Error: Check failed: (reporter->Assert(begin < data->shape[axis])) is false: The sum of sections must match the input.shape[axis]

Troubleshooting
#1

Hello: I want to optimize a yolov8s model to tvm model, have the following steps:

  1. torch.jit.trace
  2. relay.frontend.from_pytorch
  3. relay.build(opt_level=3)
  4. runner = autotvm.LocalRunner(…), tasks = autotvm.task.extract_from_program(…), autotvm.tuner.XGBTuner(…)
  5. finally relay.build the best model Input size is 1, 3, 540, 960 same in all steps. I use the standard yolov8s model with tvm api in docs. How to solve it? Thanks !