Upload 9 files
Browse files- config.json +164 -0
- honkai_zunda_e131_s210000.safetensors +3 -0
- honkai_zunda_e251_s402000.safetensors +3 -0
- honkai_zunda_e435_s696000.safetensors +3 -0
- honkai_zunda_e465_s744000.safetensors +3 -0
- honkai_zunda_e499_s798000.safetensors +3 -0
- mos_result.png +0 -0
- mos_smoothed.png +0 -0
- style_vectors.npy +3 -0
config.json
ADDED
@@ -0,0 +1,164 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"model_name": "honkai_zunda",
|
3 |
+
"train": {
|
4 |
+
"log_interval": 50,
|
5 |
+
"eval_interval": 2000,
|
6 |
+
"seed": 42,
|
7 |
+
"epochs": 500,
|
8 |
+
"learning_rate": 0.0001,
|
9 |
+
"betas": [
|
10 |
+
0.8,
|
11 |
+
0.99
|
12 |
+
],
|
13 |
+
"eps": 1e-09,
|
14 |
+
"batch_size": 8,
|
15 |
+
"bf16_run": false,
|
16 |
+
"fp16_run": false,
|
17 |
+
"lr_decay": 0.99996,
|
18 |
+
"segment_size": 16384,
|
19 |
+
"init_lr_ratio": 1,
|
20 |
+
"warmup_epochs": 0,
|
21 |
+
"c_mel": 45,
|
22 |
+
"c_kl": 1.0,
|
23 |
+
"c_commit": 100,
|
24 |
+
"skip_optimizer": false,
|
25 |
+
"freeze_ZH_bert": false,
|
26 |
+
"freeze_JP_bert": false,
|
27 |
+
"freeze_EN_bert": false,
|
28 |
+
"freeze_emo": false,
|
29 |
+
"freeze_style": false,
|
30 |
+
"freeze_decoder": false
|
31 |
+
},
|
32 |
+
"data": {
|
33 |
+
"use_jp_extra": true,
|
34 |
+
"training_files": "Data/honkai_zunda/train.list",
|
35 |
+
"validation_files": "Data/honkai_zunda/val.list",
|
36 |
+
"max_wav_value": 32768.0,
|
37 |
+
"sampling_rate": 44100,
|
38 |
+
"filter_length": 2048,
|
39 |
+
"hop_length": 512,
|
40 |
+
"win_length": 2048,
|
41 |
+
"n_mel_channels": 128,
|
42 |
+
"mel_fmin": 0.0,
|
43 |
+
"mel_fmax": null,
|
44 |
+
"add_blank": true,
|
45 |
+
"n_speakers": 545,
|
46 |
+
"cleaned_text": true,
|
47 |
+
"spk2id": {
|
48 |
+
"zundamon": 0,
|
49 |
+
"Guinaifen": 1,
|
50 |
+
"Natasha": 2,
|
51 |
+
"Clockie": 3,
|
52 |
+
"Yanqing": 4,
|
53 |
+
"March 7th": 9,
|
54 |
+
"Black Swan": 10,
|
55 |
+
"Hook": 12,
|
56 |
+
"Sampo": 13,
|
57 |
+
"Seele": 15,
|
58 |
+
"Dan Heng": 16,
|
59 |
+
"Sushang": 17,
|
60 |
+
"Jing Yuan": 19,
|
61 |
+
"Topaz": 20,
|
62 |
+
"Firefly": 21,
|
63 |
+
"Sparkle": 22,
|
64 |
+
"Aventurine": 23,
|
65 |
+
"Welt": 24,
|
66 |
+
"Hanya": 25,
|
67 |
+
"Argenti": 26,
|
68 |
+
"Fu Xuan": 27,
|
69 |
+
"Gepard": 29,
|
70 |
+
"Boothill": 30,
|
71 |
+
"Luka": 31,
|
72 |
+
"Serval": 32,
|
73 |
+
"Himeko": 33,
|
74 |
+
"Luocha": 36,
|
75 |
+
"Gallagher": 37,
|
76 |
+
"Robin": 38,
|
77 |
+
"Acheron": 39,
|
78 |
+
"Jade": 41,
|
79 |
+
"Clara": 43,
|
80 |
+
"Kafka": 44,
|
81 |
+
"Pela": 46,
|
82 |
+
"Bronya": 48,
|
83 |
+
"Lynx": 53,
|
84 |
+
"Jingliu": 55,
|
85 |
+
"Bailu": 60,
|
86 |
+
"Trustful Boy": 61,
|
87 |
+
"Yukong": 68,
|
88 |
+
"Blade": 76,
|
89 |
+
"Silver Wolf": 81,
|
90 |
+
"Misha": 93,
|
91 |
+
"Tingyun": 99,
|
92 |
+
"Huohuo": 103,
|
93 |
+
"Xueyi": 132,
|
94 |
+
"Qingque": 163
|
95 |
+
},
|
96 |
+
"num_styles": 1,
|
97 |
+
"style2id": {
|
98 |
+
"Neutral": 0
|
99 |
+
}
|
100 |
+
},
|
101 |
+
"model": {
|
102 |
+
"use_spk_conditioned_encoder": true,
|
103 |
+
"use_noise_scaled_mas": true,
|
104 |
+
"use_mel_posterior_encoder": false,
|
105 |
+
"use_duration_discriminator": false,
|
106 |
+
"use_wavlm_discriminator": true,
|
107 |
+
"inter_channels": 192,
|
108 |
+
"hidden_channels": 192,
|
109 |
+
"filter_channels": 768,
|
110 |
+
"n_heads": 2,
|
111 |
+
"n_layers": 6,
|
112 |
+
"kernel_size": 3,
|
113 |
+
"p_dropout": 0.1,
|
114 |
+
"resblock": "1",
|
115 |
+
"resblock_kernel_sizes": [
|
116 |
+
3,
|
117 |
+
7,
|
118 |
+
11
|
119 |
+
],
|
120 |
+
"resblock_dilation_sizes": [
|
121 |
+
[
|
122 |
+
1,
|
123 |
+
3,
|
124 |
+
5
|
125 |
+
],
|
126 |
+
[
|
127 |
+
1,
|
128 |
+
3,
|
129 |
+
5
|
130 |
+
],
|
131 |
+
[
|
132 |
+
1,
|
133 |
+
3,
|
134 |
+
5
|
135 |
+
]
|
136 |
+
],
|
137 |
+
"upsample_rates": [
|
138 |
+
8,
|
139 |
+
8,
|
140 |
+
2,
|
141 |
+
2,
|
142 |
+
2
|
143 |
+
],
|
144 |
+
"upsample_initial_channel": 512,
|
145 |
+
"upsample_kernel_sizes": [
|
146 |
+
16,
|
147 |
+
16,
|
148 |
+
8,
|
149 |
+
2,
|
150 |
+
2
|
151 |
+
],
|
152 |
+
"n_layers_q": 3,
|
153 |
+
"use_spectral_norm": false,
|
154 |
+
"gin_channels": 512,
|
155 |
+
"slm": {
|
156 |
+
"model": "./slm/wavlm-base-plus",
|
157 |
+
"sr": 16000,
|
158 |
+
"hidden": 768,
|
159 |
+
"nlayers": 13,
|
160 |
+
"initial_channel": 64
|
161 |
+
}
|
162 |
+
},
|
163 |
+
"version": "2.6.0-JP-Extra"
|
164 |
+
}
|
honkai_zunda_e131_s210000.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:af199f74db66d175290c298f9eb0d4361ce0b5a014c05318ea6ed5321882a876
|
3 |
+
size 252265100
|
honkai_zunda_e251_s402000.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5a46ff6ab5fa9cdcf9e4c7ca76984a9a1e3448662a7fe163b67abec1833996f8
|
3 |
+
size 252265100
|
honkai_zunda_e435_s696000.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5f52cc82a27ab2b4747f1ed8a8b8654ce440102629eaa2d581873762b4e99753
|
3 |
+
size 252265100
|
honkai_zunda_e465_s744000.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ab9b626c83b0d607678a07b219091c1a075dc430afb435a3278d5752bc235841
|
3 |
+
size 252265100
|
honkai_zunda_e499_s798000.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cce68acfb061675e035623a3eb9d747761cfc26d46f9611aa577ab334fd2f5b8
|
3 |
+
size 252265100
|
mos_result.png
ADDED
![]() |
mos_smoothed.png
ADDED
![]() |
style_vectors.npy
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d906502e5bfb18ca9a272f07af15e6c1f14240d565ddaf078e6b9762348dd31b
|
3 |
+
size 1152
|