SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Fall back to slow image processor in ImageProcessingAuto when no fast processor available (#34785)

Browse Source 
* refactor image_processing_auto logic

* fix fast image processor tests

* Fix tests fast vit image processor

* Add safeguard when use_fast True and torchvision not available

* change default use_fast back to None, add warnings

* remove debugging print

* call get_image_processor_class_from_name once
This commit is contained in:
Yoni Gozlan
2024-12-15 14:00:36 -05:00
committed by GitHub
parent ca03842cdc
commit 5615a39369
8 changed files with 72 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
tests/models/vision_text_dual_encoder/test_processor_vision_text_dual_encoder.py
View File

@@ -21,13 +21,13 @@ import unittest
from transformers import BertTokenizerFast
from transformers.models.bert.tokenization_bert import VOCAB_FILES_NAMES, BertTokenizer
from transformers.testing_utils import require_tokenizers, require_vision
from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available
from transformers.utils import IMAGE_PROCESSOR_NAME, is_torchvision_available, is_vision_available
from ...test_processing_common import ProcessorTesterMixin
if is_vision_available():
from transformers import VisionTextDualEncoderProcessor, ViTImageProcessor
from transformers import VisionTextDualEncoderProcessor, ViTImageProcessor, ViTImageProcessorFast
@require_tokenizers
@@ -63,6 +63,8 @@ class VisionTextDualEncoderProcessorTest(ProcessorTesterMixin, unittest.TestCase
return BertTokenizer.from_pretrained(self.tmpdirname, **kwargs)
def get_image_processor(self, **kwargs):
if is_torchvision_available():
return ViTImageProcessorFast.from_pretrained(self.tmpdirname, **kwargs)
return ViTImageProcessor.from_pretrained(self.tmpdirname, **kwargs)
def tearDown(self):
@@ -81,7 +83,7 @@ class VisionTextDualEncoderProcessorTest(ProcessorTesterMixin, unittest.TestCase
self.assertIsInstance(processor.tokenizer, (BertTokenizer, BertTokenizerFast))
self.assertEqual(processor.image_processor.to_json_string(), image_processor.to_json_string())
self.assertIsInstance(processor.image_processor, ViTImageProcessor)
self.assertIsInstance(processor.image_processor, (ViTImageProcessor, ViTImageProcessorFast))
def test_save_load_pretrained_additional_features(self):
processor = VisionTextDualEncoderProcessor(
@@ -100,7 +102,7 @@ class VisionTextDualEncoderProcessorTest(ProcessorTesterMixin, unittest.TestCase
self.assertIsInstance(processor.tokenizer, (BertTokenizer, BertTokenizerFast))
self.assertEqual(processor.image_processor.to_json_string(), image_processor_add_kwargs.to_json_string())
self.assertIsInstance(processor.image_processor, ViTImageProcessor)
self.assertIsInstance(processor.image_processor, (ViTImageProcessor, ViTImageProcessorFast))
def test_image_processor(self):
image_processor = self.get_image_processor()
@@ -110,8 +112,8 @@ class VisionTextDualEncoderProcessorTest(ProcessorTesterMixin, unittest.TestCase
image_input = self.prepare_image_inputs()
input_feat_extract = image_processor(image_input, return_tensors="np")
input_processor = processor(images=image_input, return_tensors="np")
input_feat_extract = image_processor(image_input, return_tensors="pt")
input_processor = processor(images=image_input, return_tensors="pt")
for key in input_feat_extract.keys():
self.assertAlmostEqual(input_feat_extract[key].sum(), input_processor[key].sum(), delta=1e-2)