Fast imports part 3 (#9474)

* New intermediate inits

* Update template

* Avoid importing torch/tf/flax in tokenization unless necessary

* Styling

* Shutup flake8

* Better python version check

src/transformers/models/rag/__init__.py

@@ -16,11 +16,43 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from ...file_utils import is_torch_available
-from .configuration_rag import RagConfig
-from .retrieval_rag import RagRetriever
-from .tokenization_rag import RagTokenizer
+from typing import TYPE_CHECKING
 
+from ...file_utils import _BaseLazyModule, is_torch_available
+
+
+_import_structure = {
+    "configuration_rag": ["RagConfig"],
+    "retrieval_rag": ["RagRetriever"],
+    "tokenization_rag": ["RagTokenizer"],
+}
 
 if is_torch_available():
-    from .modeling_rag import RagModel, RagSequenceForGeneration, RagTokenForGeneration
+    _import_structure["modeling_rag"] = ["RagModel", "RagSequenceForGeneration", "RagTokenForGeneration"]
+
+
+if TYPE_CHECKING:
+    from .configuration_rag import RagConfig
+    from .retrieval_rag import RagRetriever
+    from .tokenization_rag import RagTokenizer
+
+    if is_torch_available():
+        from .modeling_rag import RagModel, RagSequenceForGeneration, RagTokenForGeneration
+
+else:
+    import importlib
+    import os
+    import sys
+
+    class _LazyModule(_BaseLazyModule):
+        """
+        Module class that surfaces all objects but only performs associated imports when the objects are requested.
+        """
+
+        __file__ = globals()["__file__"]
+        __path__ = [os.path.dirname(__file__)]
+
+        def _get_module(self, module_name: str):
+            return importlib.import_module("." + module_name, self.__name__)
+
+    sys.modules[__name__] = _LazyModule(__name__, _import_structure)