You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
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!
The reason will be displayed to describe this comment to others. Learn more.
FWIW, this sequence works in Google Colab. There are still issues with scipy, numpy and llvmlite. But generation works.
models_path = MYDRIVE_PATH / "AI" / "models"
%cd {str(models_path)}
!git clone https://github.com/neonbjb/tortoise-tts.git
%cd tortoise-tts
!pip install -U scipy
# Imports used through the rest of the notebook.
!pip install llvmlite==0.39.1
!pip install torchaudio==0.13.1
!pip install torchvision==0.14.1
!pip install torchtext==0.14.1
!pip install torchdata==0.5.1
!pip install transformers==4.19.0
!pip install tqdm
!pip install rotary_embedding_torch
!pip install tokenizers
!pip install inflect
!pip install progressbar
!pip install einops==0.4.1
!pip install unidecode
!pip install scipy==0.10.1
!pip install librosa==0.9.1
!pip install numba==0.48.0
!pip install ffmpeg
!pip install numpy==1.20.0
!pip install numba==0.48.0
!pip install threadpoolctl
!pip install appdirs
!python3 setup.py install
import torch
import torchaudio
import torch.nn as nn
import torch.nn.functional as F
import IPython
from tortoise.api import TextToSpeech
from tortoise.utils.audio import load_audio, load_voice, load_voices
# This will download all the models used by Tortoise from the HuggingFace hub.
print("Downloading models from HuggingFace")
tts = TextToSpeech()
# generate test
voice_list = ['tom', 'daniel',]
voice_samples, conditioning_latents = load_voices(voice_list)
preset = "fast" # for testing
gen = tts.tts_with_preset("They used to say that if man was meant to fly, he’d have wings. But he did fly. He discovered he had to.",
voice_samples=voice_samples, conditioning_latents=conditioning_latents,
preset=preset)
torchaudio.save('/tmp/captain_kirkard.wav', gen.squeeze(0).cpu(), 24000)
IPython.display.Audio('/tmp/captain_kirkard.wav')
c9951acThere was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That's really too bad. This breaks a whole workflow for me. I hope you or someone can help me lock the versions to get it working in colab again.
c9951acThere was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
FWIW, this sequence works in Google Colab. There are still issues with scipy, numpy and llvmlite. But generation works.