Amadeus
Amadeus
Same issue when trying to convert palm_detection_full.
It maybe a PReLu related issue https://github.com/NVIDIA/TensorRT/issues/310
This issue also happens to me (Win10, firefox103, 5.1.2, with 21:9 screen), setting the video alignment still work after I choose "Never" on "Enable This extension".
In my case, Reload through F5 doesn't disable the extension. The video alignment still effect.
[ultrawidify-settings.json.txt](https://github.com/tamius-han/ultrawidify/files/9274609/ultrawidify-settings.json.txt) firefox 104 / win10 yes, the issue persist in a new tab
Memory leaks on torchvision-0.14.0+cu117 (torchvision-0.14.0%2Bcu117-cp37-cp37m-win_amd64.whl). When will this be fixed? easy to reproduce: ```python for i in range(10000): torchvision.io.decode_jpeg(torch.frombuffer(jpeg_bytes,dtype=torch.uint8), device='cuda') ``` Memory leaks didn't happen when using pynvjpeg 0.0.13, which...
We have to think about how to deal with the max token issue. The current max_token is 4096, means that if we keep chatting, the system role will eventually disappear,...
just got approved
Try change ``` torch.distributed.init_process_group("nccl") ``` to ``` torch.distributed.init_process_group("gloo") ```
@maple3142 I also encountered this issue.