Skip to content

Conversation

@tip3x
Copy link
Collaborator

@tip3x tip3x commented Jun 25, 2025

No description provided.

Copy link
Member

@doronHarnoy doronHarnoy left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Cool

lstm_tensor_in_onnx_order = tf_transpose(lstm_tensor, perm=[1, 0, 2], tf_name=f"{params['cleaned_name']}_lstm_transpose")

# Add identical dense contains the lstm tensor for easy fetch of latent space
input_dim = int(lstm_tensor.shape[2])
Copy link
Collaborator

@tomkoren21 tomkoren21 Jun 25, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggest to raise an Exception here if shape is dynamic? This line would fail if lstm_tensor shape is None
Or else replace with tf.shape(lstm_tensor)[2]

Copy link
Collaborator

@tomkoren21 tomkoren21 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, suggest to add Exception or treat the case of a dynamic shape

@tip3x tip3x merged commit 93f9160 into master Jun 25, 2025
3 checks passed
@tip3x tip3x deleted the latent-space-for-lstm branch June 25, 2025 10:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants