Upload 3 files
Browse files
pretrained_models/G_latest.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ad9bc43c8ef538f5e7864cd219b4614916c5a4d79833b2d360401a4deb7202de
|
3 |
+
size 158887401
|
pretrained_models/config.json
ADDED
@@ -0,0 +1,147 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"train": {
|
3 |
+
"log_interval": 100,
|
4 |
+
"eval_interval": 1000,
|
5 |
+
"seed": 1234,
|
6 |
+
"epochs": 10000,
|
7 |
+
"learning_rate": 0.0002,
|
8 |
+
"betas": [
|
9 |
+
0.8,
|
10 |
+
0.99
|
11 |
+
],
|
12 |
+
"eps": 1e-09,
|
13 |
+
"batch_size": 16,
|
14 |
+
"fp16_run": true,
|
15 |
+
"lr_decay": 0.999875,
|
16 |
+
"segment_size": 8192,
|
17 |
+
"init_lr_ratio": 1,
|
18 |
+
"warmup_epochs": 0,
|
19 |
+
"c_mel": 45,
|
20 |
+
"c_kl": 1.0
|
21 |
+
},
|
22 |
+
"data": {
|
23 |
+
"training_files": "final_annotation_train.txt",
|
24 |
+
"validation_files": "final_annotation_val.txt",
|
25 |
+
"text_cleaners": [
|
26 |
+
"zh_ja_mixture_cleaners"
|
27 |
+
],
|
28 |
+
"max_wav_value": 32768.0,
|
29 |
+
"sampling_rate": 22050,
|
30 |
+
"filter_length": 1024,
|
31 |
+
"hop_length": 256,
|
32 |
+
"win_length": 1024,
|
33 |
+
"n_mel_channels": 80,
|
34 |
+
"mel_fmin": 0.0,
|
35 |
+
"mel_fmax": null,
|
36 |
+
"add_blank": true,
|
37 |
+
"n_speakers": 3,
|
38 |
+
"cleaned_text": true
|
39 |
+
},
|
40 |
+
"model": {
|
41 |
+
"inter_channels": 192,
|
42 |
+
"hidden_channels": 192,
|
43 |
+
"filter_channels": 768,
|
44 |
+
"n_heads": 2,
|
45 |
+
"n_layers": 6,
|
46 |
+
"kernel_size": 3,
|
47 |
+
"p_dropout": 0.1,
|
48 |
+
"resblock": "1",
|
49 |
+
"resblock_kernel_sizes": [
|
50 |
+
3,
|
51 |
+
7,
|
52 |
+
11
|
53 |
+
],
|
54 |
+
"resblock_dilation_sizes": [
|
55 |
+
[
|
56 |
+
1,
|
57 |
+
3,
|
58 |
+
5
|
59 |
+
],
|
60 |
+
[
|
61 |
+
1,
|
62 |
+
3,
|
63 |
+
5
|
64 |
+
],
|
65 |
+
[
|
66 |
+
1,
|
67 |
+
3,
|
68 |
+
5
|
69 |
+
]
|
70 |
+
],
|
71 |
+
"upsample_rates": [
|
72 |
+
8,
|
73 |
+
8,
|
74 |
+
2,
|
75 |
+
2
|
76 |
+
],
|
77 |
+
"upsample_initial_channel": 512,
|
78 |
+
"upsample_kernel_sizes": [
|
79 |
+
16,
|
80 |
+
16,
|
81 |
+
4,
|
82 |
+
4
|
83 |
+
],
|
84 |
+
"n_layers_q": 3,
|
85 |
+
"use_spectral_norm": false,
|
86 |
+
"gin_channels": 256
|
87 |
+
},
|
88 |
+
"speakers": [
|
89 |
+
"wavs",
|
90 |
+
"specialweek",
|
91 |
+
"zhongli"
|
92 |
+
],
|
93 |
+
"symbols": [
|
94 |
+
"_",
|
95 |
+
",",
|
96 |
+
".",
|
97 |
+
"!",
|
98 |
+
"?",
|
99 |
+
"-",
|
100 |
+
"~",
|
101 |
+
"\u2026",
|
102 |
+
"A",
|
103 |
+
"E",
|
104 |
+
"I",
|
105 |
+
"N",
|
106 |
+
"O",
|
107 |
+
"Q",
|
108 |
+
"U",
|
109 |
+
"a",
|
110 |
+
"b",
|
111 |
+
"d",
|
112 |
+
"e",
|
113 |
+
"f",
|
114 |
+
"g",
|
115 |
+
"h",
|
116 |
+
"i",
|
117 |
+
"j",
|
118 |
+
"k",
|
119 |
+
"l",
|
120 |
+
"m",
|
121 |
+
"n",
|
122 |
+
"o",
|
123 |
+
"p",
|
124 |
+
"r",
|
125 |
+
"s",
|
126 |
+
"t",
|
127 |
+
"u",
|
128 |
+
"v",
|
129 |
+
"w",
|
130 |
+
"y",
|
131 |
+
"z",
|
132 |
+
"\u0283",
|
133 |
+
"\u02a7",
|
134 |
+
"\u02a6",
|
135 |
+
"\u026f",
|
136 |
+
"\u0279",
|
137 |
+
"\u0259",
|
138 |
+
"\u0265",
|
139 |
+
"\u207c",
|
140 |
+
"\u02b0",
|
141 |
+
"`",
|
142 |
+
"\u2192",
|
143 |
+
"\u2193",
|
144 |
+
"\u2191",
|
145 |
+
" "
|
146 |
+
]
|
147 |
+
}
|
pretrained_models/finetune_speaker.json
ADDED
@@ -0,0 +1,147 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"train": {
|
3 |
+
"log_interval": 100,
|
4 |
+
"eval_interval": 1000,
|
5 |
+
"seed": 1234,
|
6 |
+
"epochs": 10000,
|
7 |
+
"learning_rate": 0.0002,
|
8 |
+
"betas": [
|
9 |
+
0.8,
|
10 |
+
0.99
|
11 |
+
],
|
12 |
+
"eps": 1e-09,
|
13 |
+
"batch_size": 16,
|
14 |
+
"fp16_run": true,
|
15 |
+
"lr_decay": 0.999875,
|
16 |
+
"segment_size": 8192,
|
17 |
+
"init_lr_ratio": 1,
|
18 |
+
"warmup_epochs": 0,
|
19 |
+
"c_mel": 45,
|
20 |
+
"c_kl": 1.0
|
21 |
+
},
|
22 |
+
"data": {
|
23 |
+
"training_files": "final_annotation_train.txt",
|
24 |
+
"validation_files": "final_annotation_val.txt",
|
25 |
+
"text_cleaners": [
|
26 |
+
"zh_ja_mixture_cleaners"
|
27 |
+
],
|
28 |
+
"max_wav_value": 32768.0,
|
29 |
+
"sampling_rate": 22050,
|
30 |
+
"filter_length": 1024,
|
31 |
+
"hop_length": 256,
|
32 |
+
"win_length": 1024,
|
33 |
+
"n_mel_channels": 80,
|
34 |
+
"mel_fmin": 0.0,
|
35 |
+
"mel_fmax": null,
|
36 |
+
"add_blank": true,
|
37 |
+
"n_speakers": 3,
|
38 |
+
"cleaned_text": true
|
39 |
+
},
|
40 |
+
"model": {
|
41 |
+
"inter_channels": 192,
|
42 |
+
"hidden_channels": 192,
|
43 |
+
"filter_channels": 768,
|
44 |
+
"n_heads": 2,
|
45 |
+
"n_layers": 6,
|
46 |
+
"kernel_size": 3,
|
47 |
+
"p_dropout": 0.1,
|
48 |
+
"resblock": "1",
|
49 |
+
"resblock_kernel_sizes": [
|
50 |
+
3,
|
51 |
+
7,
|
52 |
+
11
|
53 |
+
],
|
54 |
+
"resblock_dilation_sizes": [
|
55 |
+
[
|
56 |
+
1,
|
57 |
+
3,
|
58 |
+
5
|
59 |
+
],
|
60 |
+
[
|
61 |
+
1,
|
62 |
+
3,
|
63 |
+
5
|
64 |
+
],
|
65 |
+
[
|
66 |
+
1,
|
67 |
+
3,
|
68 |
+
5
|
69 |
+
]
|
70 |
+
],
|
71 |
+
"upsample_rates": [
|
72 |
+
8,
|
73 |
+
8,
|
74 |
+
2,
|
75 |
+
2
|
76 |
+
],
|
77 |
+
"upsample_initial_channel": 512,
|
78 |
+
"upsample_kernel_sizes": [
|
79 |
+
16,
|
80 |
+
16,
|
81 |
+
4,
|
82 |
+
4
|
83 |
+
],
|
84 |
+
"n_layers_q": 3,
|
85 |
+
"use_spectral_norm": false,
|
86 |
+
"gin_channels": 256
|
87 |
+
},
|
88 |
+
"speakers": {
|
89 |
+
"wavs": 0,
|
90 |
+
"specialweek": 1,
|
91 |
+
"zhongli": 2
|
92 |
+
},
|
93 |
+
"symbols": [
|
94 |
+
"_",
|
95 |
+
",",
|
96 |
+
".",
|
97 |
+
"!",
|
98 |
+
"?",
|
99 |
+
"-",
|
100 |
+
"~",
|
101 |
+
"\u2026",
|
102 |
+
"A",
|
103 |
+
"E",
|
104 |
+
"I",
|
105 |
+
"N",
|
106 |
+
"O",
|
107 |
+
"Q",
|
108 |
+
"U",
|
109 |
+
"a",
|
110 |
+
"b",
|
111 |
+
"d",
|
112 |
+
"e",
|
113 |
+
"f",
|
114 |
+
"g",
|
115 |
+
"h",
|
116 |
+
"i",
|
117 |
+
"j",
|
118 |
+
"k",
|
119 |
+
"l",
|
120 |
+
"m",
|
121 |
+
"n",
|
122 |
+
"o",
|
123 |
+
"p",
|
124 |
+
"r",
|
125 |
+
"s",
|
126 |
+
"t",
|
127 |
+
"u",
|
128 |
+
"v",
|
129 |
+
"w",
|
130 |
+
"y",
|
131 |
+
"z",
|
132 |
+
"\u0283",
|
133 |
+
"\u02a7",
|
134 |
+
"\u02a6",
|
135 |
+
"\u026f",
|
136 |
+
"\u0279",
|
137 |
+
"\u0259",
|
138 |
+
"\u0265",
|
139 |
+
"\u207c",
|
140 |
+
"\u02b0",
|
141 |
+
"`",
|
142 |
+
"\u2192",
|
143 |
+
"\u2193",
|
144 |
+
"\u2191",
|
145 |
+
" "
|
146 |
+
]
|
147 |
+
}
|