Update README.md
Browse files
README.md
CHANGED
@@ -3,203 +3,479 @@ datasets:
|
|
3 |
- google/MusicCaps
|
4 |
language:
|
5 |
- ja
|
6 |
-
pipeline_tag: text-to-speech
|
7 |
tags:
|
8 |
- music
|
9 |
---
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
# Model Details
|
17 |
-
|
18 |
-
## Model Description
|
19 |
-
|
20 |
-
<!-- Provide a longer summary of what this model is. -->
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
- **Developed by:** [More Information Needed]
|
25 |
-
- **Shared by [optional]:** [More Information Needed]
|
26 |
-
- **Model type:** [More Information Needed]
|
27 |
-
- **Language(s) (NLP):** [More Information Needed]
|
28 |
-
- **License:** [More Information Needed]
|
29 |
-
- **Finetuned from model [optional]:** [More Information Needed]
|
30 |
-
|
31 |
-
## Model Sources [optional]
|
32 |
-
|
33 |
-
<!-- Provide the basic links for the model. -->
|
34 |
-
|
35 |
-
- **Repository:** [More Information Needed]
|
36 |
-
- **Paper [optional]:** [More Information Needed]
|
37 |
-
- **Demo [optional]:** [More Information Needed]
|
38 |
-
|
39 |
-
# Uses
|
40 |
-
|
41 |
-
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
42 |
-
|
43 |
-
## Direct Use
|
44 |
-
|
45 |
-
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
46 |
-
|
47 |
-
[More Information Needed]
|
48 |
-
|
49 |
-
## Downstream Use [optional]
|
50 |
-
|
51 |
-
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
52 |
-
|
53 |
-
[More Information Needed]
|
54 |
-
|
55 |
-
## Out-of-Scope Use
|
56 |
-
|
57 |
-
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
58 |
-
|
59 |
-
[More Information Needed]
|
60 |
-
|
61 |
-
# Bias, Risks, and Limitations
|
62 |
-
|
63 |
-
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
64 |
-
|
65 |
-
[More Information Needed]
|
66 |
-
|
67 |
-
## Recommendations
|
68 |
-
|
69 |
-
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
70 |
-
|
71 |
-
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
72 |
-
|
73 |
-
## How to Get Started with the Model
|
74 |
-
|
75 |
-
Use the code below to get started with the model.
|
76 |
-
|
77 |
-
[More Information Needed]
|
78 |
-
|
79 |
-
# Training Details
|
80 |
-
|
81 |
-
## Training Data
|
82 |
-
|
83 |
-
<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
84 |
-
|
85 |
-
[More Information Needed]
|
86 |
-
|
87 |
-
## Training Procedure
|
88 |
-
|
89 |
-
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
90 |
-
|
91 |
-
### Preprocessing [optional]
|
92 |
-
|
93 |
-
[More Information Needed]
|
94 |
-
|
95 |
-
|
96 |
-
### Training Hyperparameters
|
97 |
-
|
98 |
-
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
99 |
-
|
100 |
-
### Speeds, Sizes, Times [optional]
|
101 |
-
|
102 |
-
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
103 |
-
|
104 |
-
[More Information Needed]
|
105 |
-
|
106 |
-
# Evaluation
|
107 |
-
|
108 |
-
<!-- This section describes the evaluation protocols and provides the results. -->
|
109 |
-
|
110 |
-
## Testing Data, Factors & Metrics
|
111 |
-
|
112 |
-
### Testing Data
|
113 |
-
|
114 |
-
<!-- This should link to a Data Card if possible. -->
|
115 |
-
|
116 |
-
[More Information Needed]
|
117 |
-
|
118 |
-
### Factors
|
119 |
-
|
120 |
-
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
121 |
-
|
122 |
-
[More Information Needed]
|
123 |
-
|
124 |
-
### Metrics
|
125 |
-
|
126 |
-
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
127 |
-
|
128 |
-
[More Information Needed]
|
129 |
-
|
130 |
-
## Results
|
131 |
-
|
132 |
-
[More Information Needed]
|
133 |
-
|
134 |
-
### Summary
|
135 |
-
|
136 |
-
|
137 |
-
|
138 |
-
# Model Examination [optional]
|
139 |
-
|
140 |
-
<!-- Relevant interpretability work for the model goes here -->
|
141 |
-
|
142 |
-
[More Information Needed]
|
143 |
-
|
144 |
-
# Environmental Impact
|
145 |
-
|
146 |
-
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
147 |
-
|
148 |
-
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
149 |
-
|
150 |
-
- **Hardware Type:** [More Information Needed]
|
151 |
-
- **Hours used:** [More Information Needed]
|
152 |
-
- **Cloud Provider:** [More Information Needed]
|
153 |
-
- **Compute Region:** [More Information Needed]
|
154 |
-
- **Carbon Emitted:** [More Information Needed]
|
155 |
-
|
156 |
-
# Technical Specifications [optional]
|
157 |
-
|
158 |
-
## Model Architecture and Objective
|
159 |
-
|
160 |
-
[More Information Needed]
|
161 |
-
|
162 |
-
## Compute Infrastructure
|
163 |
-
|
164 |
-
[More Information Needed]
|
165 |
-
|
166 |
-
### Hardware
|
167 |
-
|
168 |
-
[More Information Needed]
|
169 |
-
|
170 |
-
### Software
|
171 |
-
|
172 |
-
[More Information Needed]
|
173 |
-
|
174 |
-
# Citation [optional]
|
175 |
-
|
176 |
-
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
177 |
-
|
178 |
-
**BibTeX:**
|
179 |
-
|
180 |
-
[More Information Needed]
|
181 |
-
|
182 |
-
**APA:**
|
183 |
-
|
184 |
-
[More Information Needed]
|
185 |
-
|
186 |
-
# Glossary [optional]
|
187 |
-
|
188 |
-
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
189 |
-
|
190 |
-
[More Information Needed]
|
191 |
-
|
192 |
-
# More Information [optional]
|
193 |
-
|
194 |
-
[More Information Needed]
|
195 |
-
|
196 |
-
# Model Card Authors [optional]
|
197 |
-
|
198 |
-
[More Information Needed]
|
199 |
-
|
200 |
-
# Model Card Contact
|
201 |
-
|
202 |
-
[More Information Needed]
|
203 |
-
|
204 |
|
205 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
- google/MusicCaps
|
4 |
language:
|
5 |
- ja
|
|
|
6 |
tags:
|
7 |
- music
|
8 |
---
|
9 |
+
--
|
10 |
+
datasets:
|
11 |
+
- google/MusicCaps
|
12 |
+
language:
|
13 |
+
- ja
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
|
15 |
+
## ESPnet2 TTS model
|
16 |
+
|
17 |
+
### `megumi/asaoka`
|
18 |
+
|
19 |
+
This model was trained by mio using [amadeus recipe](https://github.com/mio2333/espnet/tree/master/egs2/amadeus/tts1) in [espnet](https://github.com/espnet/espnet/).
|
20 |
+
|
21 |
+
|
22 |
+
### Demo: How to use in ESPnet2
|
23 |
+
|
24 |
+
Follow the [ESPnet installation instructions](https://espnet.github.io/espnet/installation.html)
|
25 |
+
if you haven't done that already.
|
26 |
+
|
27 |
+
```bash
|
28 |
+
cd espnet
|
29 |
+
git checkout d5b5ec7b2e77bd3e10707141818b7e6c57ac6b3f
|
30 |
+
pip install -e .
|
31 |
+
cd egs2/amadeus/tts1
|
32 |
+
./run.sh --skip_data_prep false --skip_train true --download_model mio/amadeus
|
33 |
+
```
|
34 |
+
|
35 |
+
|
36 |
+
|
37 |
+
## TTS config
|
38 |
+
|
39 |
+
<details><summary>expand</summary>
|
40 |
+
|
41 |
+
```
|
42 |
+
config: conf/tuning/finetune_vits.yaml
|
43 |
+
print_config: false
|
44 |
+
log_level: INFO
|
45 |
+
dry_run: false
|
46 |
+
iterator_type: sequence
|
47 |
+
output_dir: exp/tts_amadeus_vits_finetune_from_jsut_32_sentence
|
48 |
+
ngpu: 1
|
49 |
+
seed: 777
|
50 |
+
num_workers: 4
|
51 |
+
num_att_plot: 3
|
52 |
+
dist_backend: nccl
|
53 |
+
dist_init_method: env://
|
54 |
+
dist_world_size: null
|
55 |
+
dist_rank: null
|
56 |
+
local_rank: 0
|
57 |
+
dist_master_addr: null
|
58 |
+
dist_master_port: null
|
59 |
+
dist_launcher: null
|
60 |
+
multiprocessing_distributed: false
|
61 |
+
unused_parameters: true
|
62 |
+
sharded_ddp: false
|
63 |
+
cudnn_enabled: true
|
64 |
+
cudnn_benchmark: false
|
65 |
+
cudnn_deterministic: false
|
66 |
+
collect_stats: false
|
67 |
+
write_collected_feats: false
|
68 |
+
max_epoch: 2000
|
69 |
+
patience: null
|
70 |
+
val_scheduler_criterion:
|
71 |
+
- valid
|
72 |
+
- loss
|
73 |
+
early_stopping_criterion:
|
74 |
+
- valid
|
75 |
+
- loss
|
76 |
+
- min
|
77 |
+
best_model_criterion:
|
78 |
+
- - train
|
79 |
+
- total_count
|
80 |
+
- max
|
81 |
+
keep_nbest_models: 3
|
82 |
+
nbest_averaging_interval: 0
|
83 |
+
grad_clip: -1
|
84 |
+
grad_clip_type: 2.0
|
85 |
+
grad_noise: false
|
86 |
+
accum_grad: 1
|
87 |
+
no_forward_run: false
|
88 |
+
resume: true
|
89 |
+
train_dtype: float32
|
90 |
+
use_amp: false
|
91 |
+
log_interval: 50
|
92 |
+
use_matplotlib: true
|
93 |
+
use_tensorboard: true
|
94 |
+
create_graph_in_tensorboard: false
|
95 |
+
use_wandb: true
|
96 |
+
wandb_project: amadeus
|
97 |
+
wandb_id: null
|
98 |
+
wandb_entity: null
|
99 |
+
wandb_name: null
|
100 |
+
wandb_model_log_interval: -1
|
101 |
+
detect_anomaly: false
|
102 |
+
pretrain_path: null
|
103 |
+
init_param:
|
104 |
+
- downloads/f3698edf589206588f58f5ec837fa516/exp/tts_train_vits_raw_phn_jaconv_pyopenjtalk_accent_with_pause/train.total_count.ave_10best.pth:tts:tts
|
105 |
+
ignore_init_mismatch: false
|
106 |
+
freeze_param: []
|
107 |
+
num_iters_per_epoch: null
|
108 |
+
batch_size: 20
|
109 |
+
valid_batch_size: null
|
110 |
+
batch_bins: 5000000
|
111 |
+
valid_batch_bins: null
|
112 |
+
train_shape_file:
|
113 |
+
- exp/tts_stats_raw_linear_spectrogram_phn_jaconv_pyopenjtalk_accent_with_pause/train/text_shape.phn
|
114 |
+
- exp/tts_stats_raw_linear_spectrogram_phn_jaconv_pyopenjtalk_accent_with_pause/train/speech_shape
|
115 |
+
valid_shape_file:
|
116 |
+
- exp/tts_stats_raw_linear_spectrogram_phn_jaconv_pyopenjtalk_accent_with_pause/valid/text_shape.phn
|
117 |
+
- exp/tts_stats_raw_linear_spectrogram_phn_jaconv_pyopenjtalk_accent_with_pause/valid/speech_shape
|
118 |
+
batch_type: numel
|
119 |
+
valid_batch_type: null
|
120 |
+
fold_length:
|
121 |
+
- 150
|
122 |
+
- 204800
|
123 |
+
sort_in_batch: descending
|
124 |
+
sort_batch: descending
|
125 |
+
multiple_iterator: false
|
126 |
+
chunk_length: 500
|
127 |
+
chunk_shift_ratio: 0.5
|
128 |
+
num_cache_chunks: 1024
|
129 |
+
train_data_path_and_name_and_type:
|
130 |
+
- - dump/22k/raw/train/text
|
131 |
+
- text
|
132 |
+
- text
|
133 |
+
- - dump/22k/raw/train/wav.scp
|
134 |
+
- speech
|
135 |
+
- sound
|
136 |
+
valid_data_path_and_name_and_type:
|
137 |
+
- - dump/22k/raw/dev/text
|
138 |
+
- text
|
139 |
+
- text
|
140 |
+
- - dump/22k/raw/dev/wav.scp
|
141 |
+
- speech
|
142 |
+
- sound
|
143 |
+
allow_variable_data_keys: false
|
144 |
+
max_cache_size: 0.0
|
145 |
+
max_cache_fd: 32
|
146 |
+
valid_max_cache_size: null
|
147 |
+
optim: adamw
|
148 |
+
optim_conf:
|
149 |
+
lr: 0.0001
|
150 |
+
betas:
|
151 |
+
- 0.8
|
152 |
+
- 0.99
|
153 |
+
eps: 1.0e-09
|
154 |
+
weight_decay: 0.0
|
155 |
+
scheduler: exponentiallr
|
156 |
+
scheduler_conf:
|
157 |
+
gamma: 0.999875
|
158 |
+
optim2: adamw
|
159 |
+
optim2_conf:
|
160 |
+
lr: 0.0001
|
161 |
+
betas:
|
162 |
+
- 0.8
|
163 |
+
- 0.99
|
164 |
+
eps: 1.0e-09
|
165 |
+
weight_decay: 0.0
|
166 |
+
scheduler2: exponentiallr
|
167 |
+
scheduler2_conf:
|
168 |
+
gamma: 0.999875
|
169 |
+
generator_first: false
|
170 |
+
token_list:
|
171 |
+
- <blank>
|
172 |
+
- <unk>
|
173 |
+
- '1'
|
174 |
+
- '2'
|
175 |
+
- '0'
|
176 |
+
- '3'
|
177 |
+
- '4'
|
178 |
+
- '-1'
|
179 |
+
- '5'
|
180 |
+
- a
|
181 |
+
- o
|
182 |
+
- '-2'
|
183 |
+
- i
|
184 |
+
- '-3'
|
185 |
+
- u
|
186 |
+
- e
|
187 |
+
- k
|
188 |
+
- n
|
189 |
+
- t
|
190 |
+
- '6'
|
191 |
+
- r
|
192 |
+
- '-4'
|
193 |
+
- s
|
194 |
+
- N
|
195 |
+
- m
|
196 |
+
- pau
|
197 |
+
- '7'
|
198 |
+
- sh
|
199 |
+
- d
|
200 |
+
- g
|
201 |
+
- w
|
202 |
+
- '8'
|
203 |
+
- U
|
204 |
+
- '-5'
|
205 |
+
- I
|
206 |
+
- cl
|
207 |
+
- h
|
208 |
+
- y
|
209 |
+
- b
|
210 |
+
- '9'
|
211 |
+
- j
|
212 |
+
- ts
|
213 |
+
- ch
|
214 |
+
- '-6'
|
215 |
+
- z
|
216 |
+
- p
|
217 |
+
- '-7'
|
218 |
+
- f
|
219 |
+
- ky
|
220 |
+
- ry
|
221 |
+
- '-8'
|
222 |
+
- gy
|
223 |
+
- '-9'
|
224 |
+
- hy
|
225 |
+
- ny
|
226 |
+
- '-10'
|
227 |
+
- by
|
228 |
+
- my
|
229 |
+
- '-11'
|
230 |
+
- '-12'
|
231 |
+
- '-13'
|
232 |
+
- py
|
233 |
+
- '-14'
|
234 |
+
- '-15'
|
235 |
+
- v
|
236 |
+
- '10'
|
237 |
+
- '-16'
|
238 |
+
- '-17'
|
239 |
+
- '11'
|
240 |
+
- '-21'
|
241 |
+
- '-20'
|
242 |
+
- '12'
|
243 |
+
- '-19'
|
244 |
+
- '13'
|
245 |
+
- '-18'
|
246 |
+
- '14'
|
247 |
+
- dy
|
248 |
+
- '15'
|
249 |
+
- ty
|
250 |
+
- '-22'
|
251 |
+
- '16'
|
252 |
+
- '18'
|
253 |
+
- '19'
|
254 |
+
- '17'
|
255 |
+
- <sos/eos>
|
256 |
+
odim: null
|
257 |
+
model_conf: {}
|
258 |
+
use_preprocessor: true
|
259 |
+
token_type: phn
|
260 |
+
bpemodel: null
|
261 |
+
non_linguistic_symbols: null
|
262 |
+
cleaner: jaconv
|
263 |
+
g2p: pyopenjtalk_accent_with_pause
|
264 |
+
feats_extract: linear_spectrogram
|
265 |
+
feats_extract_conf:
|
266 |
+
n_fft: 1024
|
267 |
+
hop_length: 256
|
268 |
+
win_length: null
|
269 |
+
normalize: null
|
270 |
+
normalize_conf: {}
|
271 |
+
tts: vits
|
272 |
+
tts_conf:
|
273 |
+
generator_type: vits_generator
|
274 |
+
generator_params:
|
275 |
+
hidden_channels: 192
|
276 |
+
spks: -1
|
277 |
+
global_channels: -1
|
278 |
+
segment_size: 32
|
279 |
+
text_encoder_attention_heads: 2
|
280 |
+
text_encoder_ffn_expand: 4
|
281 |
+
text_encoder_blocks: 6
|
282 |
+
text_encoder_positionwise_layer_type: conv1d
|
283 |
+
text_encoder_positionwise_conv_kernel_size: 3
|
284 |
+
text_encoder_positional_encoding_layer_type: rel_pos
|
285 |
+
text_encoder_self_attention_layer_type: rel_selfattn
|
286 |
+
text_encoder_activation_type: swish
|
287 |
+
text_encoder_normalize_before: true
|
288 |
+
text_encoder_dropout_rate: 0.1
|
289 |
+
text_encoder_positional_dropout_rate: 0.0
|
290 |
+
text_encoder_attention_dropout_rate: 0.1
|
291 |
+
use_macaron_style_in_text_encoder: true
|
292 |
+
use_conformer_conv_in_text_encoder: false
|
293 |
+
text_encoder_conformer_kernel_size: -1
|
294 |
+
decoder_kernel_size: 7
|
295 |
+
decoder_channels: 512
|
296 |
+
decoder_upsample_scales:
|
297 |
+
- 8
|
298 |
+
- 8
|
299 |
+
- 2
|
300 |
+
- 2
|
301 |
+
decoder_upsample_kernel_sizes:
|
302 |
+
- 16
|
303 |
+
- 16
|
304 |
+
- 4
|
305 |
+
- 4
|
306 |
+
decoder_resblock_kernel_sizes:
|
307 |
+
- 3
|
308 |
+
- 7
|
309 |
+
- 11
|
310 |
+
decoder_resblock_dilations:
|
311 |
+
- - 1
|
312 |
+
- 3
|
313 |
+
- 5
|
314 |
+
- - 1
|
315 |
+
- 3
|
316 |
+
- 5
|
317 |
+
- - 1
|
318 |
+
- 3
|
319 |
+
- 5
|
320 |
+
use_weight_norm_in_decoder: true
|
321 |
+
posterior_encoder_kernel_size: 5
|
322 |
+
posterior_encoder_layers: 16
|
323 |
+
posterior_encoder_stacks: 1
|
324 |
+
posterior_encoder_base_dilation: 1
|
325 |
+
posterior_encoder_dropout_rate: 0.0
|
326 |
+
use_weight_norm_in_posterior_encoder: true
|
327 |
+
flow_flows: 4
|
328 |
+
flow_kernel_size: 5
|
329 |
+
flow_base_dilation: 1
|
330 |
+
flow_layers: 4
|
331 |
+
flow_dropout_rate: 0.0
|
332 |
+
use_weight_norm_in_flow: true
|
333 |
+
use_only_mean_in_flow: true
|
334 |
+
stochastic_duration_predictor_kernel_size: 3
|
335 |
+
stochastic_duration_predictor_dropout_rate: 0.5
|
336 |
+
stochastic_duration_predictor_flows: 4
|
337 |
+
stochastic_duration_predictor_dds_conv_layers: 3
|
338 |
+
vocabs: 85
|
339 |
+
aux_channels: 513
|
340 |
+
discriminator_type: hifigan_multi_scale_multi_period_discriminator
|
341 |
+
discriminator_params:
|
342 |
+
scales: 1
|
343 |
+
scale_downsample_pooling: AvgPool1d
|
344 |
+
scale_downsample_pooling_params:
|
345 |
+
kernel_size: 4
|
346 |
+
stride: 2
|
347 |
+
padding: 2
|
348 |
+
scale_discriminator_params:
|
349 |
+
in_channels: 1
|
350 |
+
out_channels: 1
|
351 |
+
kernel_sizes:
|
352 |
+
- 15
|
353 |
+
- 41
|
354 |
+
- 5
|
355 |
+
- 3
|
356 |
+
channels: 128
|
357 |
+
max_downsample_channels: 1024
|
358 |
+
max_groups: 16
|
359 |
+
bias: true
|
360 |
+
downsample_scales:
|
361 |
+
- 2
|
362 |
+
- 2
|
363 |
+
- 4
|
364 |
+
- 4
|
365 |
+
- 1
|
366 |
+
nonlinear_activation: LeakyReLU
|
367 |
+
nonlinear_activation_params:
|
368 |
+
negative_slope: 0.1
|
369 |
+
use_weight_norm: true
|
370 |
+
use_spectral_norm: false
|
371 |
+
follow_official_norm: false
|
372 |
+
periods:
|
373 |
+
- 2
|
374 |
+
- 3
|
375 |
+
- 5
|
376 |
+
- 7
|
377 |
+
- 11
|
378 |
+
period_discriminator_params:
|
379 |
+
in_channels: 1
|
380 |
+
out_channels: 1
|
381 |
+
kernel_sizes:
|
382 |
+
- 5
|
383 |
+
- 3
|
384 |
+
channels: 32
|
385 |
+
downsample_scales:
|
386 |
+
- 3
|
387 |
+
- 3
|
388 |
+
- 3
|
389 |
+
- 3
|
390 |
+
- 1
|
391 |
+
max_downsample_channels: 1024
|
392 |
+
bias: true
|
393 |
+
nonlinear_activation: LeakyReLU
|
394 |
+
nonlinear_activation_params:
|
395 |
+
negative_slope: 0.1
|
396 |
+
use_weight_norm: true
|
397 |
+
use_spectral_norm: false
|
398 |
+
generator_adv_loss_params:
|
399 |
+
average_by_discriminators: false
|
400 |
+
loss_type: mse
|
401 |
+
discriminator_adv_loss_params:
|
402 |
+
average_by_discriminators: false
|
403 |
+
loss_type: mse
|
404 |
+
feat_match_loss_params:
|
405 |
+
average_by_discriminators: false
|
406 |
+
average_by_layers: false
|
407 |
+
include_final_outputs: true
|
408 |
+
mel_loss_params:
|
409 |
+
fs: 22050
|
410 |
+
n_fft: 1024
|
411 |
+
hop_length: 256
|
412 |
+
win_length: null
|
413 |
+
window: hann
|
414 |
+
n_mels: 80
|
415 |
+
fmin: 0
|
416 |
+
fmax: null
|
417 |
+
log_base: null
|
418 |
+
lambda_adv: 1.0
|
419 |
+
lambda_mel: 45.0
|
420 |
+
lambda_feat_match: 2.0
|
421 |
+
lambda_dur: 1.0
|
422 |
+
lambda_kl: 1.0
|
423 |
+
sampling_rate: 22050
|
424 |
+
cache_generator_outputs: true
|
425 |
+
pitch_extract: null
|
426 |
+
pitch_extract_conf: {}
|
427 |
+
pitch_normalize: null
|
428 |
+
pitch_normalize_conf: {}
|
429 |
+
energy_extract: null
|
430 |
+
energy_extract_conf: {}
|
431 |
+
energy_normalize: null
|
432 |
+
energy_normalize_conf: {}
|
433 |
+
required:
|
434 |
+
- output_dir
|
435 |
+
- token_list
|
436 |
+
version: '202207'
|
437 |
+
distributed: false
|
438 |
+
```
|
439 |
+
|
440 |
+
</details>
|
441 |
+
|
442 |
+
|
443 |
+
|
444 |
+
### Citing ESPnet
|
445 |
+
|
446 |
+
```BibTex
|
447 |
+
@inproceedings{watanabe2018espnet,
|
448 |
+
author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
|
449 |
+
title={{ESPnet}: End-to-End Speech Processing Toolkit},
|
450 |
+
year={2018},
|
451 |
+
booktitle={Proceedings of Interspeech},
|
452 |
+
pages={2207--2211},
|
453 |
+
doi={10.21437/Interspeech.2018-1456},
|
454 |
+
url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
|
455 |
+
}
|
456 |
+
|
457 |
+
|
458 |
+
|
459 |
+
|
460 |
+
@inproceedings{hayashi2020espnet,
|
461 |
+
title={{Espnet-TTS}: Unified, reproducible, and integratable open source end-to-end text-to-speech toolkit},
|
462 |
+
author={Hayashi, Tomoki and Yamamoto, Ryuichi and Inoue, Katsuki and Yoshimura, Takenori and Watanabe, Shinji and Toda, Tomoki and Takeda, Kazuya and Zhang, Yu and Tan, Xu},
|
463 |
+
booktitle={Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
|
464 |
+
pages={7654--7658},
|
465 |
+
year={2020},
|
466 |
+
organization={IEEE}
|
467 |
+
}
|
468 |
+
```
|
469 |
+
|
470 |
+
or arXiv:
|
471 |
+
|
472 |
+
```bibtex
|
473 |
+
@misc{watanabe2018espnet,
|
474 |
+
title={ESPnet: End-to-End Speech Processing Toolkit},
|
475 |
+
author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
|
476 |
+
year={2018},
|
477 |
+
eprint={1804.00015},
|
478 |
+
archivePrefix={arXiv},
|
479 |
+
primaryClass={cs.CL}
|
480 |
+
}
|
481 |
+
```
|