forked from AIGC-Audio/AudioGPT
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathaudio-chatgpt.py
1381 lines (1281 loc) · 75 KB
/
audio-chatgpt.py
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
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
import sys
import os
sys.path.append(os.path.dirname(os.path.realpath(__file__)))
sys.path.append(os.path.dirname(os.path.dirname(os.path.realpath(__file__))))
sys.path.append(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'NeuralSeq'))
sys.path.append(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'text_to_audio/Make_An_Audio'))
sys.path.append(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'audio_detection'))
sys.path.append(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'mono2binaural'))
import gradio as gr
import matplotlib
import librosa
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
from diffusers import StableDiffusionPipeline
from langchain.agents.initialize import initialize_agent
from langchain.agents.tools import Tool
from langchain.chains.conversation.memory import ConversationBufferMemory
from langchain.llms.openai import OpenAI
import re
import uuid
import soundfile
from PIL import Image
import numpy as np
from omegaconf import OmegaConf
from transformers import pipeline, BlipProcessor, BlipForConditionalGeneration
from einops import repeat
from ldm.util import instantiate_from_config
from ldm.data.extract_mel_spectrogram import TRANSFORMS_16000
from vocoder.bigvgan.models import VocoderBigVGAN
from ldm.models.diffusion.ddim import DDIMSampler
from wav_evaluation.models.CLAPWrapper import CLAPWrapper
from audio_to_text.inference_waveform import AudioCapModel
import whisper
from inference.svs.ds_e2e import DiffSingerE2EInfer
from inference.tts.GenerSpeech import GenerSpeechInfer
from inference.tts.PortaSpeech import TTSInference
from utils.hparams import set_hparams
from utils.hparams import hparams as hp
import scipy.io.wavfile as wavfile
import librosa
from audio_infer.utils import config as detection_config
from audio_infer.pytorch.models import PVT
from src.models import BinauralNetwork
from sound_extraction.model.LASSNet import LASSNet
from sound_extraction.utils.stft import STFT
from sound_extraction.utils.wav_io import load_wav, save_wav
from target_sound_detection.src import models as tsd_models
from target_sound_detection.src.models import event_labels
from target_sound_detection.src.utils import median_filter, decode_with_timestamps
from espnet2.bin.svs_inference import SingingGenerate
import clip
import numpy as np
AUDIO_CHATGPT_PREFIX = """AudioGPT
AudioGPT can not directly read audios, but it has a list of tools to finish different speech, audio, and singing voice tasks. Each audio will have a file name formed as "audio/xxx.wav". When talking about audios, AudioGPT is very strict to the file name and will never fabricate nonexistent files.
AudioGPT is able to use tools in a sequence, and is loyal to the tool observation outputs rather than faking the audio content and audio file name. It will remember to provide the file name from the last tool observation, if a new audio is generated.
Human may provide new audios to AudioGPT with a description. The description helps AudioGPT to understand this audio, but AudioGPT should use tools to finish following tasks, rather than directly imagine from the description.
Overall, AudioGPT is a powerful audio dialogue assistant tool that can help with a wide range of tasks and provide valuable insights and information on a wide range of topics.
TOOLS:
------
AudioGPT has access to the following tools:"""
AUDIO_CHATGPT_FORMAT_INSTRUCTIONS = """To use a tool, please use the following format:
```
Thought: Do I need to use a tool? Yes
Action: the action to take, should be one of [{tool_names}]
Action Input: the input to the action
Observation: the result of the action
```
When you have a response to say to the Human, or if you do not need to use a tool, you MUST use the format:
```
Thought: Do I need to use a tool? No
{ai_prefix}: [your response here]
```
"""
AUDIO_CHATGPT_SUFFIX = """You are very strict to the filename correctness and will never fake a file name if not exists.
You will remember to provide the audio file name loyally if it's provided in the last tool observation.
Begin!
Previous conversation history:
{chat_history}
New input: {input}
Thought: Do I need to use a tool? {agent_scratchpad}"""
def cut_dialogue_history(history_memory, keep_last_n_words = 500):
tokens = history_memory.split()
n_tokens = len(tokens)
print(f"history_memory:{history_memory}, n_tokens: {n_tokens}")
if n_tokens < keep_last_n_words:
return history_memory
else:
paragraphs = history_memory.split('\n')
last_n_tokens = n_tokens
while last_n_tokens >= keep_last_n_words:
last_n_tokens = last_n_tokens - len(paragraphs[0].split(' '))
paragraphs = paragraphs[1:]
return '\n' + '\n'.join(paragraphs)
def initialize_model(config, ckpt, device):
config = OmegaConf.load(config)
model = instantiate_from_config(config.model)
model.load_state_dict(torch.load(ckpt,map_location='cpu')["state_dict"], strict=False)
model = model.to(device)
model.cond_stage_model.to(model.device)
model.cond_stage_model.device = model.device
sampler = DDIMSampler(model)
return sampler
def initialize_model_inpaint(config, ckpt):
config = OmegaConf.load(config)
model = instantiate_from_config(config.model)
model.load_state_dict(torch.load(ckpt,map_location='cpu')["state_dict"], strict=False)
device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
model = model.to(device)
print(model.device,device,model.cond_stage_model.device)
sampler = DDIMSampler(model)
return sampler
def select_best_audio(prompt,wav_list):
clap_model = CLAPWrapper('useful_ckpts/CLAP/CLAP_weights_2022.pth','useful_ckpts/CLAP/config.yml',use_cuda=torch.cuda.is_available())
text_embeddings = clap_model.get_text_embeddings([prompt])
score_list = []
for data in wav_list:
sr,wav = data
audio_embeddings = clap_model.get_audio_embeddings([(torch.FloatTensor(wav),sr)], resample=True)
score = clap_model.compute_similarity(audio_embeddings, text_embeddings,use_logit_scale=False).squeeze().cpu().numpy()
score_list.append(score)
max_index = np.array(score_list).argmax()
print(score_list,max_index)
return wav_list[max_index]
def merge_audio(audio_path_1, audio_path_2):
merged_signal = []
sr_1, signal_1 = wavfile.read(audio_path_1)
sr_2, signal_2 = wavfile.read(audio_path_2)
merged_signal.append(signal_1)
merged_signal.append(signal_2)
merged_signal = np.hstack(merged_signal)
merged_signal = np.asarray(merged_signal, dtype=np.int16)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
wavfile.write(audio_filename, sr_2, merged_signal)
return audio_filename
class T2I:
def __init__(self, device):
print("Initializing T2I to %s" % device)
self.device = device
self.pipe = StableDiffusionPipeline.from_pretrained("runwayml/stable-diffusion-v1-5", torch_dtype=torch.float16)
self.text_refine_tokenizer = AutoTokenizer.from_pretrained("Gustavosta/MagicPrompt-Stable-Diffusion")
self.text_refine_model = AutoModelForCausalLM.from_pretrained("Gustavosta/MagicPrompt-Stable-Diffusion")
self.text_refine_gpt2_pipe = pipeline("text-generation", model=self.text_refine_model, tokenizer=self.text_refine_tokenizer, device=self.device)
self.pipe.to(device)
def inference(self, text):
image_filename = os.path.join('image', str(uuid.uuid4())[0:8] + ".png")
refined_text = self.text_refine_gpt2_pipe(text)[0]["generated_text"]
print(f'{text} refined to {refined_text}')
image = self.pipe(refined_text).images[0]
image.save(image_filename)
print(f"Processed T2I.run, text: {text}, image_filename: {image_filename}")
return image_filename
class ImageCaptioning:
def __init__(self, device):
print("Initializing ImageCaptioning to %s" % device)
self.device = device
self.processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
self.model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base").to(self.device)
def inference(self, image_path):
inputs = self.processor(Image.open(image_path), return_tensors="pt").to(self.device)
out = self.model.generate(**inputs)
captions = self.processor.decode(out[0], skip_special_tokens=True)
return captions
class T2A:
def __init__(self, device):
print("Initializing Make-An-Audio to %s" % device)
self.device = device
self.sampler = initialize_model('text_to_audio/Make_An_Audio/configs/text_to_audio/txt2audio_args.yaml', 'text_to_audio/Make_An_Audio/useful_ckpts/ta40multi_epoch=000085.ckpt', device=device)
self.vocoder = VocoderBigVGAN('text_to_audio/Make_An_Audio/vocoder/logs/bigv16k53w',device=device)
def txt2audio(self, text, seed = 55, scale = 1.5, ddim_steps = 100, n_samples = 3, W = 624, H = 80):
SAMPLE_RATE = 16000
prng = np.random.RandomState(seed)
start_code = prng.randn(n_samples, self.sampler.model.first_stage_model.embed_dim, H // 8, W // 8)
start_code = torch.from_numpy(start_code).to(device=self.device, dtype=torch.float32)
uc = self.sampler.model.get_learned_conditioning(n_samples * [""])
c = self.sampler.model.get_learned_conditioning(n_samples * [text])
shape = [self.sampler.model.first_stage_model.embed_dim, H//8, W//8] # (z_dim, 80//2^x, 848//2^x)
samples_ddim, _ = self.sampler.sample(S = ddim_steps,
conditioning = c,
batch_size = n_samples,
shape = shape,
verbose = False,
unconditional_guidance_scale = scale,
unconditional_conditioning = uc,
x_T = start_code)
x_samples_ddim = self.sampler.model.decode_first_stage(samples_ddim)
x_samples_ddim = torch.clamp((x_samples_ddim+1.0)/2.0, min=0.0, max=1.0) # [0, 1]
wav_list = []
for idx,spec in enumerate(x_samples_ddim):
wav = self.vocoder.vocode(spec)
wav_list.append((SAMPLE_RATE,wav))
best_wav = select_best_audio(text, wav_list)
return best_wav
def inference(self, text, seed = 55, scale = 1.5, ddim_steps = 100, n_samples = 3, W = 624, H = 80):
melbins,mel_len = 80,624
with torch.no_grad():
result = self.txt2audio(
text = text,
H = melbins,
W = mel_len
)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
soundfile.write(audio_filename, result[1], samplerate = 16000)
print(f"Processed T2I.run, text: {text}, audio_filename: {audio_filename}")
return audio_filename
class I2A:
def __init__(self, device):
print("Initializing Make-An-Audio-Image to %s" % device)
self.device = device
self.sampler = initialize_model('text_to_audio/Make_An_Audio_img/configs/img_to_audio/img2audio_args.yaml', 'text_to_audio/Make_An_Audio_img/useful_ckpts/ta54_epoch=000216.ckpt', device=device)
self.vocoder = VocoderBigVGAN('text_to_audio/Make_An_Audio_img/vocoder/logs/bigv16k53w',device=device)
def img2audio(self, image, seed = 55, scale = 3, ddim_steps = 100, W = 624, H = 80):
SAMPLE_RATE = 16000
n_samples = 1 # only support 1 sample
prng = np.random.RandomState(seed)
start_code = prng.randn(n_samples, self.sampler.model.first_stage_model.embed_dim, H // 8, W // 8)
start_code = torch.from_numpy(start_code).to(device=self.device, dtype=torch.float32)
uc = self.sampler.model.get_learned_conditioning(n_samples * [""])
#image = Image.fromarray(image)
image = Image.open(image)
image = self.sampler.model.cond_stage_model.preprocess(image).unsqueeze(0)
image_embedding = self.sampler.model.cond_stage_model.forward_img(image)
c = image_embedding.repeat(n_samples, 1, 1)
shape = [self.sampler.model.first_stage_model.embed_dim, H//8, W//8] # (z_dim, 80//2^x, 848//2^x)
samples_ddim, _ = self.sampler.sample(S=ddim_steps,
conditioning=c,
batch_size=n_samples,
shape=shape,
verbose=False,
unconditional_guidance_scale=scale,
unconditional_conditioning=uc,
x_T=start_code)
x_samples_ddim = self.sampler.model.decode_first_stage(samples_ddim)
x_samples_ddim = torch.clamp((x_samples_ddim+1.0)/2.0, min=0.0, max=1.0) # [0, 1]
wav_list = []
for idx,spec in enumerate(x_samples_ddim):
wav = self.vocoder.vocode(spec)
wav_list.append((SAMPLE_RATE,wav))
best_wav = wav_list[0]
return best_wav
def inference(self, image, seed = 55, scale = 3, ddim_steps = 100, W = 624, H = 80):
melbins,mel_len = 80,624
with torch.no_grad():
result = self.img2audio(
image=image,
H=melbins,
W=mel_len
)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
soundfile.write(audio_filename, result[1], samplerate = 16000)
print(f"Processed I2a.run, image_filename: {image}, audio_filename: {audio_filename}")
return audio_filename
class TTS:
def __init__(self, device=None):
if device is None:
device = 'cuda' if torch.cuda.is_available() else 'cpu'
print("Initializing PortaSpeech to %s" % device)
self.device = device
self.exp_name = 'checkpoints/ps_adv_baseline'
self.set_model_hparams()
self.inferencer = TTSInference(self.hp, device)
def set_model_hparams(self):
set_hparams(exp_name=self.exp_name, print_hparams=False)
self.hp = hp
def inference(self, text):
self.set_model_hparams()
inp = {"text": text}
out = self.inferencer.infer_once(inp)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
soundfile.write(audio_filename, out, samplerate=22050)
return audio_filename
class T2S:
def __init__(self, device= None):
if device is None:
device = 'cuda' if torch.cuda.is_available() else 'cpu'
print("Initializing DiffSinger to %s" % device)
self.device = device
self.exp_name = 'checkpoints/0831_opencpop_ds1000'
self.config= 'NeuralSeq/egs/egs_bases/svs/midi/e2e/opencpop/ds1000.yaml'
self.set_model_hparams()
self.pipe = DiffSingerE2EInfer(self.hp, device)
self.default_inp = {
'text': '你 说 你 不 SP 懂 为 何 在 这 时 牵 手 AP',
'notes': 'D#4/Eb4 | D#4/Eb4 | D#4/Eb4 | D#4/Eb4 | rest | D#4/Eb4 | D4 | D4 | D4 | D#4/Eb4 | F4 | D#4/Eb4 | D4 | rest',
'notes_duration': '0.113740 | 0.329060 | 0.287950 | 0.133480 | 0.150900 | 0.484730 | 0.242010 | 0.180820 | 0.343570 | 0.152050 | 0.266720 | 0.280310 | 0.633300 | 0.444590'
}
def set_model_hparams(self):
set_hparams(config=self.config, exp_name=self.exp_name, print_hparams=False)
self.hp = hp
def inference(self, inputs):
self.set_model_hparams()
val = inputs.split(",")
key = ['text', 'notes', 'notes_duration']
try:
inp = {k: v for k, v in zip(key, val)}
wav = self.pipe.infer_once(inp)
except:
print('Error occurs. Generate default audio sample.\n')
inp = self.default_inp
wav = self.pipe.infer_once(inp)
#if inputs == '' or len(val) < len(key):
# inp = self.default_inp
#else:
# inp = {k:v for k,v in zip(key,val)}
#wav = self.pipe.infer_once(inp)
wav *= 32767
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
wavfile.write(audio_filename, self.hp['audio_sample_rate'], wav.astype(np.int16))
print(f"Processed T2S.run, audio_filename: {audio_filename}")
return audio_filename
class t2s_VISinger:
def __init__(self, device=None):
if device is None:
device = 'cuda' if torch.cuda.is_available() else 'cpu'
print("Initializing VISingere to %s" % device)
tag = 'AQuarterMile/opencpop_visinger1'
self.model = SingingGenerate.from_pretrained(
model_tag=str_or_none(tag),
device=device,
)
phn_dur = [[0. , 0.219 ],
[0.219 , 0.50599998],
[0.50599998, 0.71399999],
[0.71399999, 1.097 ],
[1.097 , 1.28799999],
[1.28799999, 1.98300004],
[1.98300004, 7.10500002],
[7.10500002, 7.60400009]]
phn = ['sh', 'i', 'q', 'v', 'n', 'i', 'SP', 'AP']
score = [[0, 0.50625, 'sh_i', 58, 'sh_i'], [0.50625, 1.09728, 'q_v', 56, 'q_v'], [1.09728, 1.9832100000000001, 'n_i', 53, 'n_i'], [1.9832100000000001, 7.105360000000001, 'SP', 0, 'SP'], [7.105360000000001, 7.604390000000001, 'AP', 0, 'AP']]
tempo = 70
tmp = {}
tmp["label"] = phn_dur, phn
tmp["score"] = tempo, score
self.default_inp = tmp
def inference(self, inputs):
val = inputs.split(",")
key = ['text', 'notes', 'notes_duration']
try: # TODO: input will be update
inp = {k: v for k, v in zip(key, val)}
wav = self.model(text=inp)["wav"]
except:
print('Error occurs. Generate default audio sample.\n')
inp = self.default_inp
wav = self.model(text=inp)["wav"]
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
soundfile.write(audio_filename, wav, samplerate=self.model.fs)
return audio_filename
class TTS_OOD:
def __init__(self, device):
if device is None:
device = 'cuda' if torch.cuda.is_available() else 'cpu'
print("Initializing GenerSpeech to %s" % device)
self.device = device
self.exp_name = 'checkpoints/GenerSpeech'
self.config = 'NeuralSeq/modules/GenerSpeech/config/generspeech.yaml'
self.set_model_hparams()
self.pipe = GenerSpeechInfer(self.hp, device)
def set_model_hparams(self):
set_hparams(config=self.config, exp_name=self.exp_name, print_hparams=False)
f0_stats_fn = f'{hp["binary_data_dir"]}/train_f0s_mean_std.npy'
if os.path.exists(f0_stats_fn):
hp['f0_mean'], hp['f0_std'] = np.load(f0_stats_fn)
hp['f0_mean'] = float(hp['f0_mean'])
hp['f0_std'] = float(hp['f0_std'])
hp['emotion_encoder_path'] = 'checkpoints/Emotion_encoder.pt'
self.hp = hp
def inference(self, inputs):
self.set_model_hparams()
key = ['ref_audio', 'text']
val = inputs.split(",")
inp = {k: v for k, v in zip(key, val)}
wav = self.pipe.infer_once(inp)
wav *= 32767
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
wavfile.write(audio_filename, self.hp['audio_sample_rate'], wav.astype(np.int16))
print(
f"Processed GenerSpeech.run. Input text:{val[1]}. Input reference audio: {val[0]}. Output Audio_filename: {audio_filename}")
return audio_filename
class Inpaint:
def __init__(self, device):
print("Initializing Make-An-Audio-inpaint to %s" % device)
self.device = device
self.sampler = initialize_model_inpaint('text_to_audio/Make_An_Audio_inpaint/configs/inpaint/txt2audio_args.yaml', 'text_to_audio/Make_An_Audio_inpaint/useful_ckpts/inpaint7_epoch00047.ckpt')
self.vocoder = VocoderBigVGAN('./vocoder/logs/bigv16k53w',device=device)
self.cmap_transform = matplotlib.cm.viridis
def make_batch_sd(self, mel, mask, num_samples=1):
mel = torch.from_numpy(mel)[None,None,...].to(dtype=torch.float32)
mask = torch.from_numpy(mask)[None,None,...].to(dtype=torch.float32)
masked_mel = (1 - mask) * mel
mel = mel * 2 - 1
mask = mask * 2 - 1
masked_mel = masked_mel * 2 -1
batch = {
"mel": repeat(mel.to(device=self.device), "1 ... -> n ...", n=num_samples),
"mask": repeat(mask.to(device=self.device), "1 ... -> n ...", n=num_samples),
"masked_mel": repeat(masked_mel.to(device=self.device), "1 ... -> n ...", n=num_samples),
}
return batch
def gen_mel(self, input_audio_path):
SAMPLE_RATE = 16000
sr, ori_wav = wavfile.read(input_audio_path)
print("gen_mel")
print(sr,ori_wav.shape,ori_wav)
ori_wav = ori_wav.astype(np.float32, order='C') / 32768.0
if len(ori_wav.shape)==2:# stereo
ori_wav = librosa.to_mono(ori_wav.T)
print(sr,ori_wav.shape,ori_wav)
ori_wav = librosa.resample(ori_wav,orig_sr = sr,target_sr = SAMPLE_RATE)
mel_len,hop_size = 848,256
input_len = mel_len * hop_size
if len(ori_wav) < input_len:
input_wav = np.pad(ori_wav,(0,mel_len*hop_size),constant_values=0)
else:
input_wav = ori_wav[:input_len]
mel = TRANSFORMS_16000(input_wav)
return mel
def gen_mel_audio(self, input_audio):
SAMPLE_RATE = 16000
sr,ori_wav = input_audio
print("gen_mel_audio")
print(sr,ori_wav.shape,ori_wav)
ori_wav = ori_wav.astype(np.float32, order='C') / 32768.0
if len(ori_wav.shape)==2:# stereo
ori_wav = librosa.to_mono(ori_wav.T)
print(sr,ori_wav.shape,ori_wav)
ori_wav = librosa.resample(ori_wav,orig_sr = sr,target_sr = SAMPLE_RATE)
mel_len,hop_size = 848,256
input_len = mel_len * hop_size
if len(ori_wav) < input_len:
input_wav = np.pad(ori_wav,(0,mel_len*hop_size),constant_values=0)
else:
input_wav = ori_wav[:input_len]
mel = TRANSFORMS_16000(input_wav)
return mel
def show_mel_fn(self, input_audio_path):
crop_len = 500
crop_mel = self.gen_mel(input_audio_path)[:,:crop_len]
color_mel = self.cmap_transform(crop_mel)
image = Image.fromarray((color_mel*255).astype(np.uint8))
image_filename = os.path.join('image', str(uuid.uuid4())[0:8] + ".png")
image.save(image_filename)
return image_filename
def inpaint(self, batch, seed, ddim_steps, num_samples=1, W=512, H=512):
model = self.sampler.model
prng = np.random.RandomState(seed)
start_code = prng.randn(num_samples, model.first_stage_model.embed_dim, H // 8, W // 8)
start_code = torch.from_numpy(start_code).to(device=self.device, dtype=torch.float32)
c = model.get_first_stage_encoding(model.encode_first_stage(batch["masked_mel"]))
cc = torch.nn.functional.interpolate(batch["mask"],
size=c.shape[-2:])
c = torch.cat((c, cc), dim=1) # (b,c+1,h,w) 1 is mask
shape = (c.shape[1]-1,)+c.shape[2:]
samples_ddim, _ = self.sampler.sample(S=ddim_steps,
conditioning=c,
batch_size=c.shape[0],
shape=shape,
verbose=False)
x_samples_ddim = model.decode_first_stage(samples_ddim)
mel = torch.clamp((batch["mel"]+1.0)/2.0,min=0.0, max=1.0)
mask = torch.clamp((batch["mask"]+1.0)/2.0,min=0.0, max=1.0)
predicted_mel = torch.clamp((x_samples_ddim+1.0)/2.0,min=0.0, max=1.0)
inpainted = (1-mask)*mel+mask*predicted_mel
inpainted = inpainted.cpu().numpy().squeeze()
inapint_wav = self.vocoder.vocode(inpainted)
return inpainted, inapint_wav
def inference(self, input_audio, mel_and_mask, seed = 55, ddim_steps = 100):
SAMPLE_RATE = 16000
torch.set_grad_enabled(False)
mel_img = Image.open(mel_and_mask['image'])
mask_img = Image.open(mel_and_mask["mask"])
show_mel = np.array(mel_img.convert("L"))/255
mask = np.array(mask_img.convert("L"))/255
mel_bins,mel_len = 80,848
input_mel = self.gen_mel_audio(input_audio)[:,:mel_len]
mask = np.pad(mask,((0,0),(0,mel_len-mask.shape[1])),mode='constant',constant_values=0)
print(mask.shape,input_mel.shape)
with torch.no_grad():
batch = self.make_batch_sd(input_mel,mask,num_samples=1)
inpainted,gen_wav = self.inpaint(
batch=batch,
seed=seed,
ddim_steps=ddim_steps,
num_samples=1,
H=mel_bins, W=mel_len
)
inpainted = inpainted[:,:show_mel.shape[1]]
color_mel = self.cmap_transform(inpainted)
input_len = int(input_audio[1].shape[0] * SAMPLE_RATE / input_audio[0])
gen_wav = (gen_wav * 32768).astype(np.int16)[:input_len]
image = Image.fromarray((color_mel*255).astype(np.uint8))
image_filename = os.path.join('image', str(uuid.uuid4())[0:8] + ".png")
image.save(image_filename)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
soundfile.write(audio_filename, gen_wav, samplerate = 16000)
return image_filename, audio_filename
class ASR:
def __init__(self, device):
print("Initializing Whisper to %s" % device)
self.device = device
self.model = whisper.load_model("base", device=device)
def inference(self, audio_path):
audio = whisper.load_audio(audio_path)
audio = whisper.pad_or_trim(audio)
mel = whisper.log_mel_spectrogram(audio).to(self.device)
_, probs = self.model.detect_language(mel)
options = whisper.DecodingOptions()
result = whisper.decode(self.model, mel, options)
return result.text
def translate_english(self, audio_path):
audio = self.model.transcribe(audio_path, language='English')
return audio['text']
class A2T:
def __init__(self, device):
print("Initializing Audio-To-Text Model to %s" % device)
self.device = device
self.model = AudioCapModel("audio_to_text/audiocaps_cntrstv_cnn14rnn_trm")
def inference(self, audio_path):
audio = whisper.load_audio(audio_path)
caption_text = self.model(audio)
return caption_text[0]
class GeneFace:
def __init__(self, device=None):
print("Initializing GeneFace model to %s" % device)
from audio_to_face.GeneFace_binding import GeneFaceInfer
if device is None:
device = 'cuda' if torch.cuda.is_available() else 'cpu'
self.device = device
self.geneface_model = GeneFaceInfer(device)
print("Loaded GeneFace model")
def inference(self, audio_path):
audio_base_name = os.path.basename(audio_path)[:-4]
out_video_name = audio_path.replace("audio","video").replace(".wav", ".mp4")
inp = {
'audio_source_name': audio_path,
'out_npy_name': f'geneface/tmp/{audio_base_name}.npy',
'cond_name': f'geneface/tmp/{audio_base_name}.npy',
'out_video_name': out_video_name,
'tmp_imgs_dir': f'video/tmp_imgs',
}
self.geneface_model.infer_once(inp)
return out_video_name
class SoundDetection:
def __init__(self, device):
self.device = device
self.sample_rate = 32000
self.window_size = 1024
self.hop_size = 320
self.mel_bins = 64
self.fmin = 50
self.fmax = 14000
self.model_type = 'PVT'
self.checkpoint_path = 'audio_detection/audio_infer/useful_ckpts/audio_detection.pth'
self.classes_num = detection_config.classes_num
self.labels = detection_config.labels
self.frames_per_second = self.sample_rate // self.hop_size
# Model = eval(self.model_type)
self.model = PVT(sample_rate=self.sample_rate, window_size=self.window_size,
hop_size=self.hop_size, mel_bins=self.mel_bins, fmin=self.fmin, fmax=self.fmax,
classes_num=self.classes_num)
checkpoint = torch.load(self.checkpoint_path, map_location=self.device)
self.model.load_state_dict(checkpoint['model'])
self.model.to(device)
def inference(self, audio_path):
# Forward
(waveform, _) = librosa.core.load(audio_path, sr=self.sample_rate, mono=True)
waveform = waveform[None, :] # (1, audio_length)
waveform = torch.from_numpy(waveform)
waveform = waveform.to(self.device)
# Forward
with torch.no_grad():
self.model.eval()
batch_output_dict = self.model(waveform, None)
framewise_output = batch_output_dict['framewise_output'].data.cpu().numpy()[0]
"""(time_steps, classes_num)"""
# print('Sound event detection result (time_steps x classes_num): {}'.format(
# framewise_output.shape))
import numpy as np
import matplotlib.pyplot as plt
sorted_indexes = np.argsort(np.max(framewise_output, axis=0))[::-1]
top_k = 10 # Show top results
top_result_mat = framewise_output[:, sorted_indexes[0 : top_k]]
"""(time_steps, top_k)"""
# Plot result
stft = librosa.core.stft(y=waveform[0].data.cpu().numpy(), n_fft=self.window_size,
hop_length=self.hop_size, window='hann', center=True)
frames_num = stft.shape[-1]
fig, axs = plt.subplots(2, 1, sharex=True, figsize=(10, 4))
axs[0].matshow(np.log(np.abs(stft)), origin='lower', aspect='auto', cmap='jet')
axs[0].set_ylabel('Frequency bins')
axs[0].set_title('Log spectrogram')
axs[1].matshow(top_result_mat.T, origin='upper', aspect='auto', cmap='jet', vmin=0, vmax=1)
axs[1].xaxis.set_ticks(np.arange(0, frames_num, self.frames_per_second))
axs[1].xaxis.set_ticklabels(np.arange(0, frames_num / self.frames_per_second))
axs[1].yaxis.set_ticks(np.arange(0, top_k))
axs[1].yaxis.set_ticklabels(np.array(self.labels)[sorted_indexes[0 : top_k]])
axs[1].yaxis.grid(color='k', linestyle='solid', linewidth=0.3, alpha=0.3)
axs[1].set_xlabel('Seconds')
axs[1].xaxis.set_ticks_position('bottom')
plt.tight_layout()
image_filename = os.path.join('image', str(uuid.uuid4())[0:8] + ".png")
plt.savefig(image_filename)
return image_filename
class SoundExtraction:
def __init__(self, device):
self.device = device
self.model_file = 'sound_extraction/useful_ckpts/LASSNet.pt'
self.stft = STFT()
import torch.nn as nn
self.model = nn.DataParallel(LASSNet(device)).to(device)
checkpoint = torch.load(self.model_file)
self.model.load_state_dict(checkpoint['model'])
self.model.eval()
def inference(self, inputs):
#key = ['ref_audio', 'text']
val = inputs.split(",")
audio_path = val[0] # audio_path, text
text = val[1]
waveform = load_wav(audio_path)
waveform = torch.tensor(waveform).transpose(1,0)
mixed_mag, mixed_phase = self.stft.transform(waveform)
text_query = ['[CLS] ' + text]
mixed_mag = mixed_mag.transpose(2,1).unsqueeze(0).to(self.device)
est_mask = self.model(mixed_mag, text_query)
est_mag = est_mask * mixed_mag
est_mag = est_mag.squeeze(1)
est_mag = est_mag.permute(0, 2, 1)
est_wav = self.stft.inverse(est_mag.cpu().detach(), mixed_phase)
est_wav = est_wav.squeeze(0).squeeze(0).numpy()
#est_path = f'output/est{i}.wav'
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
print('audio_filename ', audio_filename)
save_wav(est_wav, audio_filename)
return audio_filename
class Binaural:
def __init__(self, device):
self.device = device
self.model_file = 'mono2binaural/useful_ckpts/m2b/binaural_network.net'
self.position_file = ['mono2binaural/useful_ckpts/m2b/tx_positions.txt',
'mono2binaural/useful_ckpts/m2b/tx_positions2.txt',
'mono2binaural/useful_ckpts/m2b/tx_positions3.txt',
'mono2binaural/useful_ckpts/m2b/tx_positions4.txt',
'mono2binaural/useful_ckpts/m2b/tx_positions5.txt']
self.net = BinauralNetwork(view_dim=7,
warpnet_layers=4,
warpnet_channels=64,
)
self.net.load_from_file(self.model_file)
self.sr = 48000
def inference(self, audio_path):
mono, sr = librosa.load(path=audio_path, sr=self.sr, mono=True)
mono = torch.from_numpy(mono)
mono = mono.unsqueeze(0)
import numpy as np
import random
rand_int = random.randint(0,4)
view = np.loadtxt(self.position_file[rand_int]).transpose().astype(np.float32)
view = torch.from_numpy(view)
if not view.shape[-1] * 400 == mono.shape[-1]:
mono = mono[:,:(mono.shape[-1]//400)*400] #
if view.shape[1]*400 > mono.shape[1]:
m_a = view.shape[1] - mono.shape[-1]//400
rand_st = random.randint(0,m_a)
view = view[:,m_a:m_a+(mono.shape[-1]//400)] #
# binauralize and save output
self.net.eval().to(self.device)
mono, view = mono.to(self.device), view.to(self.device)
chunk_size = 48000 # forward in chunks of 1s
rec_field = 1000 # add 1000 samples as "safe bet" since warping has undefined rec. field
rec_field -= rec_field % 400 # make sure rec_field is a multiple of 400 to match audio and view frequencies
chunks = [
{
"mono": mono[:, max(0, i-rec_field):i+chunk_size],
"view": view[:, max(0, i-rec_field)//400:(i+chunk_size)//400]
}
for i in range(0, mono.shape[-1], chunk_size)
]
for i, chunk in enumerate(chunks):
with torch.no_grad():
mono = chunk["mono"].unsqueeze(0)
view = chunk["view"].unsqueeze(0)
binaural = self.net(mono, view).squeeze(0)
if i > 0:
binaural = binaural[:, -(mono.shape[-1]-rec_field):]
chunk["binaural"] = binaural
binaural = torch.cat([chunk["binaural"] for chunk in chunks], dim=-1)
binaural = torch.clamp(binaural, min=-1, max=1).cpu()
#binaural = chunked_forwarding(net, mono, view)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
import torchaudio
torchaudio.save(audio_filename, binaural, sr)
#soundfile.write(audio_filename, binaural, samplerate = 48000)
print(f"Processed Binaural.run, audio_filename: {audio_filename}")
return audio_filename
class TargetSoundDetection:
def __init__(self, device):
self.device = device
self.MEL_ARGS = {
'n_mels': 64,
'n_fft': 2048,
'hop_length': int(22050 * 20 / 1000),
'win_length': int(22050 * 40 / 1000)
}
self.EPS = np.spacing(1)
self.clip_model, _ = clip.load("ViT-B/32", device=self.device)
self.event_labels = event_labels
self.id_to_event = {i : label for i, label in enumerate(self.event_labels)}
config = torch.load('audio_detection/target_sound_detection/useful_ckpts/tsd/run_config.pth', map_location='cpu')
config_parameters = dict(config)
config_parameters['tao'] = 0.6
if 'thres' not in config_parameters.keys():
config_parameters['thres'] = 0.5
if 'time_resolution' not in config_parameters.keys():
config_parameters['time_resolution'] = 125
model_parameters = torch.load('audio_detection/target_sound_detection/useful_ckpts/tsd/run_model_7_loss=-0.0724.pt'
, map_location=lambda storage, loc: storage) # load parameter
self.model = getattr(tsd_models, config_parameters['model'])(config_parameters,
inputdim=64, outputdim=2, time_resolution=config_parameters['time_resolution'], **config_parameters['model_args'])
self.model.load_state_dict(model_parameters)
self.model = self.model.to(self.device).eval()
self.re_embeds = torch.load('audio_detection/target_sound_detection/useful_ckpts/tsd/text_emb.pth')
self.ref_mel = torch.load('audio_detection/target_sound_detection/useful_ckpts/tsd/ref_mel.pth')
def extract_feature(self, fname):
import soundfile as sf
y, sr = sf.read(fname, dtype='float32')
print('y ', y.shape)
ti = y.shape[0]/sr
if y.ndim > 1:
y = y.mean(1)
y = librosa.resample(y, sr, 22050)
lms_feature = np.log(librosa.feature.melspectrogram(y, **self.MEL_ARGS) + self.EPS).T
return lms_feature,ti
def build_clip(self, text):
text = clip.tokenize(text).to(self.device) # ["a diagram with dog", "a dog", "a cat"]
text_features = self.clip_model.encode_text(text)
return text_features
def cal_similarity(self, target, retrievals):
ans = []
#target =torch.from_numpy(target)
for name in retrievals.keys():
tmp = retrievals[name]
#tmp = torch.from_numpy(tmp)
s = torch.cosine_similarity(target.squeeze(), tmp.squeeze(), dim=0)
ans.append(s.item())
return ans.index(max(ans))
def inference(self, text, audio_path):
target_emb = self.build_clip(text) # torch type
idx = self.cal_similarity(target_emb, self.re_embeds)
target_event = self.id_to_event[idx]
embedding = self.ref_mel[target_event]
embedding = torch.from_numpy(embedding)
embedding = embedding.unsqueeze(0).to(self.device).float()
#print('embedding ', embedding.shape)
inputs,ti = self.extract_feature(audio_path)
#print('ti ', ti)
inputs = torch.from_numpy(inputs)
inputs = inputs.unsqueeze(0).to(self.device).float()
#print('inputs ', inputs.shape)
decision, decision_up, logit = self.model(inputs, embedding)
pred = decision_up.detach().cpu().numpy()
pred = pred[:,:,0]
frame_num = decision_up.shape[1]
time_ratio = ti / frame_num
filtered_pred = median_filter(pred, window_size=1, threshold=0.5)
#print('filtered_pred ', filtered_pred)
time_predictions = []
for index_k in range(filtered_pred.shape[0]):
decoded_pred = []
decoded_pred_ = decode_with_timestamps(target_event, filtered_pred[index_k,:])
if len(decoded_pred_) == 0: # neg deal
decoded_pred_.append((target_event, 0, 0))
decoded_pred.append(decoded_pred_)
for num_batch in range(len(decoded_pred)): # when we test our model,the batch_size is 1
cur_pred = pred[num_batch]
# Save each frame output, for later visualization
label_prediction = decoded_pred[num_batch] # frame predict
# print(label_prediction)
for event_label, onset, offset in label_prediction:
time_predictions.append({
'onset': onset*time_ratio,
'offset': offset*time_ratio,})
ans = ''
for i,item in enumerate(time_predictions):
ans = ans + 'segment' + str(i+1) + ' start_time: ' + str(item['onset']) + ' end_time: ' + str(item['offset']) + '\t'
#print(ans)
return ans
class Speech_Enh_SS_SC:
"""Speech Enhancement or Separation in single-channel
Example usage:
enh_model = Speech_Enh_SS("cuda")
enh_wav = enh_model.inference("./test_chime4_audio_M05_440C0213_PED_REAL.wav")
"""
def __init__(self, device="cuda", model_name="espnet/Wangyou_Zhang_chime4_enh_train_enh_conv_tasnet_raw"):
self.model_name = model_name
self.device = device
print("Initializing ESPnet Enh to %s" % device)
self._initialize_model()
def _initialize_model(self):
from espnet_model_zoo.downloader import ModelDownloader
from espnet2.bin.enh_inference import SeparateSpeech
d = ModelDownloader()
cfg = d.download_and_unpack(self.model_name)
self.separate_speech = SeparateSpeech(
train_config=cfg["train_config"],
model_file=cfg["model_file"],
# for segment-wise process on long speech
segment_size=2.4,
hop_size=0.8,
normalize_segment_scale=False,
show_progressbar=True,
ref_channel=None,
normalize_output_wav=True,
device=self.device,
)
def inference(self, speech_path, ref_channel=0):
speech, sr = soundfile.read(speech_path)
speech = speech[:, ref_channel]
# speech = torch.from_numpy(speech)
# assert speech.dim() == 1
enh_speech = self.separate_speech(speech[None, ...], fs=sr)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
# if len(enh_speech) == 1:
soundfile.write(audio_filename, enh_speech[0].squeeze(), samplerate=sr)
# return enh_speech[0]
# return enh_speech
# else:
# print("############")
# audio_filename_1 = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
# soundfile.write(audio_filename_1, enh_speech[0].squeeze(), samplerate=sr)
# audio_filename_2 = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
# soundfile.write(audio_filename_2, enh_speech[1].squeeze(), samplerate=sr)
# audio_filename = merge_audio(audio_filename_1, audio_filename_2)
return audio_filename
class Speech_SS:
def __init__(self, device="cuda", model_name="lichenda/wsj0_2mix_skim_noncausal"):
self.model_name = model_name
self.device = device
print("Initializing ESPnet SS to %s" % device)
self._initialize_model()
def _initialize_model(self):
from espnet_model_zoo.downloader import ModelDownloader
from espnet2.bin.enh_inference import SeparateSpeech
d = ModelDownloader()
cfg = d.download_and_unpack(self.model_name)
self.separate_speech = SeparateSpeech(
train_config=cfg["train_config"],
model_file=cfg["model_file"],
# for segment-wise process on long speech
segment_size=2.4,
hop_size=0.8,
normalize_segment_scale=False,
show_progressbar=True,
ref_channel=None,
normalize_output_wav=True,
device=self.device,
)
def inference(self, speech_path):
speech, sr = soundfile.read(speech_path)
enh_speech = self.separate_speech(speech[None, ...], fs=sr)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
if len(enh_speech) == 1:
soundfile.write(audio_filename, enh_speech[0], samplerate=sr)
else:
# print("############")
audio_filename_1 = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
soundfile.write(audio_filename_1, enh_speech[0].squeeze(), samplerate=sr)
audio_filename_2 = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
soundfile.write(audio_filename_2, enh_speech[1].squeeze(), samplerate=sr)
audio_filename = merge_audio(audio_filename_1, audio_filename_2)
return audio_filename
class Speech_Enh_SS_MC:
"""Speech Enhancement or Separation in multi-channel"""
def __init__(self, device="cuda", model_name="espnet/Wangyou_Zhang_chime4_enh_train_enh_dc_crn_mapping_snr_raw", ref_channel=4):
self.model_name = model_name
self.ref_channel = ref_channel
self.device = device
print("Initializing ESPnet Enh to %s" % device)
self._initialize_model()
def _initialize_model(self):
from espnet_model_zoo.downloader import ModelDownloader
from espnet2.bin.enh_inference import SeparateSpeech
d = ModelDownloader()
cfg = d.download_and_unpack(self.model_name)
self.separate_speech = SeparateSpeech(
train_config=cfg["train_config"],
model_file=cfg["model_file"],
# for segment-wise process on long speech
segment_size=2.4,
hop_size=0.8,
normalize_segment_scale=False,
show_progressbar=True,
ref_channel=self.ref_channel,
normalize_output_wav=True,
device=self.device,
)
def inference(self, speech_path):
speech, sr = soundfile.read(speech_path)
speech = speech.T
print(speech[None, ...])
enh_speech = self.separate_speech(speech[None, ...], fs=sr)
audio_filename = os.path.join('audio', str(uuid.uuid4())[0:8] + ".wav")
# if len(enh_speech) == 1:
soundfile.write(audio_filename, enh_speech[0].squeeze(), samplerate=sr)
# return enh_speech[0]
# return enh_speech
# else:
# soundfile.write(audio_filename, enh_speech, samplerate=sr)
return audio_filename
class ConversationBot:
def __init__(self):
print("Initializing AudioGPT")
self.llm = OpenAI(temperature=0)
self.t2i = T2I(device="cuda:1")
self.i2t = ImageCaptioning(device="cuda:0")
self.t2a = T2A(device="cuda:0")
self.tts = TTS(device="cpu")
self.t2s = T2S(device="cpu")
self.i2a = I2A(device="cuda:0")
self.a2t = A2T(device="cpu")
self.asr = ASR(device="cuda:0")