JinqingZheng
JinqingZheng
Hi, how to set int8 for tensorrt ? I don't find the parameter.
> > Hi, how to set int8 for tensorrt ? I don't find the parameter. > > fp32: https://github.com/open-mmlab/mmdeploy/blob/master/configs/mmdet/instance-seg/instance-seg_tensorrt_dynamic-320x320-1344x1344.py fp16: https://github.com/open-mmlab/mmdeploy/blob/master/configs/mmdet/instance-seg/instance-seg_tensorrt-fp16_dynamic-320x320-1344x1344.py int8: https://github.com/open-mmlab/mmdeploy/blob/master/configs/mmdet/instance-seg/instance-seg_tensorrt-int8_dynamic-320x320-1344x1344.py You can adjust final dtype by setting...
> > > > Hi, how to set int8 for tensorrt ? I don't find the parameter. > > > > > > > > > fp32: https://github.com/open-mmlab/mmdeploy/blob/master/configs/mmdet/instance-seg/instance-seg_tensorrt_dynamic-320x320-1344x1344.py fp16: https://github.com/open-mmlab/mmdeploy/blob/master/configs/mmdet/instance-seg/instance-seg_tensorrt-fp16_dynamic-320x320-1344x1344.py...
> > > > Hi, how to set int8 for tensorrt ? I don't find the parameter. > > > > > > > > > fp32: https://github.com/open-mmlab/mmdeploy/blob/master/configs/mmdet/instance-seg/instance-seg_tensorrt_dynamic-320x320-1344x1344.py fp16: https://github.com/open-mmlab/mmdeploy/blob/master/configs/mmdet/instance-seg/instance-seg_tensorrt-fp16_dynamic-320x320-1344x1344.py...
> Thanks for your error report and we appreciate it a lot. > > **Checklist** > > 1. I have searched related issues but cannot get the expected help. >...
> Please check the input of convex_iou. Illegal input causes this error. > > Refer to https://github.com/open-mmlab/mmcv/blob/699398ad8697516304a2207ef248bf8d3e6e1b7b/mmcv/ops/convex_iou.py Thanks for your reply. I will check it.
> 上图的MultiScaleFlipAug是0.3版本的配置，但是在1x中没有找到，不知道是不是这个原因。 应该不是这个原因。我是在1.x版本下复现的实验结果，使用自己训练的模型，测试结果与官方结果差距有点大，目前还没有找到与官方实现不一致的地方。不知道你用0.3版本复现的结果是多少。
> 我也是用的1x，使用自己训练的模型  跟官方也有差距 我对了一下数据增广部分的设置，flip和rotate的参数设置不一致，不知道这个影响是不是大
> 我单尺度结果没问题，你的呢 单尺度差0.56，也不太一致，加WX交流下？