init__()
+        super().__init__()
         self.level = level
         self._linear = nn.Linear(5, 2)
 

python/paddle/fluid/tests/unittests/ipu/test_modelruntime_ipu.py

@@ -23,7 +23,7 @@
 
 class SimpleLayer(paddle.nn.Layer):
     def __init__(self):
-        super(SimpleLayer, self).__init__()
+        super().__init__()
         self.conv = paddle.nn.Conv2D(
             in_channels=3, out_channels=1, kernel_size=2, stride=1
         )

python/paddle/fluid/tests/unittests/ir/inference/test_trt_inference_predictor.py

@@ -55,7 +55,7 @@ def getdtype(dtype="float32"):
 
 class BackendPaddle:
     def __init__(self):
-        super(BackendPaddle, self).__init__()
+        super().__init__()
         self.h2d_time = []
         self.compute_time = []
         self.d2h_time = []
@@ -341,7 +341,7 @@ def forward(self, inputs):
 
 class Test(nn.Layer):
     def __init__(self):
-        super(Test, self).__init__()
+        super().__init__()
         self.conv = ConvBNLayer(
             num_channels=3, num_filters=64, filter_size=3, stride=2, act='relu'
         )

python/paddle/fluid/tests/unittests/ir/inference/test_trt_support_nhwc_pass.py

@@ -27,7 +27,7 @@
 
 class SimpleNet(nn.Layer):
     def __init__(self):
-        super(SimpleNet, self).__init__()
+        super().__init__()
         self.conv1 = nn.Conv2D(
             in_channels=4,
             out_channels=4,

python/paddle/fluid/tests/unittests/prim/composite_ops/test_composite_batch_norm.py

@@ -350,7 +350,7 @@ def apply_to_static(net, use_cinn):
 
 class PrimeNet(paddle.nn.Layer):
     def __init__(self):
-        super(PrimeNet, self).__init__()
+        super().__init__()
         self.conv = nn.Conv2D(4, 2, (3, 3), bias_attr=False)
         self.bn = BatchNorm(2, act="relu")
         self.run_mean = zeros([2])

python/paddle/fluid/tests/unittests/prim/composite_ops/test_composite_layer_norm.py

@@ -212,7 +212,7 @@ def apply_to_static(net, use_cinn):
 
 class PrimeNet(paddle.nn.Layer):
     def __init__(self, n_shape):
-        super(PrimeNet, self).__init__()
+        super().__init__()
         self.ln = LayerNorm(n_shape)
 
     def forward(self, x):

python/paddle/fluid/tests/unittests/prim/composite_ops/test_composite_softmax.py

@@ -137,7 +137,7 @@ def apply_to_static(net, use_cinn):
 
 class PrimeNet(paddle.nn.Layer):
     def __init__(self):
-        super(PrimeNet, self).__init__()
+        super().__init__()
         self.sf = F.softmax
 
     def forward(self, x, current_axis):