Torch native q conformance

tests/post_training/pipelines/base.py

@@ -408,7 +408,7 @@ def save_compressed_model(self) -> None:
             self.path_compressed_ir = self.output_model_dir / "model.xml"
             ov.serialize(ov_model, self.path_compressed_ir)
         elif self.backend == BackendType.FX_TORCH:
-            exported_model = torch.export.export(self.compressed_model, (self.dummy_tensor,))
+            exported_model = torch.export.export(self.model, (self.dummy_tensor,))
             ov_model = ov.convert_model(exported_model, example_input=self.dummy_tensor.cpu(), input=self.input_size)
             self.path_compressed_ir = self.output_model_dir / "model.xml"
             ov.serialize(ov_model, self.path_compressed_ir)

tests/post_training/pipelines/image_classification_base.py

@@ -76,7 +76,8 @@ def process_result(request, userdata):
     def _validate_torch_compile(
         self, val_loader: torch.utils.data.DataLoader, predictions: np.ndarray, references: np.ndarray
     ):
-        compiled_model = torch.compile(self.compressed_model, backend="openvino")
+        # compiled_model = torch.compile(self.compressed_model, backend="openvino")
+        compiled_model = torch.compile(self.compressed_model)
         for i, (images, target) in enumerate(val_loader):
             # W/A for memory leaks when using torch DataLoader and OpenVINO
             pred = compiled_model(images)
@@ -104,3 +105,33 @@ def _validate(self):
 
         self.run_info.metric_name = "Acc@1"
         self.run_info.metric_value = acc_top1
+
+    def _compress_torch_native(self):
+        import os
+
+        os.environ["TORCHINDUCTOR_FREEZING"] = "1"
+
+        from torch.ao.quantization.quantize_pt2e import convert_pt2e
+        from torch.ao.quantization.quantize_pt2e import prepare_pt2e
+        from torch.ao.quantization.quantizer.x86_inductor_quantizer import X86InductorQuantizer
+        from torch.ao.quantization.quantizer.x86_inductor_quantizer import get_default_x86_inductor_quantization_config
+
+        quantizer = X86InductorQuantizer()
+        quantizer.set_global(get_default_x86_inductor_quantization_config())
+
+        prepared_model = prepare_pt2e(self.model, quantizer)
+        for data in self.calibration_dataset.get_inference_data():
+            prepared_model(data)
+        self.compressed_model = convert_pt2e(prepared_model)
+
+    def _compress_nncf_pt2e(self):
+        pass
+
+    def _compress(self):
+        """
+        Quantize self.model
+        """
+        if self.backend != BackendType.FX_TORCH:
+            super()._compress()
+
+        self._compress_torch_native()

tests/post_training/test_quantize_conformance.py

@@ -75,11 +75,6 @@ def fixture_run_benchmark_app(pytestconfig):
     return pytestconfig.getoption("benchmark")
 
 
-@pytest.fixture(scope="session", name="validate_in_backend")
-def fixture_validate_in_backend(pytestconfig):
-    return pytestconfig.getoption("validate_in_backend")
-
-
 @pytest.fixture(scope="session", name="extra_columns")
 def fixture_extra_columns(pytestconfig):
     return pytestconfig.getoption("extra_columns")
@@ -267,7 +262,6 @@ def test_ptq_quantization(
     run_torch_cuda_backend: bool,
     subset_size: Optional[int],
     run_benchmark_app: bool,
-    validate_in_backend: bool,
     capsys: pytest.CaptureFixture,
     extra_columns: bool,
     memory_monitor: bool,
@@ -295,7 +289,6 @@ def test_ptq_quantization(
                 "data_dir": data_dir,
                 "no_eval": no_eval,
                 "run_benchmark_app": run_benchmark_app,
-                "validate_in_backend": validate_in_backend,
                 "batch_size": batch_size,
                 "memory_monitor": memory_monitor,
             }