Ruoxi

@dunnevan  I tried with your parameters, however I didn't get satisfactory results... Is it because the different datasets of music should have different parameters...?

> Hmm, I have no idea why my result stays the same somehow. My run was `ESRNet_4xplus, nproc=1, [480x270] --> [1920x1080], noFace, noExtractFirst` on 4x A40 (48Gb VRAM each), 52...

I am running the animevideov3 model without FP32 and the outputs are correct. Could you please provide more details about your setup? I don't have RoCM available and there may...

This command is working fine on my machine: ```bash python inference_realesrgan_video_fast.py --model_name=realesr-general-x4v3 -i "videos\2022-12-24 17-53-30.mp4" -s 2 ``` Did I understand your input correctly?

It also works here with -dn 0.

> FYI, I observe `torch.compile + channel_last` provides 2x speedup (no tiling, no face enhancing, fp16) on NVIDIA A4000. > > ``` > self.model = self.model.to(memory_format=torch.channels_last) > self.model = torch.compile(self.model)...

https://github.com/pytorch/pytorch/issues/92542 I guess RRDB-based networks and VGG-based networks have different preferences for channel formats.

I have met the same problem. In a clean ubuntu docker, I apt-installed bison, flex and autoconf, then git-cloned this repository, then run ./bootstrap and ./configure, and it goes like...

After I removed the line I still cannot get it compiled. ``` daemon.c:143:45: error: 'struct sockaddr_in' has no member named 'sin6_port'; did you mean 'sin_port'? ```

I can compile 0.9.23 version, and can configure (but not compile) 0.10.0 version, and cannot configure the master version.