import tempfile
from typing import Optional
from TTS.config import load_config
import gradio as gr
import numpy as np
from TTS.utils.manage import ModelManager
from TTS.utils.synthesizer import Synthesizer
MODELS = {}
SPEAKERS = {}
MAX_TXT_LEN = 100
manager = ModelManager()
MODEL_NAMES = manager.list_tts_models()
# filter out multi-speaker models and slow wavegrad vocoders
filters = ["vctk", "your_tts", "ek1"]
MODEL_NAMES = [model_name for model_name in MODEL_NAMES if not any(f in model_name for f in filters)]
EN = [el for el in MODEL_NAMES if "/en/" in el]
OTHER = [el for el in MODEL_NAMES if "/en/" not in el]
EN[0], EN[5] = EN[5], EN[0]
MODEL_NAMES = EN + OTHER
# reorder models
print(MODEL_NAMES)
def tts(text: str, model_name: str):
if len(text) > MAX_TXT_LEN:
text = text[:MAX_TXT_LEN]
print(f"Input text was cutoff since it went over the {MAX_TXT_LEN} character limit.")
print(text, model_name)
# download model
model_path, config_path, model_item = manager.download_model(model_name)
vocoder_name: Optional[str] = model_item["default_vocoder"]
# download vocoder
vocoder_path = None
vocoder_config_path = None
if vocoder_name is not None:
vocoder_path, vocoder_config_path, _ = manager.download_model(vocoder_name)
# init synthesizer
synthesizer = Synthesizer(
model_path, config_path, None, None, vocoder_path, vocoder_config_path,
)
# synthesize
if synthesizer is None:
raise NameError("model not found")
wavs = synthesizer.tts(text, None)
# return output
with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as fp:
synthesizer.save_wav(wavs, fp)
return fp.name
title = """๐ธ๐ฌ CoquiTTS Playground
"""
with gr.Blocks(analytics_enabled=False) as demo:
with gr.Row():
with gr.Column():
gr.Markdown(
"""
## 
"""
)
gr.Markdown(
"""
## ๐ธCoqui.ai News
- ๐ฃ โTTS, our production TTS model that can speak 13 languages, is released [Blog Post](https://coqui.ai/blog/tts/open_xtts), [Demo](https://huggingface.co/spaces/coqui/xtts), [Docs](https://tts.readthedocs.io/en/dev/models/xtts.html)
- ๐ฃ [๐ถBark](https://github.com/suno-ai/bark) is now available for inference with unconstrained voice cloning. [Docs](https://tts.readthedocs.io/en/dev/models/bark.html)
- ๐ฃ You can use [~1100 Fairseq models](https://github.com/facebookresearch/fairseq/tree/main/examples/mms) with ๐ธTTS.
- ๐ฃ ๐ธTTS now supports ๐ขTortoise with faster inference. [Docs](https://tts.readthedocs.io/en/dev/models/tortoise.html)
- ๐ฃ **Coqui Studio API** is landed on ๐ธTTS. - [Example](https://github.com/coqui-ai/TTS/blob/dev/README.md#-python-api)
- ๐ฃ [**Coqui Studio API**](https://docs.coqui.ai/docs) is live.
- ๐ฃ Voice generation with prompts - **Prompt to Voice** - is live on [**Coqui Studio**](https://app.coqui.ai/auth/signin)!! - [Blog Post](https://coqui.ai/blog/tts/prompt-to-voice)
- ๐ฃ Voice generation with fusion - **Voice fusion** - is live on [**Coqui Studio**](https://app.coqui.ai/auth/signin).
- ๐ฃ Voice cloning is live on [**Coqui Studio**](https://app.coqui.ai/auth/signin).
"""
)
with gr.Column():
gr.Markdown(
"""
๐ป This space showcases some of the **[CoquiTTS](https://github.com/coqui-ai/TTS)** models.
There are > 30 languages with single and multi speaker models, all thanks to our ๐ Contributors.
Visit the links below for more.
| | |
| ------------------------------- | --------------------------------------- |
| ๐ธ๐ฌ **CoquiTTS** | [Github](https://github.com/coqui-ai/TTS) |
| ๐ผ **Documentation** | [ReadTheDocs](https://tts.readthedocs.io/en/latest/)
| ๐ฉโ๐ป **Questions** | [GitHub Discussions] |
| ๐ฏ **Community** | [](https://discord.gg/5eXr5seRrv) |
[github issue tracker]: https://github.com/coqui-ai/tts/issues
[github discussions]: https://github.com/coqui-ai/TTS/discussions
[discord]: https://discord.gg/5eXr5seRrv
"""
)
with gr.Row():
gr.Markdown(
"""
๐ Model contributors
- @nmstoker
- @kaiidams
- @WeberJulian,
- @Edresson
- @thorstenMueller
- @r-dh
- @kirianguiller
- @robinhad
- @fkarabiber
- @nicolalandro
- @a-froghyar
- @manmay-nakhashi
- @noml4u
"""
)
with gr.Row():
with gr.Column():
input_text = gr.inputs.Textbox(
label="Input Text",
default="This sentence has been generated by a speech synthesis system.",
)
model_select = gr.inputs.Dropdown(
label="Pick Model: tts_models///",
choices=MODEL_NAMES,
default="tts_models/en/jenny/jenny"
)
tts_button = gr.Button("Send", elem_id="send-btn", visible=True)
with gr.Column():
output_audio = gr.outputs.Audio(label="Output", type="filepath")
tts_button.click(
tts,
inputs=[
input_text,
model_select,
],
outputs=[output_audio],
)
demo.queue(concurrency_count=16).launch(debug=True)