diff --git a/intel_extension_for_transformers/neural_chat/utils/common.py b/intel_extension_for_transformers/neural_chat/utils/common.py
index 22c599d6215..4d25f829a16 100644
--- a/intel_extension_for_transformers/neural_chat/utils/common.py
+++ b/intel_extension_for_transformers/neural_chat/utils/common.py
@@ -69,47 +69,3 @@ def is_hf_model(model_name_or_path):
 
 def supported_gpus():
     return ['flex', 'max', 'arc']
-
-def get_gpu_family():
-    ''' Get gpu device family info.
-
-    Return 'flex'|'max'|'arc'| 'no_gpu'| assert
-
-    Note, this function need to import intel_extension_for_pytorch
-
-
-    Additional info (common gpu name):
-      'Intel(R) Data Center GPU Flex 170'
-      'Intel(R) Data Center GPU Max 1100'
-      'Intel(R) Arc(TM) A770 Graphics'
-    '''
-
-    import intel_extension_for_pytorch as ipex
-    if not (hasattr(torch, "xpu") and torch.xpu.is_available()):
-        return 'no_gpu'
-
-    name = torch.xpu.get_device_name()
-    if 'GPU Flex' in name:
-        result = 'flex'
-    elif 'GPU Max' in name:
-        result = 'max'
-    elif 'Arc(TM)' in name:
-        result = 'arc'
-    else:
-        assert False, "Unsupported GPU device: {}".format(name)
-
-    if result not in supported_gpus():
-        assert False, "Unsupported GPU device: {}".format(name)
-    else:
-        return result
-
-_autoround_available = importlib.util.find_spec("auto_round") is not None
-_autoround_version = "N/A"
-if _autoround_available:
-    try:
-        _autoround_version = importlib_metadata.version("auto_round")
-    except importlib_metadata.PackageNotFoundError:
-        _autoround_available = False
-
-def is_autoround_available():
-    return _autoround_available
diff --git a/intel_extension_for_transformers/tools/utils.py b/intel_extension_for_transformers/tools/utils.py
new file mode 100644
index 00000000000..0e466724380
--- /dev/null
+++ b/intel_extension_for_transformers/tools/utils.py
@@ -0,0 +1,101 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+#
+# 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.
+
+"""Utility."""
+
+import importlib
+import sys
+import torch
+if sys.version_info < (3, 8):
+    import importlib_metadata
+else:
+    import importlib.metadata as importlib_metadata
+
+try:
+    import habana_frameworks.torch.hpu as hthpu
+    is_hpu_available = True
+except ImportError:
+    is_hpu_available = False
+
+def supported_gpus():
+    return ['flex', 'max', 'arc']
+
+def get_gpu_family():
+    ''' Get gpu device family info.
+
+    Return 'flex'|'max'|'arc'| 'no_gpu'| assert
+
+    Note, this function need to import intel_extension_for_pytorch
+
+
+    Additional info (common gpu name):
+      'Intel(R) Data Center GPU Flex 170'
+      'Intel(R) Data Center GPU Max 1100'
+      'Intel(R) Arc(TM) A770 Graphics'
+    '''
+
+    import intel_extension_for_pytorch as ipex
+    if not (hasattr(torch, "xpu") and torch.xpu.is_available()):
+        return 'no_gpu'
+
+    name = torch.xpu.get_device_name()
+    if 'GPU Flex' in name:
+        result = 'flex'
+    elif 'GPU Max' in name:
+        result = 'max'
+    elif 'Arc(TM)' in name:
+        result = 'arc'
+    else:
+        assert False, "Unsupported GPU device: {}".format(name)
+
+    if result not in supported_gpus():
+        assert False, "Unsupported GPU device: {}".format(name)
+    else:
+        return result
+
+_ipex_available = importlib.util.find_spec("intel_extension_for_pytorch") is not None
+_ipex_version = "N/A"
+if _ipex_available:
+    try:
+        _ipex_version = importlib_metadata.version("intel_extension_for_pytorch")
+    except importlib_metadata.PackageNotFoundError:
+        _ipex_available = False
+
+def is_ipex_available():
+    return _ipex_available
+
+_autoround_available = importlib.util.find_spec("auto_round") is not None
+_autoround_version = "N/A"
+if _autoround_available:
+    try:
+        _autoround_version = importlib_metadata.version("auto_round")
+    except importlib_metadata.PackageNotFoundError:
+        _autoround_available = False
+
+def is_autoround_available():
+    return _autoround_available
+
+def get_device_type():
+    if torch.cuda.is_available():
+        device = "cuda"
+    elif is_hpu_available:
+        device = "hpu"
+    elif is_ipex_available() and torch.xpu.is_available():
+        device = "xpu"
+    else:
+        device = "cpu"
+    return device
diff --git a/intel_extension_for_transformers/transformers/dpo_trainer.py b/intel_extension_for_transformers/transformers/dpo_trainer.py
index ecb71f211f9..1f41b643aca 100644
--- a/intel_extension_for_transformers/transformers/dpo_trainer.py
+++ b/intel_extension_for_transformers/transformers/dpo_trainer.py
@@ -29,7 +29,7 @@
 from peft import PeftModel, get_peft_model, prepare_model_for_kbit_training
 import time
 import logging
-from intel_extension_for_transformers.neural_chat.utils.common import is_hpu_available
+from intel_extension_for_transformers.tools.utils import is_hpu_available
 
 logger = logging.getLogger(__name__)
 
diff --git a/intel_extension_for_transformers/transformers/llm/finetuning/finetuning.py b/intel_extension_for_transformers/transformers/llm/finetuning/finetuning.py
index 9388a7f5725..1149827c127 100644
--- a/intel_extension_for_transformers/transformers/llm/finetuning/finetuning.py
+++ b/intel_extension_for_transformers/transformers/llm/finetuning/finetuning.py
@@ -60,8 +60,8 @@
 from transformers.integrations.deepspeed import (
     is_deepspeed_available,
 )
-from intel_extension_for_transformers.neural_chat.utils.common import is_hpu_available
-from intel_extension_for_transformers.neural_chat.utils.common import get_device_type
+from intel_extension_for_transformers.tools.utils import is_hpu_available
+from intel_extension_for_transformers.tools.utils import get_device_type
 
 
 if is_bitsandbytes_available():
diff --git a/intel_extension_for_transformers/transformers/llm/quantization/utils.py b/intel_extension_for_transformers/transformers/llm/quantization/utils.py
index 4e5bad8d3a1..9e6bd1f31ac 100644
--- a/intel_extension_for_transformers/transformers/llm/quantization/utils.py
+++ b/intel_extension_for_transformers/transformers/llm/quantization/utils.py
@@ -26,7 +26,7 @@
 from neural_compressor.adaptor.torch_utils.model_wrapper import WeightOnlyLinear
 from neural_compressor.utils.utility import LazyImport
 from neural_compressor.config import PostTrainingQuantConfig
-from intel_extension_for_transformers.neural_chat.utils.common import is_ipex_available, is_autoround_available
+from intel_extension_for_transformers.tools.utils import is_ipex_available, is_autoround_available
 from transformers import AutoTokenizer
 
 if is_ipex_available():
diff --git a/intel_extension_for_transformers/transformers/modeling/trl_models/modeling_base.py b/intel_extension_for_transformers/transformers/modeling/trl_models/modeling_base.py
index fcbed718948..a8acf5f9dbf 100644
--- a/intel_extension_for_transformers/transformers/modeling/trl_models/modeling_base.py
+++ b/intel_extension_for_transformers/transformers/modeling/trl_models/modeling_base.py
@@ -32,7 +32,7 @@
 from safetensors.torch import load_file as safe_load_file
 from transformers import PreTrainedModel
 import importlib
-from intel_extension_for_transformers.neural_chat.utils.common import is_hpu_available
+from intel_extension_for_transformers.tools.utils import is_hpu_available
 
 
 def is_peft_available():
diff --git a/requirements.txt b/requirements.txt
index abb2e6fd1a1..0b45ec85d3f 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,6 +1,3 @@
-fastapi
 py-cpuinfo
 setuptools>=65
 setuptools_scm[toml]>=6.2
-uvicorn
-yacs
diff --git a/tests/CI/test_weight_only_gpu.py b/tests/CI/test_weight_only_gpu.py
index 6e72f53a9c4..1221170c51f 100644
--- a/tests/CI/test_weight_only_gpu.py
+++ b/tests/CI/test_weight_only_gpu.py
@@ -24,7 +24,7 @@
 from intel_extension_for_transformers.transformers import GPTQConfig, RtnConfig
 from math import isclose
 from transformers import AutoTokenizer
-from intel_extension_for_transformers.neural_chat.utils.common import get_gpu_family, is_ipex_available
+from intel_extension_for_transformers.tools.utils import get_gpu_family, is_ipex_available
 from torch.utils.data import DataLoader