Switch PyTorch PTQ to common implementation

nncf/quantization/quantize_model.py

@@ -59,7 +59,7 @@ def quantize(
         - `performance`: Symmetric quantization of weights and activations.
         - `mixed`: Symmetric quantization of weights and asymmetric quantization of activations.
         Default value is None. In this case, `mixed` preset is used for `transformer`
-        model type otherwise `performace`.
+        model type otherwise `performance`.
     :type  preset: nncf.QuantizationPreset
     :param target_device: A target device the specificity of which will be taken
         into account while compressing in order to obtain the best performance
@@ -185,7 +185,7 @@ def quantize_with_accuracy_control(
         - `performance`: Symmetric quantization of weights and activations.
         - `mixed`: Symmetric quantization of weights and asymmetric quantization of activations.
         Default value is None. In this case, `mixed` preset is used for `transformer`
-        model type otherwise `performace`.
+        model type otherwise `performance`.
     :type preset: nncf.QuantizationPreset
     :param target_device: A target device the specificity of which will be taken
         into account while compressing in order to obtain the best performance
@@ -317,7 +317,7 @@ def quantize_with_tune_hyperparams(
         - `performance`: Symmetric quantization of weights and activations.
         - `mixed`: Symmetric quantization of weights and asymmetric quantization of activations.
         Default value is None. In this case, `mixed` preset is used for `transformer`
-        model type otherwise `performace`.
+        model type otherwise `performance`.
     :param target_device: A target device the specificity of which will be taken
         into account while compressing in order to obtain the best performance
         for this type of device.

nncf/torch/engine.py

@@ -9,7 +9,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from typing import Any, Dict, Union
+from typing import Any, Dict, Tuple, Union
 
 import torch
 from torch import nn
@@ -32,12 +32,17 @@ def __init__(self, model: nn.Module):
         self._model = model
         self._model.eval()
 
-    def infer(self, input_data: Union[torch.Tensor, Dict[str, torch.Tensor]]) -> Union[torch.Tensor, Dict[str, Any]]:
+    def infer(
+        self, input_data: Union[torch.Tensor, Tuple[torch.Tensor], Dict[str, torch.Tensor]]
+    ) -> Union[torch.Tensor, Dict[str, Any]]:
         """
         Runs Torch model on the provided input.
 
-        :param input_data: inputs for the model
-        :return output_data: model outputs
+        :param input_data: Inputs for the model.
+        :return: Model outputs.
         """
-
+        if isinstance(input_data, dict):
+            return self._model(**input_data)
+        if isinstance(input_data, tuple):
+            return self._model(*input_data)
         return self._model(input_data)