# Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved. # Copyright 2020 The HuggingFace Team. 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. from __future__ import annotations import os import random import tempfile import unittest from typing import List import numpy as np import paddle from parameterized import parameterized, parameterized_class from paddlenlp import __version__ as current_version from paddlenlp.transformers import ( AutoModel, AutoModelForQuestionAnswering, AutoModelForTokenClassification, BertForMaskedLM, BertForMultipleChoice, BertForPretraining, BertForQuestionAnswering, BertForSequenceClassification, BertForTokenClassification, BertModel, ) from paddlenlp.transformers.bert.configuration import BertConfig from paddlenlp.transformers.model_utils import PretrainedModel from paddlenlp.utils import install_package, uninstall_package from ...testing_utils import require_package, slow from ..test_configuration_common import ConfigTester from ..test_modeling_common import ( ModelTesterMixin, ModelTesterPretrainedMixin, ids_tensor, random_attention_mask, ) class BertModelTester: def __init__( self, parent: BertModelTest, batch_size=13, seq_length=7, is_training=True, use_input_mask=True, use_token_type_ids=True, use_labels=True, vocab_size=99, hidden_size=32, num_hidden_layers=5, num_attention_heads=4, intermediate_size=37, hidden_act="gelu", hidden_dropout_prob=0.1, attention_probs_dropout_prob=0.1, max_position_embeddings=512, type_vocab_size=16, initializer_range=0.02, pad_token_id=0, pool_act="tanh", fuse=False, type_sequence_label_size=2, num_labels=3, num_choices=4, scope=None, dropout=0.56, return_dict=False, ): self.parent: BertModelTest = parent self.batch_size = batch_size self.seq_length = seq_length self.is_training = is_training self.use_input_mask = use_input_mask self.use_token_type_ids = use_token_type_ids self.use_labels = use_labels self.vocab_size = vocab_size self.hidden_size = hidden_size self.num_hidden_layers = num_hidden_layers self.num_attention_heads = num_attention_heads self.intermediate_size = intermediate_size self.hidden_act = hidden_act self.hidden_dropout_prob = hidden_dropout_prob self.attention_probs_dropout_prob = attention_probs_dropout_prob self.max_position_embeddings = max_position_embeddings self.type_vocab_size = type_vocab_size self.initializer_range = initializer_range self.pad_token_id = pad_token_id self.pool_act = pool_act self.fuse = fuse self.type_sequence_label_size = type_sequence_label_size self.num_labels = num_labels self.num_choices = num_choices self.scope = scope self.dropout = dropout self.return_dict = return_dict def prepare_config_and_inputs(self): input_ids = ids_tensor([self.batch_size, self.seq_length], self.vocab_size) input_mask = None if self.use_input_mask: input_mask = random_attention_mask([self.batch_size, self.seq_length]) token_type_ids = None if self.use_token_type_ids: token_type_ids = ids_tensor([self.batch_size, self.seq_length], self.type_vocab_size) sequence_labels = None token_labels = None choice_labels = None if self.use_labels: sequence_labels = ids_tensor([self.batch_size], self.type_sequence_label_size) token_labels = ids_tensor([self.batch_size, self.seq_length], self.num_labels) choice_labels = ids_tensor([self.batch_size], self.num_choices) config = self.get_config() return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels def get_config(self) -> BertConfig: return BertConfig( vocab_size=self.vocab_size, hidden_size=self.hidden_size, num_hidden_layers=self.num_hidden_layers, num_attention_heads=self.num_attention_heads, intermediate_size=self.intermediate_size, hidden_act=self.hidden_act, hidden_dropout_prob=self.hidden_dropout_prob, attention_probs_dropout_prob=self.attention_probs_dropout_prob, max_position_embeddings=self.max_position_embeddings, type_vocab_size=self.type_vocab_size, initializer_range=self.initializer_range, pad_token_id=self.pad_token_id, pool_act=self.pool_act, fuse=self.fuse, num_labels=self.num_labels, num_choices=self.num_choices, ) def create_and_check_model( self, config: BertConfig, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels ): model = BertModel(config) model.eval() result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids) result = model(input_ids, token_type_ids=token_type_ids) result = model(input_ids) self.parent.assertEqual(result[0].shape, [self.batch_size, self.seq_length, self.hidden_size]) self.parent.assertEqual(result[1].shape, [self.batch_size, self.hidden_size]) def create_and_check_for_masked_lm( self, config: BertConfig, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ): model = BertForMaskedLM(config) model.eval() result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, labels=token_labels) self.parent.assertEqual(result[1].shape, [self.batch_size, self.seq_length, self.vocab_size]) def create_and_check_model_past_large_inputs( self, config: BertConfig, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ): model = BertModel(config) model.eval() # first forward pass outputs = model(input_ids, attention_mask=input_mask, use_cache=True, return_dict=self.return_dict) past_key_values = outputs.past_key_values if self.return_dict else outputs[2] # create hypothetical multiple next token and extent to next_input_ids next_tokens = ids_tensor((self.batch_size, 3), self.vocab_size) next_mask = ids_tensor((self.batch_size, 3), vocab_size=2) # append to next input_ids and next_input_ids = paddle.concat([input_ids, next_tokens], axis=-1) next_attention_mask = paddle.concat([input_mask, next_mask], axis=-1) outputs = model( next_input_ids, attention_mask=next_attention_mask, output_hidden_states=True, return_dict=self.return_dict ) output_from_no_past = outputs[2][0] outputs = model( next_tokens, attention_mask=next_attention_mask, past_key_values=past_key_values, output_hidden_states=True, return_dict=self.return_dict, ) output_from_past = outputs[2][0] # select random slice random_slice_idx = ids_tensor((1,), output_from_past.shape[-1]).item() output_from_no_past_slice = output_from_no_past[:, -3:, random_slice_idx].detach() output_from_past_slice = output_from_past[:, :, random_slice_idx].detach() self.parent.assertTrue(output_from_past_slice.shape[1] == next_tokens.shape[1]) # test that outputs are equal for slice self.parent.assertTrue(paddle.allclose(output_from_past_slice, output_from_no_past_slice, atol=1e-3)) def create_and_check_for_pretraining( self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ): model = BertForPretraining(config) model.eval() result = model( input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, labels=token_labels, next_sentence_label=sequence_labels, ) self.parent.assertEqual(result[1].shape, [self.batch_size, self.seq_length, self.vocab_size]) self.parent.assertEqual(result[2].shape, [self.batch_size, 2]) def create_and_check_for_multiple_choice( self, config: BertConfig, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ): model = BertForMultipleChoice(config) model.eval() multiple_choice_inputs_ids = input_ids.unsqueeze(1).expand([-1, self.num_choices, -1]) multiple_choice_token_type_ids = token_type_ids.unsqueeze(1).expand([-1, self.num_choices, -1]) multiple_choice_input_mask = input_mask.unsqueeze(1).expand([-1, self.num_choices, -1]) result = model( multiple_choice_inputs_ids, attention_mask=multiple_choice_input_mask, token_type_ids=multiple_choice_token_type_ids, labels=choice_labels, ) self.parent.assertEqual(result[1].shape, [self.batch_size, self.num_choices]) def create_and_check_for_question_answering( self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ): model = BertForQuestionAnswering(config) model.eval() result = model( input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, start_positions=sequence_labels, end_positions=sequence_labels, return_dict=self.return_dict, ) if sequence_labels is not None: result = result[1:] self.parent.assertEqual(result[0].shape, [self.batch_size, self.seq_length]) self.parent.assertEqual(result[1].shape, [self.batch_size, self.seq_length]) def create_and_check_for_sequence_classification( self, config: BertConfig, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ): model = BertForSequenceClassification(config) model.eval() result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, labels=sequence_labels) self.parent.assertEqual(result[1].shape, [self.batch_size, self.num_labels]) def create_and_check_for_token_classification( self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ): model = BertForTokenClassification(config) model.eval() result = model(input_ids, attention_mask=input_mask, token_type_ids=token_type_ids, labels=token_labels) self.parent.assertEqual(result[1].shape, [self.batch_size, self.seq_length, self.num_labels]) def test_addition_params(self, config: BertConfig, *args, **kwargs): config.num_labels = 7 config.classifier_dropout = 0.98 model = BertForTokenClassification(config) model.eval() self.parent.assertEqual(model.classifier.weight.shape, [config.hidden_size, 7]) self.parent.assertEqual(model.dropout.p, 0.98) def prepare_config_and_inputs_for_common(self): config_and_inputs = self.prepare_config_and_inputs() ( config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels, ) = config_and_inputs inputs_dict = {"input_ids": input_ids, "token_type_ids": token_type_ids, "attention_mask": input_mask} return config, inputs_dict @parameterized_class( ("return_dict", "use_labels"), [ [False, False], [False, True], [True, False], [True, True], ], ) class BertModelTest(ModelTesterMixin, unittest.TestCase): base_model_class = BertModel return_dict = False use_labels = False test_tie_weights = True all_model_classes = ( BertModel, BertForMaskedLM, BertForMultipleChoice, BertForPretraining, BertForQuestionAnswering, BertForSequenceClassification, BertForTokenClassification, ) def setUp(self): super().setUp() self.model_tester = BertModelTester(self) self.config_tester = ConfigTester(self, config_class=BertConfig, vocab_size=256, hidden_size=24) def test_config(self): # self.config_tester.create_and_test_config_from_and_save_pretrained() self.config_tester.run_common_tests() def test_model(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*config_and_inputs) def test_for_masked_lm(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_masked_lm(*config_and_inputs) def test_decoder_model_past_with_large_inputs(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model_past_large_inputs(*config_and_inputs) def test_for_multiple_choice(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_multiple_choice(*config_and_inputs) def test_for_pretraining(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_pretraining(*config_and_inputs) def test_for_question_answering(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_question_answering(*config_and_inputs) def test_for_sequence_classification(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_sequence_classification(*config_and_inputs) def test_for_token_classification(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_for_token_classification(*config_and_inputs) def test_for_custom_params(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.test_addition_params(*config_and_inputs) def test_model_name_list(self): config = self.model_tester.get_config() model = self.base_model_class(config) self.assertTrue(len(model.model_name_list) != 0) @slow def test_params_compatibility_of_init_method(self): """test initing model with different params""" model: BertForTokenClassification = BertForTokenClassification.from_pretrained( "bert-base-uncased", num_classes=4, dropout=0.3 ) assert model.num_labels == 4 assert model.dropout.p == 0.3 class BertCompatibilityTest(unittest.TestCase): test_model_id = "hf-internal-testing/tiny-random-BertModel" @classmethod @require_package("transformers", "torch") def setUpClass(cls) -> None: from transformers import BertModel # when python application is done, `TemporaryDirectory` will be free cls.torch_model_path = tempfile.TemporaryDirectory().name model = BertModel.from_pretrained(cls.test_model_id) model.save_pretrained(cls.torch_model_path) def test_model_config_mapping(self): config = BertConfig(num_labels=22, hidden_dropout_prob=0.99) self.assertEqual(config.hidden_dropout_prob, 0.99) self.assertEqual(config.num_labels, 22) def setUp(self) -> None: self.tempdirs: List[tempfile.TemporaryDirectory] = [] def tearDown(self) -> None: for tempdir in self.tempdirs: tempdir.cleanup() def get_tempdir(self) -> str: tempdir = tempfile.TemporaryDirectory() self.tempdirs.append(tempdir) return tempdir.name def run_token_for_classification(self, version: str): install_package("paddlenlp", version=version) from paddlenlp import __version__ self.assertEqual(__version__, version) from paddlenlp.transformers import BertForTokenClassification, BertModel tempdir = self.get_tempdir() # prepare the old version of model old_model = BertModel.from_pretrained("bert-base-uncased") old_model_path = os.path.join(tempdir, "old-model") old_model.save_pretrained(old_model_path) old_model_for_token = BertForTokenClassification.from_pretrained( "bert-base-uncased", num_classes=4, dropout=0.3 ) old_model_for_token_path = os.path.join(tempdir, "old-model-for-token") old_model_for_token.save_pretrained(old_model_for_token_path) uninstall_package("paddlenlp") from paddlenlp import __version__ self.assertEqual(__version__, current_version) from paddlenlp.transformers import BertForTokenClassification, BertModel # bert: from old bert model = BertModel.from_pretrained(old_model_path) self.compare_two_model(old_model, model) # bert: from old bert-for-token model = BertModel.from_pretrained(old_model_for_token_path) self.compare_two_model(old_model, model) # bert-for-token: from old bert model = BertForTokenClassification.from_pretrained(old_model_path) self.compare_two_model(old_model_for_token, model) self.assertNotEqual(model.num_labels, 4) self.assertNotEqual(model.dropout.p, 0.3) # bert-for-token: from old bert-for-token model = BertForTokenClassification.from_pretrained(old_model_for_token_path) self.compare_two_model(old_model_for_token, model) self.assertEqual(model.num_labels, 4) self.assertEqual(model.dropout.p, 0.3) def compare_two_model(self, first_model: PretrainedModel, second_model: PretrainedModel): first_weight_name = "encoder.layers.8.linear2.weight" if first_model.__class__.__name__ != "BertModel": first_weight_name = "bert." + first_weight_name second_weight_name = "encoder.layers.8.linear2.weight" if second_model.__class__.__name__ != "BertModel": second_weight_name = "bert." + second_weight_name first_tensor = first_model.state_dict()[first_weight_name] second_tensor = second_model.state_dict()[second_weight_name] self.compare_two_weight(first_tensor, second_tensor) def compare_two_weight(self, first_tensor, second_tensor): diff = paddle.sum(first_tensor - second_tensor).item() self.assertEqual(diff, 0.0) @slow def test_paddlenlp_token_classification(self): versions = ["3.0.0b4"] for version in versions: install_package("paddlenlp", version=version) self.run_token_for_classification(version) uninstall_package("paddlenlp") @slow def test_bert_save_token_load(self): """bert -> token""" from paddlenlp.transformers import BertForTokenClassification, BertModel saved_dir = os.path.join(self.get_tempdir(), "bert-saved") bert: BertModel = BertModel.from_pretrained("bert-base-uncased") bert.save_pretrained(saved_dir) bert_for_token = BertForTokenClassification.from_pretrained(saved_dir) self.compare_two_model(bert, bert_for_token) @slow def test_bert_save_bert_load(self): """bert -> bert""" saved_dir = os.path.join(self.get_tempdir(), "bert-saved") bert: BertModel = BertModel.from_pretrained("bert-base-uncased") bert.save_pretrained(saved_dir) bert_loaded = BertModel.from_pretrained(saved_dir) self.compare_two_model(bert, bert_loaded) @slow def test_token_saved_bert_load(self): """token -> bert""" from paddlenlp.transformers import BertForTokenClassification, BertModel saved_dir = os.path.join(self.get_tempdir(), "bert-token-saved") bert_for_token = BertForTokenClassification.from_pretrained("bert-base-uncased") bert_for_token.save_pretrained(saved_dir) bert = BertModel.from_pretrained(saved_dir) self.compare_two_model(bert, bert_for_token) @slow def test_token_saved_token_load(self): """token -> token""" saved_dir = os.path.join(self.get_tempdir(), "bert-token-saved") bert_for_token = BertForTokenClassification.from_pretrained("bert-base-uncased") bert_for_token.save_pretrained(saved_dir) bert_for_token_loaded = BertForTokenClassification.from_pretrained(saved_dir) self.compare_two_model(bert_for_token, bert_for_token_loaded) @slow def test_auto_model(self): AutoModel.from_pretrained("bert-base-uncased") model = AutoModelForTokenClassification.from_pretrained("bert-base-uncased", num_classes=4, dropout=0.3) self.assertEqual(model.num_labels, 4) self.assertEqual(model.dropout.p, 0.3) model = AutoModelForQuestionAnswering.from_pretrained("bert-base-uncased", dropout=0.3) self.assertEqual(model.dropout.p, 0.3) @require_package("transformers", "torch") def test_bert_converter(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create common input input_ids = np.random.randint(100, 200, [1, 20]) # 2. forward the paddle model from paddlenlp.transformers import BertModel paddle_model = BertModel.from_pretrained( "hf-internal-testing/tiny-random-BertModel", from_hf_hub=True, cache_dir=tempdir ) paddle_model.eval() paddle_logit = paddle_model(paddle.to_tensor(input_ids))[0] # 3. forward the torch model import torch from transformers import BertModel torch_model = BertModel.from_pretrained("hf-internal-testing/tiny-random-BertModel", cache_dir=tempdir) torch_model.eval() torch_logit = torch_model(torch.tensor(input_ids), return_dict=False)[0] self.assertTrue( np.allclose( paddle_logit.detach().cpu().reshape([-1])[:9].numpy(), torch_logit.detach().cpu().reshape([-1])[:9].numpy(), rtol=1e-4, ) ) @require_package("transformers", "torch") def test_bert_converter_from_local_dir(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create common input input_ids = np.random.randint(100, 200, [1, 20]) # 2. forward the torch model import torch from transformers import BertModel torch_model = BertModel.from_pretrained("hf-internal-testing/tiny-random-BertModel") torch_model.eval() torch_model.save_pretrained(tempdir) torch_logit = torch_model(torch.tensor(input_ids), return_dict=False)[0] # 2. forward the paddle model from paddlenlp.transformers import BertModel paddle_model = BertModel.from_pretrained(tempdir, convert_from_torch=True) paddle_model.eval() paddle_logit = paddle_model(paddle.to_tensor(input_ids))[0] self.assertTrue( np.allclose( paddle_logit.detach().cpu().reshape([-1])[:9].numpy(), torch_logit.detach().cpu().reshape([-1])[:9].numpy(), rtol=1e-4, ) ) @parameterized.expand( [ ("BertModel",), # ("BertForMaskedLM",), TODO: need to tie weights # ("BertForPretraining", "BertForPreTraining"), TODO: need to tie weights ("BertForMultipleChoice",), ("BertForQuestionAnswering",), ("BertForSequenceClassification",), ("BertForTokenClassification",), ] ) @require_package("transformers", "torch") def test_bert_classes_from_local_dir(self, class_name, pytorch_class_name: str | None = None): pytorch_class_name = pytorch_class_name or class_name with tempfile.TemporaryDirectory() as tempdir: # 1. create common input input_ids = np.random.randint(100, 200, [1, 20]) # 2. forward the torch model import torch import transformers torch_model_class = getattr(transformers, pytorch_class_name) torch_model = torch_model_class.from_pretrained(self.torch_model_path) torch_model.eval() if "MultipleChoice" in class_name: # construct input for MultipleChoice Model torch_model.config.num_choices = random.randint(2, 10) input_ids = ( paddle.to_tensor(input_ids) .unsqueeze(1) .expand([-1, torch_model.config.num_choices, -1]) .cpu() .numpy() ) torch_model.save_pretrained(tempdir) torch_logit = torch_model(torch.tensor(input_ids), return_dict=False)[0] # 3. forward the paddle model from paddlenlp import transformers paddle_model_class = getattr(transformers, class_name) paddle_model = paddle_model_class.from_pretrained(tempdir, convert_from_torch=True) paddle_model.eval() paddle_logit = paddle_model(paddle.to_tensor(input_ids), return_dict=False)[0] self.assertTrue( np.allclose( paddle_logit.detach().cpu().reshape([-1])[:9].numpy(), torch_logit.detach().cpu().reshape([-1])[:9].numpy(), atol=1e-3, ) ) class BertModelIntegrationTest(ModelTesterPretrainedMixin, unittest.TestCase): base_model_class = BertModel hf_remote_test_model_path = "PaddleCI/tiny-random-bert" paddlehub_remote_test_model_path = "__internal_testing__/tiny-random-bert" @slow def test_inference_no_attention(self): model = BertModel.from_pretrained("bert-base-uncased") model.eval() input_ids = paddle.to_tensor([[0, 345, 232, 328, 740, 140, 1695, 69, 6078, 1588, 2]]) attention_mask = paddle.to_tensor([[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]) with paddle.no_grad(): output = model(input_ids, attention_mask=attention_mask)[0] expected_shape = [1, 11, 768] self.assertEqual(output.shape, expected_shape) expected_slice = paddle.to_tensor( [[[0.4249, 0.1008, 0.7531], [0.3771, 0.1188, 0.7467], [0.4152, 0.1098, 0.7108]]] ) self.assertTrue(paddle.allclose(output[:, 1:4, 1:4], expected_slice, atol=1e-4)) @slow def test_inference_with_attention(self): model = BertModel.from_pretrained("bert-base-uncased") model.eval() input_ids = paddle.to_tensor([[0, 345, 232, 328, 740, 140, 1695, 69, 6078, 1588, 2]]) attention_mask = paddle.to_tensor([[0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]) with paddle.no_grad(): output = model(input_ids, attention_mask=attention_mask)[0] expected_shape = [1, 11, 768] self.assertEqual(output.shape, expected_shape) expected_slice = paddle.to_tensor( [[[0.4249, 0.1008, 0.7531], [0.3771, 0.1188, 0.7467], [0.4152, 0.1098, 0.7108]]] ) self.assertTrue(paddle.allclose(output[:, 1:4, 1:4], expected_slice, atol=1e-4)) if __name__ == "__main__": unittest.main()