Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add auto model for image-text-to-text #32472

Merged
merged 7 commits into from
Oct 8, 2024
Merged

Add auto model for image-text-to-text #32472

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
2eeec14
Add Auto model for image-text-to-text
yonigozlan Aug 6, 2024
65a7144
Remove donut from processing auto, add chameleon ti image text to tex…
yonigozlan Aug 7, 2024
6186091
add qwen2_vl and llava_onevision
yonigozlan Sep 12, 2024
ff5d2c6
add pixtral to auto model for image-text-to-text
yonigozlan Sep 24, 2024
27f50d3
add mllama and idefics3
yonigozlan Oct 2, 2024
13387d5
remove models in IGNORE_NON_AUTO_CONFIGURED
yonigozlan Oct 4, 2024
9311ba2
add AutoModelForImageTextToText to tests and doc
yonigozlan Oct 6, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions docs/source/en/model_doc/auto.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -381,3 +381,7 @@ The following auto classes are available for the following multimodal tasks.
### FlaxAutoModelForVision2Seq

[[autodoc]] FlaxAutoModelForVision2Seq

### AutoModelForImageTextToText

[[autodoc]] AutoModelForImageTextToText
4 changes: 4 additions & 0 deletions docs/source/ja/model_doc/auto.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -368,3 +368,7 @@ AutoModel.register(NewModelConfig, NewModel)
### FlaxAutoModelForVision2Seq

[[autodoc]] FlaxAutoModelForVision2Seq

### AutoModelForImageTextToText

[[autodoc]] AutoModelForImageTextToText
4 changes: 4 additions & 0 deletions src/transformers/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1405,6 +1405,7 @@
"MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING",
"MODEL_FOR_IMAGE_MAPPING",
"MODEL_FOR_IMAGE_SEGMENTATION_MAPPING",
"MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING",
"MODEL_FOR_IMAGE_TO_IMAGE_MAPPING",
"MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING",
"MODEL_FOR_KEYPOINT_DETECTION_MAPPING",
Expand Down Expand Up @@ -1446,6 +1447,7 @@
"AutoModelForDocumentQuestionAnswering",
"AutoModelForImageClassification",
"AutoModelForImageSegmentation",
"AutoModelForImageTextToText",
"AutoModelForImageToImage",
"AutoModelForInstanceSegmentation",
"AutoModelForKeypointDetection",
Expand Down Expand Up @@ -6251,6 +6253,7 @@
MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING,
MODEL_FOR_IMAGE_MAPPING,
MODEL_FOR_IMAGE_SEGMENTATION_MAPPING,
MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING,
MODEL_FOR_IMAGE_TO_IMAGE_MAPPING,
MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING,
MODEL_FOR_KEYPOINT_DETECTION_MAPPING,
Expand Down Expand Up @@ -6292,6 +6295,7 @@
AutoModelForDocumentQuestionAnswering,
AutoModelForImageClassification,
AutoModelForImageSegmentation,
AutoModelForImageTextToText,
AutoModelForImageToImage,
AutoModelForInstanceSegmentation,
AutoModelForKeypointDetection,
Expand Down
4 changes: 4 additions & 0 deletions src/transformers/models/auto/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -74,6 +74,7 @@
"MODEL_FOR_UNIVERSAL_SEGMENTATION_MAPPING",
"MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING",
"MODEL_FOR_VISION_2_SEQ_MAPPING",
"MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING",
"MODEL_FOR_VISUAL_QUESTION_ANSWERING_MAPPING",
"MODEL_MAPPING",
"MODEL_WITH_LM_HEAD_MAPPING",
Expand Down Expand Up @@ -119,6 +120,7 @@
"AutoModelWithLMHead",
"AutoModelForZeroShotImageClassification",
"AutoModelForZeroShotObjectDetection",
"AutoModelForImageTextToText",
]

try:
Expand Down Expand Up @@ -238,6 +240,7 @@
MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING,
MODEL_FOR_IMAGE_MAPPING,
MODEL_FOR_IMAGE_SEGMENTATION_MAPPING,
MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING,
MODEL_FOR_IMAGE_TO_IMAGE_MAPPING,
MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING,
MODEL_FOR_KEYPOINT_DETECTION_MAPPING,
Expand Down Expand Up @@ -279,6 +282,7 @@
AutoModelForDocumentQuestionAnswering,
AutoModelForImageClassification,
AutoModelForImageSegmentation,
AutoModelForImageTextToText,
AutoModelForImageToImage,
AutoModelForInstanceSegmentation,
AutoModelForKeypointDetection,
Expand Down
36 changes: 36 additions & 0 deletions src/transformers/models/auto/modeling_auto.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -753,6 +753,32 @@
]
)

MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING_NAMES = OrderedDict(
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should the existing classes be deleted from MODEL_FOR_VISION_2_SEQ_MAPPING_NAMES or do we keep them there?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I just removed the ones that previously were in IGNORE_NON_AUTO_CONFIGURED.
Won't deleting them from MODEL_FOR_VISION_2_SEQ_MAPPING_NAMES be a problem for BC?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can also add llava-next-video and video-llava? Those two can work with image+text inputs, as well as video+text

[
("blip", "BlipForConditionalGeneration"),
("blip-2", "Blip2ForConditionalGeneration"),
("chameleon", "ChameleonForConditionalGeneration"),
("fuyu", "FuyuForCausalLM"),
("git", "GitForCausalLM"),
("idefics", "IdeficsForVisionText2Text"),
("idefics2", "Idefics2ForConditionalGeneration"),
("idefics3", "Idefics3ForConditionalGeneration"),
("instructblip", "InstructBlipForConditionalGeneration"),
("kosmos-2", "Kosmos2ForConditionalGeneration"),
("llava", "LlavaForConditionalGeneration"),
("llava_next", "LlavaNextForConditionalGeneration"),
("llava_onevision", "LlavaOnevisionForConditionalGeneration"),
("mllama", "MllamaForConditionalGeneration"),
("paligemma", "PaliGemmaForConditionalGeneration"),
("pix2struct", "Pix2StructForConditionalGeneration"),
("pixtral", "LlavaForConditionalGeneration"),
("qwen2_vl", "Qwen2VLForConditionalGeneration"),
("udop", "UdopForConditionalGeneration"),
("vipllava", "VipLlavaForConditionalGeneration"),
("vision-encoder-decoder", "VisionEncoderDecoderModel"),
]
)

MODEL_FOR_MASKED_LM_MAPPING_NAMES = OrderedDict(
[
# Model for Masked LM mapping
Expand Down Expand Up @@ -1413,6 +1439,9 @@
CONFIG_MAPPING_NAMES, MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING_NAMES
)
MODEL_FOR_VISION_2_SEQ_MAPPING = _LazyAutoMapping(CONFIG_MAPPING_NAMES, MODEL_FOR_VISION_2_SEQ_MAPPING_NAMES)
MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING = _LazyAutoMapping(
CONFIG_MAPPING_NAMES, MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING_NAMES
)
MODEL_FOR_VISUAL_QUESTION_ANSWERING_MAPPING = _LazyAutoMapping(
CONFIG_MAPPING_NAMES, MODEL_FOR_VISUAL_QUESTION_ANSWERING_MAPPING_NAMES
)
Expand Down Expand Up @@ -1707,6 +1736,13 @@ class AutoModelForVision2Seq(_BaseAutoModelClass):
AutoModelForVision2Seq = auto_class_update(AutoModelForVision2Seq, head_doc="vision-to-text modeling")


class AutoModelForImageTextToText(_BaseAutoModelClass):
_model_mapping = MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING


AutoModelForImageTextToText = auto_class_update(AutoModelForImageTextToText, head_doc="image-text-to-text modeling")


class AutoModelForAudioClassification(_BaseAutoModelClass):
_model_mapping = MODEL_FOR_AUDIO_CLASSIFICATION_MAPPING

Expand Down
1 change: 1 addition & 0 deletions src/transformers/models/auto/processing_auto.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,6 +99,7 @@
("trocr", "TrOCRProcessor"),
("tvlt", "TvltProcessor"),
("tvp", "TvpProcessor"),
("udop", "UdopProcessor"),
("unispeech", "Wav2Vec2Processor"),
("unispeech-sat", "Wav2Vec2Processor"),
("video_llava", "VideoLlavaProcessor"),
Expand Down
10 changes: 10 additions & 0 deletions src/transformers/utils/dummy_pt_objects.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -707,6 +707,9 @@ def __init__(self, *args, **kwargs):
MODEL_FOR_IMAGE_SEGMENTATION_MAPPING = None


MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING = None


MODEL_FOR_IMAGE_TO_IMAGE_MAPPING = None


Expand Down Expand Up @@ -874,6 +877,13 @@ def __init__(self, *args, **kwargs):
requires_backends(self, ["torch"])


class AutoModelForImageTextToText(metaclass=DummyObject):
_backends = ["torch"]

def __init__(self, *args, **kwargs):
requires_backends(self, ["torch"])


class AutoModelForImageToImage(metaclass=DummyObject):
_backends = ["torch"]

Expand Down
Loading