Code crashing without error while using pytorch profiler with_stack=True

1

I am using window 11 with A10 GPU while running code with pytorch profiler with keeping with_stack=True
my code snippet is as below:

with torch.no_grad():
with profile(activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA],profile_memory=True,record_shapes=True,with_stack=True,
on_trace_ready=torch.profiler.tensorboard_trace_handler(f’./profiler_logs_proxylessNASGPUs8{self.group}')) as prof:
model(data)

2

With Windows its a known issue.
Maybe you can try nvidia nsys profiler
nsys profile -t cuda,cudnn,cublas python your_script.py

Also check you CUDA is compatible with your torch

3

Thanks bro!! its working