File size: 27,315 Bytes
85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 85d3b29 1a7d583 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 |
import os
import subprocess
import sys
import shutil
import gradio as gr
from assets.i18n.i18n import I18nAuto
from core import (
run_preprocess_script,
run_extract_script,
run_train_script,
run_index_script,
run_prerequisites_script,
)
from rvc.configs.config import max_vram_gpu, get_gpu_info
from rvc.lib.utils import format_title
from tabs.settings.restart import restart_applio
i18n = I18nAuto()
now_dir = os.getcwd()
sys.path.append(now_dir)
pretraineds_v1 = [
(
"pretrained_v1/",
[
"D32k.pth",
"D40k.pth",
"D48k.pth",
"G32k.pth",
"G40k.pth",
"G48k.pth",
"f0D32k.pth",
"f0D40k.pth",
"f0D48k.pth",
"f0G32k.pth",
"f0G40k.pth",
"f0G48k.pth",
],
),
]
folder_mapping = {
"pretrained_v1/": "rvc/pretraineds/pretrained_v1/",
}
sup_audioext = {
"wav",
"mp3",
"flac",
"ogg",
"opus",
"m4a",
"mp4",
"aac",
"alac",
"wma",
"aiff",
"webm",
"ac3",
}
# Custom Pretraineds
pretraineds_custom_path = os.path.join(
now_dir, "rvc", "pretraineds", "pretraineds_custom"
)
pretraineds_custom_path_relative = os.path.relpath(pretraineds_custom_path, now_dir)
if not os.path.exists(pretraineds_custom_path_relative):
os.makedirs(pretraineds_custom_path_relative)
def get_pretrained_list(suffix):
return [
os.path.join(dirpath, filename)
for dirpath, _, filenames in os.walk(pretraineds_custom_path_relative)
for filename in filenames
if filename.endswith(".pth") and suffix in filename
]
pretraineds_list_d = get_pretrained_list("D")
pretraineds_list_g = get_pretrained_list("G")
def refresh_custom_pretraineds():
return (
{"choices": sorted(get_pretrained_list("G")), "__type__": "update"},
{"choices": sorted(get_pretrained_list("D")), "__type__": "update"},
)
# Dataset Creator
datasets_path = os.path.join(now_dir, "assets", "datasets")
if not os.path.exists(datasets_path):
os.makedirs(datasets_path)
datasets_path_relative = os.path.relpath(datasets_path, now_dir)
def get_datasets_list():
return [
dirpath
for dirpath, _, filenames in os.walk(datasets_path_relative)
if any(filename.endswith(tuple(sup_audioext)) for filename in filenames)
]
def refresh_datasets():
return {"choices": sorted(get_datasets_list()), "__type__": "update"}
# Model Names
models_path = os.path.join(now_dir, "logs")
def get_models_list():
return [
os.path.basename(dirpath)
for dirpath in os.listdir(models_path)
if os.path.isdir(os.path.join(models_path, dirpath))
and all(excluded not in dirpath for excluded in ["zips", "mute"])
]
def refresh_models():
return {"choices": sorted(get_models_list()), "__type__": "update"}
# Refresh Models and Datasets
def refresh_models_and_datasets():
return (
{"choices": sorted(get_models_list()), "__type__": "update"},
{"choices": sorted(get_datasets_list()), "__type__": "update"},
)
# Drop Model
def save_drop_model(dropbox):
if ".pth" not in dropbox:
gr.Info(
i18n(
"The file you dropped is not a valid pretrained file. Please try again."
)
)
else:
file_name = os.path.basename(dropbox)
pretrained_path = os.path.join(pretraineds_custom_path_relative, file_name)
if os.path.exists(pretrained_path):
os.remove(pretrained_path)
os.rename(dropbox, pretrained_path)
gr.Info(
i18n(
"Click the refresh button to see the pretrained file in the dropdown menu."
)
)
return None
# Drop Dataset
def save_drop_dataset_audio(dropbox, dataset_name):
if not dataset_name:
gr.Info("Please enter a valid dataset name. Please try again.")
return None, None
else:
file_extension = os.path.splitext(dropbox)[1][1:].lower()
if file_extension not in sup_audioext:
gr.Info("The file you dropped is not a valid audio file. Please try again.")
else:
dataset_name = format_title(dataset_name)
audio_file = format_title(os.path.basename(dropbox))
dataset_path = os.path.join(now_dir, "assets", "datasets", dataset_name)
if not os.path.exists(dataset_path):
os.makedirs(dataset_path)
destination_path = os.path.join(dataset_path, audio_file)
if os.path.exists(destination_path):
os.remove(destination_path)
os.rename(dropbox, destination_path)
gr.Info(
i18n(
"The audio file has been successfully added to the dataset. Please click the preprocess button."
)
)
dataset_path = os.path.dirname(destination_path)
relative_dataset_path = os.path.relpath(dataset_path, now_dir)
return None, relative_dataset_path
# Export
## Get Pth and Index Files
def get_pth_list():
return [
os.path.relpath(os.path.join(dirpath, filename), now_dir)
for dirpath, _, filenames in os.walk(models_path)
for filename in filenames
if filename.endswith(".pth")
]
def get_index_list():
return [
os.path.relpath(os.path.join(dirpath, filename), now_dir)
for dirpath, _, filenames in os.walk(models_path)
for filename in filenames
if filename.endswith(".index") and "trained" not in filename
]
def refresh_pth_and_index_list():
return (
{"choices": sorted(get_pth_list()), "__type__": "update"},
{"choices": sorted(get_index_list()), "__type__": "update"},
)
## Export Pth and Index Files
def export_pth(pth_path):
if pth_path and os.path.exists(pth_path):
return pth_path
return None
def export_index(index_path):
if index_path and os.path.exists(index_path):
return index_path
return None
## Upload to Google Drive
def upload_to_google_drive(pth_path, index_path):
def upload_file(file_path):
if file_path:
try:
gr.Info(f"Uploading {pth_path} to Google Drive...")
google_drive_folder = "/content/drive/MyDrive/ApplioExported"
if not os.path.exists(google_drive_folder):
os.makedirs(google_drive_folder)
google_drive_file_path = os.path.join(
google_drive_folder, os.path.basename(file_path)
)
if os.path.exists(google_drive_file_path):
os.remove(google_drive_file_path)
shutil.copy2(file_path, google_drive_file_path)
gr.Info("File uploaded successfully.")
except Exception as error:
print(error)
gr.Info("Error uploading to Google Drive")
upload_file(pth_path)
upload_file(index_path)
# Train Tab
def train_tab():
with gr.Accordion(i18n("Preprocess")):
with gr.Row():
with gr.Column():
model_name = gr.Dropdown(
label=i18n("Model Name"),
info=i18n("Name of the new model."),
choices=get_models_list(),
value="my-project",
interactive=True,
allow_custom_value=True,
)
dataset_path = gr.Dropdown(
label=i18n("Dataset Path"),
info=i18n("Path to the dataset folder."),
# placeholder=i18n("Enter dataset path"),
choices=get_datasets_list(),
allow_custom_value=True,
interactive=True,
)
refresh = gr.Button(i18n("Refresh"))
dataset_creator = gr.Checkbox(
label=i18n("Dataset Creator"),
value=False,
interactive=True,
visible=True,
)
with gr.Column(visible=False) as dataset_creator_settings:
with gr.Accordion(i18n("Dataset Creator")):
dataset_name = gr.Textbox(
label=i18n("Dataset Name"),
info=i18n("Name of the new dataset."),
placeholder=i18n("Enter dataset name"),
interactive=True,
)
upload_audio_dataset = gr.File(
label=i18n("Upload Audio Dataset"),
type="filepath",
interactive=True,
)
with gr.Column():
sampling_rate = gr.Radio(
label=i18n("Sampling Rate"),
info=i18n("The sampling rate of the audio files."),
choices=["32000", "40000", "48000"],
value="40000",
interactive=True,
)
rvc_version = gr.Radio(
label=i18n("RVC Version"),
info=i18n("The RVC version of the model."),
choices=["v1", "v2"],
value="v2",
interactive=True,
)
preprocess_output_info = gr.Textbox(
label=i18n("Output Information"),
info=i18n("The output information will be displayed here."),
value="",
max_lines=8,
interactive=False,
)
with gr.Row():
preprocess_button = gr.Button(i18n("Preprocess Dataset"))
preprocess_button.click(
run_preprocess_script,
[model_name, dataset_path, sampling_rate],
preprocess_output_info,
api_name="preprocess_dataset",
)
with gr.Accordion(i18n("Extract")):
with gr.Row():
hop_length = gr.Slider(
1,
512,
128,
step=1,
label=i18n("Hop Length"),
info=i18n(
"Denotes the duration it takes for the system to transition to a significant pitch change. Smaller hop lengths require more time for inference but tend to yield higher pitch accuracy."
),
interactive=True,
visible=False,
)
with gr.Row():
with gr.Column():
f0method = gr.Radio(
label=i18n("Pitch extraction algorithm"),
info=i18n(
"Pitch extraction algorithm to use for the audio conversion. The default algorithm is rmvpe, which is recommended for most cases."
),
choices=["pm", "dio", "crepe", "crepe-tiny", "harvest", "rmvpe"],
value="rmvpe",
interactive=True,
)
extract_output_info = gr.Textbox(
label=i18n("Output Information"),
info=i18n("The output information will be displayed here."),
value="",
max_lines=8,
interactive=False,
)
extract_button = gr.Button(i18n("Extract Features"))
extract_button.click(
run_extract_script,
[model_name, rvc_version, f0method, hop_length, sampling_rate],
extract_output_info,
api_name="extract_features",
)
with gr.Accordion(i18n("Train")):
with gr.Row():
batch_size = gr.Slider(
1,
50,
max_vram_gpu(0),
step=1,
label=i18n("Batch Size"),
info=i18n(
"It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results."
),
interactive=True,
)
save_every_epoch = gr.Slider(
1,
100,
10,
step=1,
label=i18n("Save Every Epoch"),
info=i18n("Determine at how many epochs the model will saved at."),
interactive=True,
)
total_epoch = gr.Slider(
1,
10000,
500,
step=1,
label=i18n("Total Epoch"),
info=i18n(
"Specifies the overall quantity of epochs for the model training process."
),
interactive=True,
)
with gr.Row():
pitch_guidance = gr.Checkbox(
label=i18n("Pitch Guidance"),
info=i18n(
"By employing pitch guidance, it becomes feasible to mirror the intonation of the original voice, including its pitch. This feature is particularly valuable for singing and other scenarios where preserving the original melody or pitch pattern is essential."
),
value=True,
interactive=True,
)
pretrained = gr.Checkbox(
label=i18n("Pretrained"),
info=i18n(
"Utilize pretrained models when training your own. This approach reduces training duration and enhances overall quality."
),
value=True,
interactive=True,
)
save_only_latest = gr.Checkbox(
label=i18n("Save Only Latest"),
info=i18n(
"Enabling this setting will result in the G and D files saving only their most recent versions, effectively conserving storage space."
),
value=False,
interactive=True,
)
save_every_weights = gr.Checkbox(
label=i18n("Save Every Weights"),
info=i18n(
"This setting enables you to save the weights of the model at the conclusion of each epoch."
),
value=True,
interactive=True,
)
custom_pretrained = gr.Checkbox(
label=i18n("Custom Pretrained"),
info=i18n(
"Utilizing custom pretrained models can lead to superior results, as selecting the most suitable pretrained models tailored to the specific use case can significantly enhance performance."
),
value=False,
interactive=True,
)
multiple_gpu = gr.Checkbox(
label=i18n("GPU Settings"),
info=(
i18n(
"Sets advanced GPU settings, recommended for users with better GPU architecture."
)
),
value=False,
interactive=True,
)
overtraining_detector = gr.Checkbox(
label=i18n("Overtraining Detector"),
info=i18n(
"Detect overtraining to prevent the model from learning the training data too well and losing the ability to generalize to new data."
),
value=False,
interactive=True,
)
with gr.Row():
with gr.Column(visible=False) as pretrained_custom_settings:
with gr.Accordion(i18n("Pretrained Custom Settings")):
upload_pretrained = gr.File(
label=i18n("Upload Pretrained Model"),
type="filepath",
interactive=True,
)
refresh_custom_pretaineds_button = gr.Button(
i18n("Refresh Custom Pretraineds")
)
g_pretrained_path = gr.Dropdown(
label=i18n("Custom Pretrained G"),
info=i18n(
"Select the custom pretrained model for the generator."
),
choices=sorted(pretraineds_list_g),
interactive=True,
allow_custom_value=True,
)
d_pretrained_path = gr.Dropdown(
label=i18n("Custom Pretrained D"),
info=i18n(
"Select the custom pretrained model for the discriminator."
),
choices=sorted(pretraineds_list_d),
interactive=True,
allow_custom_value=True,
)
with gr.Column(visible=False) as gpu_custom_settings:
with gr.Accordion(i18n("GPU Settings")):
gpu = gr.Textbox(
label=i18n("GPU Number"),
info=i18n(
"Specify the number of GPUs you wish to utilize for training by entering them separated by hyphens (-)."
),
placeholder=i18n("0 to ∞ separated by -"),
value="0",
interactive=True,
)
gr.Textbox(
label=i18n("GPU Information"),
info=i18n("The GPU information will be displayed here."),
value=get_gpu_info(),
interactive=False,
)
with gr.Column(visible=False) as overtraining_settings:
with gr.Accordion(i18n("Overtraining Detector Settings")):
overtraining_threshold = gr.Slider(
1,
100,
50,
step=1,
label=i18n("Overtraining Threshold"),
info=i18n(
"Set the maximum number of epochs you want your model to stop training if no improvement is detected."
),
interactive=True,
)
with gr.Row():
train_output_info = gr.Textbox(
label=i18n("Output Information"),
info=i18n("The output information will be displayed here."),
value="",
max_lines=8,
interactive=False,
)
with gr.Row():
train_button = gr.Button(i18n("Start Training"))
train_button.click(
run_train_script,
[
model_name,
rvc_version,
save_every_epoch,
save_only_latest,
save_every_weights,
total_epoch,
sampling_rate,
batch_size,
gpu,
pitch_guidance,
overtraining_detector,
overtraining_threshold,
pretrained,
custom_pretrained,
g_pretrained_path,
d_pretrained_path,
],
train_output_info,
api_name="start_training",
)
stop_train_button = gr.Button(
i18n("Stop Training & Restart Applio"), visible=False
)
stop_train_button.click(
fn=restart_applio,
inputs=[],
outputs=[],
)
index_button = gr.Button(i18n("Generate Index"))
index_button.click(
run_index_script,
[model_name, rvc_version],
train_output_info,
api_name="generate_index",
)
with gr.Accordion(i18n("Export Model"), open=False):
if not os.name == "nt":
gr.Markdown(
i18n(
"The button 'Upload' is only for google colab: Uploads the exported files to the ApplioExported folder in your Google Drive."
)
)
with gr.Row():
with gr.Column():
pth_file_export = gr.File(
label=i18n("Exported Pth file"),
type="filepath",
value=None,
interactive=False,
)
pth_dropdown_export = gr.Dropdown(
label=i18n("Pth file"),
info=i18n("Select the pth file to be exported"),
choices=get_pth_list(),
value=None,
interactive=True,
allow_custom_value=True,
)
with gr.Column():
index_file_export = gr.File(
label=i18n("Exported Index File"),
type="filepath",
value=None,
interactive=False,
)
index_dropdown_export = gr.Dropdown(
label=i18n("Index File"),
info=i18n("Select the index file to be exported"),
choices=get_index_list(),
value=None,
interactive=True,
allow_custom_value=True,
)
with gr.Row():
with gr.Column():
refresh_export = gr.Button(i18n("Refresh"))
if not os.name == "nt":
upload_exported = gr.Button(i18n("Upload"), variant="primary")
upload_exported.click(
fn=upload_to_google_drive,
inputs=[pth_dropdown_export, index_dropdown_export],
outputs=[],
)
def toggle_visible(checkbox):
return {"visible": checkbox, "__type__": "update"}
def toggle_visible_hop_length(f0method):
if f0method == "crepe" or f0method == "crepe-tiny":
return {"visible": True, "__type__": "update"}
return {"visible": False, "__type__": "update"}
def toggle_pretrained(pretrained, custom_pretrained):
if custom_pretrained == False:
return {"visible": pretrained, "__type__": "update"}, {
"visible": False,
"__type__": "update",
}
else:
return {"visible": pretrained, "__type__": "update"}, {
"visible": pretrained,
"__type__": "update",
}
def enable_stop_train_button():
return {"visible": False, "__type__": "update"}, {
"visible": True,
"__type__": "update",
}
def disable_stop_train_button():
return {"visible": True, "__type__": "update"}, {
"visible": False,
"__type__": "update",
}
def download_prerequisites(version):
for remote_folder, file_list in pretraineds_v1:
local_folder = folder_mapping.get(remote_folder, "")
missing = False
for file in file_list:
destination_path = os.path.join(local_folder, file)
if not os.path.exists(destination_path):
missing = True
if version == "v1" and missing == True:
gr.Info(
"Downloading prerequisites... Please wait till it finishes to start preprocessing."
)
run_prerequisites_script("True", "False", "True", "True")
gr.Info(
"Prerequisites downloaded successfully, you may now start preprocessing."
)
rvc_version.change(
fn=download_prerequisites,
inputs=[rvc_version],
outputs=[],
)
refresh.click(
fn=refresh_models_and_datasets,
inputs=[],
outputs=[model_name, dataset_path],
)
dataset_creator.change(
fn=toggle_visible,
inputs=[dataset_creator],
outputs=[dataset_creator_settings],
)
upload_audio_dataset.upload(
fn=save_drop_dataset_audio,
inputs=[upload_audio_dataset, dataset_name],
outputs=[upload_audio_dataset, dataset_path],
)
f0method.change(
fn=toggle_visible_hop_length,
inputs=[f0method],
outputs=[hop_length],
)
pretrained.change(
fn=toggle_pretrained,
inputs=[pretrained, custom_pretrained],
outputs=[custom_pretrained, pretrained_custom_settings],
)
custom_pretrained.change(
fn=toggle_visible,
inputs=[custom_pretrained],
outputs=[pretrained_custom_settings],
)
refresh_custom_pretaineds_button.click(
fn=refresh_custom_pretraineds,
inputs=[],
outputs=[g_pretrained_path, d_pretrained_path],
)
upload_pretrained.upload(
fn=save_drop_model,
inputs=[upload_pretrained],
outputs=[upload_pretrained],
)
overtraining_detector.change(
fn=toggle_visible,
inputs=[overtraining_detector],
outputs=[overtraining_settings],
)
multiple_gpu.change(
fn=toggle_visible,
inputs=[multiple_gpu],
outputs=[gpu_custom_settings],
)
train_button.click(
fn=enable_stop_train_button,
inputs=[],
outputs=[train_button, stop_train_button],
)
train_output_info.change(
fn=disable_stop_train_button,
inputs=[],
outputs=[train_button, stop_train_button],
)
pth_dropdown_export.change(
fn=export_pth,
inputs=[pth_dropdown_export],
outputs=[pth_file_export],
)
index_dropdown_export.change(
fn=export_index,
inputs=[index_dropdown_export],
outputs=[index_file_export],
)
refresh_export.click(
fn=refresh_pth_and_index_list,
inputs=[],
outputs=[pth_dropdown_export, index_dropdown_export],
)
|