yolov7 onnx to trt core dump

[qrsssh]

python tensorrt-python/export.py -o mfd-20230613.onnx -e a.trt -p fp32 --end2end --v8
Namespace(calib_batch_size=8, calib_cache='./calibration.cache', calib_input=None, calib_num_images=5000, conf_thres=0.4, end2end=True, engine='a.trt', iou_thres=0.5, max_det=100, onnx='a.onnx', precision='fp32', v10=False, v8=True, verbose=False, workspace=1)
[07/10/2024-02:16:51] [TRT] [W] Unable to determine GPU memory usage: CUDA driver version is insufficient for CUDA runtime version
[07/10/2024-02:16:51] [TRT] [W] Unable to determine GPU memory usage: CUDA driver version is insufficient for CUDA runtime version
[07/10/2024-02:16:51] [TRT] [I] [MemUsageChange] Init CUDA: CPU +0, GPU +0, now: CPU 10, GPU 0 (MiB)
terminate called without an active exception
已放弃 (核心已转储

[qrsssh]

@Linaom1214