chore: import upstream snapshot with attribution
This commit is contained in:
@@ -0,0 +1,77 @@
|
||||
# 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
|
||||
|
||||
from test_case_base import (
|
||||
TestCaseBase,
|
||||
test_instruction_translator_cache_context,
|
||||
)
|
||||
|
||||
import paddle
|
||||
from paddle.jit import sot
|
||||
|
||||
|
||||
class EmbeddingLayer(paddle.nn.Layer):
|
||||
def __init__(self):
|
||||
super().__init__()
|
||||
self.embedding = paddle.nn.Embedding(10, 10)
|
||||
|
||||
def forward(self, x):
|
||||
x = x + 1 - 1
|
||||
x = self.embedding(x)
|
||||
return x + 1
|
||||
|
||||
|
||||
def call_embedding_layer(x: paddle.Tensor, layer: paddle.nn.Layer):
|
||||
return layer(x)
|
||||
|
||||
|
||||
def call_functional_embedding(x: paddle.Tensor, weight: paddle.Tensor):
|
||||
x = x + 1 - 1
|
||||
x = paddle.nn.functional.embedding(x, weight)
|
||||
return x + 1
|
||||
|
||||
|
||||
class TestForceDynamic(TestCaseBase):
|
||||
def test_embedding_layer(self):
|
||||
paddle.jit.marker.force_dynamic(paddle.nn.Embedding)
|
||||
|
||||
layer = EmbeddingLayer()
|
||||
with test_instruction_translator_cache_context() as ctx:
|
||||
self.assertEqual(ctx.translate_count, 0)
|
||||
self.assert_results(
|
||||
call_embedding_layer,
|
||||
paddle.randint(0, 10, [1, 3, 224, 224], dtype='int64'),
|
||||
layer,
|
||||
)
|
||||
self.assertGreater(ctx.translate_count, 1)
|
||||
|
||||
sot.utils.paddle_api_config.break_graph_layer_classes.clear()
|
||||
|
||||
def test_functional_embedding(self):
|
||||
paddle.jit.marker.force_dynamic(paddle.nn.functional.embedding)
|
||||
|
||||
weight = paddle.randn([10, 10])
|
||||
x = paddle.randint(0, 10, [1, 3, 224, 224], dtype='int64')
|
||||
with test_instruction_translator_cache_context() as ctx:
|
||||
self.assertEqual(ctx.translate_count, 0)
|
||||
self.assert_results(call_functional_embedding, x, weight)
|
||||
self.assertGreater(ctx.translate_count, 1)
|
||||
|
||||
sot.utils.paddle_api_config.break_graph_functions.clear()
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
unittest.main()
|
||||
Reference in New Issue
Block a user