Hugging Face's logo

ayousanz
/
piper-ljspeech

Text-to-Speech
TensorBoard
ONNX
English
tts
piper
vits
ljspeech
Model card Files
xet
Metrics Community
ayousanz commited on
Commit
bfd65f0
·
verified ·
1 Parent(s): fc77535

Upload folder using huggingface_hub

Browse files
Files changed (6) hide show
  1. version_0/checkpoints/epoch=1-step=3112.ckpt +3 -0
  2. version_0/events.out.tfevents.1747994945.piper-0.6987.0 +3 -0
  3. version_0/hparams.yaml +126 -0
  4. version_1/checkpoints/epoch=499-step=519000.ckpt +3 -0
  5. version_1/events.out.tfevents.1747996302.piper-0.7423.0 +3 -0
  6. version_1/hparams.yaml +126 -0
version_0/checkpoints/epoch=1-step=3112.ckpt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20a5c961c41d5a05dab111728890fdfed9e382bbb137d4f2b8a35e897d56546f
3
+ size 845889993
version_0/events.out.tfevents.1747994945.piper-0.6987.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25ed49b3eafd945c623305bad099fe27c03533bb959644ca6e1a02232f7fa56a
3
+ size 209402246
version_0/hparams.yaml ADDED
@@ -0,0 +1,126 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ accelerator: gpu
2
+ accumulate_grad_batches: null
3
+ amp_backend: null
4
+ amp_level: null
5
+ auto_lr_find: false
6
+ auto_scale_batch_size: false
7
+ auto_select_gpus: null
8
+ batch_size: 16
9
+ benchmark: null
10
+ betas: !!python/tuple
11
+ - 0.8
12
+ - 0.99
13
+ c_kl: 1.0
14
+ c_mel: 45
15
+ channels: 1
16
+ check_val_every_n_epoch: 1
17
+ checkpoint_epochs: 1
18
+ dataset:
19
+ - !!python/object/apply:pathlib.PosixPath
20
+ - /
21
+ - data
22
+ - piper_ljspeech_training
23
+ - dataset.jsonl
24
+ dataset_dir: &id001 !!python/object/apply:pathlib.PosixPath
25
+ - /
26
+ - data
27
+ - piper_ljspeech_training
28
+ default_root_dir: *id001
29
+ detect_anomaly: false
30
+ devices: '1'
31
+ enable_checkpointing: true
32
+ enable_model_summary: true
33
+ enable_progress_bar: true
34
+ eps: 1.0e-09
35
+ fast_dev_run: false
36
+ filter_channels: 768
37
+ filter_length: 1024
38
+ gin_channels: 0
39
+ gpus: null
40
+ grad_clip: null
41
+ gradient_clip_algorithm: null
42
+ gradient_clip_val: null
43
+ hidden_channels: 192
44
+ hop_length: 256
45
+ inference_mode: true
46
+ init_lr_ratio: 1.0
47
+ inter_channels: 192
48
+ ipus: null
49
+ kernel_size: 3
50
+ learning_rate: 0.0002
51
+ limit_predict_batches: null
52
+ limit_test_batches: null
53
+ limit_train_batches: null
54
+ limit_val_batches: null
55
+ log_every_n_steps: 50
56
+ logger: true
57
+ lr_decay: 0.999875
58
+ max_epochs: 500
59
+ max_phoneme_ids: null
60
+ max_steps: -1
61
+ max_time: null
62
+ mel_channels: 80
63
+ mel_fmax: null
64
+ mel_fmin: 0.0
65
+ min_epochs: null
66
+ min_steps: null
67
+ move_metrics_to_cpu: false
68
+ multiple_trainloader_mode: max_size_cycle
69
+ n_heads: 2
70
+ n_layers: 6
71
+ n_layers_q: 3
72
+ num_nodes: 1
73
+ num_processes: null
74
+ num_sanity_val_steps: 2
75
+ num_speakers: 1
76
+ num_symbols: 256
77
+ num_test_examples: 10
78
+ num_workers: 1
79
+ overfit_batches: 0.0
80
+ p_dropout: 0.1
81
+ plugins: null
82
+ precision: 32
83
+ profiler: null
84
+ quality: medium
85
+ reload_dataloaders_every_n_epochs: 0
86
+ replace_sampler_ddp: true
87
+ resblock: '2'
88
+ resblock_dilation_sizes: !!python/tuple
89
+ - !!python/tuple
90
+ - 1
91
+ - 2
92
+ - !!python/tuple
93
+ - 2
94
+ - 6
95
+ - !!python/tuple
96
+ - 3
97
+ - 12
98
+ resblock_kernel_sizes: !!python/tuple
99
+ - 3
100
+ - 5
101
+ - 7
102
+ resume_from_checkpoint: null
103
+ resume_from_single_speaker_checkpoint: null
104
+ sample_bytes: 2
105
+ sample_rate: 22050
106
+ seed: 1234
107
+ segment_size: 8192
108
+ strategy: null
109
+ sync_batchnorm: false
110
+ tpu_cores: null
111
+ track_grad_norm: -1
112
+ upsample_initial_channel: 256
113
+ upsample_kernel_sizes: !!python/tuple
114
+ - 16
115
+ - 16
116
+ - 8
117
+ upsample_rates: !!python/tuple
118
+ - 8
119
+ - 8
120
+ - 4
121
+ use_sdp: true
122
+ use_spectral_norm: false
123
+ val_check_interval: null
124
+ validation_split: 0.05
125
+ warmup_epochs: 0
126
+ win_length: 1024
version_1/checkpoints/epoch=499-step=519000.ckpt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7eef3f4bba366b7653bc34353d1309c575d68fb9ef1cad0c97abde0a5e481e9
3
+ size 845890057
version_1/events.out.tfevents.1747996302.piper-0.7423.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b626d5533834283d2eb378b70619c676cb2806b32b1a4a8bab1edac7790a1219
3
+ size 38924888081
version_1/hparams.yaml ADDED
@@ -0,0 +1,126 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ accelerator: gpu
2
+ accumulate_grad_batches: null
3
+ amp_backend: null
4
+ amp_level: null
5
+ auto_lr_find: false
6
+ auto_scale_batch_size: false
7
+ auto_select_gpus: null
8
+ batch_size: 24
9
+ benchmark: null
10
+ betas: !!python/tuple
11
+ - 0.8
12
+ - 0.99
13
+ c_kl: 1.0
14
+ c_mel: 45
15
+ channels: 1
16
+ check_val_every_n_epoch: 1
17
+ checkpoint_epochs: 1
18
+ dataset:
19
+ - !!python/object/apply:pathlib.PosixPath
20
+ - /
21
+ - data
22
+ - piper_ljspeech_training
23
+ - dataset.jsonl
24
+ dataset_dir: &id001 !!python/object/apply:pathlib.PosixPath
25
+ - /
26
+ - data
27
+ - piper_ljspeech_training
28
+ default_root_dir: *id001
29
+ detect_anomaly: false
30
+ devices: '1'
31
+ enable_checkpointing: true
32
+ enable_model_summary: true
33
+ enable_progress_bar: true
34
+ eps: 1.0e-09
35
+ fast_dev_run: false
36
+ filter_channels: 768
37
+ filter_length: 1024
38
+ gin_channels: 0
39
+ gpus: null
40
+ grad_clip: null
41
+ gradient_clip_algorithm: null
42
+ gradient_clip_val: null
43
+ hidden_channels: 192
44
+ hop_length: 256
45
+ inference_mode: true
46
+ init_lr_ratio: 1.0
47
+ inter_channels: 192
48
+ ipus: null
49
+ kernel_size: 3
50
+ learning_rate: 0.0002
51
+ limit_predict_batches: null
52
+ limit_test_batches: null
53
+ limit_train_batches: null
54
+ limit_val_batches: null
55
+ log_every_n_steps: 50
56
+ logger: true
57
+ lr_decay: 0.999875
58
+ max_epochs: 500
59
+ max_phoneme_ids: null
60
+ max_steps: -1
61
+ max_time: null
62
+ mel_channels: 80
63
+ mel_fmax: null
64
+ mel_fmin: 0.0
65
+ min_epochs: null
66
+ min_steps: null
67
+ move_metrics_to_cpu: false
68
+ multiple_trainloader_mode: max_size_cycle
69
+ n_heads: 2
70
+ n_layers: 6
71
+ n_layers_q: 3
72
+ num_nodes: 1
73
+ num_processes: null
74
+ num_sanity_val_steps: 2
75
+ num_speakers: 1
76
+ num_symbols: 256
77
+ num_test_examples: 10
78
+ num_workers: 1
79
+ overfit_batches: 0.0
80
+ p_dropout: 0.1
81
+ plugins: null
82
+ precision: 32
83
+ profiler: null
84
+ quality: medium
85
+ reload_dataloaders_every_n_epochs: 0
86
+ replace_sampler_ddp: true
87
+ resblock: '2'
88
+ resblock_dilation_sizes: !!python/tuple
89
+ - !!python/tuple
90
+ - 1
91
+ - 2
92
+ - !!python/tuple
93
+ - 2
94
+ - 6
95
+ - !!python/tuple
96
+ - 3
97
+ - 12
98
+ resblock_kernel_sizes: !!python/tuple
99
+ - 3
100
+ - 5
101
+ - 7
102
+ resume_from_checkpoint: null
103
+ resume_from_single_speaker_checkpoint: null
104
+ sample_bytes: 2
105
+ sample_rate: 22050
106
+ seed: 1234
107
+ segment_size: 8192
108
+ strategy: null
109
+ sync_batchnorm: false
110
+ tpu_cores: null
111
+ track_grad_norm: -1
112
+ upsample_initial_channel: 256
113
+ upsample_kernel_sizes: !!python/tuple
114
+ - 16
115
+ - 16
116
+ - 8
117
+ upsample_rates: !!python/tuple
118
+ - 8
119
+ - 8
120
+ - 4
121
+ use_sdp: true
122
+ use_spectral_norm: false
123
+ val_check_interval: null
124
+ validation_split: 0.05
125
+ warmup_epochs: 0
126
+ win_length: 1024