saurabh-shandilya commented on issue #13296:
URL: https://github.com/apache/tvm/issues/13296#issuecomment-1306378686
@masahi Added logs based on your suggestion and it seems to be
RelayToTIRTargetHook -
[15:32:52] K:\code\tvm_gpu\src\ir\transform.cc:431: RelayToTIRTargetHook
Traceback (most recent call last):
File "K:/code/tvm_gpu/python/object_detection_copy.py", line 201, in
<module>
verify_model_in_relay(script_module,img,relay_target)
File "K:/code/tvm_gpu/python/object_detection_copy.py", line 148, in
verify_model_in_relay
vm_exec = relay.vm.compile(mod, target=target, params=params)
File "K:\code\tvm_gpu\python\tvm\relay\backend\vm.py", line 79, in compile
compiler.lower(mod, target)
File "K:\code\tvm_gpu\python\tvm\relay\backend\vm.py", line 155, in lower
self._lower(mod, target, target_host)
File "K:\code\tvm_gpu\python\tvm\_ffi\_ctypes\packed_func.py", line 233,
in __call__
ctypes.byref(ret_tcode),
OSError: exception: stack overflow
What's the way past this issue then?
For debugging is there a way to dump a specific instance of an op or for
every op in the ir?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
