File size: 1,530 Bytes
c852ee2
 
bcfd9f0
 
 
 
 
 
 
 
 
 
 
 
 
 
c852ee2
bcfd9f0
 
c852ee2
bcfd9f0
 
 
c852ee2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bcfd9f0
c852ee2
bcfd9f0
 
c852ee2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
from flask import Flask, redirect, url_for, request, send_from_directory
from flask import Flask, render_template, request, send_file
import gradio as gr
from synthesize import synthesize, load_model
from synthesis.vocoders import Hifigan
import sounddevice as sd
import soundfile as sf

model = load_model("checkpoints/checkpoint_9000.zip")
vocoder = Hifigan("weights/custom_pctest/model.pt", "weights/custom_pctest/config.json")

def inference(text: str):
    synthesize(
        model=model,
        text=text,
        graph_path="graph.png",
        audio_path="tts.wav",
        vocoder=vocoder,
    )
    return True

app = Flask(__name__)


@app.route("/")
def index():
    return render_template(
    "index.html",
    show_details=False,
    use_multi_speaker=False,
    speaker_ids=None,
    use_gst=False)

@app.route("/details")
def details():
    vocoder_config = None

    return render_template(
        "details.html",
        show_details=False,
        model_config=None,
        vocoder_config=None,
        args=None,
    )


# This format is similar to  coqui-ai TTS api
@app.route('/api/tts', methods = ['GET']) #?text=this%20is%20a%20tses&speaker_id=&style_wav=
def get_file():
    text = request.args.get('text')
    print(request.args)
    if not text:
        print('empty text')
        return {'message':'textempty'}
    else:
        inference(text)
        return send_from_directory('./', 'tts.wav', as_attachment=True)

if __name__ == '__main__':
   app.run(debug = True, host='0.0.0.0', port=5002)