len

[Feature Request] LoRa fine-tuning?

1

I propose to add support for LoRa fine-tuning of the model, as it has been shown to be effective for GPT models based on community feedback. Additionally, there have been...

[Feature] non-blocking mode for polling

4

When using the Telegraf library to implement a Telegram bot, I noticed that my bot would frequently hang and appear unresponsive, seemingly due to network issues. At first it was...

### verify - [X] 我已经仔细阅读[项目文档](https://github.com/tindy2013/subconverter/blob/master/README-cn.md)，确认现有功能无法解决我的需求 - [X] 我已经检索过现有[issue](https://github.com/tindy2013/subconverter/issues)，确认与现有issue的内容并不重复 - [X] 我已经尝试自行解决，确认自己没有能力解决 ### 功能描述 ### 需求 希望sub接口支持post请求 类似这样 ``` POST /sub HTTP/1.1 ... { "target": "clash", "new_name": true, "url": [ ......

Corrected `api.py` to utilize models from the project directory to prevent unnecessary re-downloading models.

My requirement is similar to the issue #133 mentioned. I need to generate a .d.ts file representing only the public API of a large TypeScript codebase, excluding private/internal symbols and...

https://swivid.github.io/F5-TTS/ ## TODOs - [ ] pin code - [ ] f5 infer/pipeline - [ ] tests

完善 ModelZoo

2

因为开始支持多种模型，并且模型还有不同版本，配置起来有点麻烦，想增加新模型有点繁琐。特别 gpt-sovits 很多自定义的模型也很需要不同模型的配置和加载流程。所以得重构下。 ~~本来最好的是直接用 transformer 流水线，但是，貌似很多开源都不支持...~~ ## TODOs - [ ] model config file - [ ] refactor model zoo - [ ] impl dl/loader/infer/runner

目前 `EncoderConfig` 只能调用代码的时候配置，没有释放出来接口 可以把 bitrate 、 acodec 配置释放到 api 和 webui 中 - [ ] 增加 api encode config 参数 - [ ] 增加 webui encoder 配置 ref #163

# background 另外 有個 From Subtitle 的想法， 希望能在導入字幕文件後，加入表格式的參數註釋，方才輸出到SSML模塊 現在From Subtitle頁面只有音色選擇，其餘參數需在輸出後修改，較為無序 _Originally posted by @hymch in https://github.com/lenML/Speech-AI-Forge/discussions/151#discussioncomment-10597244_ # plan - [ ] 重构 ssml webui， 增加 Scripts tab 单独处理脚本 - [...

### Discussed in https://github.com/lenML/ChatTTS-Forge/discussions/151 Originally posted by **hymch** September 10, 2024 使用 From Subtitle 加載 字幕文件轉入SSML模塊後 如果改動