torch_npu.profiler.schedule (wait, active, warmup = 0, repeat = 0, skip_first = 0)
设置不同step的行为。用于构造torch_npu.profiler.profile的schedule参数。
with torch_npu.profiler.profile(
schedule=torch_npu.profiler.schedule(wait=1, warmup=1, active=2, repeat=2, skip_first=10),
on_trace_ready=torch_npu.profiler.tensorboard_trace_handler("./result"),
with_stack=True) as prof:
for step in range(steps):
train_one_step(step, steps, train_loader, model, optimizer, criterion)
prof.step()