DDXDB

> @DDXDB Currently, SYCL backend don't support iq2_xxs. Please try with other model file, like https://huggingface.co/TheBloke/Llama-2-7B-GGUF/blob/main/llama-2-7b.Q4_0.gguf > > If you try llama.cpp for SYCL, please try with the [guide](https://github.com/ggerganov/llama.cpp/blob/master/README-sycl.md) to...

> @DDXDB Looks like it can't detect the device correctly. Could you try with latest code? And run ./build/bin/ls-sycl-device.exe The latest master code works for me! By the way, does...

> @DDXDB It's great to see your result. Yes, support multiple GPUs as default. But it supports more GPUs with same max compute units. > > If you have iGPU...

> 用det_by_fps，然后修改 config.py 能细说一下么，菜逼只会gui ：(

### new Log 2024-01-02 00:40:33,227 (GuiMain.py) (Line 1412) [INFO] : Python version: 3.11.7 | packaged by conda-forge | (main, Dec 23 2023, 14:27:59) [MSC v.1937 64 bit (AMD64)] 2024-01-02 00:40:33,228...

No longer crashes, but seems too slow, 5 minutes is still 0%. Still sometimes the GPU is not recognized.

Correction: After a long 0%, 8 minutes of audio was completed in about 30 seconds

Uses the original Releases still, no replacement files. 13 minutes 40 seconds of video, holding 0% after about 6 minutes the GPU starts working, and it finishes in 30 seconds...

> Yes, it's due to the AOT issue. When the progress gets stuck at 0%, torch (IPEX, precisely) is compiling the acceleration codes. The replacement contains the compiled codes so...

> What about 12.55.8? 12.55.0 is what the compiler outputs when entering the PCI ID of A770 (0x56A0), but according to the architecture of A770 it should be 12.55.8 Work...