JiaPai12138
JiaPai12138
weights: [yolov4-csp-s-mish](https://github.com/AlexeyAB/darknet/releases/download/yolov4/yolov4-csp-s-mish.weights) cfg: [yolov4-csp-s-mish](https://github.com/AlexeyAB/darknet/blob/master/cfg/yolov4-csp-s-mish.cfg) Boxes are larger and confidences are dropping for example, trained with a same dateset to generate both yolov4-tiny and yolov4-csp-s-mish weights convert to onnx using demo_darknet2onnx...
目前 cuda v11.6 + cudnn v8.3.3 会报错 (已添加到path) ``` F:\tensorrt-python-main>python export.py -o v6n.onnx -e v6n.trt -p fp16 [07/04/2022-19:11:00] [TRT] [I] [MemUsageChange] Init CUDA: CPU +291, GPU +0, now: CPU 4590,...
classes, scores, boxes = model.detect(frames, conf_thd, nms_thd) 其中model这个是来自： `net = cv2.dnn.readNetFromDarknet(CONFIG_FILE, WEIGHT_FILE) # 读取权重与配置文件` `net.setPreferableBackend(cv2.dnn.DNN_BACKEND_CUDA)` `net.setPreferableTarget(cv2.dnn.DNN_TARGET_CUDA)` `model = cv2.dnn_DetectionModel(net)` `model.setInputParams(size=(side_width, side_height), scale=1/255, swapRB=True)`
rt 使用官方导出源码中的--include-nms和--end2end命令后 只能使用onnxruntime 的 CUDAExecutionProvider 才能读取成功么
您好 在直接使用py代码时出现以下问题 2022-06-10 16:25:20.0589966 [E:onnxruntime:, sequential_executor.cc:368 onnxruntime::SequentialExecutor::Execute] Non-zero status code returned while running Resize node. Name:'Resize_0' Status Message: D:\a\_work\1\s\onnxruntime\core\providers\dml\DmlExecutionProvider\src\MLOperatorAuthorImpl.cpp(1856)\onnxruntime_pybind11_state.pyd!00007FFF12A2ABAA: (caller: 00007FFF12A2A8B5) Exception(1) tid(59c) 80070057 The parameter is incorrect. Traceback (most...