New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Error in PredictCost() for the op: op: "Softmax" attr #1044

Open

krishnanNuance opened this issue Oct 14, 2022 · 0 comments

Contributor

krishnanNuance commented Oct 14, 2022 •

edited

Loading

Describe the bug

The model converted from ONNX to TF on validation returns NaN as result.

To Reproduce

export fairseq's transformer lm model to onnx
export the onnx model to TF
load the saved_model and validate with any src_token value
the output array contains NaN

Python, ONNX, ONNX-TF, Tensorflow version

Python version: 1.11.0+cu102
ONNX version: 1.10.2
ONNX-TF version: 1.10.0 + implementation for squeeze and unsqueeze for opset v 13
Tensorflow version: 2.8.0

krishnanNuance mentioned this issue

Grappler error when Softmax input has a variable dimension tensorflow/tensorflow#50575

Closed

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment