-
Notifications
You must be signed in to change notification settings - Fork 450
Description
After conversion, onn.checker throws error "No Op registered for RFFT2D with domain_version of 18"
Code
import tf2onnx
tflite_path = "/kaggle/input/birdnet/tensorflow2/default/1/CustomClassifier.tflite"
onnx_path = "/kaggle/working/bpi.onnx"
model_proto, external_tensor_storage = tf2onnx.convert.from_tflite(tflite_path, opset=18)
onnx.checker.check_model(onnx_path, True)
Output
ValidationError Traceback (most recent call last)
/tmp/ipykernel_36/227921013.py in <cell line: 0>()
1 import onnx
----> 2 onnx.checker.check_model(onnx_path, True)
/usr/local/lib/python3.11/dist-packages/onnx/checker.py in check_model(model, full_check, skip_opset_compatibility_check, check_custom_domain)
161 # If model is a path instead of ModelProto
162 if isinstance(model, (str, os.PathLike)):
--> 163 C.check_model_path(
164 os.fspath(model),
165 full_check,
ValidationError: No Op registered for RFFT2D with domain_version of 18
==> Context: Bad node spec for node. Name: stft/rfft;stft/rfft7 OpType: RFFT2D