Proposed fix for TF example now running on safetensors. (#23208)

* Proposed fix for TF example now running on safetensors.

* Adding more warnings and returning keys.

* Trigger CI

* Trigger CI

---------

Co-authored-by: Sylvain Gugger <Sylvain.gugger@gmail.com>

examples/tensorflow/test_tensorflow_examples.py

@@ -297,7 +297,6 @@ class ExamplesTests(TestCasePlus):
             result = get_results(tmp_dir)
             self.assertGreaterEqual(result["bleu"], 30)
 
-    @skip("Fix me Matt")
     def test_run_image_classification(self):
         tmp_dir = self.get_auto_remove_tmp_dir()
         testargs = f"""

src/transformers/modeling_tf_pytorch_utils.py

@@ -246,6 +246,7 @@ def load_pytorch_state_dict_in_tf2_model(
     output_loading_info=False,
     _prefix=None,
     tf_to_pt_weight_rename=None,
+    ignore_mismatched_sizes=False,
 ):
     """Load a pytorch state_dict in a TF 2.0 model."""
     import tensorflow as tf
@@ -297,6 +298,7 @@ def load_pytorch_state_dict_in_tf2_model(
     weight_value_tuples = []
     all_pytorch_weights = set(pt_state_dict.keys())
     missing_keys = []
+    mismatched_keys = []
     for symbolic_weight in symbolic_weights:
         sw_name = symbolic_weight.name
         name, transpose = convert_tf_weight_name_to_pt_weight_name(
@@ -319,7 +321,18 @@ def load_pytorch_state_dict_in_tf2_model(
                     continue
             raise AttributeError(f"{name} not found in PyTorch model")
 
-        array = apply_transpose(transpose, pt_state_dict[name], symbolic_weight.shape)
+        try:
+            array = apply_transpose(transpose, pt_state_dict[name], symbolic_weight.shape)
+        except tf.errors.InvalidArgumentError as e:
+            if not ignore_mismatched_sizes:
+                error_msg = str(e)
+                error_msg += (
+                    "\n\tYou may consider adding `ignore_mismatched_sizes=True` in the model `from_pretrained` method."
+                )
+                raise tf.errors.InvalidArgumentError(error_msg)
+            else:
+                mismatched_keys.append((name, pt_state_dict[name].shape, symbolic_weight.shape))
+                continue
 
         tf_loaded_numel += tensor_size(array)
 
@@ -367,8 +380,26 @@ def load_pytorch_state_dict_in_tf2_model(
             f"you can already use {tf_model.__class__.__name__} for predictions without further training."
         )
 
+    if len(mismatched_keys) > 0:
+        mismatched_warning = "\n".join(
+            [
+                f"- {key}: found shape {shape1} in the checkpoint and {shape2} in the model instantiated"
+                for key, shape1, shape2 in mismatched_keys
+            ]
+        )
+        logger.warning(
+            f"Some weights of {tf_model.__class__.__name__} were not initialized from the model checkpoint"
+            f" are newly initialized because the shapes did not"
+            f" match:\n{mismatched_warning}\nYou should probably TRAIN this model on a down-stream task to be able"
+            " to use it for predictions and inference."
+        )
+
     if output_loading_info:
-        loading_info = {"missing_keys": missing_keys, "unexpected_keys": unexpected_keys}
+        loading_info = {
+            "missing_keys": missing_keys,
+            "unexpected_keys": unexpected_keys,
+            "mismatched_keys": mismatched_keys,
+        }
         return tf_model, loading_info
 
     return tf_model

src/transformers/modeling_tf_utils.py

@@ -2820,6 +2820,7 @@ class TFPreTrainedModel(tf.keras.Model, TFModelUtilsMixin, TFGenerationMixin, Pu
                 allow_missing_keys=True,
                 output_loading_info=output_loading_info,
                 _prefix=load_weight_prefix,
+                ignore_mismatched_sizes=ignore_mismatched_sizes,
             )
 
         # 'by_name' allow us to do transfer learning by skipping/adding layers