from TTS.api import TTS
import torch
from datetime import datetime
tts = TTS("tts_models/en/multi-dataset/tortoise-v2")
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
tts.to(device)
text="There is, therefore, an increasing need to understand BEVs from a systems perspective. This involves an in-depth consideration of the environmental impact of the product using life cycle assessment (LCA) as well as taking a broader 'circular economy' approach. On the one hand, LCA is a means of assessing the environmental impact associated with all stages of a product's life from cradle to grave: from raw material extraction and processing to the product's manufacture to its use in everyday life and finally to its end of life."


# cloning `lj` voice from `TTS/tts/utils/assets/tortoise/voices/lj`
# with custom inference settings overriding defaults.
time_now = datetime.now().strftime("%Y%m%d%H%M%S")
output_path = f"output/tortoise_{time_now}.wav"
tts.tts_to_file(text,
                file_path=output_path,
                voice_dir="voices",
                speaker="test",
                split_sentences=False, # Change to True if context is not enough
                num_autoregressive_samples=20,
                diffusion_iterations=50)

# # Using presets with the same voice
# tts.tts_to_file(text,
#                 file_path="output.wav",
#                 voice_dir="path/to/tortoise/voices/dir/",
#                 speaker="lj",
#                 preset="ultra_fast")

# # Random voice generation
# tts.tts_to_file(text,
#                 file_path="output.wav")