[Bug-Report|缺陷反馈]: 安装后运行命令报错：ModuleNotFoundError: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti

已关闭

[Bug-Report|缺陷反馈]: 安装后运行命令报错：ModuleNotFoundError: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils' #379

tangpeiqi96创建于 2025年12月30日关闭于 1月27日

tangpeiqi96

2025年12月30日创建

Thanks for sending an issue! Please fill in the following template to help quickly solve your problem.

Describe the current behavior / 问题描述 (Mandatory / 必填)

(base) [root@localhost ]# python -c "from msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils import quant_tensor, dequant_tensor; print('OK')"
2025-12-30 20:43:59,336 - msmodelslim - WARNING - Unable to import torch_npu.
2025-12-30 20:43:59,337 - msmodelslim - WARNING - Unable to import torch_npu.
Traceback (most recent call last):
File "", line 1, in
from msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils import quant_tensor, dequant_tensor; print('OK')
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/pytorch/llm_ptq/anti_outlier/init.py", line 5, in
from .anti_outlier import AntiOutlier
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/pytorch/llm_ptq/anti_outlier/anti_outlier.py", line 42, in
from msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils import (
...<4 lines>...
)
ModuleNotFoundError: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils'

Environment / 环境信息 (Mandatory / 必填)

CANN版本8.3.RC2

Steps to reproduce the issue / 重现步骤 (Mandatory / 必填)

按照指导安装后执行python -c "from msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils import quant_tensor, dequant_tensor; print('OK')"

Describe the expected behavior / 预期结果 (Mandatory / 必填)

预期不报错

Special notes for this issue/备注 (Optional / 选填)

anreywmh成员

2025年12月31日评论：

您好，首先确定您的CANN版本大于8.2RC1，之后执行source /usr/local/Ascend/ascend-toolkit/set_env.sh
记得进入msmodelslim目录下执行bash install.sh

tangpeiqi96

2025年12月31日评论：

已确认CANN版本版本为8.3.RC2
(base) [root@localhost test]# cd /usr/local/Ascend/ascend-toolkit/latest/$(uname -m)-linux
(base) [root@localhost aarch64-linux]# cat ascend_toolkit_install.info
package_name=Ascend-cann-toolkit
version=8.3.RC2
innerversion=V100R001C23SPC002B210
compatible_version=[V100R001C15],[V100R001C18],[V100R001C19],[V100R001C20],[V100R001C21],[V100R001C23]
arch=aarch64
os=linux
path=/usr/local/Ascend/ascend-toolkit/8.3.RC2/aarch64-linux

执行了source /usr/local/Ascend/ascend-toolkit/set_env.sh，并进入msmodelslim目录下执行bash install.sh，问题仍然存在：

(base) [root@localhost msmodelslim]# source /usr/local/Ascend/ascend-toolkit/set_env.sh
(base) [root@localhost msmodelslim]# cd ..
(base) [root@localhost msit]# cd ..
(base) [root@localhost test]# msmodelslim quant --model_path /home/weight/Qwen3-32B --save_path /home/weight/Qwen3-32B-W8A8 --device npu --model_type Qwen3-32B --quant_type w8a8 --trust_remote_code True
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - WARNING - write directory not exists, creating directory '/home/weight/Qwen3-32B-W8A8'
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - quantization with following parameters:
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - model_type: Qwen3-32B
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - model_path: /home/weight/Qwen3-32B
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - save_path: /home/weight/Qwen3-32B-W8A8
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - device_type: DeviceType.NPU
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - using single device (default): npu
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - quant_type: QuantType.W8A8
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - trust_remote_code: True
2025-12-31 09:56:59,114 - msmodelslim.app.naive_quantization - INFO - ===========ANALYSE MODEL===========
2025-12-31 09:56:59,120 - msmodelslim.app.naive_quantization - INFO - Found 37 model adapters: ['DeepSeek-R1', 'DeepSeek-R1-0528', 'DeepSeek-R1-Distill-Qwen-1.5B', 'DeepSeek-R1-Distill-Qwen-7B', 'DeepSeek-V3', 'DeepSeek-V3-0324', 'DeepSeek-V3.1', 'DeepSeek-V3.2', 'DeepSeek-V3.2-Exp', 'HunyuanVideo', 'Kimi-K2-Instruct-0905', 'Kimi-K2-Thinking', 'QwQ-32B', 'Qwen-QwQ-32B', 'Qwen1.5-110B', 'Qwen2-72B', 'Qwen2-7B', 'Qwen2.5-32B-Instruct', 'Qwen2.5-72B-Instruct', 'Qwen2.5-7B-Instruct', 'Qwen2.5-Coder-7B-Instruct', 'Qwen3-14B', 'Qwen3-235B', 'Qwen3-30B', 'Qwen3-32B', 'Qwen3-8B', 'Qwen3-Next-80B-A3B-Instruct', 'Qwen3-VL-235B-A22B', 'Qwen3-VL-30B-A3B', 'Wan2.1', 'Wan2.2', 'Wan2_1', 'Wan2_2', 'default', 'hunyuan-video', 'hunyuan_video', 'hunyuanvideo']
WARNING:root:Overwriting existing implementation for <class 'msmodelslim.core.quantizer.base.AutoWeightQuantizer'> with dispatch key: (QScheme(scope=per_channel, dtype=int4, symmetric=True), 'minmax')
2025-12-31 09:57:00,036 - msmodelslim.app.naive_quantization - INFO - Using model adapter Qwen3ModelAdapter.
2025-12-31 09:57:00,036 - msmodelslim.app.naive_quantization - INFO - ===========GET BEST PRACTICE===========
2025-12-31 09:57:00,055 - msmodelslim.app.naive_quantization - INFO - Naive Quant apply config_id: qwen3-32b-dense-w8a8
2025-12-31 09:57:00,055 - msmodelslim.app.naive_quantization - INFO - Get best practice qwen3-32b-dense-w8a8 success.
2025-12-31 09:57:00,055 - msmodelslim.app.naive_quantization - INFO - ===========QUANTIZE MODEL===========
2025-12-31 09:57:00,065 - msmodelslim - WARNING - Unable to import torch_npu.
2025-12-31 09:57:00,065 - msmodelslim - WARNING - Unable to import torch_npu.
2025-12-31 09:57:00,070 - msmodelslim.core.quant_service.proxy - WARNING - Failed to load plugin modelslim_v0: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils'
2025-12-31 09:57:00,070 - msmodelslim.core.quant_service.proxy - INFO - Load quant service plugin modelslim_v1 success!
2025-12-31 09:57:00,070 - msmodelslim.core.quant_service.proxy - INFO - Load quant service plugin multimodal_sd_modelslim_v1 success!
2025-12-31 09:57:00,073 - msmodelslim.core.quant_service.proxy - INFO - Load quant service plugin multimodal_vlm_modelslim_v1 success!
2025-12-31 09:57:00,073 - msmodelslim.app.naive_quantization - ERROR - [EnvError] Code: 100, Message: Quant service plugin for backend 'modelslim_v0' failed to load: Exception: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils' Traceback: Traceback (most recent call last): File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/core/quant_service/proxy.py", line 81, in load_plugins plugin_class = entry.load() File "/root/miniconda3/lib/python3.13/importlib/metadata/init.py", line 179, in load module = import_module(match.group('module')) File "/root/miniconda3/lib/python3.13/importlib/init.py", line 88, in import_module return _bootstrap._gcd_import(name[level:], package, level) ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "", line 1387, in _gcd_import File "", line 1360, in _find_and_load File "", line 1331, in _find_and_load_unlocked File "", line 935, in _load_unlocked File "", line 1023, in exec_module File "", line 488, in _call_with_frames_removed File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/core/quant_service/modelslim_v0/quant_service.py", line 10, in from msmodelslim.pytorch.llm_ptq.anti_outlier import AntiOutlier, AntiOutlierConfig File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/pytorch/llm_ptq/anti_outlier/init.py", line 5, in from .anti_outlier import AntiOutlier File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/pytorch/llm_ptq/anti_outlier/anti_outlier.py", line 42, in from msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils import ( ...<4 lines>... ) ModuleNotFoundError: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils'
Traceback (most recent call last):
File "/root/miniconda3/bin/msmodelslim", line 7, in
sys.exit(main())
~~~~^^
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/cli/main.py", line 99, in main
quant_main(args)
~~~~~~~~~~^^^^^^
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/cli/naive_quantization/main.py", line 50, in main
app.quant(
~~~~~~~~~^
model_type=args.model_type,
^^^^^^^^^^^^^^^^^^^^^^^^^^^
...<6 lines>...
trust_remote_code=args.trust_remote_code
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/utils/logging.py", line 289, in wrapper
return func(*args, **kwargs)
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/utils/exception_decorator.py", line 159, in wrapper
return func(*args, **kwargs)
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/app/naive_quantization/application.py", line 250, in quant
self._quant(
~~~~~~~~~~~^
model_type, model_path, save_path, device_type,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
device_index, quant_type, config_path, trust_remote_code
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/utils/logging.py", line 289, in wrapper
return func(*args, **kwargs)
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/app/naive_quantization/application.py", line 281, in _quant
self.quant_service.quantize(
~~~~~~~~~~~~~~~~~~~~~~~~~~~^
quant_config=practice_config.extract_quant_config(),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
...❤️ lines>...
device_indices=device_index
^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/utils/logging.py", line 289, in wrapper
return func(*args, **kwargs)
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/core/quant_service/proxy.py", line 39, in quantize
self.quant_service = load_quant_service_cls(quant_config.apiversion)(self.dataset_loader)
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/core/quant_service/proxy.py", line 99, in load_quant_service_cls
raise EnvError(
f"Quant service plugin for backend '{backend_name}' failed to load:\n{_FAILED_PLUGINS[backend_name]}")
msmodelslim.utils.exception.EnvError: Code: 100, Message: Quant service plugin for backend 'modelslim_v0' failed to load:
Exception: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils'
Traceback: Traceback (most recent call last):
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/core/quant_service/proxy.py", line 81, in load_plugins
plugin_class = entry.load()
File "/root/miniconda3/lib/python3.13/importlib/metadata/init.py", line 179, in load
module = import_module(match.group('module'))
File "/root/miniconda3/lib/python3.13/importlib/init.py", line 88, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "", line 1387, in _gcd_import
File "", line 1360, in _find_and_load
File "", line 1331, in _find_and_load_unlocked
File "", line 935, in _load_unlocked
File "", line 1023, in exec_module
File "", line 488, in _call_with_frames_removed
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/core/quant_service/modelslim_v0/quant_service.py", line 10, in
from msmodelslim.pytorch.llm_ptq.anti_outlier import AntiOutlier, AntiOutlierConfig
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/pytorch/llm_ptq/anti_outlier/init.py", line 5, in
from .anti_outlier import AntiOutlier
File "/root/miniconda3/lib/python3.13/site-packages/msmodelslim/pytorch/llm_ptq/anti_outlier/anti_outlier.py", line 42, in
from msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils import (
...<4 lines>...
)
ModuleNotFoundError: No module named 'msmodelslim.pytorch.llm_ptq.anti_outlier.anti_utils'

anreywmh成员

1月5日评论：

您好，python3.13不支持，您可以选择换个python版本或者采用我们最新的V1一键量化框架
推荐您使用V1一键量化框架，可参考该yaml：https://gitcode.com/Ascend/msit/pull/4814/diffs
msmodelslim quant --model_path /home/weight/Qwen3-32B --save_path /home/weight/Qwen3-32B-W8A8 --device npu --model_type Qwen3-32B --config_path ${config_path} --trust_remote_code True

A AiBoGe成员

1月6日关联了看板：msit

A anreywmh成员

1月27日 issue状态由 TODO 改变为 DONE