wehub-resource-sync
225 lines
8.9 KiB
Python
225 lines
8.9 KiB
Python
# Copyright (c) 2025 PaddlePaddle Authors. All Rights Reserved.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
import unittest
|
|
|
|
import numpy as np
|
|
|
|
import paddle
|
|
|
|
|
|
class TestExpOutAndParamDecorator(unittest.TestCase):
|
|
def setUp(self):
|
|
paddle.disable_static()
|
|
self.x_np = np.random.rand(3, 4).astype(np.float32)
|
|
self.test_types = ["decorator", "out", "out_decorator"]
|
|
|
|
def do_test(self, test_type):
|
|
x = paddle.to_tensor(self.x_np, stop_gradient=False)
|
|
if test_type == 'raw':
|
|
result = paddle.exp(x)
|
|
result.mean().backward()
|
|
return result, x.grad
|
|
elif test_type == 'decorator':
|
|
result = paddle.exp(input=x)
|
|
result.mean().backward()
|
|
return result, x.grad
|
|
elif test_type == 'out':
|
|
out = paddle.empty_like(x)
|
|
out.stop_gradient = False
|
|
paddle.exp(x, out=out)
|
|
out.mean().backward()
|
|
return out, x.grad
|
|
elif test_type == 'out_decorator':
|
|
out = paddle.empty_like(x)
|
|
out.stop_gradient = False
|
|
paddle.exp(input=x, out=out)
|
|
out.mean().backward()
|
|
return out, x.grad
|
|
else:
|
|
raise ValueError(f"Unknown test type: {test_type}")
|
|
|
|
def test_all(self):
|
|
out_std, grad_std = self.do_test('raw')
|
|
for test_type in self.test_types:
|
|
out, grad = self.do_test(test_type)
|
|
np.testing.assert_allclose(out.numpy(), out_std.numpy(), rtol=1e-7)
|
|
np.testing.assert_allclose(
|
|
grad.numpy(), grad_std.numpy(), rtol=1e-7
|
|
)
|
|
|
|
|
|
class TestExpSleefVectorized(unittest.TestCase):
|
|
"""Test exp with shapes that exercise Sleef vectorized paths.
|
|
|
|
For AVX2:
|
|
- float32: VEC_SIZE = 8, so shapes >= 8 trigger vectorized path
|
|
- float64: VEC_SIZE = 4, so shapes >= 4 trigger vectorized path
|
|
(Covers vexp_avx2_f64 in sleef_vectorized_math.h lines 335-348)
|
|
|
|
Test both:
|
|
1. Shapes that are exact multiples of VEC_SIZE (only vectorized loop)
|
|
2. Shapes with remainder (vectorized loop + scalar tail)
|
|
|
|
Note: If MKL is available at runtime, the mkl_exp path (lines 704-706, 722-724)
|
|
will be triggered instead. Both paths are tested through this test.
|
|
"""
|
|
|
|
def setUp(self):
|
|
paddle.disable_static()
|
|
|
|
def test_exp_float32_vectorized_exact(self):
|
|
"""Test float32 exp with shape that's exact multiple of 8.
|
|
Covers vexp_avx2_f32 main loop (lines 323-327).
|
|
"""
|
|
# Shape 16 = 8 * 2, exercises only vectorized loop
|
|
x_np = np.random.uniform(-2, 2, size=(16,)).astype(np.float32)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-5, atol=1e-5
|
|
)
|
|
|
|
def test_exp_float32_vectorized_with_tail(self):
|
|
"""Test float32 exp with shape that has remainder when divided by 8.
|
|
Covers vexp_avx2_f32 both main loop (323-327) and scalar tail (329-331).
|
|
"""
|
|
# Shape 13 = 8 + 5, exercises both vectorized loop and scalar tail
|
|
x_np = np.random.uniform(-2, 2, size=(13,)).astype(np.float32)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-5, atol=1e-5
|
|
)
|
|
|
|
def test_exp_float64_vectorized_exact(self):
|
|
"""Test float64 exp with shape that's exact multiple of 4.
|
|
Covers vexp_avx2_f64 main loop (lines 339-343).
|
|
This specifically tests the code at sleef_vectorized_math.h L335-348.
|
|
"""
|
|
# Shape 12 = 4 * 3, exercises only vectorized loop
|
|
x_np = np.random.uniform(-2, 2, size=(12,)).astype(np.float64)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-10, atol=1e-10
|
|
)
|
|
|
|
def test_exp_float64_vectorized_with_tail(self):
|
|
"""Test float64 exp with shape that has remainder when divided by 4.
|
|
Covers vexp_avx2_f64 both main loop (339-343) and scalar tail (345-347).
|
|
This specifically tests the code at sleef_vectorized_math.h L335-348.
|
|
"""
|
|
# Shape 11 = 4 * 2 + 3, exercises both vectorized loop and scalar tail
|
|
x_np = np.random.uniform(-2, 2, size=(11,)).astype(np.float64)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-10, atol=1e-10
|
|
)
|
|
|
|
def test_exp_float32_large_shape(self):
|
|
"""Test float32 exp with large shape for comprehensive coverage.
|
|
Tests MKL VML path (mkl_exp) if available, otherwise Sleef path.
|
|
"""
|
|
x_np = np.random.uniform(-2, 2, size=(1024,)).astype(np.float32)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-5, atol=1e-5
|
|
)
|
|
|
|
def test_exp_float64_large_shape(self):
|
|
"""Test float64 exp with large shape for comprehensive coverage.
|
|
Tests MKL VML path (mkl_exp) if available, otherwise Sleef path.
|
|
This specifically tests vexp_avx2_f64 at sleef_vectorized_math.h L335-348.
|
|
"""
|
|
x_np = np.random.uniform(-2, 2, size=(1024,)).astype(np.float64)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-10, atol=1e-10
|
|
)
|
|
|
|
def test_exp_float32_2d_shape(self):
|
|
"""Test float32 exp with 2D shape to verify flattened processing."""
|
|
# Shape (4, 5) = 20 elements, exercises vectorized path
|
|
x_np = np.random.uniform(-2, 2, size=(4, 5)).astype(np.float32)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-5, atol=1e-5
|
|
)
|
|
|
|
def test_exp_float64_2d_shape(self):
|
|
"""Test float64 exp with 2D shape to verify flattened processing.
|
|
Covers vexp_avx2_f64 at sleef_vectorized_math.h L335-348.
|
|
"""
|
|
# Shape (3, 5) = 15 elements, exercises vectorized path with tail
|
|
x_np = np.random.uniform(-2, 2, size=(3, 5)).astype(np.float64)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-10, atol=1e-10
|
|
)
|
|
|
|
def test_exp_float32_small_shape_fallback(self):
|
|
"""Test float32 exp with small shape (numel < 8) to cover Eigen fallback path."""
|
|
# Shape 5 < 8, triggers Eigen fallback instead of SIMD
|
|
x_np = np.random.uniform(-2, 2, size=(5,)).astype(np.float32)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-5, atol=1e-5
|
|
)
|
|
|
|
def test_exp_float64_small_shape_fallback(self):
|
|
"""Test float64 exp with small shape (numel < 8) to cover Eigen fallback path."""
|
|
# Shape 3 < 8, triggers Eigen fallback instead of SIMD
|
|
x_np = np.random.uniform(-2, 2, size=(3,)).astype(np.float64)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-10, atol=1e-10
|
|
)
|
|
|
|
def test_exp_float64_boundary_values(self):
|
|
"""Test float64 exp with boundary values to ensure numerical stability.
|
|
Covers vexp_avx2_f64 at sleef_vectorized_math.h L335-348.
|
|
"""
|
|
# Test values near boundaries
|
|
x_np = np.array(
|
|
[0.0, 1.0, -1.0, 10.0, -10.0, 50.0, -50.0, 100.0], dtype=np.float64
|
|
)
|
|
x = paddle.to_tensor(x_np, place=paddle.CPUPlace())
|
|
result = paddle.exp(x)
|
|
expected = np.exp(x_np)
|
|
np.testing.assert_allclose(
|
|
result.numpy(), expected, rtol=1e-10, atol=1e-10
|
|
)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
unittest.main()
|