# coding=utf-8 # Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved. # Copyright 2021 The HuggingFace Inc. 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. """ Testing suite for the Paddle CLIP model. """ import copy import inspect import tempfile import unittest import numpy as np import paddle from paddle import nn from PIL import Image from paddlenlp.transformers import ( CLIPConfig, CLIPModel, CLIPProcessor, CLIPTextConfig, CLIPTextModel, CLIPTextModelWithProjection, CLIPVisionConfig, CLIPVisionModel, CLIPVisionModelWithProjection, CLIPVisionTransformer, ) from paddlenlp.transformers.clip.modeling import CLIP_PRETRAINED_MODEL_ARCHIVE_LIST from ...testing_utils import get_tests_dir, require_package, slow from ..test_configuration_common import ConfigTester from ..test_modeling_common import ( ModelTesterMixin, floats_tensor, ids_tensor, random_attention_mask, ) def _config_zero_init(config): configs_no_init = copy.deepcopy(config) for key in configs_no_init.__dict__.keys(): if "_range" in key or "_std" in key or "initializer_factor" in key or "layer_scale" in key: setattr(configs_no_init, key, 1e-10) return configs_no_init class CLIPVisionModelTester: def __init__( self, parent, batch_size=12, image_size=30, patch_size=2, num_channels=3, is_training=True, hidden_size=32, projection_dim=32, num_hidden_layers=5, num_attention_heads=4, intermediate_size=37, dropout=0.1, attention_dropout=0.1, initializer_range=0.02, scope=None, ): self.parent = parent self.batch_size = batch_size self.image_size = image_size self.patch_size = patch_size self.num_channels = num_channels self.is_training = is_training self.hidden_size = hidden_size self.projection_dim = projection_dim self.num_hidden_layers = num_hidden_layers self.num_attention_heads = num_attention_heads self.intermediate_size = intermediate_size self.dropout = dropout self.attention_dropout = attention_dropout self.initializer_range = initializer_range self.scope = scope # in ViT, the seq length equals the number of patches + 1 (we add 1 for the [CLS] token) num_patches = (image_size // patch_size) ** 2 self.seq_length = num_patches + 1 def prepare_config_and_inputs(self): pixel_values = floats_tensor([self.batch_size, self.num_channels, self.image_size, self.image_size]) config = self.get_config() return config, pixel_values def get_config(self): return CLIPVisionConfig( image_size=self.image_size, patch_size=self.patch_size, num_channels=self.num_channels, hidden_size=self.hidden_size, projection_dim=self.projection_dim, num_hidden_layers=self.num_hidden_layers, num_attention_heads=self.num_attention_heads, intermediate_size=self.intermediate_size, dropout=self.dropout, attention_dropout=self.attention_dropout, initializer_range=self.initializer_range, ) def create_and_check_model(self, config, pixel_values): model = CLIPVisionModel(config=config) model.eval() with paddle.no_grad(): result = model(pixel_values) # expected sequence length = num_patches + 1 (we add 1 for the [CLS] token) image_size = (self.image_size, self.image_size) patch_size = (self.patch_size, self.patch_size) num_patches = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0]) self.parent.assertEqual(result.last_hidden_state.shape, [self.batch_size, num_patches + 1, self.hidden_size]) self.parent.assertEqual(result.pooler_output.shape, [self.batch_size, self.hidden_size]) def create_and_check_model_with_projection(self, config, pixel_values): model = CLIPVisionModelWithProjection(config=config) model.eval() with paddle.no_grad(): result = model(pixel_values) # expected sequence length = num_patches + 1 (we add 1 for the [CLS] token) image_size = (self.image_size, self.image_size) patch_size = (self.patch_size, self.patch_size) num_patches = (image_size[1] // patch_size[1]) * (image_size[0] // patch_size[0]) self.parent.assertEqual(result.last_hidden_state.shape, [self.batch_size, num_patches + 1, self.hidden_size]) self.parent.assertEqual(result.image_embeds.shape, [self.batch_size, self.projection_dim]) def prepare_config_and_inputs_for_common(self): config_and_inputs = self.prepare_config_and_inputs() config, pixel_values = config_and_inputs inputs_dict = {"pixel_values": pixel_values} return config, inputs_dict class CLIPVisionModelTest(ModelTesterMixin, unittest.TestCase): """ Here we also overwrite some of the tests of test_modeling_common.py, as CLIP does not use input_ids, inputs_embeds, attention_mask and seq_length. """ all_model_classes = (CLIPVisionModel, CLIPVisionModelWithProjection) test_resize_embeddings = False use_test_model_name_list = False def setUp(self): self.model_tester = CLIPVisionModelTester(self) self.config_tester = ConfigTester(self, config_class=CLIPVisionConfig, has_text_modality=False, hidden_size=37) def test_config(self): self.config_tester.run_common_tests() @unittest.skip(reason="CLIP does not use inputs_embeds") def test_inputs_embeds(self): pass def test_model_common_attributes(self): config, _ = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: model = model_class(config) self.assertIsInstance(model.get_input_embeddings(), (nn.Layer)) x = model.get_output_embeddings() self.assertTrue(x is None or isinstance(x, nn.Linear)) def test_forward_signature(self): config, _ = self.model_tester.prepare_config_and_inputs_for_common() for model_class in self.all_model_classes: model = model_class(config) signature = inspect.signature(model.forward) # signature.parameters is an OrderedDict => so arg_names order is deterministic arg_names = [*signature.parameters.keys()] expected_arg_names = ["pixel_values"] self.assertListEqual(arg_names[:1], expected_arg_names) 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_model_with_projection(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model_with_projection(*config_and_inputs) def test_training(self): pass def test_training_gradient_checkpointing(self): pass @unittest.skip(reason="CLIPVisionModel has no base class and is not available in MODEL_MAPPING") def test_save_load_fast_init_from_base(self): pass @unittest.skip(reason="CLIPVisionModel has no base class and is not available in MODEL_MAPPING") def test_save_load_fast_init_to_base(self): pass @slow def test_model_from_pretrained(self): for model_name in CLIP_PRETRAINED_MODEL_ARCHIVE_LIST[:1]: model = CLIPVisionModel.from_pretrained(model_name) self.assertIsNotNone(model) @slow def test_model_with_projection_from_pretrained(self): for model_name in CLIP_PRETRAINED_MODEL_ARCHIVE_LIST[:1]: model = CLIPVisionModelWithProjection.from_pretrained(model_name) self.assertIsNotNone(model) if isinstance(model.vision_model, CLIPVisionTransformer): self.assertTrue(hasattr(model, "vision_projection")) class CLIPTextModelTester: def __init__( self, parent, batch_size=12, seq_length=7, is_training=True, use_input_mask=True, use_labels=True, vocab_size=99, hidden_size=32, projection_dim=32, num_hidden_layers=5, num_attention_heads=4, intermediate_size=37, dropout=0.1, attention_dropout=0.1, max_position_embeddings=512, initializer_range=0.02, scope=None, ): self.parent = 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_labels = use_labels self.vocab_size = vocab_size self.hidden_size = hidden_size self.projection_dim = projection_dim self.num_hidden_layers = num_hidden_layers self.num_attention_heads = num_attention_heads self.intermediate_size = intermediate_size self.dropout = dropout self.attention_dropout = attention_dropout self.max_position_embeddings = max_position_embeddings self.initializer_range = initializer_range self.scope = scope 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]) if input_mask is not None: batch_size, seq_length = input_mask.shape rnd_start_indices = np.random.randint(1, seq_length - 1, size=(batch_size,)) for batch_idx, start_index in enumerate(rnd_start_indices): input_mask[batch_idx, :start_index] = 1 input_mask[batch_idx, start_index:] = 0 config = self.get_config() return config, input_ids, input_mask def get_config(self): return CLIPTextConfig( vocab_size=self.vocab_size, hidden_size=self.hidden_size, projection_dim=self.projection_dim, num_hidden_layers=self.num_hidden_layers, num_attention_heads=self.num_attention_heads, intermediate_size=self.intermediate_size, dropout=self.dropout, attention_dropout=self.attention_dropout, max_position_embeddings=self.max_position_embeddings, initializer_range=self.initializer_range, ) def create_and_check_model(self, config, input_ids, input_mask): model = CLIPTextModel(config=config) model.eval() with paddle.no_grad(): result = model(input_ids, attention_mask=input_mask) self.parent.assertEqual(result.last_hidden_state.shape, [self.batch_size, self.seq_length, self.hidden_size]) self.parent.assertEqual(result.pooler_output.shape, [self.batch_size, self.hidden_size]) def create_and_check_model_with_projection(self, config, input_ids, input_mask): model = CLIPTextModelWithProjection(config=config) model.eval() with paddle.no_grad(): result = model(input_ids, attention_mask=input_mask) self.parent.assertEqual(result.last_hidden_state.shape, [self.batch_size, self.seq_length, self.hidden_size]) self.parent.assertEqual(result.text_embeds.shape, [self.batch_size, self.projection_dim]) def prepare_config_and_inputs_for_common(self): config_and_inputs = self.prepare_config_and_inputs() config, input_ids, input_mask = config_and_inputs inputs_dict = {"input_ids": input_ids, "attention_mask": input_mask} return config, inputs_dict class CLIPTextModelTest(ModelTesterMixin, unittest.TestCase): all_model_classes = (CLIPTextModel, CLIPTextModelWithProjection) use_test_model_name_list = False def setUp(self): self.model_tester = CLIPTextModelTester(self) self.config_tester = ConfigTester(self, config_class=CLIPTextConfig, hidden_size=37) def test_config(self): 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_model_with_projection(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model_with_projection(*config_and_inputs) def test_training(self): pass def test_training_gradient_checkpointing(self): pass @unittest.skip(reason="CLIP does not use inputs_embeds") def test_inputs_embeds(self): pass @unittest.skip(reason="CLIPTextModel has no base class and is not available in MODEL_MAPPING") def test_save_load_fast_init_from_base(self): pass @unittest.skip(reason="CLIPTextModel has no base class and is not available in MODEL_MAPPING") def test_save_load_fast_init_to_base(self): pass @slow def test_model_from_pretrained(self): for model_name in CLIP_PRETRAINED_MODEL_ARCHIVE_LIST[:1]: model = CLIPTextModel.from_pretrained(model_name) self.assertIsNotNone(model) @slow def test_model_with_projection_from_pretrained(self): for model_name in CLIP_PRETRAINED_MODEL_ARCHIVE_LIST[:1]: model = CLIPTextModelWithProjection.from_pretrained(model_name) self.assertIsNotNone(model) self.assertTrue(hasattr(model, "text_projection")) class CLIPModelTester: def __init__(self, parent, text_kwargs=None, vision_kwargs=None, is_training=True): if text_kwargs is None: text_kwargs = {} if vision_kwargs is None: vision_kwargs = {} self.parent = parent self.text_model_tester = CLIPTextModelTester(parent, **text_kwargs) self.vision_model_tester = CLIPVisionModelTester(parent, **vision_kwargs) self.is_training = is_training def prepare_config_and_inputs(self): text_config, input_ids, attention_mask = self.text_model_tester.prepare_config_and_inputs() vision_config, pixel_values = self.vision_model_tester.prepare_config_and_inputs() config = self.get_config() return config, input_ids, attention_mask, pixel_values def get_config(self): return CLIPConfig.from_text_vision_configs( self.text_model_tester.get_config(), self.vision_model_tester.get_config(), projection_dim=64 ) def create_and_check_model(self, config, input_ids, attention_mask, pixel_values): model = CLIPModel(config) model.eval() with paddle.no_grad(): result = model(input_ids, pixel_values, attention_mask=attention_mask) self.parent.assertEqual( result.logits_per_image.shape, [self.vision_model_tester.batch_size, self.text_model_tester.batch_size] ) self.parent.assertEqual( result.logits_per_text.shape, [self.text_model_tester.batch_size, self.vision_model_tester.batch_size] ) def prepare_config_and_inputs_for_common(self): config_and_inputs = self.prepare_config_and_inputs() config, input_ids, attention_mask, pixel_values = config_and_inputs inputs_dict = { "input_ids": input_ids, "attention_mask": attention_mask, "pixel_values": pixel_values, "return_loss": True, } return config, inputs_dict class CLIPModelTest(ModelTesterMixin, unittest.TestCase): all_model_classes = (CLIPModel,) test_resize_embeddings = False test_attention_outputs = False use_test_model_name_list = False def setUp(self): self.model_tester = CLIPModelTester(self) def test_model(self): config_and_inputs = self.model_tester.prepare_config_and_inputs() self.model_tester.create_and_check_model(*config_and_inputs) @unittest.skip(reason="Hidden_states is tested in individual model tests") def test_hidden_states_output(self): pass @unittest.skip(reason="Inputs_embeds is tested in individual model tests") def test_inputs_embeds(self): pass @unittest.skip(reason="Retain_grad is tested in individual model tests") def test_retain_grad_hidden_states_attentions(self): pass @unittest.skip(reason="CLIPModel does not have input/output embeddings") def test_model_common_attributes(self): pass # override as the `logit_scale` parameter initialization is different for CLIP def test_initialization(self): config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common() configs_no_init = _config_zero_init(config) for model_class in self.all_model_classes: model = model_class(config=configs_no_init) for name, param in model.named_parameters(): if not param.stop_gradient: # check if `logit_scale` is initialized as per the original implementation if name == "logit_scale": self.assertAlmostEqual( param.item(), np.log(1 / 0.07), delta=1e-3, msg=f"Parameter {name} of model {model_class} seems not properly initialized", ) else: self.assertIn( ((param.mean() * 1e9).round() / 1e9).item(), [0.0, 1.0], msg=f"Parameter {name} of model {model_class} seems not properly initialized", ) def test_load_vision_text_config(self): config, inputs_dict = self.model_tester.prepare_config_and_inputs_for_common() # Save CLIPConfig and check if we can load CLIPVisionConfig from it with tempfile.TemporaryDirectory() as tmp_dir_name: config.save_pretrained(tmp_dir_name) vision_config = CLIPVisionConfig.from_pretrained(tmp_dir_name) self.assertDictEqual(config.vision_config.to_dict(), vision_config.to_dict()) # Save CLIPConfig and check if we can load CLIPTextConfig from it with tempfile.TemporaryDirectory() as tmp_dir_name: config.save_pretrained(tmp_dir_name) text_config = CLIPTextConfig.from_pretrained(tmp_dir_name) self.assertDictEqual(config.text_config.to_dict(), text_config.to_dict()) @slow def test_model_from_pretrained(self): for model_name in CLIP_PRETRAINED_MODEL_ARCHIVE_LIST[:1]: model = CLIPModel.from_pretrained(model_name) self.assertIsNotNone(model) # We will verify our results on an image of cute cats def prepare_img(): CUTE_CATS = get_tests_dir("fixtures/tests_samples/COCO/000000039769.png") image = Image.open(CUTE_CATS) return image class CLIPModelCompatibilityTest(unittest.TestCase): model_id = "hf-internal-testing/tiny-random-CLIPModel" def setUp(self): # 1. create input processor = CLIPProcessor.from_pretrained(self.model_id, from_hf_hub=True) image = prepare_img() self.inputs = processor( text=["a photo of a cat", "a photo of a dog"], images=image, padding=True, return_tensors="np" ) @unittest.skip("model diff exists, need to be fixed") @require_package("transformers", "torch") def test_clip_converter(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create input inputs = self.inputs # 2. forward the paddle model from paddlenlp.transformers import CLIPModel paddle_model = CLIPModel.from_pretrained(self.model_id, from_hf_hub=True, cache_dir=tempdir) paddle_model.eval() paddle_logit = paddle_model( input_ids=paddle.to_tensor(inputs["input_ids"]), pixel_values=paddle.to_tensor(inputs["pixel_values"]) ) # 3. forward the torch model import torch from transformers import CLIPModel torch_model = CLIPModel.from_pretrained(self.model_id, cache_dir=tempdir) torch_model.eval() torch_logit = torch_model( input_ids=torch.tensor(inputs["input_ids"]), pixel_values=torch.tensor(inputs["pixel_values"]) ) # 4. compare results self.assertTrue( np.allclose( paddle_logit.logits_per_image.detach().cpu().numpy(), torch_logit.logits_per_image.detach().cpu().numpy(), rtol=1e-4, ) ) self.assertTrue( np.allclose( paddle_logit.logits_per_text.detach().cpu().numpy(), torch_logit.logits_per_text.detach().cpu().numpy(), rtol=1e-4, ) ) @unittest.skip("model diff exists, need to be fixed") @require_package("transformers", "torch") def test_clip_converter_from_local_dir(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create input inputs = self.inputs # 2. forward the torch model import torch from transformers import CLIPModel torch_model = CLIPModel.from_pretrained(self.model_id) torch_model.eval() torch_model.save_pretrained(tempdir) torch_logit = torch_model( input_ids=torch.tensor(inputs["input_ids"]), pixel_values=torch.tensor(inputs["pixel_values"]) ) # 3. forward the paddle model from paddlenlp.transformers import CLIPModel paddle_model = CLIPModel.from_pretrained(tempdir, convert_from_torch=True) paddle_model.eval() paddle_logit = paddle_model( input_ids=paddle.to_tensor(inputs["input_ids"]), pixel_values=paddle.to_tensor(inputs["pixel_values"]) ) # 4. compare results self.assertTrue( np.allclose( paddle_logit.logits_per_image.detach().cpu().numpy(), torch_logit.logits_per_image.detach().cpu().numpy(), rtol=1e-4, ) ) self.assertTrue( np.allclose( paddle_logit.logits_per_text.detach().cpu().numpy(), torch_logit.logits_per_text.detach().cpu().numpy(), rtol=1e-4, ) ) @require_package("transformers", "torch") def test_clip_text_converter(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create input inputs = self.inputs # 2. forward the paddle model from paddlenlp.transformers import CLIPTextModel paddle_model = CLIPTextModel.from_pretrained(self.model_id, from_hf_hub=True, cache_dir=tempdir) paddle_model.eval() paddle_logit = paddle_model(input_ids=paddle.to_tensor(inputs["input_ids"])) # 3. forward the torch model import torch from transformers import CLIPTextModel torch_model = CLIPTextModel.from_pretrained(self.model_id, cache_dir=tempdir) torch_model.eval() torch_logit = torch_model(input_ids=torch.tensor(inputs["input_ids"])) # 4. compare results np.testing.assert_equal( paddle_logit.last_hidden_state.shape, torch_logit.last_hidden_state.shape, ) np.testing.assert_equal( paddle_logit.pooler_output.shape, torch_logit.pooler_output.shape, ) @require_package("transformers", "torch") def test_clip_vision_converter(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create input inputs = self.inputs # 2. forward the paddle model from paddlenlp.transformers import CLIPVisionModel paddle_model = CLIPVisionModel.from_pretrained(self.model_id, from_hf_hub=True, cache_dir=tempdir) paddle_model.eval() paddle_logit = paddle_model(pixel_values=paddle.to_tensor(inputs["pixel_values"])) # 3. forward the torch model import torch from transformers import CLIPVisionModel torch_model = CLIPVisionModel.from_pretrained(self.model_id, cache_dir=tempdir) torch_model.eval() torch_logit = torch_model(pixel_values=torch.tensor(inputs["pixel_values"])) # 4. compare results np.testing.assert_equal( paddle_logit.last_hidden_state.shape, torch_logit.last_hidden_state.shape, ) np.testing.assert_equal( paddle_logit.pooler_output.shape, torch_logit.pooler_output.shape, ) @require_package("transformers", "torch") def test_clip_text_with_projection_converter(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create input inputs = self.inputs # 2. forward the paddle model from paddlenlp.transformers import CLIPTextModelWithProjection paddle_model = CLIPTextModelWithProjection.from_pretrained( self.model_id, from_hf_hub=True, cache_dir=tempdir ) paddle_model.eval() paddle_logit = paddle_model(input_ids=paddle.to_tensor(inputs["input_ids"])) # 3. forward the torch model import torch from transformers import CLIPTextModelWithProjection torch_model = CLIPTextModelWithProjection.from_pretrained( self.model_id, cache_dir=tempdir, ignore_mismatched_sizes=True ) torch_model.eval() torch_logit = torch_model(input_ids=torch.tensor(inputs["input_ids"])) # 4. compare results np.testing.assert_equal( paddle_logit.last_hidden_state.shape, torch_logit.last_hidden_state.shape, ) @require_package("transformers", "torch") def test_clip_vision_with_projection_converter(self): with tempfile.TemporaryDirectory() as tempdir: # 1. create input inputs = self.inputs # 2. forward the paddle model from paddlenlp.transformers import CLIPVisionModelWithProjection paddle_model = CLIPVisionModelWithProjection.from_pretrained( self.model_id, from_hf_hub=True, cache_dir=tempdir ) paddle_model.eval() paddle_logit = paddle_model(pixel_values=paddle.to_tensor(inputs["pixel_values"])) # 3. forward the torch model import torch from transformers import CLIPVisionModelWithProjection torch_model = CLIPVisionModelWithProjection.from_pretrained( self.model_id, cache_dir=tempdir, ignore_mismatched_sizes=True ) torch_model.eval() torch_logit = torch_model(pixel_values=torch.tensor(inputs["pixel_values"])) # 4. compare results np.testing.assert_equal( paddle_logit.last_hidden_state.shape, torch_logit.last_hidden_state.shape, ) class CLIPModelIntegrationTest(unittest.TestCase): @slow def test_inference(self): model_name = "openai/clip-vit-base-patch32" model = CLIPModel.from_pretrained(model_name) model.eval() processor = CLIPProcessor.from_pretrained(model_name) image = prepare_img() inputs = processor( text=["a photo of a cat", "a photo of a dog"], images=image, padding=True, return_tensors="pd" ) # forward pass with paddle.no_grad(): outputs = model(**inputs) # verify the logits self.assertEqual( outputs.logits_per_image.shape, [inputs.pixel_values.shape[0], inputs.input_ids.shape[0]], ) self.assertEqual( outputs.logits_per_text.shape, [inputs.input_ids.shape[0], inputs.pixel_values.shape[0]], ) expected_logits = paddle.to_tensor([[24.5701, 19.3049]]) self.assertTrue(paddle.allclose(outputs.logits_per_image, expected_logits, atol=1e-3))