Open deltabytesyt opened 1 year ago
Hi , getting same issue .
Colab The original colab no longer works by a combination of Google's tendency to forward-break things and Python's package management system. I do not intend to keep fixing it so it has been removed. Apologies!
follow the new guide or use colab fixes in other issues/prs
ModuleNotFoundError Traceback (most recent call last) in <cell line: 5>() 3 voice = 'tom' 4 ----> 5 from einops import rearrange, repeat 6 from tortoise.api import TextToSpeech 7 from tortoise.utils.audio import load_audio, load_voice, load_voices
ModuleNotFoundError: No module named 'einops'
Requirements seem borked and do not work out of the box
In colab first: !pip3 install --force einops==0.4.1
!git clone https://github.com/neonbjb/tortoise-tts.git %cd tortoise-tts
!pip3 install tqdm !pip3 install rotary_embedding_torch !pip3 install transformers==4.19 !pip3 install tokenizers !pip3 install inflect !pip3 install progressbar !pip3 install einops==0.4.1 !pip3 install unidecode !pip3 install scipy==0.10.1 !pip3 install librosa==0.9.1 !pip3 install numba==0.48.0 !pip3 install ffmpeg
!pip3 install numpy==1.22 !pip3 install numba==0.48.0 !pip3 install torchaudio !pip3 install threadpoolctl !pip3 install llvmlite !pip3 install appdirs
This will throw some errors but should pass. I can generate audio.
***Colab ram limits will be eventually exceeded running tts to generate >3 clips in my experience. A single generation will push the environment to 11.7gb ram/12.7
... Requirements seem borked and do not work out of the box ...
Indeed, it's a shame. Thanks for your answer it's the only useful one.
ModuleNotFoundError Traceback (most recent call last) in <cell line: 5>() 3 voice = 'tom' 4 ----> 5 from einops import rearrange, repeat 6 from tortoise.api import TextToSpeech 7 from tortoise.utils.audio import load_audio, load_voice, load_voices ModuleNotFoundError: No module named 'einops'
Requirements seem borked and do not work out of the box
In colab first: !pip3 install --force einops==0.4.1
!git clone https://github.com/neonbjb/tortoise-tts.git %cd tortoise-tts
!pip3 install tqdm !pip3 install rotary_embedding_torch !pip3 install transformers==4.19 !pip3 install tokenizers !pip3 install inflect !pip3 install progressbar !pip3 install einops==0.4.1 !pip3 install unidecode !pip3 install scipy==0.10.1 !pip3 install librosa==0.9.1 !pip3 install numba==0.48.0 !pip3 install ffmpeg #!pip3 install numpy==1.20.3 !pip3 install numpy==1.22 !pip3 install numba==0.48.0 !pip3 install torchaudio !pip3 install threadpoolctl !pip3 install llvmlite !pip3 install appdirs
This will throw some errors but should pass. I can generate audio.
***Colab ram limits will be eventually exceeded running tts to generate >3 clips in my experience. A single generation will push the environment to 11.7gb ram/12.7
alas. rotary_embedding_torch now requires >=0.7 einops. Attempted newer version. Many breaks followed. 😥
ModuleNotFoundError Traceback (most recent call last) in <cell line: 5>()
3 voice = 'tom'
4
----> 5 from einops import rearrange, repeat
6 from tortoise.api import TextToSpeech
7 from tortoise.utils.audio import load_audio, load_voice, load_voices
ModuleNotFoundError: No module named 'einops'