Meiqi Zhao

Results 3 comments of Meiqi Zhao

❓ [Question] compiled ExportedProgram is slower than uncompiled model

Update: In comparison, `trt_model = torch.compile(model, backend='tensorrt', options={'truncate_long_and_double': True, 'enabled_precisions': {torch.half}})` reduces the inference time to 0.3ms. But JIT compilation would be infeasible for my use case since the time...

[Computer Vision exercises] Unable to launch Computer vision exercises that rely on webcam for input

I have the same issue. You could run `v4l2-ctl --list-devices` and see a list of your video devices, make sure the camera you use is listed as /dev/video0

❓ [Question] how to specify dynamic shape when using torch_tensorrt.save

@peri044 Thank you for clarifying!