Issues with einops on Windows #74

New Issue

Dziombolong · 2023-03-06T14:54:07Z

Dziombolong commented

2023-03-06 14:54:07 +00:00

Hello,

When running setup-cuda.bat, this message comes up after all is finished:

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
vector-quantize-pytorch 1.0.7 requires einops>=0.6, but you have einops 0.3.2 which is incompatible.
local-attention 1.8.4 requires einops>=0.6.0, but you have einops 0.3.2 which is incompatible.

This makes me unable to train my models, as this error comes up after I try to start training:

D:\ai-voice-cloning>train.bat D:\ai-voice-cloning\training\Kalabala\train.yaml

D:\ai-voice-cloning>call .\venv\Scripts\activate.bat
Traceback (most recent call last):
  File "D:\ai-voice-cloning\src\train.py", line 49, in <module>
    from codes import train as tr
  File "D:\ai-voice-cloning\./dlas\codes\train.py", line 16, in <module>
    from trainer.ExtensibleTrainer import ExtensibleTrainer
  File "D:\ai-voice-cloning\./dlas/codes\trainer\ExtensibleTrainer.py", line 17, in <module>
    from trainer.injectors.audio_injectors import normalize_mel
  File "D:\ai-voice-cloning\./dlas/codes\trainer\injectors\audio_injectors.py", line 7, in <module>
    from models.audio.music.cheater_gen_ar import ConditioningAR
  File "D:\ai-voice-cloning\./dlas/codes\models\audio\music\cheater_gen_ar.py", line 8, in <module>
    from models.audio.tts.lucidrains_dvae import DiscreteVAE
  File "D:\ai-voice-cloning\./dlas/codes\models\audio\tts\lucidrains_dvae.py", line 10, in <module>
    from vector_quantize_pytorch import VectorQuantize
  File "D:\ai-voice-cloning\venv\lib\site-packages\vector_quantize_pytorch\__init__.py", line 2, in <module>
    from vector_quantize_pytorch.residual_vq import ResidualVQ
  File "D:\ai-voice-cloning\venv\lib\site-packages\vector_quantize_pytorch\residual_vq.py", line 10, in <module>
    from einops import rearrange, repeat, pack, unpack
ImportError: cannot import name 'pack' from 'einops' (D:\ai-voice-cloning\venv\lib\site-packages\einops\__init__.py)
Press any key to continue . . .

Additional Info:
Python 3.9.13
RTX 2060 6GB
Windows 10

Hello, When running setup-cuda.bat, this message comes up after all is finished: ``` ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. vector-quantize-pytorch 1.0.7 requires einops>=0.6, but you have einops 0.3.2 which is incompatible. local-attention 1.8.4 requires einops>=0.6.0, but you have einops 0.3.2 which is incompatible. ``` This makes me unable to train my models, as this error comes up after I try to start training: ``` D:\ai-voice-cloning>train.bat D:\ai-voice-cloning\training\Kalabala\train.yaml D:\ai-voice-cloning>call .\venv\Scripts\activate.bat Traceback (most recent call last): File "D:\ai-voice-cloning\src\train.py", line 49, in <module> from codes import train as tr File "D:\ai-voice-cloning\./dlas\codes\train.py", line 16, in <module> from trainer.ExtensibleTrainer import ExtensibleTrainer File "D:\ai-voice-cloning\./dlas/codes\trainer\ExtensibleTrainer.py", line 17, in <module> from trainer.injectors.audio_injectors import normalize_mel File "D:\ai-voice-cloning\./dlas/codes\trainer\injectors\audio_injectors.py", line 7, in <module> from models.audio.music.cheater_gen_ar import ConditioningAR File "D:\ai-voice-cloning\./dlas/codes\models\audio\music\cheater_gen_ar.py", line 8, in <module> from models.audio.tts.lucidrains_dvae import DiscreteVAE File "D:\ai-voice-cloning\./dlas/codes\models\audio\tts\lucidrains_dvae.py", line 10, in <module> from vector_quantize_pytorch import VectorQuantize File "D:\ai-voice-cloning\venv\lib\site-packages\vector_quantize_pytorch\__init__.py", line 2, in <module> from vector_quantize_pytorch.residual_vq import ResidualVQ File "D:\ai-voice-cloning\venv\lib\site-packages\vector_quantize_pytorch\residual_vq.py", line 10, in <module> from einops import rearrange, repeat, pack, unpack ImportError: cannot import name 'pack' from 'einops' (D:\ai-voice-cloning\venv\lib\site-packages\einops\__init__.py) Press any key to continue . . . ``` Additional Info: Python 3.9.13 RTX 2060 6GB Windows 10

deviandice commented

2023-03-06 15:55:39 +00:00

You can jury rig it a bit. It's to do with whisperX. If you're not using it just do the following in powershell. Had the same issue on linux and this worked for me.

./venv/scripts/activate.bat 
pip install einops==0.6 --no-deps
deactivate

You can jury rig it a bit. It's to do with whisperX. If you're not using it just do the following in powershell. Had the same issue on linux and this worked for me. ``` ./venv/scripts/activate.bat pip install einops==0.6 --no-deps deactivate ```

mrq commented

2023-03-06 16:45:29 +00:00

Remedied in 11a1f6a00e. I forgot to commit it at 1AM last night.

WhisperX will break the dependencies needed for DLAS, so installing DLAS's dependencies will fix it.

Remedied in 11a1f6a00ea3a2408d7592537a748433ad3b5496. I forgot to commit it at 1AM last night. WhisperX will break the dependencies needed for DLAS, so installing DLAS's dependencies will fix it.

mrq closed this issue

2023-03-06 16:45:29 +00:00

mrq referenced this issue

2023-03-06 21:51:17 +00:00

Just some questions from a newbie... #69

nirurin referenced this issue

2023-03-06 22:01:41 +00:00

Just some questions from a newbie... #69

Sign in to join this conversation.