[pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

neural_compressor/__init__.py

@@ -25,10 +25,7 @@
     QuantizationAwareTrainingConfig,
     MixedPrecisionConfig,
 )
-from .transformers import(
-    GPTQConfig,
-    RtnConfig
-)
+from .transformers import GPTQConfig, RtnConfig
 from .contrib import *
 from .model import *
 from .metric import *

neural_compressor/transformers/__init__.py

@@ -1 +1,15 @@
+# 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.
+
 from .utils.quantization_config import GPTQConfig, RtnConfig

neural_compressor/transformers/models/__init__.py

@@ -1 +1,15 @@
+# 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.
+
 from .modeling_auto import _BaseQBitsAutoModelClass

neural_compressor/transformers/models/modeling_auto.py

@@ -41,29 +41,30 @@
 import torch
 import transformers
 from accelerate import init_empty_weights
-
 from transformers import AutoConfig
 from transformers.configuration_utils import PretrainedConfig
 from transformers.modeling_utils import load_state_dict
-from transformers.utils import has_file,  is_safetensors_available
-from neural_compressor.transformers.quantization.utils import save_low_bit, replace_linear
-from neural_compressor.torch.algorithms.weight_only.modules import INCWeightOnlyLinear
-
-from ..quantization.utils import (
-    convert_dtype_torch2str,
-    replace_linear,
+from transformers.utils import (
+    SAFE_WEIGHTS_INDEX_NAME,
+    SAFE_WEIGHTS_NAME,
+    WEIGHTS_INDEX_NAME,
+    WEIGHTS_NAME,
+    has_file,
+    is_safetensors_available,
 )
+
+from neural_compressor.torch.algorithms.weight_only.modules import INCWeightOnlyLinear
+from neural_compressor.torch.utils import is_ipex_available
 from neural_compressor.transformers import GPTQConfig, RtnConfig
-from neural_compressor.utils.utility import LazyImport, CpuInfo
+from neural_compressor.transformers.quantization.utils import replace_linear, save_low_bit
 from neural_compressor.utils import logger
-from neural_compressor.torch.utils import is_ipex_available
-from transformers.utils import SAFE_WEIGHTS_INDEX_NAME, SAFE_WEIGHTS_NAME, WEIGHTS_INDEX_NAME, WEIGHTS_NAME
+from neural_compressor.utils.utility import CpuInfo, LazyImport
 
+from ..quantization.utils import convert_dtype_torch2str, replace_linear
 
 torch = LazyImport("torch")
 
 
-
 def build_woq_model(model, quantization_config):
     from neural_compressor.adaptor.torch_utils.util import set_module
 
@@ -523,9 +524,7 @@ def load_low_bit(cls, pretrained_model_name_or_path, *model_args, **kwargs):
             else:
                 logger.warning("bits number only supports 4, 8.")
                 quantization_config.weight_dtype = "int4"
-                logger.warning(
-                    "int4 weight_dtype is used, please change the config.json if you don't want to use it."
-                )
+                logger.warning("int4 weight_dtype is used, please change the config.json if you don't want to use it.")
         else:
             if quantization_config.weight_dtype not in [
                 "int4_fullrange",
@@ -554,7 +553,6 @@ def load_low_bit(cls, pretrained_model_name_or_path, *model_args, **kwargs):
 
         model = build_woq_model(model, quantization_config)
 
-
         if is_sharded:
             loaded_state_dict_keys = sharded_metadata["all_checkpoint_keys"]
         else:
@@ -708,4 +706,4 @@ class AutoModel(_BaseQBitsAutoModelClass):
 
 
 class AutoModelForSeq2SeqLM(_BaseQBitsAutoModelClass):
-    ORIG_MODEL = transformers.AutoModelForSeq2SeqLM
+    ORIG_MODEL = transformers.AutoModelForSeq2SeqLM

neural_compressor/transformers/quantization/__init__.py

@@ -1 +1,15 @@
+# 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.
+
 from .utils import convert_to_quantized_model, save_low_bit

neural_compressor/transformers/quantization/utils.py

@@ -18,26 +18,25 @@
 import logging
 import math
 import os
+
 <<<<<<< Updated upstream
 
 from accelerate import init_empty_weights
 from datasets import load_dataset
+
 =======
-import types
 import json
+import types
 
 from datasets import load_dataset
+
 from neural_compressor.torch.algorithms.weight_only.modules import INCWeightOnlyLinear
-from neural_compressor.torch.quantization import (
-    GPTQConfig,
-    RTNConfig,
-    convert,
-    prepare,
-)
-from neural_compressor.utils.utility import LazyImport, CpuInfo
+from neural_compressor.torch.quantization import GPTQConfig, RTNConfig, convert, prepare
+from neural_compressor.utils.utility import CpuInfo, LazyImport
+
 >>>>>>> Stashed changes
 from transformers import AutoTokenizer
-from transformers.utils import SAFE_WEIGHTS_NAME, WEIGHTS_NAME 
+from transformers.utils import SAFE_WEIGHTS_NAME, WEIGHTS_NAME
 
 from neural_compressor.torch.algorithms.weight_only.modules import INCWeightOnlyLinear as WeightOnlyLinear
 from neural_compressor.torch.quantization import GPTQConfig, RTNConfig, convert, prepare
@@ -298,9 +297,7 @@ def _replace_linear(
                     or device == "auto"
 >>>>>>> Stashed changes
                 ):
-                    from intel_extension_for_pytorch.nn.modules import (
-                        WeightOnlyQuantizedLinear as ipex_linear,
-                    )
+                    from intel_extension_for_pytorch.nn.modules import WeightOnlyQuantizedLinear as ipex_linear
                     from intel_extension_for_pytorch.utils.weight_only_quantization import (
                         _convert_optimum_format_to_desired,
                     )
@@ -355,8 +352,9 @@ def _replace_linear(
                     )
 
                 elif device == "xpu" or device == torch.device("xpu"):
-                    from intel_extension_for_pytorch.nn.utils._quantize_convert import \
-                        WeightOnlyQuantizedLinear as ipex_linear # pylint: disable=E0401
+                    from intel_extension_for_pytorch.nn.utils._quantize_convert import (
+                        WeightOnlyQuantizedLinear as ipex_linear,  # pylint: disable=E0401
+                    )
                     model._modules[name] = ipex_linear(
                         in_features,
                         out_features,