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

Add support for args to ProcessorMixin for backward compatibility (#33479)

Browse Source 
* add check and prepare args for BC to ProcessorMixin, improve ProcessorTesterMixin

* change size and crop_size in processor kwargs tests to do_rescale and rescale_factor

* remove unnecessary llava processor kwargs test overwrite

* nit

* change data_arg_name to input_name

* Remove unnecessary test override

* Remove unnecessary tests Paligemma

* Move test_prepare_and_validate_optional_call_args to TesterMixin, add docstring
This commit is contained in:
Yoni Gozlan
2024-09-20 11:40:59 -04:00
committed by GitHub
parent 31caf0b95f
commit c0c6815dc9
10 changed files with 173 additions and 812 deletions
Download Patch File Download Diff File Expand all files Collapse all files

203
tests/models/llava_onevision/test_processing_llava_onevision.py
View File

@@ -16,7 +16,7 @@ import shutil
import tempfile
import unittest
from transformers.testing_utils import require_torch, require_vision
from transformers.testing_utils import require_vision
from transformers.utils import is_vision_available
from ...test_processing_common import ProcessorTesterMixin
@@ -100,204 +100,3 @@ class LlavaOnevisionProcessorTest(ProcessorTesterMixin, unittest.TestCase):
formatted_prompt = processor.apply_chat_template(messages, add_generation_prompt=True)
self.assertEqual(expected_prompt, formatted_prompt)
@require_torch
@require_vision
def test_image_processor_defaults_preserved_by_image_kwargs(self):
# Rewrite as llava-next image processor return pixel values with an added dimesion for image patches
if "image_processor" not in self.processor_class.attributes:
self.skipTest(f"image_processor attribute not present in {self.processor_class}")
image_processor = self.get_component("image_processor", size=(234, 234))
video_processor = self.get_component("video_processor", size=(234, 234))
tokenizer = self.get_component("tokenizer", max_length=117)
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = "lower newer"
image_input = self.prepare_image_inputs()
inputs = processor(text=input_str, images=image_input)
# added dimension for image patches
self.assertEqual(len(inputs["pixel_values"][0][0][0]), 234)
@require_torch
@require_vision
def test_kwargs_overrides_default_image_processor_kwargs(self):
if "image_processor" not in self.processor_class.attributes:
self.skipTest(f"image_processor attribute not present in {self.processor_class}")
image_processor = self.get_component("image_processor", crop_size=(234, 234))
video_processor = self.get_component("video_processor", size=(234, 234))
tokenizer = self.get_component("tokenizer", max_length=117)
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = "lower newer"
image_input = self.prepare_image_inputs()
inputs = processor(text=input_str, images=image_input, size=[224, 224])
# added dimension for image patches
self.assertEqual(len(inputs["pixel_values"][0][0][0]), 224)
@require_torch
@require_vision
def test_unstructured_kwargs(self):
image_processor = self.get_component("image_processor")
video_processor = self.get_component("video_processor")
tokenizer = self.get_component("tokenizer")
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = "lower newer"
image_input = self.prepare_image_inputs()
inputs = processor(
text=input_str,
images=image_input,
return_tensors="pt",
size={"height": 214, "width": 214},
padding="max_length",
max_length=76,
)
# added dimension for image patches
self.assertEqual(inputs["pixel_values"].shape[3], 214)
self.assertEqual(len(inputs["input_ids"][0]), 76)
@require_torch
@require_vision
def test_unstructured_kwargs_batched(self):
image_processor = self.get_component("image_processor")
video_processor = self.get_component("video_processor")
tokenizer = self.get_component("tokenizer")
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = ["lower newer", "upper older longer string"]
image_input = self.prepare_image_inputs() * 2
inputs = processor(
text=input_str,
images=image_input,
return_tensors="pt",
size={"height": 214, "width": 214},
padding="longest",
max_length=76,
)
self.assertEqual(inputs["pixel_values"].shape[3], 214)
self.assertEqual(len(inputs["input_ids"][0]), 4)
@require_torch
@require_vision
def test_structured_kwargs_nested(self):
image_processor = self.get_component("image_processor")
video_processor = self.get_component("video_processor")
tokenizer = self.get_component("tokenizer")
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = "lower newer"
image_input = self.prepare_image_inputs()
# Define the kwargs for each modality
all_kwargs = {
"common_kwargs": {"return_tensors": "pt"},
"images_kwargs": {"size": {"height": 214, "width": 214}},
"text_kwargs": {"padding": "max_length", "max_length": 76},
}
inputs = processor(text=input_str, images=image_input, **all_kwargs)
self.skip_processor_without_typed_kwargs(processor)
self.assertEqual(inputs["pixel_values"].shape[3], 214)
self.assertEqual(len(inputs["input_ids"][0]), 76)
@require_torch
@require_vision
def test_structured_kwargs_nested_from_dict(self):
image_processor = self.get_component("image_processor")
video_processor = self.get_component("video_processor")
tokenizer = self.get_component("tokenizer")
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = "lower newer"
image_input = self.prepare_image_inputs()
# Define the kwargs for each modality
all_kwargs = {
"common_kwargs": {"return_tensors": "pt"},
"images_kwargs": {"size": {"height": 214, "width": 214}},
"text_kwargs": {"padding": "max_length", "max_length": 76},
}
inputs = processor(text=input_str, images=image_input, **all_kwargs)
self.assertEqual(inputs["pixel_values"].shape[3], 214)
self.assertEqual(len(inputs["input_ids"][0]), 76)
@require_torch
@require_vision
def test_doubly_passed_kwargs(self):
image_processor = self.get_component("image_processor")
video_processor = self.get_component("video_processor")
tokenizer = self.get_component("tokenizer")
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = ["lower newer"]
image_input = self.prepare_image_inputs()
with self.assertRaises(ValueError):
_ = processor(
text=input_str,
images=image_input,
images_kwargs={"size": {"height": 222, "width": 222}},
size={"height": 214, "width": 214},
)
@require_vision
@require_torch
def test_kwargs_overrides_default_tokenizer_kwargs(self):
image_processor = self.get_component("image_processor")
video_processor = self.get_component("video_processor")
tokenizer = self.get_component("tokenizer", max_length=117)
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = "lower newer"
image_input = self.prepare_image_inputs()
inputs = processor(text=input_str, images=image_input, return_tensors="pt", max_length=112)
self.assertEqual(len(inputs["input_ids"][0]), 2)
@require_vision
@require_torch
def test_tokenizer_defaults_preserved_by_kwargs(self):
image_processor = self.get_component("image_processor")
video_processor = self.get_component("video_processor")
tokenizer = self.get_component("tokenizer", max_length=117)
processor = self.processor_class(
tokenizer=tokenizer, image_processor=image_processor, video_processor=video_processor
)
self.skip_processor_without_typed_kwargs(processor)
input_str = "lower newer"
image_input = self.prepare_image_inputs()
inputs = processor(text=input_str, images=image_input, return_tensors="pt")
self.assertEqual(len(inputs["input_ids"][0]), 2)