laugh12321 / TensorRT-YOLO

🚀 你的YOLO部署神器。TensorRT Plugin、CUDA Kernel、CUDA Graphs三管齐下,享受闪电般的推理速度。| Your YOLO Deployment Powerhouse. With the synergy of TensorRT Plugins, CUDA Kernels, and CUDA Graphs, experience lightning-fast inference speeds.
https://github.com/laugh12321/TensorRT-YOLO
GNU General Public License v3.0
638 stars 78 forks source link

[Help]: 无法导出onnx模型 #24

Closed liautumn closed 6 months ago

liautumn commented 6 months ago

image image

image

liautumn commented 6 months ago

执行完是空的,没生成onnx

laugh12321 commented 6 months ago

你下载模型了吗?

https://github.com/laugh12321/TensorRT-YOLO/tree/main/demo/detect#%E6%A8%A1%E5%9E%8B%E5%AF%BC%E5%87%BA

liautumn commented 6 months ago

image

laugh12321 commented 6 months ago

指令运行结果图看一下,上面没发完

liautumn commented 6 months ago

指令运行结果图看一下,上面没发完

image

laugh12321 commented 6 months ago

同步一下仓库代码,或者更新一下tensorrt-yolo

pip install tensorrt-yolo==3.0.2

实际效果应该时这样的: image

liautumn commented 6 months ago

同步一下仓库代码,或者更新一下tensorrt-yolo

pip install tensorrt-yolo==3.0.2

实际效果应该是这样的: 图像

好的,我试试

laugh12321 commented 6 months ago

@liautumn 可以导出的话这个issue就关闭了。

liautumn commented 6 months ago

@liautumn 可以导出的话这个issue就关闭了。

可以了,非常感谢

sanersbug commented 4 months ago

@laugh12321 为啥我更新到最新的tensorrt-yolo版本还是不行,结果还是空的,请问你是调整了什么才能正常导出的

laugh12321 commented 4 months ago

@sanersbug

sanersbug commented 4 months ago

@laugh12321 是我自己训练的yolov8s,用的就是官网的项目训练的,在Windows上操作的,我是运行ptq_calibration.py脚本,缺啥装啥,没有在创建新环境后直接pip install tensorrt_yolo image

laugh12321 commented 4 months ago

@sanersbug 我遇到过这个问题,创建一个新环境直接pip install tensorrt_yolo就好了。