【PIR API adaptor No.188】 Migrate paddle.regularizer.L2Decay into pir

[xiaoyewww]

PR types

Others

PR changes

APIs

Description

PIR API 推全升级

#58067
No.187 将 paddle.regularizer.L2Decay 迁移升级至 pir，并更新单测，单测覆盖率：1/2。

bow_net中paddle.static.nn.sequence_lod.sequence_pool没有对应的pir实现op：

1869: ERROR: test_l2 (test_regularizer_api.TestRegularizer)
1869: ----------------------------------------------------------------------
1869: Traceback (most recent call last):
1869:   File "/wuzp/Paddle/build/python/paddle/pir_utils.py", line 113, in impl
1869:     func(*args, **kwargs)
1869:   File "/wuzp/Paddle/build/test/legacy_test/test_regularizer_api.py", line 167, in test_l2
1869:     framework_l2 = self.check_l2decay_regularizer(place, model)
1869:   File "/wuzp/Paddle/build/python/paddle/pir_utils.py", line 115, in impl
1869:     func(*args, **kwargs)
1869:   File "/wuzp/Paddle/build/test/legacy_test/test_regularizer_api.py", line 120, in check_l2decay_regularizer
1869:     avg_cost = model(data, label, self.word_len)
1869:   File "/wuzp/Paddle/build/test/legacy_test/test_regularizer_api.py", line 47, in bow_net
1869:     bow = paddle.static.nn.sequence_lod.sequence_pool(
1869:   File "/wuzp/Paddle/build/python/paddle/static/nn/sequence_lod.py", line 348, in sequence_pool
1869:     pool_out = helper.create_variable_for_type_inference(dtype)
1869:   File "/wuzp/Paddle/build/python/paddle/base/layer_helper_base.py", line 468, in create_variable_for_type_inference
1869:     return self.main_program.current_block().create_var(
1869:   File "/wuzp/Paddle/build/python/paddle/base/framework.py", line 4207, in create_var
1869:     var = Variable(block=self, *args, **kwargs)
1869:   File "/wuzp/Paddle/build/python/paddle/base/framework.py", line 1455, in __init__
1869:     dtype = convert_np_dtype_to_dtype_(dtype)
1869:   File "/wuzp/Paddle/build/python/paddle/base/framework.py", line 1157, in convert_np_dtype_to_dtype_
1869:     return pir.core.convert_np_dtype_to_dtype_(np_dtype)
1869:   File "/wuzp/Paddle/build/python/paddle/pir/core.py", line 75, in convert_np_dtype_to_dtype_
1869:     dtype = np.dtype(np_dtype)
1869: TypeError: Cannot interpret '<DataType.FLOAT32: 10>' as a data type
1869: 
1869: ----------------------------------------------------------------------

[paddle-bot]

你的PR提交成功，感谢你对开源项目的贡献!
请关注后续CI自动化测试结果，详情请参考Paddle-CI手册。
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

[MarioLulab]

LGTM

[0x45f]

Suggested change

	linear = paddle.nn.Linear(1, 5, weight_attr=fc_param_attr)
	linear = paddle.nn.Linear(3, 5, weight_attr=fc_param_attr)

[xiaoyewww]

感谢，已修改，麻烦重新review~