macOS 生成的时候报错 Initializing libomp.dylib, but found libiomp5.dylib already initialized

[cdk8s-zelda]

先感谢作者提供支持

系统环境

• macOS 12.6.7 Intel CPU
• Python 3.9.13

报错截图

2024-05-31 21:33:36,503 - modelscope - INFO - PyTorch version 2.2.2 Found.
2024-05-31 21:33:36,503 - modelscope - INFO - Loading ast index from /Users/meek/.cache/modelscope/ast_indexer
2024-05-31 21:33:36,539 - modelscope - INFO - Loading done! Current index file version is 1.14.0, with md5 54c70b20f857389f69e0735c6ef5281c and a total number of 976 components indexed
INFO:ChatTTS.core:Load from local: /Users/meek/my-software/chatTTS-ui/models/pzc163/chatTTS
WARNING:ChatTTS.utils.gpu_utils:No GPU found, use CPU instead
INFO:ChatTTS.core:use cpu
INFO:ChatTTS.core:vocos loaded.
INFO:ChatTTS.core:dvae loaded.
INFO:ChatTTS.core:gpt loaded.
INFO:ChatTTS.core:decoder loaded.
INFO:ChatTTS.core:tokenizer loaded.
INFO:ChatTTS.core:All initialized.
启动:['127.0.0.1', '9966']
  0%|                                                                                                                                                                                    | 0/384 [00:00<?, ?it/s]OMP: Error #15: Initializing libomp.dylib, but found libiomp5.dylib already initialized.
OMP: Hint This means that multiple copies of the OpenMP runtime have been linked into the program. That is dangerous, since it can degrade performance or cause incorrect results. The best thing to do is to ensure that only a single OpenMP runtime is linked into the process, e.g. by avoiding static linking of the OpenMP runtime in any library. As an unsafe, unsupported, undocumented workaround you can set the environment variable KMP_DUPLICATE_LIB_OK=TRUE to allow the program to continue to execute, but that may cause crashes or silently produce incorrect results. For more information, please see http://openmp.llvm.org/
[1]    7798 abort      python3 app.py
/Users/meek/opt/anaconda3/lib/python3.9/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 1 leaked semaphore objects to clean up at shutdown
  warnings.warn('resource_tracker: There appear to be %d

尝试过一些措施，但是都没有效果

export KMP_DUPLICATE_LIB_OK=FALSE

export KMP_DUPLICATE_LIB_OK=TRUE

export OMP_DYNAMIC=FALSE

DYLD_INSERT_LIBRARIES=/usr/local/Cellar/libomp/18.1.6/lib/libomp.dylib python3 app.py

[jianchang512]

在app.py的 import os 的下一行，添加代码

os.environ['KMP_DUPLICATE_LIB_OK'] = 'True'

试试

[cdk8s-zelda]

在app.py的 import os 的下一行，添加代码

os.environ['KMP_DUPLICATE_LIB_OK'] = 'True'

试试

• 感谢，现在不报错了。但是过去了20分钟，进度一直为0。
• 我就只生成这样一句话而已：这个报错是因为在你的项目中同时初始化了两个不同版本的 OpenMP 运行时库
• 所以应该不是这个问题吧 https://github.com/jianchang512/ChatTTS-ui/issues/19

2024-05-31 22:01:36,693 - modelscope - INFO - PyTorch version 2.2.2 Found.
2024-05-31 22:01:36,693 - modelscope - INFO - Loading ast index from /Users/meek/.cache/modelscope/ast_indexer
2024-05-31 22:01:36,728 - modelscope - INFO - Loading done! Current index file version is 1.14.0, with md5 54c70b20f857389f69e0735c6ef5281c and a total number of 976 components indexed
INFO:ChatTTS.core:Load from local: /Users/meek/my-software/chatTTS-ui/models/pzc163/chatTTS
WARNING:ChatTTS.utils.gpu_utils:No GPU found, use CPU instead
INFO:ChatTTS.core:use cpu
INFO:ChatTTS.core:vocos loaded.
INFO:ChatTTS.core:dvae loaded.
INFO:ChatTTS.core:gpt loaded.
INFO:ChatTTS.core:decoder loaded.
INFO:ChatTTS.core:tokenizer loaded.
INFO:ChatTTS.core:All initialized.
启动:['127.0.0.1', '9966']
  0%|

[jianchang512]

app.py中

chat.load_models(source="local",local_path=CHATTTS_DIR)

改为

chat.load_models(source="local",local_path=CHATTTS_DIR,compile=False)

试试

[cdk8s-zelda]

app.py中

chat.load_models(source="local",local_path=CHATTTS_DIR)

改为

chat.load_models(source="local",local_path=CHATTTS_DIR,compile=False)

试试

• 非常非常感谢，改完代码就可以了。
• 经过测试，生成 这个报错是因为在你的项目中同时初始化了两个不同版本的 OpenMP 运行时库 这句话用时 14 秒。
• 环境：CPU i9-10900K