我代码动了，我不写了？ while 循环莫名其妙运行俩次，找个赛博驱魔师

代码在下面

# pt 文件，pytorch 运行

import sys
from time import time
from ultralytics import YOLO
import cv2

# Load a model
model = YOLO("yolov8n.pt")  # load an official model
model = model.to("cuda")

video_path = "C:/Users/username/Desktop/ 学姐向你发出了约会邀请❤怦然心动❤.1542979357.mp4"
cap = cv2.VideoCapture(video_path)

total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
total_time = 0
while True:
    ret, frame = cap.read()
    if not ret:
        break  # Exit the loop if no more frames are available
    # frame = cv2.resize(frame, (640, 640))
    start_time = time()
    results = model(frame, verbose=False, classes=[0])
    total_time += time() - start_time
    cv2.imshow("frame", results[0].plot())
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
    # 打印进度
    finished_frames = int(cap.get(cv2.CAP_PROP_POS_FRAMES))
    sys.stdout.write("r 进度: %.2f %%" % (finished_frames / total_frames * 100))
    sys.stdout.flush()
    # results = model(frame, verbose=False,classes=[0]) CPU FPS:  14.687764775223318
    # results = model(frame, verbose=False,classes=[0]) FPS:  64.14847574515913
    # results = model(frame, verbose=False) FPS:  60.23244793078058

    # 2024-05-22 23:32:05 FPS:  60.21048703460454 CUDA

print("nFPS:", total_frames / total_time)

这段代码在 22 号还是正常的，输出就是“进度：”滚动，然后最后输出一行 FPS，cv2.imshow 那三行还没写。

刚才去洗了个澡，加上了 cv2.imshow 那三行，又跑了一次，输出变成了

python yolo-torch.py
C:UsersusernameDocumentscodeproject.venvLibsite-packagestorchnnmodulesconv.py:456: UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR: cudnnFinalize Descriptor Failed cudnn_status: CUDNN_STATUS_NOT_SUPPORTED (Triggered internally at ..atensrcATennativecudnnConv_v8.cpp:919.)
  return F.conv2d(input, weight, bias, self.stride,
进度: 99.85 %
FPS:  75.04395604987724
进度: 99.85 %
FPS:  21.148407567206448

终端先是滚动一边进度，非常快，然后打印了一行 FPS，但是这个时候 cv2 的 imshow 没有展示出来，
在第一个 FPS 打印完之后，又开始滚动输出“进度：”，这次 cv2.imshow 会正常显示，然后最后输出一个 FPS。

但是这俩个 FPS 都和 22 号的值差了不少，22 号的时候一直稳定 60 上下

然后我注释了 # 打印进度 下面的三行，输出变成了

python yolo-torch.py
C:UsersusernameDocumentscodeproject.venvLibsite-packagestorchnnmodulesconv.py:456: UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR: cudnnFinalize Descriptor Failed cudnn_status: CUDNN_STATUS_NOT_SUPPORTED (Triggered internally at ..atensrcATennativecudnnConv_v8.cpp:919.)
  return F.conv2d(input, weight, bias, self.stride,
进度: 99.85 %
FPS:  72.13356492043005

FPS:  21.089225686164987

为什么啊。我代码一行一行看过去，也看不出来为什么 while 循环会运行俩次，第一次还是我昨天的版本？