clean fluid task: delete sampling_id api

python/paddle/fluid/layers/nn.py

@@ -96,7 +96,6 @@
     'elementwise_sub',
     'elementwise_mul',
     'gaussian_random',
-    'sampling_id',
     'clip',
     'clip_by_norm',
     'mean',
@@ -4970,45 +4969,6 @@ def gaussian_random(
     return out
 
 
-@templatedoc()
-def sampling_id(x, min=0.0, max=1.0, seed=0, dtype='float32'):
-    """
-    This op is used for sampling id from multinomial distribution from the input, sampling one id for one sample.
-
-    Parameters:
-        x (Variable): 2-D tensor, [batch_size, input_feature_dimensions]
-        min (Float): minimum , default 0.0.
-        max (Float): maximum, default 1.0.
-        seed (Float): Random seed, default 0. if seed is not 0, will generate same number every time.
-        dtype(np.dtype|core.VarDesc.VarType|str): The type of output data : float32, float_16, int etc
-
-    Returns:
-        Variable: sampling tensor.
-
-    Examples:
-        .. code-block:: python
-
-            import paddle.fluid as fluid
-            x = fluid.data(
-                name="X",
-                shape=[13, 11],
-                dtype='float32')
-
-            out = fluid.layers.sampling_id(x)
-    """
-
-    helper = LayerHelper('sampling_id', **locals())
-    out = helper.create_variable_for_type_inference(dtype)
-    helper.append_op(
-        type='sampling_id',
-        inputs={'X': x},
-        outputs={'Out': out},
-        attrs={'min': min, 'max': max, 'seed': seed},
-    )
-
-    return out
-
-
 def _elementwise_op(helper):
     op_type = helper.layer_type
     x = helper.kwargs.get('x', None)

python/paddle/fluid/layers/rnn.py

@@ -2306,40 +2306,6 @@ def __init__(
         )
         self.seed = seed
 
-    def sample(self, time, outputs, states):
-        r"""
-        Perform sampling from a categorical distribution, and the distribution
-        is computed by `softmax(outputs/softmax_temperature)`.
-
-        Parameters:
-            time(Variable): An `int64` tensor with shape `[1]` provided by the
-                caller, representing the current time step number of decoding.
-            outputs(Variable): A tensor variable. Usually it's data type is float32
-                or float64, and it's shape is `[batch_size, vocabulary_size]`,
-                representing the predicted logits of current step. It is same as
-                `outputs` returned by `BasicDecoder.output_fn(BasicDecoder.cell.call())`.
-            states(Variable): A (possibly nested structure of) tensor variable[s].
-                It is same as `new_states` returned by `BasicDecoder.cell.call()`.
-
-        Returns:
-            Variable: An `int64` tensor with shape `[batch_size]`, representing \
-                the sampled ids.
-        """
-        logits = (
-            (outputs / self.softmax_temperature)
-            if self.softmax_temperature is not None
-            else outputs
-        )
-        probs = paddle.nn.functional.softmax(logits)
-        # TODO: remove this stop_gradient. The stop_gradient of sample_ids can
-        # not pass to probs, since sampling_id op does not have corresponding
-        # grad op and thus can not pass.
-        probs.stop_gradient = True
-        sample_ids = nn.sampling_id(
-            probs, seed=self.seed, dtype=self.start_tokens.dtype
-        )
-        return sample_ids
-
 
 class BasicDecoder(Decoder):
     """

python/paddle/fluid/tests/unittests/test_layers.py

@@ -2779,7 +2779,6 @@ def setUp(self):
             {
                 "make_gaussian_random",
                 "make_kldiv_loss",
-                "make_sampling_id",
                 "make_uniform_random_batch_size_like",
             }
         )
@@ -3362,20 +3361,6 @@ def make_gaussian_random(self):
             out = layers.gaussian_random(shape=[20, 30])
             return out
 
-    def make_sampling_id(self):
-        with program_guard(
-            fluid.default_main_program(), fluid.default_startup_program()
-        ):
-            x = self._get_data(
-                name="X",
-                shape=[13, 11],
-                dtype='float32',
-                append_batch_size=False,
-            )
-
-            out = layers.sampling_id(x)
-            return out
-
     def make_sum(self):
         with program_guard(
             fluid.default_main_program(), fluid.default_startup_program()

python/paddle/fluid/tests/unittests/test_random_seed.py

@@ -372,83 +372,6 @@ def test_generator_randperm_static(self):
                 np.testing.assert_allclose(out1_res2, out2_res2, rtol=1e-05)
                 self.assertTrue(not np.allclose(out1_res2, out1_res1))
 
-    def test_generator_sampling_id_dygraph(self):
-        """Test Generator seed."""
-        gen = paddle.seed(12312321111)
-
-        fluid.enable_dygraph()
-
-        gen.manual_seed(12312321111)
-        x = fluid.layers.uniform_random(
-            [10, 10], dtype="float32", min=0.0, max=1.0
-        )
-        y = fluid.layers.sampling_id(x)
-
-        st1 = gen.get_state()
-        x1 = fluid.layers.uniform_random(
-            [10, 10], dtype="float32", min=0.0, max=1.0
-        )
-        y1 = fluid.layers.sampling_id(x)
-
-        gen.set_state(st1)
-        x2 = fluid.layers.uniform_random(
-            [10, 10], dtype="float32", min=0.0, max=1.0
-        )
-        y2 = fluid.layers.sampling_id(x)
-
-        gen.manual_seed(12312321111)
-        x3 = fluid.layers.uniform_random(
-            [10, 10], dtype="float32", min=0.0, max=1.0
-        )
-        y3 = fluid.layers.sampling_id(x)
-
-        x_np = y.numpy()
-        x1_np = y1.numpy()
-        x2_np = y2.numpy()
-        x3_np = y3.numpy()
-
-        if not core.is_compiled_with_cuda():
-            print(">>>>>>> sampling id dygraph >>>>>>>")
-            np.testing.assert_allclose(x1_np, x2_np, rtol=1e-05)
-            np.testing.assert_allclose(x_np, x3_np, rtol=1e-05)
-
-    def test_generator_randperm_static_1(self):
-
-        fluid.disable_dygraph()
-
-        paddle.seed(123123143)
-
-        startup_program = fluid.Program()
-        train_program = fluid.Program()
-        with fluid.program_guard(train_program, startup_program):
-            # example 1:
-            # attr shape is a list which doesn't contain tensor Variable.
-            x = fluid.layers.uniform_random(shape=[10, 10])
-            result_1 = fluid.layers.sampling_id(x)
-            result_2 = fluid.layers.sampling_id(x)
-
-            exe = fluid.Executor(fluid.CPUPlace())
-            exe.run(startup_program)
-            out1 = exe.run(
-                train_program, feed={}, fetch_list=[result_1, result_2]
-            )
-
-            paddle.seed(123123143)
-            out2 = exe.run(
-                train_program, feed={}, fetch_list=[result_1, result_2]
-            )
-
-            out1_res1 = np.array(out1[0])
-            out1_res2 = np.array(out1[1])
-            out2_res1 = np.array(out2[0])
-            out2_res2 = np.array(out2[1])
-
-            if not core.is_compiled_with_cuda():
-                print(">>>>>>> sampling id static >>>>>>>")
-                np.testing.assert_allclose(out1_res1, out2_res1, rtol=1e-05)
-                np.testing.assert_allclose(out1_res2, out2_res2, rtol=1e-05)
-                self.assertTrue(not np.allclose(out1_res2, out1_res1))
-
     def test_gen_TruncatedNormal_initializer(self):
         fluid.disable_dygraph()