KwamaEggStew/tortoise-tts
1
0
Fork 0
forked from mrq/tortoise-tts
Code Pull Requests Activity

Added tab to read and copy settings from a voice clip (in the future, I'll see about enmbedding the latent used to generate the voice)

Browse Source
This commit is contained in:
mrq 2023-02-06 16:00:44 +00:00
parent edb6a173d3
commit e25ec325fe
1 changed files with 134 additions and 80 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

208
app.py
View File

@ -130,7 +130,7 @@ def generate(text, delimiter, emotion, prompt, voice, mic_audio, preset, seed, c
info = { info = {
'text': text, 'text': text,
'delimiter': delimiter, 'delimiter': '\\n' if delimiter == "\n" else delimiter,
'emotion': emotion, 'emotion': emotion,
'prompt': prompt, 'prompt': prompt,
'voice': voice, 'voice': voice,
@ -185,89 +185,121 @@ def update_presets(value):
else: else:
return (gr.update(), gr.update()) return (gr.update(), gr.update())
def read_metadata(file):
j = None
if file is not None:
metadata = music_tag.load_file(file.name)
if 'lyrics' in metadata:
j = json.loads(str(metadata['lyrics']))
print(j)
return j
def copy_settings(file):
metadata = read_metadata(file)
if metadata is None:
return None
return (
metadata['text'],
metadata['delimiter'],
metadata['emotion'],
metadata['prompt'],
metadata['voice'],
metadata['mic_audio'],
metadata['preset'],
metadata['seed'],
metadata['candidates'],
metadata['num_autoregressive_samples'],
metadata['diffusion_iterations'],
metadata['temperature'],
metadata['diffusion_sampler'],
metadata['breathing_room'],
metadata['experimentals'],
)
def update_voices(): def update_voices():
return gr.Dropdown.update(choices=os.listdir(os.path.join("tortoise", "voices")) + ["microphone"]) return gr.Dropdown.update(choices=os.listdir(os.path.join("tortoise", "voices")) + ["microphone"])
def main(): def main():
with gr.Blocks() as demo: with gr.Blocks() as webui:
with gr.Row(): with gr.Tab("Generate"):
with gr.Column(): with gr.Row():
text = gr.Textbox(lines=4, label="Prompt") with gr.Column():
delimiter = gr.Textbox(lines=1, label="Line Delimiter", placeholder="\\n") text = gr.Textbox(lines=4, label="Prompt")
delimiter = gr.Textbox(lines=1, label="Line Delimiter", placeholder="\\n")
emotion = gr.Radio( emotion = gr.Radio(
["None", "Happy", "Sad", "Angry", "Disgusted", "Arrogant", "Custom"], ["None", "Happy", "Sad", "Angry", "Disgusted", "Arrogant", "Custom"],
value="None", value="None",
label="Emotion", label="Emotion",
type="value", type="value",
interactive=True interactive=True
) )
prompt = gr.Textbox(lines=1, label="Custom Emotion + Prompt (if selected)") prompt = gr.Textbox(lines=1, label="Custom Emotion + Prompt (if selected)")
voice = gr.Dropdown( voice = gr.Dropdown(
os.listdir(os.path.join("tortoise", "voices")) + ["microphone"], os.listdir(os.path.join("tortoise", "voices")) + ["microphone"],
label="Voice", label="Voice",
type="value", type="value",
) )
mic_audio = gr.Audio( mic_audio = gr.Audio(
label="Microphone Source", label="Microphone Source",
source="microphone", source="microphone",
type="filepath", type="filepath",
) )
refresh_voices = gr.Button(value="Refresh Voice List") refresh_voices = gr.Button(value="Refresh Voice List")
refresh_voices.click(update_voices, refresh_voices.click(update_voices,
inputs=None, inputs=None,
outputs=voice outputs=voice
) )
prompt.change(fn=lambda value: gr.update(value="Custom"), prompt.change(fn=lambda value: gr.update(value="Custom"),
inputs=prompt, inputs=prompt,
outputs=emotion outputs=emotion
) )
mic_audio.change(fn=lambda value: gr.update(value="microphone"), mic_audio.change(fn=lambda value: gr.update(value="microphone"),
inputs=mic_audio, inputs=mic_audio,
outputs=voice outputs=voice
) )
with gr.Column(): with gr.Column():
candidates = gr.Slider(value=1, minimum=1, maximum=6, step=1, label="Candidates") candidates = gr.Slider(value=1, minimum=1, maximum=6, step=1, label="Candidates")
seed = gr.Number(value=0, precision=0, label="Seed") seed = gr.Number(value=0, precision=0, label="Seed")
preset = gr.Radio( preset = gr.Radio(
["Ultra Fast", "Fast", "Standard", "High Quality", "None"], ["Ultra Fast", "Fast", "Standard", "High Quality", "None"],
value="None", value="None",
label="Preset", label="Preset",
type="value", type="value",
) )
num_autoregressive_samples = gr.Slider(value=128, minimum=0, maximum=512, step=1, label="Samples") num_autoregressive_samples = gr.Slider(value=128, minimum=0, maximum=512, step=1, label="Samples")
diffusion_iterations = gr.Slider(value=128, minimum=0, maximum=512, step=1, label="Iterations") diffusion_iterations = gr.Slider(value=128, minimum=0, maximum=512, step=1, label="Iterations")
temperature = gr.Slider(value=0.2, minimum=0, maximum=1, step=0.1, label="Temperature") temperature = gr.Slider(value=0.2, minimum=0, maximum=1, step=0.1, label="Temperature")
breathing_room = gr.Slider(value=12, minimum=1, maximum=32, step=1, label="Pause Size") breathing_room = gr.Slider(value=12, minimum=1, maximum=32, step=1, label="Pause Size")
diffusion_sampler = gr.Radio( diffusion_sampler = gr.Radio(
["P", "DDIM"], # + ["K_Euler_A", "DPM++2M"], ["P", "DDIM"], # + ["K_Euler_A", "DPM++2M"],
value="P", value="P",
label="Diffusion Samplers", label="Diffusion Samplers",
type="value", type="value",
) )
experimentals = gr.CheckboxGroup(["Half Precision", "Conditioning-Free"], value=["Conditioning-Free"], label="Experimental Flags") experimentals = gr.CheckboxGroup(["Half Precision", "Conditioning-Free"], value=["Conditioning-Free"], label="Experimental Flags")
preset.change(fn=update_presets, preset.change(fn=update_presets,
inputs=preset, inputs=preset,
outputs=[ outputs=[
num_autoregressive_samples, num_autoregressive_samples,
diffusion_iterations, diffusion_iterations,
], ],
) )
with gr.Column(): with gr.Column():
selected_voice = gr.Audio(label="Source Sample") selected_voice = gr.Audio(label="Source Sample")
output_audio = gr.Audio(label="Output") output_audio = gr.Audio(label="Output")
usedSeed = gr.Textbox(label="Seed", placeholder="0", interactive=False) usedSeed = gr.Textbox(label="Seed", placeholder="0", interactive=False)
submit = gr.Button(value="Generate") submit = gr.Button(value="Generate")
#stop = gr.Button(value="Stop") #stop = gr.Button(value="Stop")
submit_event = submit.click(generate, input_settings = [
inputs=[
text, text,
delimiter, delimiter,
emotion, emotion,
@ -283,13 +315,34 @@ def main():
diffusion_sampler, diffusion_sampler,
breathing_room, breathing_room,
experimentals, experimentals,
], ]
outputs=[selected_voice, output_audio, usedSeed],
)
#stop.click(fn=None, inputs=None, outputs=None, cancels=[submit_event]) submit_event = submit.click(generate,
inputs=input_settings,
outputs=[selected_voice, output_audio, usedSeed],
)
demo.queue().launch(share=args.share) #stop.click(fn=None, inputs=None, outputs=None, cancels=[submit_event])
with gr.Tab("Utilities"):
with gr.Row():
with gr.Column():
audio_in = gr.File(type="file", label="Audio Input", file_types=["audio"])
copy_button = gr.Button(value="Copy Settings")
with gr.Column():
metadata_out = gr.JSON(label="Audio Metadata")
audio_in.upload(
fn=read_metadata,
inputs=audio_in,
outputs=metadata_out,
)
copy_button.click(copy_settings,
inputs=audio_in, # JSON elements cannt be used as inputs
outputs=input_settings
)
webui.queue().launch(share=args.share)
if __name__ == "__main__": if __name__ == "__main__":
@ -299,6 +352,7 @@ if __name__ == "__main__":
parser.add_argument("--cond-latent-max-chunk-size", type=int, default=1000000, help="Sets an upper limit to audio chunk size when computing conditioning latents") parser.add_argument("--cond-latent-max-chunk-size", type=int, default=1000000, help="Sets an upper limit to audio chunk size when computing conditioning latents")
args = parser.parse_args() args = parser.parse_args()
print("Initializating TorToiSe...")
tts = TextToSpeech(minor_optimizations=not args.low_vram) tts = TextToSpeech(minor_optimizations=not args.low_vram)
main() main()