Hugging Face's logo

anvuew
/
dereverb_room

Model card Files
xet
 Community
dereverb_room
File size: 1,991 Bytes
97e7a63
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1d8af0b
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
audio:
  chunk_size: 384000
  dim_f: 1024
  dim_t: 801 # don't work (use in model)
  hop_length: 441 # don't work (use in model)
  n_fft: 2048
  num_channels: 1
  sample_rate: 44100
  min_mean_abs: 0.000

model:
  dim: 128
  depth: 16
  stereo: false
  num_stems: 1
  time_transformer_depth: 1
  freq_transformer_depth: 1
  linear_transformer_depth: 0
  freqs_per_bands: !!python/tuple
    - 2
    - 2
    - 2
    - 2
    - 2
    - 2
    - 3
    - 3
    - 3
    - 3
    - 3
    - 4
    - 4
    - 4
    - 4
    - 4
    - 5
    - 5
    - 5
    - 5
    - 6
    - 6
    - 6
    - 6
    - 7
    - 7
    - 7
    - 8
    - 8
    - 8
    - 9
    - 9
    - 10
    - 10
    - 11
    - 12
    - 13
    - 14
    - 15
    - 16
    - 17
    - 18
    - 19
    - 20
    - 21
    - 22
    - 23
    - 24
    - 25
    - 27
    - 29
    - 31
    - 33
    - 35
    - 37
    - 39
    - 41
    - 43
    - 45
    - 48
    - 52
    - 57
    - 64
  dim_head: 16
  heads: 8
  attn_dropout: 0.0
  ff_dropout: 0.0
  flash_attn: true
  dim_freqs_in: 1025
  stft_n_fft: 2048
  stft_hop_length: 512
  stft_win_length: 2048
  stft_normalized: False
  mask_estimator_depth: 3
  multi_stft_resolution_loss_weight: 1.0
  multi_stft_resolutions_window_sizes: !!python/tuple
  - 4096
  - 2048
  - 1024
  - 512
  - 256
  multi_stft_hop_size: 147
  multi_stft_normalized: False
  mlp_expansion_factor: 4
  use_torch_checkpoint: True 
  skip_connection: False 


training:
  batch_size: 4
  gradient_accumulation_steps: 1
  grad_clip: 1000.0
  instruments: ['noreverb', 'reverb']
  lr: 5.0e-5
  patience: 5
  reduce_factor: 0.75
  target_instrument: noreverb
  num_epochs: 1000
  num_steps: 1000
  q: 0.95
  coarse_loss_clip: true
  ema_momentum: 0.999
  optimizer: adam
  
  other_fix: False # it's needed for checking on multisong dataset if other is actually instrumental
  use_amp: true # enable or disable usage of mixed precision (float16) - usually it must be true



inference:
  batch_size: 1
  dim_t: 871
  num_overlap: 2