add ONNX support for LeVit (#18154)

Co-authored-by: Guilhem Chéron <guilhemc@authentifier.com>

docs/source/en/serialization.mdx

@@ -72,6 +72,7 @@ Ready-made configurations include the following architectures:
 - I-BERT
 - LayoutLM
 - LayoutLMv3
+- LeViT
 - LongT5
 - M2M100
 - Marian

src/transformers/models/levit/__init__.py

@@ -20,7 +20,7 @@ from typing import TYPE_CHECKING
 from ...utils import OptionalDependencyNotAvailable, _LazyModule, is_torch_available, is_vision_available
 
 
-_import_structure = {"configuration_levit": ["LEVIT_PRETRAINED_CONFIG_ARCHIVE_MAP", "LevitConfig"]}
+_import_structure = {"configuration_levit": ["LEVIT_PRETRAINED_CONFIG_ARCHIVE_MAP", "LevitConfig", "LevitOnnxConfig"]}
 
 try:
     if not is_vision_available():
@@ -46,7 +46,7 @@ else:
 
 
 if TYPE_CHECKING:
-    from .configuration_levit import LEVIT_PRETRAINED_CONFIG_ARCHIVE_MAP, LevitConfig
+    from .configuration_levit import LEVIT_PRETRAINED_CONFIG_ARCHIVE_MAP, LevitConfig, LevitOnnxConfig
 
     try:
         if not is_vision_available():

src/transformers/models/levit/configuration_levit.py

@@ -14,7 +14,13 @@
 # limitations under the License.
 """ LeViT model configuration"""
 
+from collections import OrderedDict
+from typing import Mapping
+
+from packaging import version
+
 from ...configuration_utils import PretrainedConfig
+from ...onnx import OnnxConfig
 from ...utils import logging
 
 
@@ -120,3 +126,21 @@ class LevitConfig(PretrainedConfig):
             ["Subsample", key_dim[0], hidden_sizes[0] // key_dim[0], 4, 2, 2],
             ["Subsample", key_dim[0], hidden_sizes[1] // key_dim[0], 4, 2, 2],
         ]
+
+
+# Copied from transformers.models.vit.configuration_vit.ViTOnnxConfig
+class LevitOnnxConfig(OnnxConfig):
+
+    torch_onnx_minimum_version = version.parse("1.11")
+
+    @property
+    def inputs(self) -> Mapping[str, Mapping[int, str]]:
+        return OrderedDict(
+            [
+                ("pixel_values", {0: "batch", 1: "sequence"}),
+            ]
+        )
+
+    @property
+    def atol_for_validation(self) -> float:
+        return 1e-4

src/transformers/onnx/features.py

@@ -333,6 +333,9 @@ class FeaturesManager:
             "token-classification",
             onnx_config_cls="models.layoutlmv3.LayoutLMv3OnnxConfig",
         ),
+        "levit": supported_features_mapping(
+            "default", "image-classification", onnx_config_cls="models.levit.LevitOnnxConfig"
+        ),
         "longt5": supported_features_mapping(
             "default",
             "default-with-past",

tests/onnx/test_onnx_v2.py

@@ -196,6 +196,7 @@ PYTORCH_EXPORT_MODELS = {
     ("xlm-roberta", "xlm-roberta-base"),
     ("layoutlm", "microsoft/layoutlm-base-uncased"),
     ("layoutlmv3", "microsoft/layoutlmv3-base"),
+    ("levit", "facebook/levit-128S"),
     ("vit", "google/vit-base-patch16-224"),
     ("deit", "facebook/deit-small-patch16-224"),
     ("beit", "microsoft/beit-base-patch16-224"),