发表于：2023-7-13 16:41:010次点击

安装部署Stable Diffusion的时候遇到困难[阿里云机器学习PAI]

我在安装部署Stable Diffusion的时候遇到困难。在执行 python launch.py --listen --lowvram --no-half --skip-torch-cuda-test 报错，好像是 conda 里面pytorch配置有问题，但是我不知道要怎么解决。

附上详细的错误信息：

Python 3.10.6 (main, Oct 24 2022, 16:07:47) [GCC 11.2.0]
Commit hash: 0cc0ee1bcb4c24a8c9715f66cede06601bfc00c8
Cloning Stable Diffusion into repositories/stable-diffusion-stability-ai...
Cloning Taming Transformers into repositories/taming-transformers...
Cloning K-diffusion into repositories/k-diffusion...
Cloning CodeFormer into repositories/CodeFormer...
Cloning BLIP into repositories/BLIP...
Installing requirements for CodeFormer
Installing requirements for Web UI
Launching Web UI with arguments: --listen --lowvram --no-half --ckpt /root/models/Stable-diffusion/v1-5-pruned-emaonly.safetensors --lora-dir /root/models/Lora
/root/.conda/envs/aigc/lib/python3.10/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: '/root/.conda/envs/aigc/lib/python3.10/site-packages/torchvision/image.so: undefined symbol: _ZN5torch3jit17parseSchemaOrNameERKSs'If you don't plan on using image functionality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torchvision` from source?
  warn(
/root/.conda/envs/aigc/lib/python3.10/site-packages/torchvision/transforms/functional_tensor.py:5: UserWarning: The torchvision.transforms.functional_tensor module is deprecated in 0.15 and will be **removed in 0.17**. Please don't rely on it. You probably just need to use APIs in torchvision.transforms.functional or in torchvision.transforms.v2.functional.
  warnings.warn(
Warning: caught exception 'Torch not compiled with CUDA enabled', memory monitor disabled
Traceback (most recent call last):
  File "/root/stable-diffusion-webui/launch.py", line 361, in 
    start()
  File "/root/stable-diffusion-webui/launch.py", line 352, in start
    import webui
  File "/root/stable-diffusion-webui/webui.py", line 15, in 
    from modules import import_hook, errors, extra_networks, ui_extra_networks_checkpoints
  File "/root/stable-diffusion-webui/modules/ui_extra_networks_checkpoints.py", line 6, in 
    from modules import shared, ui_extra_networks, sd_models
  File "/root/stable-diffusion-webui/modules/sd_models.py", line 17, in 
    from modules.sd_hijack_inpainting import do_inpainting_hijack
  File "/root/stable-diffusion-webui/modules/sd_hijack_inpainting.py", line 7, in 
    import ldm.models.diffusion.ddpm
  File "/root/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/models/diffusion/ddpm.py", line 12, in 
    import pytorch_lightning as pl
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/__init__.py", line 34, in 
    from pytorch_lightning.callbacks import Callback  # noqa: E402
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/callbacks/__init__.py", line 14, in 
    from pytorch_lightning.callbacks.callback import Callback
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/callbacks/callback.py", line 25, in 
    from pytorch_lightning.utilities.types import STEP_OUTPUT
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/utilities/types.py", line 28, in 
    from torchmetrics import Metric
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/__init__.py", line 14, in 
    from torchmetrics import functional  # noqa: E402
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/functional/__init__.py", line 14, in 
    from torchmetrics.functional.audio._deprecated import _permutation_invariant_training as permutation_invariant_training
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/functional/audio/__init__.py", line 14, in 
    from torchmetrics.functional.audio.pit import permutation_invariant_training, pit_permutate
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/functional/audio/pit.py", line 23, in 
    from torchmetrics.utilities import rank_zero_warn
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/__init__.py", line 14, in 
    from torchmetrics.utilities.checks import check_forward_full_state_property
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/checks.py", line 25, in 
    from torchmetrics.metric import Metric
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/metric.py", line 30, in 
    from torchmetrics.utilities.data import (
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/data.py", line 22, in 
    from torchmetrics.utilities.imports import _TORCH_GREATER_EQUAL_1_12, _XLA_AVAILABLE
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/imports.py", line 48, in 
    _TORCHAUDIO_GREATER_EQUAL_0_10: Optional[bool] = compare_version("torchaudio", operator.ge, "0.10.0")
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/lightning_utilities/core/imports.py", line 73, in compare_version
    pkg = importlib.import_module(package)
  File "/root/.conda/envs/aigc/lib/python3.10/importlib/__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/__init__.py", line 1, in 
    from torchaudio import (  # noqa: F401
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/_extension/__init__.py", line 43, in 
    _load_lib("libtorchaudio")
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/_extension/utils.py", line 61, in _load_lib
    torch.ops.load_library(path)
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torch/_ops.py", line 643, in load_library
    ctypes.CDLL(path)
  File "/root/.conda/envs/aigc/lib/python3.10/ctypes/__init__.py", line 374, in __init__
    self._handle = _dlopen(self._name, mode)
OSError: /root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/lib/libtorchaudio.so: undefined symbol: _ZNK5torch8autograd4Node4nameEv
(aigc) [root@iZuf61kljas3cjo5gb9jauZ stable-diffusion-webui]# python launch.py --listen --lowvram --no-half 
Python 3.10.6 (main, Oct 24 2022, 16:07:47) [GCC 11.2.0]
Commit hash: 0cc0ee1bcb4c24a8c9715f66cede06601bfc00c8
Traceback (most recent call last):
  File "/root/stable-diffusion-webui/launch.py", line 360, in 
    prepare_environment()
  File "/root/stable-diffusion-webui/launch.py", line 272, in prepare_environment
    run_python("import torch; assert torch.cuda.is_available(), 'Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check'")
  File "/root/stable-diffusion-webui/launch.py", line 129, in run_python
    return run(f'"{python}" -c "{code}"', desc, errdesc)
  File "/root/stable-diffusion-webui/launch.py", line 105, in run
    raise RuntimeError(message)
RuntimeError: Error running command.
Command: "/root/.conda/envs/aigc/bin/python" -c "import torch; assert torch.cuda.is_available(), 'Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check'"
Error code: 1
stdout: 
stderr: Traceback (most recent call last):
  File "", line 1, in 
AssertionError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check


(aigc) [root@iZuf61kljas3cjo5gb9jauZ stable-diffusion-webui]# python launch.py --listen --lowvram --no-half --skip-torch-cuda-test
Python 3.10.6 (main, Oct 24 2022, 16:07:47) [GCC 11.2.0]
Commit hash: 0cc0ee1bcb4c24a8c9715f66cede06601bfc00c8
Installing requirements for Web UI
Launching Web UI with arguments: --listen --lowvram --no-half --ckpt /root/models/Stable-diffusion/v1-5-pruned-emaonly.safetensors --lora-dir /root/models/Lora
/root/.conda/envs/aigc/lib/python3.10/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: '/root/.conda/envs/aigc/lib/python3.10/site-packages/torchvision/image.so: undefined symbol: _ZN5torch3jit17parseSchemaOrNameERKSs'If you don't plan on using image functionality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torchvision` from source?
  warn(
/root/.conda/envs/aigc/lib/python3.10/site-packages/torchvision/transforms/functional_tensor.py:5: UserWarning: The torchvision.transforms.functional_tensor module is deprecated in 0.15 and will be **removed in 0.17**. Please don't rely on it. You probably just need to use APIs in torchvision.transforms.functional or in torchvision.transforms.v2.functional.
  warnings.warn(
Warning: caught exception 'Torch not compiled with CUDA enabled', memory monitor disabled
Traceback (most recent call last):
  File "/root/stable-diffusion-webui/launch.py", line 361, in 
    start()
  File "/root/stable-diffusion-webui/launch.py", line 352, in start
    import webui
  File "/root/stable-diffusion-webui/webui.py", line 15, in 
    from modules import import_hook, errors, extra_networks, ui_extra_networks_checkpoints
  File "/root/stable-diffusion-webui/modules/ui_extra_networks_checkpoints.py", line 6, in 
    from modules import shared, ui_extra_networks, sd_models
  File "/root/stable-diffusion-webui/modules/sd_models.py", line 17, in 
    from modules.sd_hijack_inpainting import do_inpainting_hijack
  File "/root/stable-diffusion-webui/modules/sd_hijack_inpainting.py", line 7, in 
    import ldm.models.diffusion.ddpm
  File "/root/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/models/diffusion/ddpm.py", line 12, in 
    import pytorch_lightning as pl
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/__init__.py", line 34, in 
    from pytorch_lightning.callbacks import Callback  # noqa: E402
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/callbacks/__init__.py", line 14, in 
    from pytorch_lightning.callbacks.callback import Callback
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/callbacks/callback.py", line 25, in 
    from pytorch_lightning.utilities.types import STEP_OUTPUT
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/pytorch_lightning/utilities/types.py", line 28, in 
    from torchmetrics import Metric
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/__init__.py", line 14, in 
    from torchmetrics import functional  # noqa: E402
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/functional/__init__.py", line 14, in 
    from torchmetrics.functional.audio._deprecated import _permutation_invariant_training as permutation_invariant_training
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/functional/audio/__init__.py", line 14, in 
    from torchmetrics.functional.audio.pit import permutation_invariant_training, pit_permutate
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/functional/audio/pit.py", line 23, in 
    from torchmetrics.utilities import rank_zero_warn
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/__init__.py", line 14, in 
    from torchmetrics.utilities.checks import check_forward_full_state_property
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/checks.py", line 25, in 
    from torchmetrics.metric import Metric
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/metric.py", line 30, in 
    from torchmetrics.utilities.data import (
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/data.py", line 22, in 
    from torchmetrics.utilities.imports import _TORCH_GREATER_EQUAL_1_12, _XLA_AVAILABLE
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchmetrics/utilities/imports.py", line 48, in 
    _TORCHAUDIO_GREATER_EQUAL_0_10: Optional[bool] = compare_version("torchaudio", operator.ge, "0.10.0")
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/lightning_utilities/core/imports.py", line 73, in compare_version
    pkg = importlib.import_module(package)
  File "/root/.conda/envs/aigc/lib/python3.10/importlib/__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/__init__.py", line 1, in 
    from torchaudio import (  # noqa: F401
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/_extension/__init__.py", line 43, in 
    _load_lib("libtorchaudio")
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/_extension/utils.py", line 61, in _load_lib
    torch.ops.load_library(path)
  File "/root/.conda/envs/aigc/lib/python3.10/site-packages/torch/_ops.py", line 643, in load_library
    ctypes.CDLL(path)
  File "/root/.conda/envs/aigc/lib/python3.10/ctypes/__init__.py", line 374, in __init__
    self._handle = _dlopen(self._name, mode)
OSError: /root/.conda/envs/aigc/lib/python3.10/site-packages/torchaudio/lib/libtorchaudio.so: undefined symbol: _ZNK5torch8autograd4Node4nameEv

「点点赞赏，手留余香」

还没有人赞赏，快来当第一个赞赏的人吧！

=====这是一个广告位，招租中，联系qq 78315851====

6 条回复 A 作者 M 管理员

玥轩AM 2023-11-28 2:55:58 1
根据您提供的信息，可能是由于conda环境中的PyTorch版本与Stable Diffusion所需的版本不兼容导致的。您可以尝试以下方法解决问题：

创建一个新的conda环境，并在其中安装Stable Diffusion所需的依赖项。这可以确保您的环境中没有其他版本的库干扰。您可以使用以下命令创建新环境并激活它：

conda create -n stable_diffusion python=3.8conda activate stable_diffusion

在激活的环境中，尝试使用以下命令安装Stable Diffusion及其依赖项：

pip install -r requirements.txt

如果仍然遇到问题，请检查requirements.txt文件中列出的所有依赖项是否与您的conda环境中的其他库兼容。如果不兼容，您可能需要手动更新或降级这些依赖项的版本。

如果问题仍然存在，您可以尝试在conda中卸载当前的PyTorch版本，然后重新安装一个与Stable Diffusion兼容的版本。首先，使用以下命令卸载当前的PyTorch版本：

conda uninstall pytorch torchvision torchaudio cudatoolkit=10.2 -c pytorch

然后，根据您的系统和CUDA版本，从PyTorch官方网站获取适当的安装命令。例如，对于Windows系统和CUDA 10.2，您可以使用以下命令安装PyTorch：

conda install pytorch torchvision torchaudio cudatoolkit=10.2 -c pytorch -c conda-forge

希望这些建议能帮助您解决问题。如果您还有其他问题，请随时提问。
已喜欢已反对
游客i3d3sqh6dybboAM 2023-11-28 2:55:58 2

重装torchaudio
先卸载
conda uninstall torchaudio
再重装
conda install torchaudio

已喜欢已反对
魏红斌AM 2023-11-28 2:55:58 3
可能是由于您的conda环境中的PyTorch版本和CUDA支持存在问题导致的。为了解决这个问题，您可以尝试以下几个步骤：

确保您的conda环境中已经安装了最新版本的PyTorch和CUDA。您可以通过在conda命令行中执行以下命令来检查您的PyTorch和CUDA版本：

conda list | grep torchconda list | grep cuda

如果您的PyTorch和CUDA版本不是最新的，您可以通过以下命令来更新它们：

conda install pytorch torchvision cudatoolkit=11.0

如果您已经安装了最新版本的PyTorch和CUDA，但是仍然遇到问题，您可以尝试重新编译PyTorch。您可以通过以下命令来编译PyTorch：

torchvision.utils.build torchvision()

如果您仍然无法解决问题，您可以尝试重新安装PyTorch和CUDA，并确保它们的版本和环境变量都正确配置。
已喜欢已反对
ReaganYoungAM 2023-11-28 2:55:58 4
在安装和部署 Stable Diffusion 时遇到问题可能是由于 PyTorch 配置或环境问题导致的。下面是一些可能的解决方案：

确保已经正确安装了 conda 和 PyTorch。可以通过运行 conda list 命令来查看已安装的包列表，确认 PyTorch 是否已经正确安装。

如果你尚未安装 PyTorch 或者想重新安装，可以尝试使用 conda 来安装。可以使用以下命令创建一个新的 conda 环境，并在环境中安装 PyTorch：

conda create -n myenv python=3.7conda activate myenvconda install pytorch torchvision torchaudio cpuonly -c pytorch

请注意，上述命令假定你希望安装 CPU 版本的 PyTorch。如果你需要 GPU 版本，请根据你的 GPU 类型和 CUDA 版本选择合适的安装命令。

如果你已经正确安装了 PyTorch，但仍然遇到问题，可以尝试更新 PyTorch 到最新版本。使用以下命令更新 PyTorch：

conda update pytorch torchvision torchaudio -c pytorch

如果问题仍然存在，可能是由于其他依赖项或环境配置问题导致的。你可以尝试查看错误消息中的详细信息，并根据错误信息进一步调查和解决。

请注意，以上解决方案提供了一些常见的解决方法，但具体情况可能因操作系统、环境和其他因素而有所不同。如果问题仍然存在，我建议你参考相关文档、社区论坛或与开发者进行沟通，以获取更具体的帮助和支持。
已喜欢已反对
Star时光AM 2023-11-28 2:55:58 5
在安装和部署 Stable Diffusion 时遇到问题可能有多种原因。关于您提到的 conda 和 PyTorch 配置问题，以下是一些常见的解决方法：

确认环境：确保您使用的环境符合 Stable Diffusion 的要求。首先，请确认您已经安装了适当版本的 Python 和 conda，并且配置了正确的环境变量。

更新 conda：尝试更新 conda 到最新版本，可以使用以下命令：

conda update conda

清理环境：如果之前安装过其他版本的 PyTorch 或相关依赖项，可能会导致冲突。您可以尝试卸载现有的 PyTorch 及其依赖项，并重新安装 Stable Diffusion 所需的依赖项。

创建虚拟环境：为了避免与现有环境的冲突，您可以创建一个新的虚拟环境，并在其中安装 Stable Diffusion 所需的依赖项。可以使用以下命令创建虚拟环境：

conda create -n stable-diffusion python=3.8

安装 PyTorch：确保您安装了兼容版本的 PyTorch。根据您的系统和硬件配置，可以从 PyTorch 的官方网站或 conda 命令进行安装。例如，对于 CPU 版本可以使用以下命令：

conda install pytorch cpuonly -c pytorch

依赖项解决：在安装 Stable Diffusion 之前，请确保已经安装了它所需的所有依赖项。您可以查看 Stable Diffusion 的文档或 README 文件，找到相关依赖项的列表，并逐个进行安装。
已喜欢已反对
小周sirAM 2023-11-28 2:55:58 6
在安装部署 Stable Diffusion 时遇到 conda 中 PyTorch 配置的问题，下面是一些可能的解决方法：

更新 conda：首先，确保您正在使用最新版本的 conda。在命令行中运行以下命令可以更新 conda：

conda update -n base -c defaults conda

创建新的 conda 环境：尝试创建一个新的 conda 环境，并在该环境中安装 Stable Diffusion 和相应的依赖项。您可以使用以下命令创建新环境：

conda create -n sd-env python=3.7conda activate sd-env

安装 PyTorch：在新的 conda 环境中，尝试使用 conda 或 pip 安装正确版本的 PyTorch。例如，使用以下命令安装 PyTorch CPU 版本：

conda install pytorch cpuonly -c pytorch

检查 CUDA 配置：如果您的系统具有 NVIDIA GPU 并要使用 GPU 加速，确保已正确配置 CUDA。请确认您的 GPU 驱动程序已正确安装，并且与所安装的 PyTorch 版本兼容。

运行脚本：在完成以上步骤后，再次运行 python launch.py --listen --lowvram --no-half --skip-torch-cuda-test 命令，看是否还会出现错误。
已喜欢已反对
算精通AM 2023-11-28 2:55:58 7

如果在安装部署 Stable Diffusion 时遇到了问题，可能是由于 pytorch 配置出现了问题。以下是一些可能的解决步骤：

检查 pytorch 是否安装成功：在执行 python launch.py 命令之前，请确保已经成功安装了 pytorch，并且 pytorch 版本与 Stable Diffusion 要求的版本一致。可以尝试在终端中输入 python 进入 Python 解释器，然后输入 import torch，查看是否有报错提示。

确认 conda 环境正确：在执行 python launch.py 命令之前，需要确保当前的 conda 环境已经正确激活。可以尝试在终端中输入 conda activate your_env，将 your_env 替换为正确的 conda 环境名称，查看是否激活成功。

更新 pytorch 和 torchvision：如果已经安装了 pytorch 和 torchvision，可以尝试更新这两个包到最新版本，以确保与 Stable Diffusion 兼容。可以使用 conda update pytorch torchvision 命令进行更新。

检查 CUDA 版本：如果您的系统上安装了 CUDA，需要确保 CUDA 版本与 Stable Diffusion 支持的 CUDA 版本一致。可以尝试在终端中输入 nvcc –version，查看 CUDA 版本信息。

检查 GPU 驱动程序：如果您的系统上安装了 GPU 驱动程序，请确保驱动程序版本与 Stable Diffusion 支持的版本一致。可以尝试在终端中输入 nvidia-smi，查看 GPU 驱动程序版本信息

已喜欢已反对

2024 年 12 月
一	二	三	四	五	六	日
						1
2	3	4	5	6	7	8
9	10	11	12	13	14	15
16	17	18	19	20	21	22
23	24	25	26	27	28	29
30	31