【Hackathon4 No58】fix exponential and pad (#51300)

paddle/phi/kernels/cpu/exponential_kernel.cc

@@ -17,10 +17,10 @@
 #include <random>
 
 #include "paddle/phi/backends/cpu/cpu_context.h"
+#include "paddle/phi/common/amp_type_traits.h"
 #include "paddle/phi/core/generator.h"
 #include "paddle/phi/core/kernel_registry.h"
 #include "paddle/phi/kernels/funcs/distribution_helper.h"
-
 namespace phi {
 
 template <typename T, typename Context>

paddle/phi/kernels/exponential_kernel.h

@@ -15,7 +15,7 @@
 #pragma once
 
 #include "paddle/phi/core/dense_tensor.h"
-
+#include "paddle/phi/core/device_context.h"
 namespace phi {
 
 template <typename T, typename Context>

paddle/phi/kernels/gpu/exponential_kernel.cu

@@ -13,8 +13,8 @@
 // limitations under the License.
 
 #include "paddle/phi/kernels/exponential_kernel.h"
-
 #include "paddle/phi/backends/gpu/gpu_context.h"
+#include "paddle/phi/common/amp_type_traits.h"
 #include "paddle/phi/core/kernel_registry.h"
 #include "paddle/phi/kernels/funcs/distribution_helper.h"
 
@@ -25,12 +25,19 @@ void ExponentialKernel(const Context &dev_ctx,
                        const DenseTensor &x,
                        float lambda,
                        DenseTensor *out) {
-  phi::funcs::uniform_distribution<T> dist;
-  phi::funcs::exponential_transform<T> trans(lambda);
+  using MT = typename kps::details::MPTypeTrait<T>::Type;
+  phi::funcs::uniform_distribution<MT> dist;
+  phi::funcs::exponential_transform<MT> trans(lambda);
   phi::funcs::distribution_and_transform<T>(dev_ctx, out, dist, trans);
 }
 
 }  // namespace phi
 
-PD_REGISTER_KERNEL(
-    exponential, GPU, ALL_LAYOUT, phi::ExponentialKernel, float, double) {}
+PD_REGISTER_KERNEL(exponential,
+                   GPU,
+                   ALL_LAYOUT,
+                   phi::ExponentialKernel,
+                   float,
+                   double,
+                   phi::dtype::float16,
+                   phi::dtype::bfloat16) {}

paddle/phi/kernels/gpu/pad_grad_kernel.cu

@@ -25,5 +25,6 @@ PD_REGISTER_KERNEL(pad_grad,
                    float,
                    double,
                    phi::dtype::float16,
+                   phi::dtype::bfloat16,
                    phi::dtype::complex<float>,
                    phi::dtype::complex<double>) {}

paddle/phi/kernels/gpu/pad_kernel.cu

@@ -28,5 +28,6 @@ PD_REGISTER_KERNEL(pad,
                    int,
                    int64_t,
                    phi::dtype::float16,
+                   phi::dtype::bfloat16,
                    phi::dtype::complex<float>,
                    phi::dtype::complex<double>) {}

paddle/phi/kernels/pad_grad_kernel.h

@@ -17,6 +17,7 @@
 
 #include "paddle/phi/common/scalar.h"
 #include "paddle/phi/core/dense_tensor.h"
+#include "paddle/phi/core/device_context.h"
 
 namespace phi {
 

paddle/phi/kernels/pad_kernel.h

@@ -17,6 +17,7 @@
 
 #include "paddle/phi/common/scalar.h"
 #include "paddle/phi/core/dense_tensor.h"
+#include "paddle/phi/core/device_context.h"
 
 namespace phi {
 

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

@@ -15,9 +15,14 @@
 import unittest
 
 import numpy as np
-from eager_op_test import OpTest
+from eager_op_test import (
+    OpTest,
+    convert_float_to_uint16,
+    convert_uint16_to_float,
+)
 
 import paddle
+from paddle.fluid import core
 
 
 class TestExponentialOp1(OpTest):
@@ -344,6 +349,104 @@ class TestExponentialAPI(unittest.TestCase):
         paddle.enable_static()
 
 
+class TestExponentialFP16Op(OpTest):
+    def setUp(self):
+        paddle.enable_static()
+        self.op_type = "exponential"
+        self.python_api = paddle.tensor.exponential_
+        self.config()
+        self.attrs = {"lambda": self.lam}
+        self.inputs = {'X': np.empty([1024, 1024], dtype=self.dtype)}
+        self.outputs = {'Out': np.ones([1024, 1024], dtype=self.dtype)}
+
+    def config(self):
+        self.lam = 0.5
+        self.dtype = np.float16
+
+    def test_check_output(self):
+        self.check_output_customized(self.verify_output)
+
+    def verify_output(self, outs):
+        hist1, _ = np.histogram(outs[0], range=(0, 5))
+        hist1 = hist1.astype(np.float16)
+        hist1 = hist1 / float(outs[0].size)
+
+        data_np = np.random.exponential(1.0 / self.lam, [1024, 1024])
+        hist2, _ = np.histogram(data_np, range=(0, 5))
+        hist2 = hist2.astype(np.float16)
+        hist2 = hist2 / float(data_np.size)
+
+        np.testing.assert_allclose(hist1, hist2, rtol=0.05)
+
+    def test_check_grad_normal(self):
+        self.check_grad(
+            ['X'],
+            'Out',
+            in_place=True,
+            user_defined_grads=[np.zeros([1024, 1024], dtype=self.dtype)],
+            user_defined_grad_outputs=[
+                np.random.rand(1024, 1024).astype(self.dtype)
+            ],
+            check_dygraph=False,  # inplace can not call paddle.grad
+        )
+
+
+@unittest.skipIf(
+    not core.is_compiled_with_cuda()
+    or not core.is_bfloat16_supported(core.CUDAPlace(0)),
+    "core is not compiled with CUDA and not support the bfloat16",
+)
+class TestExponentialBP16Op(OpTest):
+    def setUp(self):
+        paddle.enable_static()
+        self.op_type = "exponential"
+        self.python_api = paddle.tensor.exponential_
+        self.config()
+        x = np.empty([1024, 1024]).astype('float32')
+        out = np.ones([1024, 1024]).astype('float32')
+        self.inputs = {'X': convert_float_to_uint16(x)}
+        self.attrs = {"lambda": self.lam}
+        self.outputs = {'Out': convert_float_to_uint16(out)}
+
+    def config(self):
+        self.lam = 0.5
+        self.dtype = np.uint16
+
+    def test_check_output(self):
+        place = core.CUDAPlace(0)
+        self.check_output_with_place_customized(
+            checker=self.verify_output, place=place
+        )
+
+    def verify_output(self, outs):
+        outs = convert_uint16_to_float(outs)
+        self.assertEqual(outs[0].shape, (1024, 1024))
+        hist1, _ = np.histogram(outs[0], range=(-3, 5))
+        hist1 = hist1.astype("float32")
+        hist1 = hist1 / float(outs[0].size)
+
+        data_np = np.random.exponential(1.0 / self.lam, [1024, 1024])
+        hist2, _ = np.histogram(data_np, range=(-3, 5))
+        hist2 = hist2.astype("float32")
+        hist2 = hist2 / float(data_np.size)
+
+        np.testing.assert_allclose(hist1, hist2, rtol=0.05)
+
+    def test_check_grad_normal(self):
+        place = core.CUDAPlace(0)
+        self.check_grad_with_place(
+            place,
+            ['X'],
+            'Out',
+            in_place=True,
+            user_defined_grads=[np.zeros([1024, 1024], dtype=self.dtype)],
+            user_defined_grad_outputs=[
+                np.random.rand(1024, 1024).astype(self.dtype)
+            ],
+            check_dygraph=False,  # inplace can not call paddle.grad
+        )
+
+
 if __name__ == "__main__":
     paddle.enable_static()
     unittest.main()

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

@@ -16,7 +16,7 @@ import os
 import unittest
 
 import numpy as np
-from eager_op_test import OpTest
+from eager_op_test import OpTest, convert_float_to_uint16
 from test_attribute_var import UnittestBase
 
 import paddle
@@ -96,7 +96,7 @@ def create_test_fp16(parent):
             return np.float16
 
         def test_check_grad_normal(self):
-            self.check_grad(['X'], 'Out', max_relative_error=0.3)
+            self.check_grad(['X'], 'Out')
 
     cls_name = "{}_{}".format(parent.__name__, "Fp16")
     TestPadFp16.__name__ = cls_name
@@ -202,6 +202,41 @@ class TestPaddingValueTensor3(unittest.TestCase):
         np.testing.assert_allclose(pd_out, np_out)
 
 
+@unittest.skipIf(
+    not core.is_compiled_with_cuda()
+    or not core.is_bfloat16_supported(core.CUDAPlace(0)),
+    "core is not compiled with CUDA and not support the bfloat16",
+)
+class TestPadBP16Op(OpTest):
+    def setUp(self):
+        self.initTestCase()
+        self.dtype = np.uint16
+        self.op_type = "pad"
+        self.python_api = pad_wrapper
+        x = np.random.random(self.shape).astype(np.float32)
+        self.attrs = {}
+        self.attrs['paddings'] = np.array(self.paddings).flatten()
+        self.attrs['pad_value'] = self.pad_value
+        out = np.pad(
+            x, self.paddings, mode='constant', constant_values=self.pad_value
+        )
+        self.inputs = {'X': convert_float_to_uint16(x)}
+        self.outputs = {'Out': convert_float_to_uint16(out)}
+
+    def initTestCase(self):
+        self.shape = (16, 16)
+        self.paddings = [(0, 1), (2, 3)]
+        self.pad_value = 0.0
+
+    def test_check_output(self):
+        place = core.CUDAPlace(0)
+        self.check_output_with_place(place)
+
+    def test_check_grad(self):
+        place = core.CUDAPlace(0)
+        self.check_grad_with_place(place, ['X'], 'Out')
+
+
 if __name__ == '__main__':
     paddle.enable_static()
     unittest.main()

python/paddle/tensor/random.py

@@ -1154,7 +1154,9 @@ def exponential_(x, lam=1.0, name=None):
     if in_dygraph_mode():
         return _C_ops.exponential_(x, lam)
     else:
-        check_variable_and_dtype(x, "x", ["float32", "float64"], "exponential")
+        check_variable_and_dtype(
+            x, "x", ["float16", "float32", "float64", "uint16"], "exponential"
+        )
 
         helper = LayerHelper("exponential", **locals())
         helper.append_op(