File size: 11,344 Bytes
e71a85c
 
 
 
 
 
 
 
 
 
 
 
 
c80814f
e71a85c
 
 
21d0d32
 
 
e71a85c
6666073
e71a85c
 
 
 
 
 
 
 
21d0d32
e71a85c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c80814f
e71a85c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
37d9c3a
 
 
 
 
 
 
 
 
 
 
 
 
 
e71a85c
 
 
 
 
 
 
37d9c3a
4ed0876
 
 
 
e71a85c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bd8aed8
 
 
e71a85c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6666073
848e5d3
c80814f
e71a85c
c80814f
 
01ea85b
aa4679f
3e65c2d
aa4679f
e71a85c
 
 
 
 
aa4679f
01ea85b
 
848e5d3
e71a85c
 
 
 
 
aa4679f
2cee98f
 
 
e71a85c
aa4679f
 
 
 
 
 
 
e71a85c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c80814f
 
e71a85c
 
26ceaf8
e71a85c
 
 
c80814f
e71a85c
 
 
 
 
 
 
0fa03a6
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
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
import gradio as gr
import requests
import random
import os
import zipfile 
import librosa
import time
from infer_rvc_python import BaseLoader
from pydub import AudioSegment
from tts_voice import tts_order_voice
import edge_tts
import tempfile
from audio_separator.separator import Separator
from animalesepy import *
import model_handler
import psutil
import cpuinfo
import re
import numpy as np
import wave
language_dict = tts_order_voice
from blaise import *

try:
    import spaces
    spaces_status = True
except ImportError:
    spaces_status = False

separator = Separator()
converter = BaseLoader(only_cpu=True, hubert_path=None, rmvpe_path=None)

global pth_file
global index_file

pth_file = "model.pth"
index_file = "model.index"

#CONFIGS
TEMP_DIR = "temp"
MODEL_PREFIX = "model"
PITCH_ALGO_OPT = [
    "pm",
    "harvest",
    "crepe",
    "rmvpe",
    "rmvpe+",
]

MODELS = [
    {"model": "model.pth", "index": "model.index", "model_name": "Test Model"},
]

os.makedirs(TEMP_DIR, exist_ok=True)

def unzip_file(file):
    filename = os.path.basename(file).split(".")[0] 
    with zipfile.ZipFile(file, 'r') as zip_ref:
        zip_ref.extractall(os.path.join(TEMP_DIR, filename)) 
    return True
    

def progress_bar(total, current):
    return "[" + "=" * int(current / total * 20) + ">" + " " * (20 - int(current / total * 20)) + "] " + str(int(current / total * 100)) + "%"

def contains_bad_word(text, bad_words):
    text_lower = text.lower()
    for word in bad_words:
        if word.lower() in text_lower:
            return True
    return False

bad_words = ['puttana', 'whore', 'badword3', 'badword4']

class BadWordError(Exception):
    def __init__(self, msg):
        super().__init__(msg)
        self.word = word

def download_from_url(url, name=None):
    if name is None:
        raise ValueError("The model name must be provided")
    if "/blob/" in url:
        url = url.replace("/blob/", "/resolve/") 
    if "huggingface" not in url:
        return ["The URL must be from huggingface", "Failed", "Failed"]
    if contains_bad_word(url, bad_words):
        return BadWordError("The file url has a bad word.")
    if contains_bad_word(name, bad_words):
        return BadWordError("The file name has a bad word.")
    filename = os.path.join(TEMP_DIR, MODEL_PREFIX + str(random.randint(1, 1000)) + ".zip")
    response = requests.get(url)
    total = int(response.headers.get('content-length', 0)) 
    if total > 500000000:

        return ["The file is too large. You can only download files up to 500 MB in size.", "Failed", "Failed"]
    current = 0
    with open(filename, "wb") as f:
        for data in response.iter_content(chunk_size=4096): 
            f.write(data)
            current += len(data)
            print(progress_bar(total, current), end="\r") #
    
    

    try:
        unzip_file(filename)
    except Exception as e:
        return ["Failed to unzip the file", "Failed", "Failed"] 
    unzipped_dir = os.path.join(TEMP_DIR, os.path.basename(filename).split(".")[0])
    pth_files = []
    index_files = []
    for root, dirs, files in os.walk(unzipped_dir): 
        for file in files:
            if file.endswith(".pth"):
                pth_files.append(os.path.join(root, file))
            elif file.endswith(".index"):
                index_files.append(os.path.join(root, file))
    
    print(pth_files, index_files) 
    global pth_file
    global index_file
    pth_file = pth_files[0]
    index_file = index_files[0]

    print(pth_file)
    print(index_file)

    if name == "":
        name = pth_file.split(".")[0]

    MODELS.append({"model": pth_file, "index": index_file, "model_name": name})
    return ["Downloaded as " + name, pth_files[0], index_files[0]]

def inference(audio, model_name):
        output_data = inf_handler(audio, model_name)
        vocals = output_data[0]
        inst = output_data[1]

        return vocals, inst

if spaces_status:
    @spaces.GPU()
    def convert_now(audio_files, random_tag, converter):
        return converter(
            audio_files,
            random_tag,
            overwrite=False,
            parallel_workers=8
        )

        
else:
    def convert_now(audio_files, random_tag, converter):
        return converter(
            audio_files,
            random_tag,
            overwrite=False,
            parallel_workers=8
        )


    
def run(
    model,
    audio_files,
    pitch_alg,
    pitch_lvl,
    index_inf,
    r_m_f,
    e_r,
    c_b_p,
):
    if not audio_files:
        raise ValueError("The audio pls")

    if isinstance(audio_files, str):
        audio_files = [audio_files]

    try:
        duration_base = librosa.get_duration(filename=audio_files[0])
        print("Duration:", duration_base)
    except Exception as e:
        print(e)

    random_tag = "USER_"+str(random.randint(10000000, 99999999))

    file_m = model
    print("File model:", file_m)

    # get from MODELS
    for model in MODELS:
        if model["model_name"] == file_m:
            print(model)
            file_m = model["model"]
            file_index = model["index"]
            break
    
    if not file_m.endswith(".pth"):
        raise ValueError("The model file must be a .pth file")


    print("Random tag:", random_tag)
    print("File model:", file_m)
    print("Pitch algorithm:", pitch_alg)
    print("Pitch level:", pitch_lvl)
    print("File index:", file_index)
    print("Index influence:", index_inf)
    print("Respiration median filtering:", r_m_f)
    print("Envelope ratio:", e_r)

    converter.apply_conf(
        tag=random_tag,
        file_model=file_m,
        pitch_algo=pitch_alg,
        pitch_lvl=pitch_lvl,
        file_index=file_index,
        index_influence=index_inf,
        respiration_median_filtering=r_m_f,
        envelope_ratio=e_r,
        consonant_breath_protection=c_b_p,
        resample_sr=44100 if audio_files[0].endswith('.mp3') else 0, 
    )
    time.sleep(0.1)

    result = convert_now(audio_files, random_tag, converter)
    print("Result:", result)

    return result[0]

def upload_model(index_file, pth_file, model_name):
    pth_file = pth_file.name
    index_file = index_file.name
    MODELS.append({"model": pth_file, "index": index_file, "model_name": model_name})
    return "Uploaded!"  

with gr.Blocks(theme=secret, title="Animalese RVC 🔶") as app:
    gr.Markdown("## Animalese RVC 🔶")
    gr.Markdown("**this project is forked of Ilaria RVC!**")
    with gr.Tab("Inference"):
        text_input = gr.Textbox(label="Input Text", placeholder="Enter text to convert to Animalese")
        shorten_input = gr.Checkbox(label="Shorten Words")
        pitch_input = gr.Slider(minimum=0.2, maximum=2.0, step=0.1, value=1.0, label="Pitch", visible=False)
        
        sound_gui = gr.Audio(type="filepath",autoplay=False,visible=True)
        
        def update():
            print(MODELS)
            return gr.Dropdown(label="Model",choices=[model["model_name"] for model in MODELS],visible=True,interactive=True, value=MODELS[0]["model_name"],)
        with gr.Row():
            models_dropdown = gr.Dropdown(label="Model",choices=[model["model_name"] for model in MODELS],visible=True,interactive=True, value=MODELS[0]["model_name"],)
            
        pitch_lvl_conf = gr.Slider(label="Pitch level (lower -> 'male' while higher -> 'female')",minimum=-12,maximum=12,step=1,value=0,visible=True,interactive=True,)
            
        with gr.Accordion("Settings", open=False, visible=False):
            pitch_algo_conf = gr.Dropdown(PITCH_ALGO_OPT,value=PITCH_ALGO_OPT[4],label="Pitch algorithm",visible=True,interactive=True,)
            index_inf_conf =  gr.Slider(minimum=0,maximum=1,label="Index influence -> How much accent is applied",value=0.75,)
            respiration_filter_conf = gr.Slider(minimum=0,maximum=7,label="Respiration median filtering",value=3,step=1,interactive=True,)
            envelope_ratio_conf = gr.Slider(minimum=0,maximum=1,label="Envelope ratio",value=0.25,interactive=True,)
            consonant_protec_conf = gr.Slider(minimum=0,maximum=0.5,label="Consonant breath protection",value=0.5,interactive=True,)
        with gr.Group():
            refresh_button = gr.Button("1. Refresh Models")
            preview_button = gr.Button("2. Preview!")
            button_conf = gr.Button("3. Convert",variant="primary",)
        output_conf = gr.Audio(type="filepath",label="Output",)

        
        refresh_button.click(update, outputs=[models_dropdown])

        preview_button.click(fn=lambda text, shorten, pitch: preview_audio(generate_audio(text, shorten, pitch)),
                         inputs=[text_input, shorten_input, pitch_input],
                         outputs=sound_gui)
        button_conf.click(lambda :None, None, output_conf)
        button_conf.click(
            run,
            inputs=[
                models_dropdown,
                sound_gui,
                pitch_algo_conf,
                pitch_lvl_conf,
                index_inf_conf,
                respiration_filter_conf,
                envelope_ratio_conf,
                consonant_protec_conf,
            ],
            outputs=[output_conf],
        )


    with gr.Tab("Model Loader (Download and Upload)"):
        with gr.Accordion("Model Downloader", open=False):
            gr.Markdown(
                "Download the model from the following URL and upload it here. (Huggingface RVC model)"
            )
            model = gr.Textbox(lines=1, label="Model URL")
            name = gr.Textbox(lines=1, label="Model Name", placeholder="Model Name")
            download_button = gr.Button("Download Model")
            status = gr.Textbox(lines=1, label="Status", placeholder="Waiting....", interactive=False)
            model_pth = gr.Textbox(lines=1, label="Model pth file", placeholder="Waiting....", interactive=False)
            index_pth = gr.Textbox(lines=1, label="Index pth file", placeholder="Waiting....", interactive=False)
            download_button.click(download_from_url, [model, name], outputs=[status, model_pth, index_pth])
        with gr.Accordion("Upload A Model", open=False):
            index_file_upload = gr.File(label="Index File (.index)")
            pth_file_upload = gr.File(label="Model File (.pth)")

            model_name = gr.Textbox(label="Model Name", placeholder="Model Name")
            upload_button = gr.Button("Upload Model")
            upload_status = gr.Textbox(lines=1, label="Status", placeholder="Waiting....", interactive=False)

            upload_button.click(upload_model, [index_file_upload, pth_file_upload, model_name], upload_status)
    

    with gr.Tab("Credits"):
        gr.Markdown(
            """
            Animalese RVC made by [Blane187](https://huggingface.co/Blane187)
           
            Ilaria RVC made by [Ilaria](https://huggingface.co/TheStinger) suport her on [ko-fi](https://ko-fi.com/ilariaowo)
            
            The modules made by [r3gm](https://huggingface.co/r3gm) 

            made with ❤️ by [mikus](https://github.com/cappuch) - made the ui!

            
            """
        )
    with gr.Tab(("")):
        gr.Markdown('''
            ![ilaria](https://i.ytimg.com/vi/5PWqt2Wg-us/maxresdefault.jpg)
        ''')

app.queue(api_open=False).launch(show_api=False)