File size: 40,576 Bytes
3f9bd99 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 |
import shutil
import gradio as gr
import os
import webbrowser
import subprocess
import json
import yaml
import argparse
#########################################################
py_dir=r"venv\python.exe" # SET PYTHON PATH HERE!
#########################################################
current_directory = os.path.dirname(os.path.abspath(__file__))
if not os.path.exists('./Data'):
os.mkdir('./Data')
################### Load md
try:
with open('docs/file_structure.md', mode="r", encoding="utf-8") as f:
file_structure_md=f.read()
except Exception as e:
file_structure_md='读取错误'+ str(e)
try:
with open('docs/README.md', mode="r", encoding="utf-8") as f:
readme_md=f.read()
except Exception as e:
readme_md='读取错误'+ str(e)
try:
with open('docs/command_usage.md', mode="r", encoding="utf-8") as f:
cmd_md=f.read()
except Exception as e:
cmd_md='读取错误'+ str(e)
try:
with open('docs/update_info.md', mode="r", encoding="utf-8") as f:
update_info=f.read()
except Exception as e:
update_info='读取错误'+ str(e)
try:
with open('docs/errors.md', mode="r", encoding="utf-8") as f:
errors=f.read()
except Exception as e:
errors='读取错误'+ str(e)
###################
current_page=1
class_info={}
emo_speaker_list=[]
emo_class_list=[]
emo_file_name_list=[]
emo_audio_list=[]
emo_tr_list=[]
emo_checkbox_list=[]
emo_current_proj_name=None
def load_emo_clu(proj_name):
if proj_name=="" or proj_name=='null':
return "请选择!",emo_speaker.update(choices=[],value=None)
global emo_current_proj_name
global class_info
global emo_speaker_list
emo_current_proj_name=proj_name
try:
class_info = yaml.load(open(os.path.join('Data',proj_name,'emo_clustering.yml')),Loader=yaml.FullLoader)
emo_speaker_list=list(class_info.keys())
return emo_status_text.update(value=f"当前的实验:{proj_name}"),emo_speaker.update(choices=emo_speaker_list,value=emo_speaker_list[0])#,*emo_change_speaker(emo_speaker_list[0])#,*load_page(emo_speaker_list[0],list(class_info[emo_speaker_list[0]].keys())[0])
except Exception as e:
class_info={}
emo_speaker_list=[]
return emo_status_text.update(value='加载错误'+str(e)),emo_speaker.update(choices=[],value=None)
def get_transcription(proj_name,spk,wav):#transciption and wav path
y=yaml.load(open(os.path.join('Data',proj_name,'config.yml')),Loader=yaml.FullLoader)
wav_path=os.path.normpath(os.path.join(y["dataset_path"],y["resample"]["out_dir"],spk,wav))
list_path=os.path.join(y["dataset_path"],y["preprocess_text"]["cleaned_path"])
with open(list_path, mode="r", encoding="utf-8", errors='ignore') as f:
for lines in f:
if os.path.normpath(lines.split("|")[0])==wav_path:
return lines.split("|")[3],wav_path
return "加载失败",""
def emo_change_speaker(speaker):
if speaker is None:
return emo_chosen_class.update(choices=[],value=None)
global current_page
current_page=1
lst=list(class_info[speaker].keys())
return emo_chosen_class.update(choices=lst,value=lst[0]),*switch_page(speaker,lst[0])
def emo_change_class(speaker,class_):
if (speaker is None) or (class_ is None):
return emo_ret_none()
global current_page
current_page=1
return switch_page(speaker,class_)
def emo_ret_none():
global current_page
wavname_list=[]
transcription_list=[]
audio_list=[]
checkbox_list=[]
for i in range(10):
wavname_list.append(gr.Textbox().update(value=None,label='None',visible=False))
audio_list.append(gr.Audio.update(label="情感参考音频",value=None,interactive=False,visible=False))
transcription_list.append(gr.Textbox().update(value=None,visible=False))
checkbox_list.append(gr.Checkbox().update(value=False,visible=False))
wavname_list[0]=gr.Textbox().update(value='看起来列表似乎是空的',label='None',visible=True)
list_all=wavname_list+transcription_list+audio_list+checkbox_list
return *list_all,previous_btn.update(interactive=False),emo_page_index.update(value=current_page),next_btn.update(interactive=False)
def load_page(page_start,page_end,spk,class_name):#load lines of content
global emo_current_proj_name
wavname_list=[]
transcription_list=[]
audio_list=[]
checkbox_list=[]
#page_start,page_end,_,_=calculate_page(spk,class_name)
for i in range(page_start, page_end+1):
wav_name=class_info[spk][class_name][i]
tr,wavpath=get_transcription(emo_current_proj_name,spk,wav_name)
if wavpath=="":
audio_list.append(gr.Audio.update(label="情感参考音频",value=None,interactive=False,visible=True))
else:
audio_list.append(gr.Audio.update(label="情感参考音频",value=wavpath,interactive=False,visible=True))
wavname_list.append(gr.Textbox().update(value=wav_name,label=i,visible=True))
transcription_list.append(gr.Textbox().update(value=tr,visible=True))
checkbox_list.append(gr.Checkbox().update(value=False,visible=True))
length=page_end-page_start
if length<10:
for i in range(9-length):
wavname_list.append(gr.Textbox().update(value=None,label='None',visible=False))
audio_list.append(gr.Audio.update(label="情感参考音频",value=None,interactive=False,visible=False))
transcription_list.append(gr.Textbox().update(value=None,visible=False))
checkbox_list.append(gr.Checkbox().update(value=False,visible=False))
list_all=wavname_list+transcription_list+audio_list+checkbox_list
return list_all
def calculate_page(length):#calculate the index and return
global current_page
if length<=10 and current_page==1:
return 0,length-1,False,False#range,left,right
else:
end=10*current_page-1
if end > length-1:#num<10
end=length-1
return 10*(current_page-1),end,True,False
else:#full
if current_page==1:
return 10*(current_page-1),end,False,True
return 10*(current_page-1),end,True,True
def switch_page(speaker,class_name):#switch to the specified page or adjust page index , then load page.
global current_page
global class_info
if current_page<1:
current_page=1
length=len(class_info[speaker][class_name])
if length==0:
return emo_ret_none()
if current_page>1 and (current_page-1)*10>=length:#e.g [0-9],page==2-> [10] -> out of range
current_page-=1
#print(length)
page_start,page_end,left,right=calculate_page(length)
#print(page_start,page_end)
return *load_page(page_start,page_end,speaker,class_name),previous_btn.update(interactive=left),emo_page_index.update(value=current_page),next_btn.update(interactive=right)
def reverse_selection(*checkbox_list):
rt_list = [not i if i is True else True for i in checkbox_list]
return rt_list
def del_wav_in_class(speaker,class_name,*checkbox_list):
length=len(class_info[speaker][class_name])
page_start,page_end,_,_=calculate_page(length)
id=0
checkbox_list=list(checkbox_list)
for i in range(page_start, page_end+1):
if checkbox_list[id]:
class_info[speaker][class_name][id]=None
id+=1
while None in class_info[speaker][class_name]:
class_info[speaker][class_name].remove(None)
return switch_page(speaker,class_name)
def del_rename_class(speaker,class_,new_name=None):
global class_info
global current_page
if new_name is not None:
if new_name in class_info[speaker]:
return emo_chosen_class.update(value=new_name,choices=list(class_info[speaker].keys())),gr.Textbox().update(placeholder='重名了!',value=None)
class_info[speaker][new_name]=class_info[speaker][class_]
del class_info[speaker][class_]
return emo_chosen_class.update(value=new_name,choices=list(class_info[speaker].keys())),gr.Textbox().update(placeholder='输入重命名',value=None)
else:
del class_info[speaker][class_]
lst=list(class_info[speaker].keys())
current_page=0
return emo_chosen_class.update(choices=lst,value=lst[0])
current_yml=None
def get_status():
global current_yml
try:
cfg = yaml.load(open('config.yml'),Loader=yaml.FullLoader)
current_yml='当前的训练: '+os.path.basename(cfg["dataset_path"])+"\n\n以下是配置文件内容:\n\n"
with open('config.yml', mode="r", encoding="utf-8", errors='ignore') as f:
current_y=f.read()
current_yml+=current_y
except Exception as error:
current_yml=error
get_status()
def p0_write_yml(name,val_per_spk,max_val_total,bert_num_processes,emo_num_processes,num_workers,keep_ckpts):
if name=='null'or name=='':
return '请选择!'
config_path=os.path.join('Data',name,'config.yml')
config_yml = yaml.load(open(config_path),Loader=yaml.FullLoader)
config_yml["preprocess_text"]["val_per_spk"] = int(val_per_spk)
config_yml["preprocess_text"]["max_val_total"] = int(max_val_total)
config_yml["bert_gen"]["num_processes"] = int(bert_num_processes)
config_yml["emo_gen"]["num_processes"] = int(emo_num_processes)
config_yml["train_ms"]["num_workers"]=int(num_workers)
config_yml["train_ms"]["keep_ckpts"]=int(keep_ckpts)
with open(config_path, 'w', encoding='utf-8') as f:
yaml.dump(config_yml, f)
return 'Success'
list_project = []
def refresh_project_list():
global list_project
list_project = []
for item in os.listdir('Data'):
item_path = os.path.join('Data', item)
if os.path.isdir(item_path):
list_project.append(item)
return (project_name.update(choices=list_project),project_name2.update(choices=list_project),project_name3.update(choices=list_project),project_name4.update(choices=list_project),'已刷新下拉列表')
for item in os.listdir('Data'):
item_path = os.path.join('Data', item)
if os.path.isdir(item_path):
list_project.append(item)
def p0_mkdir(name):
if name!='':
try:
path='Data'
path=os.path.join('Data',name)
os.mkdir(path)#path=data/xxx/
os.mkdir(os.path.join(path,'custom_character_voice'))
os.mkdir(os.path.join(path,'filelists'))
os.mkdir(os.path.join(path,'models'))
shutil.copy("./configs/config.json",os.path.join(path,"config.json"))
try:
with open('./configs/default_config.yml', mode="r", encoding="utf-8") as f:
cfg_yml=yaml.load(f,Loader=yaml.FullLoader)
except:
with open('default_config.yml', mode="r", encoding="utf-8") as f:
cfg_yml=yaml.load(f,Loader=yaml.FullLoader)
cfg_yml["dataset_path"]=path
cfg_yml["resample"]["in_dir"]="custom_character_voice"
cfg_yml["resample"]["out_dir"]="custom_character_voice"
cfg_yml["preprocess_text"]["cleaned_path"]='filelists/cleaned.list'
cfg_yml["preprocess_text"]["transcription_path"]='filelists/short_character_anno.list'
cfg_yml["preprocess_text"]["train_path"]='filelists/train.list'
cfg_yml["preprocess_text"]["val_path"]='filelists/val.list'
cfg_yml["preprocess_text"]["config_path"]='config.json'
cfg_yml["bert_gen"]["config_path"]='config.json'
cfg_yml["emo_gen"]["config_path"]='config.json'
cfg_yml["train_ms"]["config_path"]='config.json'
with open(os.path.join(path,"config.yml"), 'w', encoding='utf-8') as f:
yaml.dump(cfg_yml, f)
os.startfile(path)
refresh_project_list()
return project_name.update(choices=list_project,value=name),'Success. 已经自动打开了创建好的文件夹。请将音频按说话人分文件夹放入custom_character_voice内。然后进行下一步操作。'
except Exception as error:
return error
else:
return '请输入名称!'
def p0_load_cfg(projectname):
if projectname=='null'or projectname=='':
return p0_status.update(value=current_yml),'请选择!'
try:
shutil.copy(os.path.join('Data',projectname,'config.yml'),'config.yml')
get_status()
return p0_status.update(value=current_yml) ,'Success'
except Exception as error:
return p0_status.update(value=current_yml),error
def a1a_transcribe(size,lang):
command = f'{py_dir} short_audio_transcribe.py --languages {lang} --whisper_size {size}'
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。完成后无报错即可关闭进行下一步!'
def a1b_transcribe_genshin():
command = f"{py_dir} transcribe_genshin.py"
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。完成后无报错即可关闭进行下一步!'
def a2_preprocess_text():
command = f"{py_dir} preprocess_text.py"
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。完成后无报错即可关闭进行下一步!'
def a3_bert_gen():
command = f"{py_dir} bert_gen.py"
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。完成后无报错即可关闭进行下一步!'
def a3_emo_gen():
command = f"{py_dir} emo_gen.py"
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。完成后无报错即可关闭进行下一步!'
def a35_json(bs,lr,interval):
try:
with open('config.yml', mode="r", encoding="utf-8") as f:
cfg_yml=yaml.load(f,Loader=yaml.FullLoader)
config_path=os.path.join(cfg_yml["dataset_path"],'config.json')
configjson = json.load(open(config_path, encoding="utf-8"))
configjson["train"]["batch_size"] = int(bs)
configjson["train"]["learning_rate"] = lr
configjson["train"]["log_interval"] = int(interval)
configjson["train"]['eval_interval'] = int(interval)
with open(config_path, 'w', encoding='utf-8') as f:
json.dump(configjson, f, indent=2, ensure_ascii=False)
return 'success'
except Exception as error:
return error
def a4a_train():
command = f"{py_dir} train_ms.py"
with open('config.yml', mode="r", encoding="utf-8") as f:
configyml=yaml.load(f,Loader=yaml.FullLoader)
cfg_path=os.path.join(configyml["dataset_path"],'config.json')
configjson = json.load(open(cfg_path, encoding="utf-8"))
if not configjson["train"]["skip_optimizer"]:
configjson["train"]["skip_optimizer"]=True
with open(cfg_path, 'w', encoding='utf-8') as f:
json.dump(configjson, f, indent=2, ensure_ascii=False)
print("已经修改配置文件!\n")
shutil.copy('./pretrained_models/D_0.pth',os.path.join(os.path.join(configyml["dataset_path"],configyml["train_ms"]["model"]),'D_0.pth'))
shutil.copy('./pretrained_models/G_0.pth',os.path.join(os.path.join(configyml["dataset_path"],configyml["train_ms"]["model"]),'G_0.pth'))
shutil.copy('./pretrained_models/DUR_0.pth',os.path.join(os.path.join(configyml["dataset_path"],configyml["train_ms"]["model"]),'DUR_0.pth'))
print("已经复制了底模\n")
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
print(command+'\n\n')
return '新的命令行窗口已经打开,请关注输出信息。关闭窗口或Ctrl+C终止训练'
def a4b_train_cont():
command = f"{py_dir} train_ms.py"
with open('config.yml', mode="r", encoding="utf-8") as f:
configyml=yaml.load(f,Loader=yaml.FullLoader)
cfg_path=os.path.join(configyml["dataset_path"],'config.json')
configjson = json.load(open(cfg_path, encoding="utf-8"))
if configjson["train"]["skip_optimizer"]:
configjson["train"]["skip_optimizer"]=False
with open(cfg_path, 'w', encoding='utf-8') as f:
json.dump(configjson, f, indent=2, ensure_ascii=False)
print("已经修改配置文件!\n")
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
print(command+'\n\n')
return '新的命令行窗口已经打开,请关注输出信息。关闭窗口或Ctrl+C终止训练'
def start_tb():
with open('config.yml', mode="r", encoding="utf-8") as f:
configyml=yaml.load(f,Loader=yaml.FullLoader)
command = f"{py_dir} -m tensorboard.main --logdir="+os.path.join(configyml["dataset_path"],configyml["train_ms"]["model"])
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。'
ckpt_list = ['null']
def c2_refresh_sub_opt(name):
try:
global ckpt_list
ckpt_list=['null']
file_list = os.listdir(os.path.join("Data",name,"models"))
for ck in file_list:
if os.path.splitext(ck)[-1] == ".pth"and ck[:2] != "D_" and ck[:4] !="DUR_":
ckpt_list.append(ck)
return models_in_project.update(choices=ckpt_list,value=ckpt_list[-1])
except :
return models_in_project.update(choices=['null'],value='null')
def c2_infer(proj_name,model_name):
if proj_name=='null' or model_name=='null':
return '请选择模型!'
path=f'./Data/{proj_name}'
command = f'{py_dir} webui.py -c {path}/config.json -m {path}/models/{model_name}'
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。关闭窗口结束推理服务。'
def c2_infer_2(proj_name,model_name):
y=yaml.load(open('config.yml'),Loader=yaml.FullLoader)
if proj_name=='null' or model_name=='null':
y["server"]["models"]=[]
else:
y["server"]["models"]=[]
y["server"]["models"].append({"config":os.path.join('Data',proj_name,'config.json'),"device":'cuda',"language": 'ZH',"model":os.path.join('Data',proj_name,'models',model_name),"speakers":[]})
with open("config.yml", 'w', encoding='utf-8') as f:
yaml.dump(y, f)
subprocess.Popen(['start', 'cmd', '/k',f'{py_dir} server_fastapi.py'],cwd=current_directory,shell=True)
return '已经修改了全局配置文件。新的命令行窗口已经打开,请关注输出信息。关闭窗口结束推理服务。'
def write_version(name,version,cont):
if name=='null':
return opt_continue.update(value=False),'请选择!'
path=os.path.join('Data',name,'config.json')
try:
configjson = json.load(open(path))
if "version" in configjson:
if not cont:
return opt_continue.update(value=False),'版本信息已经存在,是不是手滑了?'
configjson["version"] = version
else:
configjson["version"] = version
with open(path, 'w', encoding='utf-8') as f:
json.dump(configjson, f, indent=2, ensure_ascii=False)
return opt_continue.update(value=False),f'Success. {version}'
except Exception as e:
return opt_continue.update(value=False),e
def switch_previous_page(speaker,class_name):
global current_page
current_page-=1
return switch_page(speaker,class_name)
def switch_next_page(speaker,class_name):
global current_page
current_page+=1
return switch_page(speaker,class_name)
def emo_write(proj_name):
global class_info
try:
with open(os.path.join('Data',proj_name,'emo_clustering.yml'), 'w', encoding='utf-8') as f:
yaml.dump(class_info, f)
return 'Success'
except Exception as e:
return '写入出错'+str(e)
def run_ec_gen(proj_name,num_clusters,num_wav):
if proj_name=="":
command = f"{py_dir} emotion_clustering.py -n {int(num_clusters)} -r {int(num_wav)}"
else:
command = f"{py_dir} emotion_clustering.py -y {os.path.join('Data',proj_name,'config.yml')} -n {int(num_clusters)} -r {int(num_wav)}"
print(command+'\n\n')
subprocess.Popen(['start', 'cmd', '/k', command],cwd=current_directory,shell=True)
return '新的命令行窗口已经打开,请关注输出信息。'
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument(
"-p", "--server_port", default=6660,type=int,help="server_port"
)
args = parser.parse_args()
with gr.Blocks(title="Bert-VITS-2-Manager-WebUI-210") as app:
gr.Markdown(value="""
Bert-VITS2训练管理器
严禁将此项目用于一切违反《中华人民共和国宪法》,《中华人民共和国刑法》,《中华人民共和国治安管理处罚法》和《中华人民共和国民法典》之用途。由使用本整合包产生的问题和作者、原作者无关!!!
作者:bilibili@数列解析几何一生之敌
适用于整合包版本V2.1,不兼容之前的版本。
WebUI更新日期:2023.11.30
""")
with gr.Tabs():
with gr.TabItem("1.创建实验文件夹和加载全局配置"):
with gr.Row():
with gr.Column():
p0_mkdir_name=gr.Textbox(label="这将创建实验文件夹,请输入实验名称,不要包含中文、特殊字符和保留字符。",
placeholder="请输入实验名称,最好不要包含中文、特殊字符和保留字符。",
value="",
lines=1,
interactive=True)
p0_mkdir_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
p0_mkdir_btn=gr.Button(value="创建!", variant="primary")
gr.Markdown(value="<br>")
project_name = gr.Dropdown(label="实验文件夹", choices=list_project, value='null'if not list_project else list_project[-1],interactive=True)
with gr.Row():
p0_val_ps = gr.Number(label="每个说话人的验证集数", value="4",interactive=True)
p0_val_tt = gr.Number(label="总的验证集数", value="8",interactive=True)
p0_bg_t = gr.Number(label="bert_gen线程数", value="2",interactive=True)
p0_emo_t = gr.Number(label="emo_gen线程数", value="2",interactive=True)
p0_dataloader = gr.Number(label="num_workers", value="4",interactive=True)
p0_keep_ckpt = gr.Number(label="模型留存个数", value="10",interactive=True)
p0_load_cfg_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
with gr.Row():
p0_write_cfg_btn=gr.Button(value="保存更改(但不会自动加载)", variant="primary")
p0_load_cfg_btn = gr.Button(value="加载训练配置", variant="primary")
p0_load_cfg_refresh_btn=gr.Button(value="刷新选项", variant="secondary")
with gr.Column():
#p0_current_proj=gr.Textbox(label="当前生效的训练",value="",interactive=False)
p0_status=gr.TextArea(label="训练前请确认当前的全局配置信息", value=current_yml,interactive=False)
with gr.TabItem("2.训练"):
with gr.Column():
whisper_size = gr.Radio(label="选择whisper大小,large需要12G显存", choices=['large','medium','small'], value="medium")
language = gr.Radio(label="选择语言(默认中日英,其他不支持的语言会被跳过)", choices=['C','CJ','CJE'], value="CJE")
with gr.Column():
with gr.Row():
a1a_btn = gr.Button(value="1.a.数据集重采样和标注(使用whisper)", variant="primary")
a1b_btn = gr.Button(value="1.b.处理下载的已标注的原神数据集", variant="primary")
a1_textbox_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮,二选一",interactive=False)
#gr.Markdown(value="<br>")
with gr.Row():
a2_btn = gr.Button(value="2.文本预处理", variant="primary")
a2_textbox_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
#gr.Markdown(value="<br>")
with gr.Row():
a3_btn = gr.Button(value="3-1.生成bert文件", variant="primary")
a3_btn_2 = gr.Button(value="3-2.生成emo文件", variant="primary")
a3_textbox_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
# gr.Markdown(value="<br>")
with gr.Row():
#with gr.Column():
a35_btn = gr.Button(value="写入配置文件", variant="primary")
with gr.Column():
with gr.Row():
a35_textbox_bs = gr.Number(label="批大小", value="8",interactive=True)
a35_textbox_lr = gr.Number(label="学习率", value="0.0001",interactive=True)
a35_textbox_save = gr.Number(label="保存间隔", value="100",interactive=True)
a35_textbox_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
#gr.Markdown(value="<br>")
with gr.Row():
with gr.Row():
a4a_btn = gr.Button(value="4a.首次训练", variant="primary")
a4b_btn = gr.Button(value="4b.继续训练", variant="primary")
with gr.Column():
a4_textbox_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
with gr.Row():
tb_btn = gr.Button(value="启动TensorBoard", variant="primary")
tb_textbox_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
with gr.TabItem("3.启动推理"):
gr.Markdown(value='工作区模型推理(Data内各实验目录下的模型)')
with gr.Row():
project_name2 = gr.Dropdown(label="选择实验名", choices=list_project, value='null',interactive=True)
models_in_project = gr.Dropdown(label="选择模型", choices=ckpt_list, value='null'if not ckpt_list else ckpt_list[0],interactive=True)
with gr.Column():
c2_btn = gr.Button(value="启动推理(Gr WebUI)", variant="primary")
c2_btn2 = gr.Button(value="启动推理(Hiyori UI)", variant="primary")
c2_btn_refresh=gr.Button(value="刷新选项", variant="secondary")
with gr.Column():
c2_textbox_output_text = gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
project_name2.change(c2_refresh_sub_opt,[project_name2],[models_in_project])
with gr.TabItem("辅助功能"):
with gr.TabItem("情感分类编辑"):
with gr.Row():
project_name4=gr.Dropdown(label="选择实验名", choices=list_project,value=''if not list_project else list_project[0],interactive=True)
emo_proj_load_btn=gr.Button(value="加载配置", variant="primary")
emo_proj_refresh=gr.Button(value="刷新选项", variant="secondary")
emo_status_text=gr.Textbox(label="状态", placeholder="点击处理按钮",interactive=False)
#
with gr.Column():
with gr.Row():
emo_speaker=gr.Dropdown(label='选择说话人',choices=emo_speaker_list,interactive=True)
emo_chosen_class=gr.Dropdown(label='选择类别',choices=emo_class_list,interactive=True)
emo_write_btn=gr.Button(value="写入(保存更改)", variant="primary")
with gr.Accordion("生成配置"):
with gr.Row():
ec_gen_btn=gr.Button(value="生成配置",variant="primary")
ec_gen_n_clu=gr.Number(label="类别数量",value=5)
ec_gen_range=gr.Number(label="每类的音频数",value=5)
with gr.Accordion("编辑区域"):
with gr.Row():
emo_enter_rename=gr.Textbox(label="重命名",placeholder="在此输入重命名,然后点击重命名按钮",scale=2)
emo_enter_rename_btn=gr.Button(value="重命名类别", variant="primary",scale=1)
emo_del_class_btn=gr.Button(value="删除此类别", variant="primary",scale=1)
emo_del_wav_btn=gr.Button(value="删除所选", variant="primary",scale=1)
emo_reverse_selection_btn=gr.Button(value="反向选择", variant="primary",scale=1)
with gr.Row():
emo_file_name=gr.Textbox(label="文件名",value='空列表',interactive=False,scale=4)
emo_tr=gr.Textbox(label="文本",value=None,interactive=False,scale=5,visible=False)
emo_wav=gr.Audio(label="情感参考音频",value=None,interactive=False,visible=False,scale=5)
emo_checkbox=gr.Checkbox(label='选择',value=False,scale=1,visible=False)
emo_tr_list.append(emo_tr)
emo_file_name_list.append(emo_file_name)
emo_audio_list.append(emo_wav)
emo_checkbox_list.append(emo_checkbox)
for i in range(9):
with gr.Row():
emo_file_name=gr.Textbox(label="文件名",value=None,interactive=False,scale=4,visible=False)
emo_tr=gr.Textbox(label="文本",value=None,interactive=False,scale=5,visible=False)
emo_wav=gr.Audio(label="情感参考音频",value=None,interactive=False,visible=False,scale=5)
emo_checkbox=gr.Checkbox(label='选择',value=False,scale=1,visible=False)
emo_tr_list.append(emo_tr)
emo_file_name_list.append(emo_file_name)
emo_audio_list.append(emo_wav)
emo_checkbox_list.append(emo_checkbox)
with gr.Row():
previous_btn=gr.Button(value="<-上一页", variant="secondary",interactive=False,scale=5)
emo_page_index=gr.Number(label='页码',value=current_page,interactive=False,scale=1)
next_btn=gr.Button(value="下一页->", variant="primary",interactive=False,scale=5)
emo_speaker.change(emo_change_speaker,[emo_speaker],[emo_chosen_class,*emo_file_name_list,*emo_tr_list,*emo_audio_list,*emo_checkbox_list,previous_btn,emo_page_index,next_btn])
emo_chosen_class.change(emo_change_class,[emo_speaker,emo_chosen_class],[*emo_file_name_list,*emo_tr_list,*emo_audio_list,*emo_checkbox_list,previous_btn,emo_page_index,next_btn])
with gr.TabItem("配置文件添加版本号"):
gr.Markdown(value='旧版本模型的配置文件添加版本号后方可在2.0版本下使用兼容推理')
gr.Markdown(value='按文件结构把配置文件和模型放到对应位置,然后开始操作。')
gr.Markdown(value='使用1.1和1.1.1版兼容推理需要安装上一个版本使用的日语bert。')
gr.Markdown(value='可选版本为1.0.1,1.1,1.1.1和2.0。旧整合包版本为1.0.1或1.1.1或2.0。')
project_name3 = gr.Dropdown(label="选择实验名", choices=list_project, value='null'if not list_project else list_project[0],interactive=True)
with gr.Row():
with gr.Column():
choose_version=gr.Dropdown(label="选择版本", choices=['1.0.1','1.1','1.1.1','2.0','2.1'], value='1.0.1',interactive=True)
opt_continue = gr.Checkbox(label="我没手滑")
write_ver_btn=gr.Button(value="写入",variant="primary")
write_ver_refresh_btn=gr.Button(value="刷新",variant="secondary")
with gr.Column():
write_ver_textbox=gr.Textbox(label="输出信息", placeholder="点击处理按钮",interactive=False)
with gr.TabItem("关于&帮助"):
with gr.TabItem("项目简介"):
with gr.Row():
gr.Markdown(value=readme_md)
gr.Markdown(value=update_info)
with gr.TabItem("训练流程和命令行使用"):
with gr.Row():
gr.Markdown(value=cmd_md)
gr.Markdown(value=file_structure_md)
with gr.TabItem("常见错误"):
with gr.Row():
gr.Markdown(value=errors)
p0_write_cfg_btn.click(p0_write_yml,
inputs=[project_name,p0_val_ps,p0_val_tt,p0_bg_t,p0_emo_t,p0_dataloader,p0_keep_ckpt],
outputs=[
p0_load_cfg_output_text,
],)
p0_mkdir_btn.click(p0_mkdir,
inputs=[p0_mkdir_name],
outputs=[
project_name,
p0_mkdir_output_text,
],)
p0_load_cfg_btn.click(p0_load_cfg,
inputs=[project_name],
outputs=[p0_status,
p0_load_cfg_output_text,
],)
p0_load_cfg_refresh_btn.click(refresh_project_list,
inputs=[],
outputs=[project_name,
project_name2,
project_name3,
project_name4,
p0_load_cfg_output_text,
],)
emo_proj_refresh.click(refresh_project_list,
inputs=[],
outputs=[project_name,
project_name2,
project_name3,
project_name4,
emo_status_text,
],)
a1a_btn.click(
a1a_transcribe,
inputs=[whisper_size,language],
outputs=[
a1_textbox_output_text,
],
)
a1b_btn.click(
a1b_transcribe_genshin,
outputs=[
a1_textbox_output_text,
],
)
a2_btn.click(
a2_preprocess_text,
outputs=[
a2_textbox_output_text,
],
)
a3_btn.click(
a3_bert_gen,
outputs=[
a3_textbox_output_text,
],
)
a3_btn_2.click(
a3_emo_gen,
outputs=[
a3_textbox_output_text,
],
)
a35_btn.click(
a35_json,
inputs=[a35_textbox_bs,a35_textbox_lr,a35_textbox_save],
outputs=[
a35_textbox_output_text,
],
)
a4a_btn.click(
a4a_train,
outputs=[
a4_textbox_output_text,
],
)
a4b_btn.click(
a4b_train_cont,
outputs=[
a4_textbox_output_text,
],
)
tb_btn.click(
start_tb,
outputs=[
tb_textbox_output_text,
],
)
c2_btn.click(
c2_infer,
inputs=[project_name2,models_in_project],
outputs=[
c2_textbox_output_text,
],
)
c2_btn2.click(
c2_infer_2,
inputs=[project_name2,models_in_project],
outputs=[
c2_textbox_output_text,
],
)
c2_btn_refresh.click(refresh_project_list,[],[project_name,project_name2,project_name3,project_name4,c2_textbox_output_text])
write_ver_refresh_btn.click(refresh_project_list,[],[project_name,project_name2,project_name3,project_name4,write_ver_textbox])
write_ver_btn.click(
write_version,
inputs=[project_name3,choose_version,opt_continue],
outputs=[opt_continue,
write_ver_textbox,
],
)
emo_proj_load_btn.click(load_emo_clu,inputs=[project_name4],outputs=[emo_status_text,emo_speaker])
#emo_del_class_btn.click()
emo_del_wav_btn.click(
del_wav_in_class,
inputs=[emo_speaker,emo_chosen_class,*emo_checkbox_list],
outputs=[*emo_file_name_list,*emo_tr_list,*emo_audio_list,*emo_checkbox_list,previous_btn,emo_page_index,next_btn]
)
emo_reverse_selection_btn.click(
reverse_selection,
inputs=[*emo_checkbox_list],
outputs=[*emo_checkbox_list]
)
emo_del_class_btn.click(
del_rename_class,
inputs=[emo_speaker,emo_chosen_class],
outputs=[emo_chosen_class]
)
emo_enter_rename_btn.click(
del_rename_class,
inputs=[emo_speaker,emo_chosen_class,emo_enter_rename],
outputs=[emo_chosen_class,emo_enter_rename]
)
next_btn.click(
switch_next_page,
inputs=[emo_speaker,emo_chosen_class],
outputs=[*emo_file_name_list,*emo_tr_list,*emo_audio_list,*emo_checkbox_list,previous_btn,emo_page_index,next_btn]
)
previous_btn.click(
switch_previous_page,
inputs=[emo_speaker,emo_chosen_class],
outputs=[*emo_file_name_list,*emo_tr_list,*emo_audio_list,*emo_checkbox_list,previous_btn,emo_page_index,next_btn]
)
emo_write_btn.click(
emo_write,
inputs=[project_name4],
outputs=[emo_status_text]
)
ec_gen_btn.click(
run_ec_gen,
inputs=[project_name4,ec_gen_n_clu,ec_gen_range],
outputs=[emo_status_text]
)
webbrowser.open(f"http://127.0.0.1:{args.server_port}")
app.launch(share=False,server_port=args.server_port)
|