Error running .pte model with executor_runner

🐛 Describe the bug

I have exported this model:

https://github.com/corehalt/share/raw/refs/heads/main/yolov8n_runtime_issue.pte

with the following code:

from executorch.exir.passes.constant_prop_pass import constant_prop_pass
from executorch.exir.passes.const_prop_pass import ConstPropPass
from executorch import exir

self.model.eval()
aten_dialect_program = torch.export.export(self.model, (self.im,), strict=True)
torch.export.save(aten_dialect_program, "ep.pt2")
aten_dialect_program = constant_prop_pass(aten_dialect_program)
edge_dialect_program = exir.to_edge_transform_and_lower(aten_dialect_program, transform_passes=[ConstPropPass()])
executorch_program = edge_dialect_program.to_executorch(
    exir.ExecutorchBackendConfig(
        passes=[],
        remove_view_copy=False
    )
)
fpte = self.file.with_suffix(".pte")
with open(str(fpte), "wb") as fil:
    fil.write(executorch_program.buffer)

Then I tried to run the model with the official C++ executor_runner and but I get the next error:

gdb --args ../build/third_party/executorch/executor_runner --model_path /tmp/yolov8/yolov8n.pte
(gdb) where
#0  main (argc=1, argv=0x7fffffffde28) at /test/third_party/executorch/examples/portable/executor_runner/executor_runner.cpp:244
(gdb) list
239       ET_LOG(Info, "Inputs prepared.");
240
241       // Run the model.
242       for (uint32_t i = 0; i < FLAGS_num_executions; i++) {
243         Error status = method->execute();
244         ET_CHECK_MSG(
245             status == Error::Ok,
246             "Execution of method %s failed with status 0x%" PRIx32,
247             method_name,
248             (uint32_t)status);
(gdb) print status
$1 = executorch::runtime::Error::InvalidArgument
(gdb) print method_name
$4 = 0x5555561e4718 "forward"

With other models and using the same code, the inference runs without problem.
I also wrote another executor based on the official one but I also get the same error there.
Other things I tried is to use strict=False on torch.export() but still it gives me the same error.

For reference, this is the corresponding output of torch.export.save():

https://github.com/corehalt/share/raw/refs/heads/main/yolov8n_runtime_issue.pt2

Versions

Versions of relevant libraries:
[pip3] executorch==0.6.0a0+7103bb3
[pip3] numpy==2.1.1
[pip3] torch==2.7.0.dev20250131+cpu
[pip3] torchao==0.8.0+git11333ba2
[pip3] torchaudio==2.6.0.dev20250131+cpu
[pip3] torchsr==1.0.4
[pip3] torchvision==0.22.0.dev20250131+cpu

cc @JacobSzwejbka