Cannot run oneflow.roi_align() properly

[xxxyyyzzz12345]

Summary

Traceback (most recent call last): File "/mnt/sda/yjy/summer/DL-autograd-torch-main/yjy_test.py", line 24, in output = oneflow.roi_align(input0,input1,input2,input3,input4,input5,input6) oneflow._oneflow_internal.exception.OpKernelNotFoundException: File "/home/ci-user/runners/release/_work/oneflow/oneflow/oneflow/core/framework/op_interpreter/op_interpreter_util.cpp", line 139, in Dispatchoneflow::one::Tensor Dispatch<TensorTuple>(op_expr, inputs, ctx) File "/home/ci-user/runners/release/_work/oneflow/oneflow/oneflow/core/framework/op_interpreter/op_interpreter_util.cpp", line 131, in Dispatchoneflow::one::TensorTuple Dispatch(op_expr, inputs, outputs.get(), ctx) File "/home/ci-user/runners/release/work/oneflow/oneflow/oneflow/core/framework/op_interpreter/op_interpreter.cpp", line 96, in Apply internal->Apply(op_expr, inputs, outputs, ctx) File "/home/ci-user/runners/release/_work/oneflow/oneflow/oneflow/core/framework/op_interpreter/eager_mirrored_op_interpreter.cpp", line 165, in NaiveInterpret PhysicalRun([&](InstructionsBuilder* builder) -> Maybe ... input_eager_blob_objects, output_eager_blob_objects, ctx, stream); }) File "/home/ci-user/runners/release/_work/oneflow/oneflow/oneflow/core/framework/instructions_builder.cpp", line 599, in PhysicalRun Build(&instructions_builder) File "/home/ci-user/runners/release/_work/oneflow/oneflow/oneflow/core/framework/instructions_builder.cpp", line 358, in LocalCallOpKernel vm::LocalCallOpKernelPhyInstrOperand::New( opkernel ... consistent_tensor_infer_result, ctx, *one::CurrentDevVmDepObjectConsumeMode()) File "/home/ci-user/runners/release/work/oneflow/oneflow/oneflow/core/eager/local_call_opkernel_phy_instr_operand.h", line 54, in New ptr->Init() File "/home/ci-user/runners/release/work/oneflow/oneflow/oneflow/core/eager/local_call_opkernel_phy_instr_operand.cpp", line 26, in Init mut_opkernel()->ChooseOpKernel(&user_opkernel, &need_temp_storage, attrs(), inputs().get(), outputs().get(), consistent_tensor_infer_result().get()) File "/home/ci-user/runners/release/work/oneflow/oneflow/oneflow/user/kernels/stateful_local_opkernel.cpp", line 453, in ChooseOpKernel user_op::UserOpRegistryMgr::Get().GetOpKernelRegistryResult(op_type_name, *reg_ctx) Cannot find the kernel matching Current OperatorConf. The Info of OperatorConf are op_name: roi_align1 op_type_name: roi_align DeviceType_Name: kCPU DataType_Name of x_0: kFloat DataType_Name of rois_0: kFloat DataType_Name of y_0: kFloat op_kernels_not_found_debug_str: "(device_type == gpu)"

Code to reproduce bug

input0 = oneflow.rand(2, 3, 64, 64, dtype=oneflow.float32) input1 = oneflow.rand(200, 5, dtype=oneflow.float32) input2 = 2.0 input3 = 14 input4 = 14 input5 = 2 input6 = True output = oneflow.roi_align(input0,input1,input2,input3,input4,input5,input6)

System Information

• What is your OneFlow installation (pip, source, dockerhub): pip
• OS: Ubuntu 20.04.2 LTS
• OneFlow version (run python3 -m oneflow --doctor): 0.7.0+cu112
• Python version: 3.8.8
• CUDA driver version: 11.4

[MARD1NO]

Currently OneFlow only support ROIAlign Operator in CUDA, you can set the tensor device property as "cuda"