mindspore/tests/ut/python/dataset/test_sentencepiece_tokenize...

233 lines
10 KiB
Python

# Copyright 2020-2022 Huawei Technologies Co., Ltd
#
# 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 copy
import numpy as np
import mindspore.dataset.text as text
import mindspore.dataset as ds
from mindspore.dataset.text import SentencePieceModel, to_str, SPieceTokenizerOutType
VOCAB_FILE = "../data/dataset/test_sentencepiece/vocab.txt"
DATA_FILE = "../data/dataset/testTokenizerData/sentencepiece_tokenizer.txt"
def test_sentence_piece_tokenizer_callable():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with eager mode
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.UNIGRAM, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
data = "123"
assert np.array_equal(tokenizer(data), ["", "1", "23"])
def test_from_vocab_to_str_unigram():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with UNIGRAM model
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.UNIGRAM, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = ["", "I", "▁use", "▁MindSpore", "", "to", "", "t", "r", "a", "i", "n", "", "m", "y", "▁model", "."]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def test_from_vocab_to_str_bpe():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with BPE model
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.BPE, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = ["", "I", "", "u", "s", "e", "", "M", "in", "d", "S", "p", "or", "e", "▁t", "o", "▁t", "ra", "in", "▁m",
"y", "▁m", "ode", "l", "."]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def test_from_vocab_to_str_char():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with CHAR model
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.CHAR, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = ["", "I", "", "u", "s", "e", "", "M", "i", "n", "d", "S", "p", "o", "r", "e", "", "t", "o", "", "t",
"r", "a", "i", "n", "", "m", "y", "", "m", "o", "d", "e", "l", "."]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def test_from_vocab_to_str_word():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with WORD model
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.WORD, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = ["▁I", "▁use", "▁MindSpore", "▁to", "▁train▁my▁model."]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def test_from_vocab_to_int():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with out_type equal to int
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.UNIGRAM, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.INT)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = [3, 41, 59, 53, 3, 29, 3, 6, 12, 99, 7, 10, 3, 11, 20, 45, 19]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = i["text"]
for key, value in enumerate(ret):
assert value == expect[key]
def test_from_file_to_str():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with out_type equal to string
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.UNIGRAM, {})
text.SentencePieceVocab.save_model(vocab, "./", "m.model")
tokenizer = text.SentencePieceTokenizer("./m.model", out_type=SPieceTokenizerOutType.STRING)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = ["", "I", "▁use", "▁MindSpore", "", "to", "", "t", "r", "a", "i", "n", "", "m", "y", "▁model", "."]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def test_from_file_to_int():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer while loading vocab model from file
Expectation: Output is equal to the expected value
"""
vocab = text.SentencePieceVocab.from_file([VOCAB_FILE], 100, 0.9995, SentencePieceModel.UNIGRAM, {})
text.SentencePieceVocab.save_model(vocab, "./", "m.model")
tokenizer = text.SentencePieceTokenizer("./m.model", out_type=SPieceTokenizerOutType.INT)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = [3, 41, 59, 53, 3, 29, 3, 6, 12, 99, 7, 10, 3, 11, 20, 45, 19]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = i["text"]
for key, value in enumerate(ret):
assert value == expect[key]
def test_build_from_dataset():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer while loading vocab model from dataset
Expectation: Output is equal to the expected value
"""
data = ds.TextFileDataset(VOCAB_FILE, shuffle=False)
vocab = text.SentencePieceVocab.from_dataset(data, ["text"], 100, 0.9995, SentencePieceModel.UNIGRAM, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer)
expect = ["", "I", "▁use", "▁MindSpore", "", "to", "", "t", "r", "a", "i", "n", "", "m", "y", "▁model", "."]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def apply_func(dataset):
input_columns = ["text"]
output_columns = ["text2"]
dataset = dataset.rename(input_columns, output_columns)
return dataset
def zip_test(dataset):
dataset_1 = copy.deepcopy(dataset)
dataset_2 = copy.deepcopy(dataset)
dataset_1 = dataset_1.apply(apply_func)
dataset_zip = ds.zip((dataset_1, dataset_2))
expect = ["", "I", "▁use", "▁MindSpore", "", "to", "", "t", "r", "a", "i", "n", "", "m", "y", "▁model", "."]
for i in dataset_zip.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def concat_test(dataset):
dataset_1 = copy.deepcopy(dataset)
dataset = dataset.concat(dataset_1)
expect = ["", "I", "▁use", "▁MindSpore", "", "to", "", "t", "r", "a", "i", "n", "", "m", "y", "▁model", "."]
for i in dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
ret = to_str(i["text"])
for key, value in enumerate(ret):
assert value == expect[key]
def test_with_zip_concat():
"""
Feature: SentencePieceTokenizer
Description: Test SentencePieceTokenizer with zip and concat operations
Expectation: Output is equal to the expected value
"""
data = ds.TextFileDataset(VOCAB_FILE, shuffle=False)
vocab = text.SentencePieceVocab.from_dataset(data, ["text"], 100, 0.9995, SentencePieceModel.UNIGRAM, {})
tokenizer = text.SentencePieceTokenizer(vocab, out_type=SPieceTokenizerOutType.STRING)
dataset = ds.TextFileDataset(DATA_FILE, shuffle=False)
dataset = dataset.map(operations=tokenizer, num_parallel_workers=2)
zip_test(dataset)
concat_test(dataset)
if __name__ == "__main__":
test_sentence_piece_tokenizer_callable()
test_from_vocab_to_str_unigram()
test_from_vocab_to_str_bpe()
test_from_vocab_to_str_char()
test_from_vocab_to_str_word()
test_from_vocab_to_int()
test_from_file_to_str()
test_from_file_to_int()
test_build_from_dataset()
test_with_zip_concat()