vision icon indicating copy to clipboard operation
vision copied to clipboard

retinanet_resnet50_fpn测试出错

Open zkyseu opened this issue 1 year ago • 0 comments

基于project里面提供的train.py运行以下代码

python train.py \
  --data-path /home/kunyangzhou/project/dataset/coco \
  --dataset coco \
  --model retinanet_resnet50_fpn \
  --batch-size 8 \
  --pretrained \
  --test-only

然后出现报错

Stack trace (most recent call last) in thread 3038061:
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f083c34fa6f, in 
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f08352cf7e0, in VirtualMachine::ScheduleLoop(std::function<void ()> const&)
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f08352ebeea, in vm::VirtualMachineEngine::Schedule(vm::ScheduleCtx const&)
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f08352ea27d, in vm::VirtualMachineEngine::DispatchAndPrescheduleInstructions(vm::ScheduleCtx const&)
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f08352e78b4, in vm::VirtualMachineEngine::DispatchInstruction(vm::Instruction*, vm::ScheduleCtx const&)
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f083528daf3, in vm::Instruction::Prepare()
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f083528f666, in vm::OpCallInstructionPolicy::Prepare(vm::Instruction*)
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f0837fdecc8, in StatefulOpKernel::InferTmpSize(eager::CallContext*, user_op::OpKernel const*) const
   Object "/home/kunyangzhou/anaconda3/envs/zky/lib/python3.8/site-packages/oneflow/../oneflow.libs/liboneflow-05bae072.so", at 0x7f083808f2a6, in 

Floating point exception (Integer divide by zero [0x7f083808f2a6])
retinanet_resnet50_fpn/infer.sh: line 8: 3037911 Floating point exception(core dumped) python train.py --data-path /home/kunyangzhou/project/dataset/coco --dataset coco --model retinanet_resnet50_fpn --batch-size 8 --pretrained --test-only

zkyseu avatar Oct 16 '23 07:10 zkyseu