Hugging Face's logo

voices
/
VCTK_British_English_Females

Text-to-Speech
Transformers
English
tts
audio
text to speech
coqui
coqui-tts
Inference Endpoints
Model card Files Community
VCTK_British_English_Females
File size: 4,517 Bytes 
6585abb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c3113ee
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6585abb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c3113ee
 
 
 
6585abb
c3113ee
 
 
 
 
6585abb
c3113ee
 
 
 
 
6585abb
c3113ee
 
6585abb
c3113ee
 
 
 
 
 
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
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
 
import json
import os
import subprocess

def generate_html_output(data, repository_path):
    with open('speakers.md', 'a') as file:
        for speaker_id, speaker_info in data.items():
            out_path = f"{repository_path}/samples/{speaker_id}.wav"
            age = speaker_info['age']
            gender = speaker_info['gender']
            if gender == 'F':
                gender = 'female'
            elif gender == "M":
                gender = 'male'
            accents = speaker_info['accents']
            region = speaker_info['region']

            file.write(f"<p>VCTK_{speaker_id}: {age} year old {gender}, {accents} accent ({region})<audio controls><source src=\"{out_path}\" type=\"audio/wav\"></audio> </p>\n")


# Load the data from the provided dictionary
data = {
    "VCTK_p225": {"age": 23, "gender": "F", "accents": "English", "region": "Southern England", "comments": ""},
    "VCTK_p228": {"age": 22, "gender": "F", "accents": "English", "region": "Southern England", "comments": ""},
    "VCTK_p229": {"age": 23, "gender": "F", "accents": "English", "region": "Southern England", "comments": ""},
    "VCTK_p230": {"age": 22, "gender": "F", "accents": "English", "region": "Stockton-on-tees", "comments": ""},
    "VCTK_p231": {"age": 23, "gender": "F", "accents": "English", "region": "Southern England", "comments": ""},
    "VCTK_p233": {"age": 23, "gender": "F", "accents": "English", "region": "Staffordshire", "comments": ""},
    "VCTK_p236": {"age": 23, "gender": "F", "accents": "English", "region": "Manchester", "comments": ""},
    "VCTK_p239": {"age": 22, "gender": "F", "accents": "English", "region": "Southwest England", "comments": ""},
    "VCTK_p240": {"age": 21, "gender": "F", "accents": "English", "region": "Southern England", "comments": ""},
    "VCTK_p244": {"age": 22, "gender": "F", "accents": "English", "region": "Manchester", "comments": ""},
    "VCTK_p250": {"age": 22, "gender": "F", "accents": "English", "region": "Southeast England", "comments": ""},
    "VCTK_p257": {"age": 24, "gender": "F", "accents": "English", "region": "Southern England", "comments": ""},
    "VCTK_p267": {"age": 23, "gender": "F", "accents": "English", "region": "Yorkshire", "comments": ""},
    "VCTK_p268": {"age": 23, "gender": "F", "accents": "English", "region": "Southern England", "comments": ""},
    "VCTK_p269": {"age": 20, "gender": "F", "accents": "English", "region": "Newcastle", "comments": ""},
    "VCTK_p276": {"age": 24, "gender": "F", "accents": "English", "region": "Oxford", "comments": ""},
    "VCTK_p277": {"age": 23, "gender": "F", "accents": "English", "region": "Northeast England", "comments": ""},
    "VCTK_p282": {"age": 23, "gender": "F", "accents": "English", "region": "Newcastle", "comments": ""}
}

# Convert the data to JSON format
json_data = json.dumps(data, indent=2)

# Save the JSON data to a file
with open('speakers-log.json', 'w') as file:
    file.write(json_data)

# Run the TTS command to get the speaker indices
command = "tts --model_path checkpoint_85000.pth --config_path config.json --list_speaker_idxs | grep -vE '^(\s*\||\s*>|\s*$)'"
output = subprocess.check_output(command, shell=True, text=True)

# Parse the JSON output into a Python dictionary
speaker_indices = eval(output)

# Load the speaker IDs from speakers.json
with open('speakers-log.json', 'r') as file:
    speaker_ids = json.load(file)


for speaker_idx in speaker_indices:
    # # Remove the 'VCTK_' prefix
    speaker_id = speaker_idx
    # speaker_id = speaker_idx.replace('VCTK_', '')

    # Lookup the speaker ID in the loaded speaker IDs
    if speaker_id in speaker_ids:
        speaker_id_json = speaker_ids[speaker_id]
    else:
        continue

    # # Generate the TTS command to create the audio file
    # text = f"Hello, I am from {speaker_id_json['region']}. I hope that you will select my voice for your project. Thank you."
    # # make samples directory if it doesn't exist
    # if not os.path.exists("samples"):
    #     os.makedirs("samples")

    # out_path = f"samples/{speaker_id}.wav"
    # tts_command = f"tts --text \"{text}\" --model_path checkpoint_85000.pth --language_idx en --config_path config.json --speaker_idx \"VCTK_{speaker_id}\" --out_path {out_path}"

    # Execute the TTS command
    # os.system(tts_command)

    # Write the speaker information to the speakers.md file
    generate_html_output({speaker_id: speaker_id_json}, "https://huggingface.co/voices/VCTK_European_English_Females/resolve/main")