File size: 21,547 Bytes
a7d3800 a7867ab a7d3800 7a8f8e3 a7d3800 f2cd40c a7d3800 f2cd40c a7d3800 7a8f8e3 a7d3800 7a8f8e3 a7d3800 7a8f8e3 a7d3800 1c2c969 a7d3800 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 |
from original import *
import shutil, glob
from easyfuncs import download_from_url, CachedModels
os.makedirs("dataset",exist_ok=True)
model_library = CachedModels()
with gr.Blocks(title="RVC V2",theme="Blane187/fuchsia") as app:
with gr.Row():
gr.HTML("<img src='file/a.png' alt='image'>")
with gr.Tabs():
with gr.TabItem("Inference"):
with gr.Row():
voice_model = gr.Dropdown(label="Model Voice", choices=sorted(names), value=lambda:sorted(names)[0] if len(sorted(names)) > 0 else '', interactive=True)
file_index2 = gr.Dropdown(label="Change Index",choices=sorted(index_paths), interactive=True,value=sorted(index_paths)[0] if len(sorted(index_paths)) > 0 else '')
with gr.Row():
refresh_button = gr.Button("Refresh", variant="primary")
spk_item = gr.Slider(
minimum=0,
maximum=2333,
step=1,
label="Speaker ID",
value=0,
visible=False,
interactive=False,
)
vc_transform0 = gr.Number(
label="Pitch",
value=0
)
with gr.Row():
with gr.Column():
with gr.Row():
dropbox = gr.File(label="Drop your audio here & hit the Reload button.")
with gr.Row():
record_button=gr.Audio(source="microphone", label="OR Record audio.", type="filepath")
with gr.Row():
paths_for_files = lambda path:[os.path.abspath(os.path.join(path, f)) for f in os.listdir(path) if os.path.splitext(f)[1].lower() in ('.mp3', '.wav', '.flac', '.ogg')]
input_audio0 = gr.Dropdown(
label="Input Path",
value=paths_for_files('audios')[0] if len(paths_for_files('audios')) > 0 else '',
choices=paths_for_files('audios'), # Only show absolute paths for audio files ending in .mp3, .wav, .flac or .ogg
allow_custom_value=True
)
with gr.Row():
audio_player = gr.Audio()
input_audio0.change(
inputs=[input_audio0],
outputs=[audio_player],
fn=lambda path: {"value":path,"__type__":"update"} if os.path.exists(path) else None
)
record_button.stop_recording(
fn=lambda audio:audio, #TODO save wav lambda
inputs=[record_button],
outputs=[input_audio0])
dropbox.upload(
fn=lambda audio:audio.name,
inputs=[dropbox],
outputs=[input_audio0])
with gr.Column():
with gr.Accordion("General Settings", open=False):
f0method0 = gr.Radio(
label="Method",
choices=["pm", "harvest", "crepe", "rmvpe"]
if config.dml == False
else ["pm", "harvest", "rmvpe"],
value="rmvpe",
interactive=True,
)
filter_radius0 = gr.Slider(
minimum=0,
maximum=7,
label="Breathiness Reduction (Harvest only)",
value=3,
step=1,
interactive=True,
)
resample_sr0 = gr.Slider(
minimum=0,
maximum=48000,
label="Resample",
value=0,
step=1,
interactive=True,
visible=False
)
rms_mix_rate0 = gr.Slider(
minimum=0,
maximum=1,
label="Volume Normalization",
value=0,
interactive=True,
)
protect0 = gr.Slider(
minimum=0,
maximum=0.5,
label="Breathiness Protection (0 is enabled, 0.5 is disabled)",
value=0.33,
step=0.01,
interactive=True,
)
if voice_model != None: vc.get_vc(voice_model.value,protect0,protect0)
file_index1 = gr.Textbox(
label="Index Path",
interactive=True,
visible=False#Not used here
)
refresh_button.click(
fn=change_choices,
inputs=[],
outputs=[voice_model, file_index2],
api_name="infer_refresh",
)
refresh_button.click(
fn=lambda:{"choices":paths_for_files('audios'),"__type__":"update"}, #TODO check if properly returns a sorted list of audio files in the 'audios' folder that have the extensions '.wav', '.mp3', '.ogg', or '.flac'
inputs=[],
outputs = [input_audio0],
)
refresh_button.click(
fn=lambda:{"value":paths_for_files('audios')[0],"__type__":"update"} if len(paths_for_files('audios')) > 0 else {"value":"","__type__":"update"}, #TODO check if properly returns a sorted list of audio files in the 'audios' folder that have the extensions '.wav', '.mp3', '.ogg', or '.flac'
inputs=[],
outputs = [input_audio0],
)
with gr.Accordion("Change Index", open=False):
index_rate1 = gr.Slider(
minimum=0,
maximum=1,
label="Index Strength",
value=0.5,
interactive=True,
)
with gr.Row():
f0_file = gr.File(label="F0 Path", visible=False)
with gr.Row():
vc_output2 = gr.Audio(label="Output", scale=5)
with gr.Row():
vc_output1 = gr.Textbox(label="Information")
with gr.Row():
but0 = gr.Button(value="Convert", variant="primary")
but0.click(
vc.vc_single,
[
spk_item,
input_audio0,
vc_transform0,
f0_file,
f0method0,
file_index1,
file_index2,
index_rate1,
filter_radius0,
resample_sr0,
rms_mix_rate0,
protect0,
],
[vc_output1, vc_output2],
api_name="infer_convert",
)
voice_model.change(
fn=vc.get_vc,
inputs=[voice_model, protect0, protect0],
outputs=[spk_item, protect0, protect0, file_index2, file_index2],
api_name="infer_change_voice",
)
with gr.TabItem("Download Models"):
with gr.Row():
url_input = gr.Textbox(label="URL to model", value="",placeholder="https://...", scale=6)
name_output = gr.Textbox(label="Save as", value="",placeholder="MyModel",scale=2)
url_download = gr.Button(value="Download Model",scale=2)
url_download.click(
inputs=[url_input,name_output],
outputs=[url_input],
fn=download_from_url,
)
with gr.Row():
model_browser = gr.Dropdown(choices=list(model_library.models.keys()),label="OR Search Models (Quality UNKNOWN)",scale=5)
with gr.Row():
download_from_browser = gr.Button(value="Get",scale=2)
download_from_browser.click(
inputs=[model_browser],
outputs=[model_browser],
fn=lambda model: download_from_url(model_library.models[model],model),
)
with gr.TabItem("Train"):
with gr.Row():
with gr.Column():
training_name = gr.Textbox(label="Name your model", value="My-Voice",placeholder="My-Voice")
np7 = gr.Slider(
minimum=0,
maximum=config.n_cpu,
step=1,
label="Number of CPU processes used to extract pitch features",
value=int(np.ceil(config.n_cpu / 1.5)),
interactive=True,
)
sr2 = gr.Radio(
label="Sampling Rate",
choices=["40k", "32k"],
value="32k",
interactive=True,
)
if_f0_3 = gr.Radio(
label="Will your model be used for singing? If not, you can ignore this.",
choices=[True, False],
value=True,
interactive=True,
visible=False
)
version19 = gr.Radio(
label="Version",
choices=["v1", "v2"],
value="v2",
interactive=True,
visible=False,
)
dataset_folder = gr.Textbox(
label="dataset folder", value='dataset'
)
easy_uploader = gr.Files(label="Drop your audio files here",file_types=['audio'])
with gr.Accordion(label="button if you don't set your training settings", open=False):
but1 = gr.Button("1. Process", variant="primary")
but2 = gr.Button("2. Extract Features", variant="primary")
but4 = gr.Button("3. Train Index", variant="primary")
but3 = gr.Button("4. Train Model", variant="primary")
Information_box = gr.Textbox(label="Information", value="",visible=True)
easy_uploader.upload(inputs=[dataset_folder],outputs=[],fn=lambda folder:os.makedirs(folder,exist_ok=True))
easy_uploader.upload(
fn=lambda files,folder: [shutil.copy2(f.name,os.path.join(folder,os.path.split(f.name)[1])) for f in files] if folder != "" else gr.Warning('Please enter a folder name for your dataset'),
inputs=[easy_uploader, dataset_folder],
outputs=[])
gpus6 = gr.Textbox(
label="Enter the GPU numbers to use separated by -, (e.g. 0-1-2)",
value=gpus,
interactive=True,
visible=F0GPUVisible,
)
gpu_info9 = gr.Textbox(
label="GPU Info", value=gpu_info, visible=F0GPUVisible
)
spk_id5 = gr.Slider(
minimum=0,
maximum=4,
step=1,
label="Speaker ID",
value=0,
interactive=True,
visible=False
)
but1.click(
preprocess_dataset,
[dataset_folder, training_name, sr2, np7],
[Information_box],
api_name="train_preprocess",
)
with gr.Column():
f0method8 = gr.Radio(
label="F0 extraction method",
choices=["pm", "harvest", "dio", "rmvpe", "rmvpe_gpu"],
value="rmvpe_gpu",
interactive=True,
)
gpus_rmvpe = gr.Textbox(
label="GPU numbers to use separated by -, (e.g. 0-1-2)",
value="%s-%s" % (gpus, gpus),
interactive=True,
visible=F0GPUVisible,
)
f0method8.change(
fn=change_f0_method,
inputs=[f0method8],
outputs=[gpus_rmvpe],
)
but2.click(
extract_f0_feature,
[
gpus6,
np7,
f0method8,
if_f0_3,
training_name,
version19,
gpus_rmvpe,
],
[Information_box],
api_name="train_extract_f0_feature",
)
with gr.Column():
total_epoch11 = gr.Slider(
minimum=2,
maximum=1000,
step=1,
label="Epochs (more epochs may improve quality but takes longer)",
value=150,
interactive=True,
)
with gr.Accordion(label="General Settings", open=False):
gpus16 = gr.Textbox(
label="GPUs separated by -, (e.g. 0-1-2)",
value="0",
interactive=True,
visible=True
)
save_epoch10 = gr.Slider(
minimum=1,
maximum=50,
step=1,
label="Weight Saving Frequency",
value=25,
interactive=True,
)
batch_size12 = gr.Slider(
minimum=1,
maximum=40,
step=1,
label="Batch Size",
value=default_batch_size,
interactive=True,
)
if_save_latest13 = gr.Radio(
label="Only save the latest model",
choices=["yes", "no"],
value="yes",
interactive=True,
visible=False
)
if_cache_gpu17 = gr.Radio(
label="If your dataset is UNDER 10 minutes, cache it to train faster",
choices=["yes", "no"],
value="no",
interactive=True,
)
if_save_every_weights18 = gr.Radio(
label="Save small model at every save point",
choices=["yes", "no"],
value="yes",
interactive=True,
)
with gr.Accordion(label="Change pretrains", open=False):
pretrained = lambda sr, letter: [os.path.abspath(os.path.join('assets/pretrained_v2', file)) for file in os.listdir('assets/pretrained_v2') if file.endswith('.pth') and sr in file and letter in file]
pretrained_G14 = gr.Dropdown(
label="pretrained G",
# Get a list of all pretrained G model files in assets/pretrained_v2 that end with .pth
choices = pretrained(sr2.value, 'G'),
value=pretrained(sr2.value, 'G')[0] if len(pretrained(sr2.value, 'G')) > 0 else '',
interactive=True,
visible=True
)
pretrained_D15 = gr.Dropdown(
label="pretrained D",
choices = pretrained(sr2.value, 'D'),
value= pretrained(sr2.value, 'D')[0] if len(pretrained(sr2.value, 'G')) > 0 else '',
visible=True,
interactive=True
)
with gr.Row():
download_model = gr.Button('5.Download Model')
with gr.Row():
model_files = gr.Files(label='Your Model and Index file can be downloaded here:')
download_model.click(
fn=lambda name: os.listdir(f'assets/weights/{name}') + glob.glob(f'logs/{name.split(".")[0]}/added_*.index'),
inputs=[training_name],
outputs=[model_files, Information_box])
with gr.Row():
sr2.change(
change_sr2,
[sr2, if_f0_3, version19],
[pretrained_G14, pretrained_D15],
)
version19.change(
change_version19,
[sr2, if_f0_3, version19],
[pretrained_G14, pretrained_D15, sr2],
)
if_f0_3.change(
change_f0,
[if_f0_3, sr2, version19],
[f0method8, pretrained_G14, pretrained_D15],
)
with gr.Row():
but5 = gr.Button("1 Click Training", variant="primary", visible=False)
but3.click(
click_train,
[
training_name,
sr2,
if_f0_3,
spk_id5,
save_epoch10,
total_epoch11,
batch_size12,
if_save_latest13,
pretrained_G14,
pretrained_D15,
gpus16,
if_cache_gpu17,
if_save_every_weights18,
version19,
],
Information_box,
api_name="train_start",
)
but4.click(train_index, [training_name, version19], Information_box)
but5.click(
train1key,
[
training_name,
sr2,
if_f0_3,
dataset_folder,
spk_id5,
np7,
f0method8,
save_epoch10,
total_epoch11,
batch_size12,
if_save_latest13,
pretrained_G14,
pretrained_D15,
gpus16,
if_cache_gpu17,
if_save_every_weights18,
version19,
gpus_rmvpe,
],
Information_box,
api_name="train_start_all",
)
if config.iscolab:
app.queue(concurrency_count=511, max_size=1022).launch(share=True)
else:
app.queue(concurrency_count=511, max_size=1022).launch(
server_name="0.0.0.0",
inbrowser=not config.noautoopen,
server_port=config.listen_port,
quiet=True,
) |