Unstage helper scripts

nncf/quantization/algorithms/weight_compression/openvino_modeling.py

@@ -9,18 +9,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-
-# Copyright (c) 2024 Intel Corporation
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#      http://www.apache.org/licenses/LICENSE-2.0
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-
 import inspect
 from dataclasses import dataclass
 from typing import List, Optional, Tuple

nncf/quantization/algorithms/weight_compression/scale_estimation.py

@@ -306,7 +306,6 @@ def calculate_quantization_params(
             factor = 1.0 - 0.05 * scale_step
             scaled_scale = factor * scale
 
-            input_tensors[1] = scaled_scale.data
             if config.mode == CompressWeightsMode.NF4:
                 out = do_nf4_quantization(original_weight, scaled_scale)
             else:
@@ -318,7 +317,6 @@ def calculate_quantization_params(
             near_to_ideal_scale = estimate_scales(original_weight, target, zero_mask, importance)
             near_to_ideal_scale = near_to_ideal_scale * scale_sign
 
-            input_tensors[1] = near_to_ideal_scale.data
             if config.mode == CompressWeightsMode.NF4:
                 g_compressed_weighs = do_nf4_quantization(original_weight, near_to_ideal_scale)
                 out = do_nf4_dequantization(g_compressed_weighs, near_to_ideal_scale)

nncf/quantization/algorithms/weight_compression/weight_lowering.py

@@ -17,8 +17,9 @@
 import nncf
 from nncf.parameters import CompressWeightsMode
 from nncf.quantization.algorithms.weight_compression.config import WeightCompressionConfig
-from nncf.quantization.algorithms.weight_compression.openvino_modeling import OVModelParameters, \
-    get_compress_decompress_weight_model, get_compress_weight_model
+from nncf.quantization.algorithms.weight_compression.openvino_modeling import OVModelParameters
+from nncf.quantization.algorithms.weight_compression.openvino_modeling import get_compress_decompress_weight_model
+from nncf.quantization.algorithms.weight_compression.openvino_modeling import get_compress_weight_model
 from nncf.quantization.fake_quantize import calculate_scale_zero_point
 from nncf.tensor import Tensor
 from nncf.tensor import functions as fns
@@ -492,8 +493,12 @@ def do_int_quantization(
 
 
 def calculate_quantized_dequantized_weight(
-    weight: Tensor, config: WeightCompressionConfig, scale: Tensor, zero_point: Optional[Tensor] = None,
-    invert_division: Optional[bool] = False, ov_model_params: Optional[OVModelParameters] = None,
+    weight: Tensor,
+    config: WeightCompressionConfig,
+    scale: Tensor,
+    zero_point: Optional[Tensor] = None,
+    invert_division: Optional[bool] = False,
+    ov_model_params: Optional[OVModelParameters] = None,
 ) -> Tensor:
     accelerate_through_ov = is_openvino_available()
 

nncf/tensor/functions/ov.py

@@ -1,3 +1,14 @@
+# Copyright (c) 2024 Intel Corporation
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#      http://www.apache.org/licenses/LICENSE-2.0
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 import numpy as np
 import openvino as ov