[PT FE] Fix aten::repeat regression (#19991)

* Revert "[PT FE] Simplify repeat operation (#19926)"

This reverts commit f926e0e392.

* Fix aten::repeats regression

* Simplify

* Update src/frontends/pytorch/src/op_table.cpp

* Add impacted model

src/frontends/pytorch/src/op_table.cpp

@@ -393,7 +393,7 @@ const std::map<std::string, CreatorFunction> get_supported_ops_ts() {
         {"aten::relu_", op::inplace_op<op::translate_1to1_match_1_inputs<opset10::Relu>>},
         {"aten::relu6", op::translate_relu6},
         {"aten::remainder", op::translate_remainder},
-        {"aten::repeat", op::inplace_op<op::translate_1to1_match_2_inputs<opset10::Tile>>},
+        {"aten::repeat", op::translate_1to1_match_2_inputs<opset10::Tile>},
         {"aten::repeat_interleave", op::translate_repeat_interleave},
         {"aten::reshape", op::translate_reshape},
         {"aten::reshape_as", op::translate_reshape_as},

tests/layer_tests/pytorch_tests/test_repeat.py

@@ -32,6 +32,7 @@ class TestRepeat(PytorchLayerTest):
     def test_repeat(self, repeats, ie_device, precision, ir_version):
         self._test(*self.create_model(repeats), ie_device, precision, ir_version)
 
+
 class TestRepeatList(PytorchLayerTest):
     def _prepare_input(self, repeats_shape):
         import numpy as np
@@ -54,4 +55,26 @@ class TestRepeatList(PytorchLayerTest):
     @pytest.mark.nightly
     @pytest.mark.precommit
     def test_repeat(self, repeats, ie_device, precision, ir_version):
-        self._test(*self.create_model(), ie_device, precision, ir_version,  kwargs_to_prepare_input={"repeats_shape": repeats})
+        self._test(*self.create_model(), ie_device, precision, ir_version,
+                   kwargs_to_prepare_input={"repeats_shape": repeats})
+
+
+class TestRepeatFromFlanT5(PytorchLayerTest):
+    def _prepare_input(self):
+        import numpy as np
+        return (np.random.randn(1, 15).astype(np.float32),)
+
+    def create_model(self):
+        import torch
+        from transformers.modeling_utils import ModuleUtilsMixin
+
+        class aten_repeat(torch.nn.Module):
+            def forward(self, x):
+                return ModuleUtilsMixin.create_extended_attention_mask_for_decoder(x.size(), x)
+
+        return aten_repeat(), None, "aten::repeat"
+
+    @pytest.mark.nightly
+    @pytest.mark.precommit
+    def test_repeat_t5(self, ie_device, precision, ir_version):
+        self._test(*self.create_model(), ie_device, precision, ir_version, trace_model=True)

tests/layer_tests/requirements.txt

@@ -4,6 +4,7 @@ numpy
 requests
 torch
 torchvision
+transformers
 pytest
 tensorflow-addons; python_version <= '3.10'
 jax; sys_platform == "linux"

tests/model_hub_tests/torch_tests/test_transformers.py

@@ -274,6 +274,7 @@ class TestTransformersModel(TestConvertModel):
 
     @pytest.mark.parametrize("name,type", [("bert-base-uncased", "bert"),
                                            ("facebook/bart-large-mnli", "bart"),
+                                           ("google/flan-t5-base","t5"),
                                            ("gpt2", "gpt2"),
                                            ("openai/clip-vit-large-patch14", "clip")])
     @pytest.mark.precommit