-
Thank you, that script is pretty cool.
It works for like 3-6 times, but after that I always get an error after starting the recording.
```
Exception in thread Thread-29 (record):
Traceback (mo…
-
Hello
How are you?
Thanks for contributing to this project.
I know that whisper-jax is much faster than faster-whisper.
Could you introduce whisper-jax backend to your project?
Thanks
-
我打开梯子，就会报SSL错误，关闭梯子会提示连接超时，我把huggingface的faster-whisper-large-v3模型下载到_model_cache文件夹中，运行程序还是回去huggingface下载模型，有什么办法可以直接本地运行，不要去huggingface下载模型嘛？
.
This seems fixable by adding
```python
"recursed_filetype": str,
"pattern": str
```
to https://github.com/thiswil…
-
# Whisper Large v3 missing `mel_filters` in `preprocessor_config.json`
## Issue Description
Unlike the Tiny model, Whisper Large v3 doesn't include `mel_filters` in its `preprocessor_config.json…
-
Keyword Research:
- ubuntu mainline
- mainline kernel
- ubuntu kernel mainline
- ubuntu mainline kernel
- mainline kernel install
- ubuntu kernel ppa
- linux damon
- cxl
- compute express l…
-
使用的是官方最新的整合包GPT-SoVITS-v2-240821
我发现生成的音频波形和频谱在分段的位置会出现断崖般的变化，比较频繁。
下图才是正常的波形变化。
voice file: [test.mp3.zip](https://github.com/guillaumekln/faster-whisper/files/11170404/test.mp3.zip)
voice file: Sentence and sentence du…
-
I use lang code de and no function with either prepare for dataset or prepare for large data set
I am able to produce tokenzier but in final training I receive one off the following errors
I am ab…
-
I use initial_prompt and Simplified Chinese character, but it did not work.
segments, _ = model.transcribe(audio=main_name+'.mp3', initial_prompt='听起来有点呆')
bk111 updated
6 months ago