ModelCloud · Qubitium · Oct 19, 2025 · Oct 19, 2025
diff --git a/gptqmodel/utils/nogil_patcher.py b/gptqmodel/utils/nogil_patcher.py
@@ -5,8 +5,8 @@
 
 """Straightforward monkey patch helpers for nogil runtimes."""
 
-import time
 import threading
+import time
 
 from .safe import ThreadSafe
 
@@ -33,8 +33,8 @@ def patch_safetensors_save_file() -> None:
 
 def patch_triton_autotuner() -> None:
     try:
-        from triton.runtime import autotuner as module
         import triton
+        from triton.runtime import autotuner as module
     except ImportError:
         return
 

diff --git a/gptqmodel/utils/stream.py b/gptqmodel/utils/stream.py
@@ -6,15 +6,16 @@
 from __future__ import annotations
 
 import threading
+from concurrent.futures import Future
 from dataclasses import dataclass
 from typing import Any, Callable, Dict, Iterable, List, Optional, Tuple
 
 import torch
-from concurrent.futures import Future
 
 from .logger import setup_logger
 from .threadx import DeviceThreadPool
 
+
 log = setup_logger()
 
 

diff --git a/gptqmodel/utils/threadx.py b/gptqmodel/utils/threadx.py
@@ -13,12 +13,13 @@
 import threading
 import time
 import traceback
-from datetime import datetime, timezone
 from concurrent.futures import Future
+from datetime import datetime, timezone
 from typing import Any, Callable, Dict, Iterable, List, Optional, Set, Tuple, Union
 
 import torch
 
+
 try:
     from device_smi import Device  # type: ignore
 except Exception:  # pragma: no cover - defensive: optional dependency may be unavailable

diff --git a/setup.py b/setup.py
@@ -3,11 +3,11 @@
 # SPDX-License-Identifier: Apache-2.0
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 import os
+import re
 import subprocess
 import sys
 from pathlib import Path
 
-import pcre as re
 from setuptools import find_packages, setup
 from setuptools.command.bdist_wheel import bdist_wheel as _bdist_wheel
 

diff --git a/tests/models/model_test.py b/tests/models/model_test.py
@@ -22,8 +22,8 @@
 
 # -- end do not touch
 
-from pathlib import Path  # noqa: E402
 from enum import Enum  # noqa: E402
+from pathlib import Path  # noqa: E402
 from typing import Dict, List  # noqa: E402
 
 from logbar import LogBar  # noqa: E402

diff --git a/tests/models/test_act_group_aware.py b/tests/models/test_act_group_aware.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_apertus.py b/tests/models/test_apertus.py
@@ -4,9 +4,9 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
-from gptqmodel.utils.eval import EVAL
 
 from gptqmodel import BACKEND
+from gptqmodel.utils.eval import EVAL
 
 
 class TestApertus(ModelTest):

diff --git a/tests/models/test_cohere.py b/tests/models/test_cohere.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_cohere2.py b/tests/models/test_cohere2.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_deci.py b/tests/models/test_deci.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_dream.py b/tests/models/test_dream.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_falcon.py b/tests/models/test_falcon.py
@@ -5,6 +5,7 @@
 
 import torch  # noqa: E402from tests.model_test import ModelTest
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_glm.py b/tests/models/test_glm.py
@@ -4,8 +4,10 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
+
 # | Metric                         |   MARLIN |
 # |--------------------------------|----------|
 # | arc_challenge :: acc,none      |   0.5154 |

diff --git a/tests/models/test_glm4_moe.py b/tests/models/test_glm4_moe.py
@@ -4,8 +4,10 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
+
 class TestGlm4Moe(ModelTest):
     NATIVE_MODEL_ID = "/monster/data/model/GLM-4.6/"
     DELETE_QUANTIZED_MODEL = False

diff --git a/tests/models/test_gpt_oss.py b/tests/models/test_gpt_oss.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_granite.py b/tests/models/test_granite.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_hymba.py b/tests/models/test_hymba.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_ling.py b/tests/models/test_ling.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_llama3_2.py b/tests/models/test_llama3_2.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_llama3_2_awq.py b/tests/models/test_llama3_2_awq.py
@@ -4,9 +4,9 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
-from gptqmodel.utils.eval import EVAL
 
 from gptqmodel.quantization import FORMAT, METHOD
+from gptqmodel.utils.eval import EVAL
 
 
 # a100:0

diff --git a/tests/models/test_llama4.py b/tests/models/test_llama4.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_longllama.py b/tests/models/test_longllama.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_mimo.py b/tests/models/test_mimo.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_nemotron_ultra.py b/tests/models/test_nemotron_ultra.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_qwen2_5.py b/tests/models/test_qwen2_5.py
@@ -4,8 +4,10 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
+
 # | Metric                         |   MARLIN |
 # |--------------------------------|----------|
 # | arc_challenge :: acc,none      |   0.2884 |

diff --git a/tests/models/test_qwen2_5_omni.py b/tests/models/test_qwen2_5_omni.py
@@ -6,9 +6,9 @@
 
 import soundfile as sf
 from model_test import ModelTest
-from gptqmodel.utils.eval import EVAL
 
 from gptqmodel.models.definitions.qwen2_5_omni import Qwen2_5_OmniGPTQ
+from gptqmodel.utils.eval import EVAL
 
 
 class TestQwen2_5_Omni(ModelTest):

diff --git a/tests/models/test_qwen2_5_vl.py b/tests/models/test_qwen2_5_vl.py
@@ -4,9 +4,9 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
-from gptqmodel.utils.eval import EVAL
 
 from gptqmodel.models.definitions.qwen2_vl import Qwen2VLQModel
+from gptqmodel.utils.eval import EVAL
 
 
 class TestQwen2_VL(ModelTest):

diff --git a/tests/models/test_qwen2_moe_quant.py b/tests/models/test_qwen2_moe_quant.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_qwen2_vl.py b/tests/models/test_qwen2_vl.py
@@ -4,9 +4,9 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
-from gptqmodel.utils.eval import EVAL
 
 from gptqmodel.models.definitions.qwen2_vl import Qwen2VLQModel
+from gptqmodel.utils.eval import EVAL
 
 
 class TestQwen2_VL(ModelTest):

diff --git a/tests/models/test_qwen3_moe.py b/tests/models/test_qwen3_moe.py
@@ -4,8 +4,10 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
+
 # | Metric                         |   MARLIN |
 # |--------------------------------|----------|
 # | arc_challenge :: acc,none      |   0.5094 |

diff --git a/tests/models/test_qwen3_next.py b/tests/models/test_qwen3_next.py
@@ -4,8 +4,10 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
+
 # | Metric                         |   MARLIN |
 # |--------------------------------|----------|
 # | arc_challenge :: acc,none      |   0.6271 |

diff --git a/tests/models/test_qwen3_omni.py b/tests/models/test_qwen3_omni.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_seed_oss.py b/tests/models/test_seed_oss.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_stablelm.py b/tests/models/test_stablelm.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/models/test_xverse.py b/tests/models/test_xverse.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/test_gptqv2.py b/tests/test_gptqv2.py
@@ -4,6 +4,7 @@
 # Contact: qubitium@modelcloud.ai, x.com/qubitium
 
 from models.model_test import ModelTest
+
 from gptqmodel.utils.eval import EVAL
 
 

diff --git a/tests/test_lm_head.py b/tests/test_lm_head.py
@@ -13,10 +13,10 @@
 os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
 # -- end do not touch
 from models.model_test import ModelTest  # noqa: E402
-from gptqmodel.utils.eval import EVAL  # noqa: E402
 
 from gptqmodel import GPTQModel, QuantizeConfig  # noqa: E402
 from gptqmodel.nn_modules.qlinear import BaseQuantLinear  # noqa: E402
+from gptqmodel.utils.eval import EVAL  # noqa: E402
 
 
 class TestLmHeadLoad(ModelTest):
Original file line number	Diff line number	Diff line change
Expand Up		@@ -4,6 +4,7 @@
		# Contact: qubitium@modelcloud.ai, x.com/qubitium

		from model_test import ModelTest

		from gptqmodel.utils.eval import EVAL


Expand Down
Original file line number	Diff line number	Diff line change
Expand Up		@@ -5,6 +5,7 @@

		import torch # noqa: E402from tests.model_test import ModelTest
		from model_test import ModelTest

		from gptqmodel.utils.eval import EVAL


Expand Down