wehub-resource-sync
289 lines
7.5 KiB
Python
289 lines
7.5 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 utils
|
|
|
|
import paddle
|
|
from paddle.base import core
|
|
|
|
|
|
class CINNConvertToFloat8e4m3Net(paddle.nn.Layer):
|
|
def __init__(self):
|
|
super().__init__()
|
|
|
|
def forward(
|
|
self,
|
|
x,
|
|
):
|
|
x_pow2 = x * x
|
|
x_fp8 = x_pow2.astype(paddle.float8_e4m3fn)
|
|
x_fp8.stop_gradient = True
|
|
return x_fp8
|
|
|
|
|
|
class TestFloat2Float8e4m3(unittest.TestCase):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def setUp(self):
|
|
self.prepare_data()
|
|
|
|
def prepare_data(self):
|
|
self.shape = [8 * 4096, 2048 * 2]
|
|
self.x = paddle.rand(shape=self.shape, dtype=paddle.float32)
|
|
self.x.stop_gradient = True
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNConvertToFloat8e4m3Net()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
def test_eval(self):
|
|
cinn_out = self.eval(use_cinn=True, use_prim=True)
|
|
dy_out = self.eval(use_cinn=False)
|
|
|
|
np.testing.assert_allclose(
|
|
cinn_out.numpy(), dy_out.numpy(), atol=1e-8, rtol=1e-4
|
|
)
|
|
|
|
|
|
class TestBfloat162Float8e4m3(TestFloat2Float8e4m3):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def prepare_data(self):
|
|
self.shape = [8 * 4096, 2048 * 2]
|
|
self.x = paddle.clip(
|
|
paddle.randn(self.shape).astype("bfloat16"),
|
|
min=-50,
|
|
max=50,
|
|
)
|
|
self.x.stop_gradient = True
|
|
|
|
|
|
class TestFloat162Float8e4m3(TestFloat2Float8e4m3):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def prepare_data(self):
|
|
self.shape = [8 * 4096, 2048 * 2]
|
|
self.x = paddle.rand(shape=self.shape, dtype=paddle.float16)
|
|
self.x.stop_gradient = True
|
|
|
|
|
|
class TestFloat642Float8e4m3(TestFloat2Float8e4m3):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def prepare_data(self):
|
|
self.shape = [8 * 4096, 2048 * 2]
|
|
self.x = paddle.rand(shape=self.shape, dtype=paddle.float64)
|
|
self.x.stop_gradient = True
|
|
|
|
|
|
class TestInt322Float8e4m3(TestFloat2Float8e4m3):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def prepare_data(self):
|
|
self.shape = [8 * 4096, 2048 * 2]
|
|
self.x = paddle.randint(
|
|
low=-50, high=50, shape=self.shape, dtype=paddle.int32
|
|
)
|
|
self.x.stop_gradient = True
|
|
|
|
|
|
class TestInt642Float8e4m3(TestFloat2Float8e4m3):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def prepare_data(self):
|
|
self.shape = [8 * 4096, 2048 * 2]
|
|
self.x = paddle.randint(
|
|
low=-50, high=50, shape=self.shape, dtype=paddle.int64
|
|
)
|
|
self.x.stop_gradient = True
|
|
|
|
|
|
class CINNFloat8e4m3ConvertToNet(paddle.nn.Layer):
|
|
def __init__(self):
|
|
super().__init__()
|
|
|
|
def forward(self, x_fp8, dtype):
|
|
x_dtype = x_fp8.astype(dtype)
|
|
x_out = x_dtype + x_dtype
|
|
x_out.stop_gradient = True
|
|
return x_out
|
|
|
|
|
|
class TestFloat8e4m32Float(unittest.TestCase):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def setUp(self):
|
|
self.prepare_data()
|
|
|
|
def prepare_data(self):
|
|
self.shape = [8 * 4096, 2048 * 2]
|
|
self.x = paddle.rand(shape=self.shape, dtype=paddle.float8_e4m3fn)
|
|
self.x.stop_gradient = True
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.float32)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
def test_eval(self):
|
|
cinn_out = self.eval(use_cinn=True, use_prim=True)
|
|
dy_out = self.eval(use_cinn=False)
|
|
|
|
np.testing.assert_allclose(
|
|
cinn_out.numpy(), dy_out.numpy(), atol=1e-8, rtol=1e-4
|
|
)
|
|
|
|
|
|
class TestFloat8e4m32Float64(TestFloat8e4m32Float):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.float64)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
|
|
class TestFloat8e4m32Float16(TestFloat8e4m32Float):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.float16)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
|
|
class TestFloat8e4m32Bfloat16(TestFloat8e4m32Float):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.bfloat16)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
|
|
class TestFloat8e4m32Int32(TestFloat8e4m32Float):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.int32)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
|
|
class TestFloat8e4m32Int64(TestFloat8e4m32Float):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.int64)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
|
|
class TestFloat8e4m32Int16(TestFloat8e4m32Float):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.int16)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
|
|
class TestFloat8e4m32Int8(TestFloat8e4m32Float):
|
|
"""
|
|
Test Pir API + @to_static + CINN.
|
|
"""
|
|
|
|
def eval(self, use_cinn, use_prim=False):
|
|
if use_prim:
|
|
core._set_prim_all_enabled(True)
|
|
net = CINNFloat8e4m3ConvertToNet()
|
|
net = utils.apply_to_static(net, use_cinn)
|
|
net.eval()
|
|
out = net(self.x, paddle.int8)
|
|
core._set_prim_all_enabled(False)
|
|
return out
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|