-
青梅IT未来塾での講義の実施経験を基に、第7回・6回・8回の教科書への意見を募集します。
曖昧な意見だと採用するのが難しくなりますので、どこをどう直したらよいのか、具体的な意見を出していただけるようお願いします。
-
テキストスピーチできるようにしたい。
以下メモ
OpenJTalkのnodejs版
https://github.com/hecomi/node-openjtalk
課題
1. voiceが尾崎っぽくない
htc voiceで声を選定する
https://ja.osdn.net/projects/sfnet_open-jtalk/releases/
1. wavをい…
-
I meet problem when building wheels for pyopenjtalk, followings are some errors during the building of openjtalk:
> here is the branch of openjtalk I use https://github.com/r9y9/open_jtalk/tree/c…
-
## 内容
windows環境で、エンジンを2つ起動するとopenjtalk周りでエラーが出ます。
pyopenjtalk.set_user_dict`したものはopenjtalk側でずっとファイルハンドラが必要で、同じファイルに書き込めないためです。
https://github.com/VOICEVOX/voicevox_engine/blob/1269fabb63743c9f7…
-
Hi, In **MoeGoe**, inappropriate dependency versioning constraints can cause risks.
Below are the dependencies and version constraints that the project is using
```
numba
librosa
numpy==1.22.…
-
环境是wsl2 + conda环境
在安装完requirement之后
ModuleNotFoundError: No module named 'phonemizer'
AttributeError: module 'gradio' has no attribute 'Blocks'
ImportError: Numba needs NumPy 1.24 or less
通…
-
「うっうが」→アクセント取得では`う っ っ が`になぜか変換され、`pyopenjtalk.run_frontend()`の結果では発音が`うっーが`になり、その不一致から`align_tones`の処理に失敗する。
おそらくopenjtalkのバグによりそうだが、どう対処するのが正解か分からない。
-
目前项目的日语语音合成前端直接使用pyopenjtalk的g2p函数来获取分词结果,然后将其直接输入模型。惊人的是,当前GPT模型已经在没有音高标注的情况下,通过自身的学习能力,实现了对音高的建模,取得了令人满意的韵律效果。
然而,日语语音合成在许多情况下需要准确的音高信息以实现更精准的语音控制。这一点在处理特殊词汇和人名时尤其重要,因为pyopenjtalk和GPT模型都可能无法准确标注这…
-
I try both python 3.10 and 3.11
```
PS D:\AI\Bot\VITS_TXT_to_Audio> pip install -r .\requirements.txt
Looking in indexes: https://mirrors.aliyun.com/pypi/simple
Collecting pyopenjtalk
Using c…
-
- [x] pingコマンド
- [x] 読み上げ機能
- [x] サーバー情報
- [x] ユーザー情報
- [x] サーバー管理関係
- [ ] webダッシュボード
- [x] minecraft関係(鯖情報とかユーザー情報)
- [ ] spamとか荒らし対策
- [x] 埋め込みの送信
- [x] 予約送信
- [x] botの情報
- [x] 翻訳
- [x] …