Skip to content

Commit

Permalink
change: docs
Browse files Browse the repository at this point in the history
  • Loading branch information
initialencounter committed Aug 7, 2023
1 parent 689fb6b commit d11671f
Show file tree
Hide file tree
Showing 229 changed files with 16,895 additions and 1,181 deletions.
11 changes: 11 additions & 0 deletions docs/DockerImages/Audiocraft.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,11 @@
# Audiocraft
Audiocraft是一个基于深度学习的音频处理和生成库。它具有最先进的编码器音频压缩器/标记器,以及MusicGen,一个简单而可控的音乐生成LM,具有文本和旋律调节。

Audiocraft is a library for audio processing and generation with deep learning. It features the state-of-the-art EnCodec audio compressor / tokenizer, along with MusicGen, a simple and controllable music generation LM with textual and melodic conditioning.


# usage

```
docker run -it -d -p 7862:7860 --gpus all initialencounter/audiocraft:latest bash -c "bash .start.sh"
```
1 change: 1 addition & 0 deletions docs/DockerImages/ChatRWKV.md
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
# ChatRWKV
4 changes: 4 additions & 0 deletions docs/DockerImages/_category_.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,4 @@
{
"label": "Docker 镜像",
"position": 5
}
12 changes: 12 additions & 0 deletions docs/DockerImages/openai-whisper-api.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,12 @@
# openai-whisper-api

将请求的格式改为base64

适用于koishi-plugin-whisper-asr [![npm](https://img.shields.io/npm/v/koishi-plugin-whisper-asr?style=flat-square)](https://www.npmjs.com/package/koishi-plugin-whisper-asr)

搭配[adapter-wechaty](https://www.npmjs.com/package/@initencounter/koishi-plugin-adapter-wechaty)[![npm](https://img.shields.io/npm/v/koishi-plugin-davinci-003?style=flat-square)](https://www.npmjs.com/package/koishi-plugin-davinci-003)可以在微信搭建语音回复的ChatGPT机器人

部署方法
```
docker run -d --gpus all -p 9000:9000 -e ASR_MODEL=base -e ASR_ENGINE=openai_whisper initialencounter/openai-whisper-asr-webservice:latest-gpu
```
60 changes: 60 additions & 0 deletions docs/DockerImages/paddlespeech.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,60 @@
# PaddleSpeech

```
docker run -it -p 8880-8889:8880-8889 --gpus all initialencounter/paddlespeech:ttsfinetune-latest /bin/bash
```
镜像集成了 jupyterlab,进入容器后,可通过以下命令开启服务。
```
jupyter lab --ip=0.0.0.0 --port=8880 --allow-root --notebook-dir=/home/aistudio
```

## Finetune 教程
参考自[【有手就行】使用自己的声音做语音合成](https://aistudio.baidu.com/aistudio/projectdetail/5003396)

- 1.将你的语音文件放到 work/init 目录下
>可通过 docker cp 或者 scp 将你的语音文件放到这个目录
>语音文件要求
- 2.开始训练
- 对于萌新
>直接运行```python main.py```
- 有一定深度学习经验的朋友
>可以编辑 main.py 里的代码
>根据自己的需要来微调模型



## 同时支持将微调后的模型部署为 FastAPI

- 修改app.py文件
```shell
cd /home/aistudio/initialencunter/PaddleSpeech
```
```python
# 将 model 替换为你的exp_name
am_inference_dir = os.path.abspath("/home/aistudio/inferrence/init")
# 如果你的 exp_name 为 "demo"
# 则将代码修改为
am_inference_dir = os.path.abspath("/home/aistudio/inferrence/demo")
```
- 启动接口
```shell
python app.py
```


## 问题反馈
* QQ群:399899914

* 小伙伴如果遇到问题或者有新的想法,欢迎到[这里](https://github.com/initialencounter/mykoishi/issues)反馈哦~

语音文件要求
> 对于语音合成任务,对数据是有一定要求的,尽可能上传干净的人声数据,比如像示例中的人声数据,在安静环境下录制,录制设备无论是手机,电脑,还是别的设备都可以,注意一定要控制噪音,或者提前使用音频剪辑软件进行降噪。(一定要是中文数据,这个项目目前只支持中文数据!!其它语言预标注环节会出错!)
- 1.音频不要太长,也不要太短,2s~10s之间,音频太长会报错!
- 2.音频尽量是干净人声,不要有BGM,不要有比较大的杂音,不要有一些奇奇怪怪的声效,比如回声等
- 3.声音的情绪尽量稳定,以说话的语料为主,不要是『嗯』『啊』『哈』之类的语气词
- 4.音频数量大于5 条!!否则会报错!!
关于录音工具的选择:
你可以使用一些在线运行的录音工具或者 【Adobe Audition】,【Cool Edit Pro】, 【Audacity】 等录音软件录制音频,保存为 24000采样率的 Wav 格式
也可以通过手机录音后,使用格式工厂等格式转换软件,将手机录音转换成 Wav 格式后上传到这个项目中。
13 changes: 13 additions & 0 deletions docs/DockerImages/sd-webui.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
# stable diffusion webui

# usage

```
docker run -it -d -p 7860:7860 --gpus all initialencounter/sd-webui:latest bash -c "bash start.sh"
```

如过你在宿主机上下载了模型可以使用以下命令,挂载目录到容器:

```
docker run -v 模型目录/models:/stable-diffusion-webui/models -it -d -p 7860:7860 --gpus all initialencounter/sd-webui:latest bash -c "bash start.sh"
```
14 changes: 14 additions & 0 deletions docs/DockerImages/text-gen-webui.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,14 @@
# text-gen-webui

## usage

```
docker run --gpus all -d -it -p 7860:7860 initialencounter/text-gen-webui:latest
```

## start
```
conda activate textgen;
cd /text-generation-webui;
python server.py --api --listen --listen-host "0.0.0.0" --listen-port 7860
```
18 changes: 18 additions & 0 deletions docs/DockerImages/wechaty.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,18 @@
# wechaty

## 准备
* openai apikey
* 开通过微信支付的微信账号,(不开通微信支付可能会登不上)

## 步骤

- 1.启动容器
- `docker run -p 5141:5140 initialencounter/koishi-wechaty:latest bash -c "cd /koimux_bot && yarn start"`
- 2.扫码
- `docker logs 容器ID`或者直接登录koishi控制台
- 3.配置davinci-003
- 输入key,启用插件

## 联系我
- qq群:399899914
- 哔哩哔哩:https://space.bilibili.com/225995995
5 changes: 0 additions & 5 deletions docs/First/Readme/README.md

This file was deleted.

3 changes: 0 additions & 3 deletions docs/First/Readme/_category_.json

This file was deleted.

25 changes: 0 additions & 25 deletions docs/First/Readme/java-an-zhuang-ji-shi-yong.md

This file was deleted.

73 changes: 0 additions & 73 deletions docs/First/Readme/pi-fu-zhan-bang-zhu-wen-dang.md

This file was deleted.

4 changes: 0 additions & 4 deletions docs/First/_category_.json

This file was deleted.

15 changes: 0 additions & 15 deletions docs/First/guan-yu-zan-zhu.md

This file was deleted.

43 changes: 43 additions & 0 deletions docs/KoishiPlugins/AI/arcadia.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,43 @@
# koishi-plugin-arcadia

[![npm](https://img.shields.io/npm/v/koishi-plugin-arcadia?style=flat-square)](https://www.npmjs.com/package/koishi-plugin-arcadia)

意间Ai绘画(yjart),支持100多个模型
因为懒只写了9个,后面会加上

![alt 示例2](https://raw.githubusercontent.com/initialencounter/mykoishi/master/screenshot/arcadia.png)
## 注意事项
> 使用前在 [yjai](http://open.yjai.art/openai-painting) 中获取apikey及apisecret
对于部署者行为及所产生的任何纠纷, Koishi 及 koishi-plugin-arcadia 概不负责。
如果有更多文本内容想要修改,可以在本地化中修改 zh 内容
发送arcadia.style,即可切换风格

# 更新日志

## v1.0.0
>
正式版
新增引擎切换,积分查询,任务查询,输出模式切换

## v0.1.3
>
修复无法自动发送图片bug

## v0.1.2
>
修复未知错误,增加风格
"梵高",
'托马斯·科尔',
'赛博朋克',
'中国画',
'蒸汽波',
'莫奈',
'二次元色彩大师',
'经典动漫',
'油画'

## v0.1.1
>
能跑了

## 感谢inisis佬对插件的支持
28 changes: 28 additions & 0 deletions docs/KoishiPlugins/AI/artwise.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,28 @@
# koishi-plugin-artwise

[![npm](https://img.shields.io/npm/v/koishi-plugin-artwise?style=flat-square)](https://www.npmjs.com/package/koishi-plugin-artwise)

知数云接口

![alt 示例](https://raw.githubusercontent.com/initialencounter/mykoishi/master/screenshot/artwise.png)
> 使用前在 知数云 中获取token
对于部署者行为及所产生的任何纠纷, Koishi 及 koishi-plugin-artwise 概不负责。
## 使用方法

| 功能 | 指令 |
| ---- | ---- |
| midjourney | mj |
| ChatGPT 4 | mai |
| ChatGPT 3.5 | wai |



## 问题反馈
* QQ群:399899914

* 小伙伴如果遇到问题或者有新的想法,欢迎到[这里](https://github.com/initialencounter/mykoishi/issues)反馈哦~

## 更新日志

- v0.0.2
- 新增mj失败后自动重试
Loading

0 comments on commit d11671f

Please sign in to comment.