lance
lance
 - 其中origin 是使用mmcls训练的mobilenetv2_1.5 - autoslim是下载的权重(计算量和参数量是拿的README),参见:  在推理速度上没有体现?还是操作有误,分别推理了10000次,预热100次取平均
Hi,i want to convert model to onnx,but i was unsuccessful! Do you have any good Suggestions? TKS!
loss nan

data set
Do I have to use segmentation data sets?but i just have box
### 是否已有关于该错误的issue或讨论? | Is there an existing issue / discussion for this? - [X] 我已经搜索过已有的issues和讨论 | I have searched the existing issues / discussions ### 该问题是否在FAQ中有解答? | Is there an...
