yumianhuli

Is there an api to get the transcription progress percent？ thanks！

Hello！ Is there a function to remove the model from VRAM？ Thanks！

Is there any parameter setting that makes the audio timestamp more accurate，Or it cannot be set at this time？ Thanks！

gpu inference speed is slower than the cpu

4

CPU inference ： | INFO | __main__:process: 67-process time: 2742.7728176116943ms GPU inference ： | INFO | __main__:process: 67-process time: 5317.4756874159873ms I use the big-lama model for inference My cpu inference...

I use set PYTHONPATH=%cd% python bin\to_jit.py model.path=%cd%\big-lama +save_path=%cd%\big-lama\big-lama.jit in the root directory of lama under windows os in order to convert ckpt format to pt format, but An error occurred：...

ckpt to pt model format

2

I use set PYTHONPATH=%cd% python bin\to_jit.py model.path=%cd%\big-lama +save_path=%cd%\big-lama\big-lama.jit in the root directory of lama under windows os in order to convert ckpt format to pt format, but An error occurred：...

Hello! Is there a method that use more Time for less VRAM usage for refinement infer？ Thanks!

我用work_crawler2.13版本下载快看漫画网站上的比如https://www.kuaikanmanhua.com/web/topic/21983/?source=comic的章节，但是下载的每张jpg图片都是不完整的，大概只有完整一张图片的1/3、1/4的样子，请问可以改进一下吗？ 谢谢！

How do I load the directory where the model is stored

2

How do I use reader = easyocr.Reader(['ch_sim'], gpu=True) and load the local specified directory where the recognition network model is stored ,thank you!

源码安装问题

6

运行python gui.py，报ModuleNotFoundError: No module named 'fsplit.filesplit' 1、安装pip install fsplit 2、源码安装https://github.com/yodeng/fsplit.git，pip install . 均无果，大佬看看。。