Skip to content
This repository was archived by the owner on Feb 3, 2025. It is now read-only.
This repository was archived by the owner on Feb 3, 2025. It is now read-only.

"Incompatible shapes" error during inference #333

Open
@bugzyz

Description

@bugzyz

Hi there,
we met this error during our inferencing but currently still don't know how to fix this. As for inferencing, the original model looks fine but the tf-trt optimized model got this error.
We noticed below warning during our conversion. Is it related to the error? And if the answer is yes, how to enable the preview feature kFASTER_DYNAMIC_SHAPES_0805?

Inference error

Graph execution error:
Detected at node 'StatefulPartitionedCall/etc_sequence_tagger/make_global_local_transformer_side_inputs/Equal' defined at (most recent call last):
Node: 'StatefulPartitionedCall/etc_sequence_tagger/make_global_local_transformer_side_inputs/Equal'
Incompatible shapes: [19,1,64] vs. [32,1024,1]
     [[{{node StatefulPartitionedCall/etc_sequence_tagger/make_global_local_transformer_side_inputs/Equal}}]] [Op:__inference_signature_wrapper_4750]

Warning during conversion

2023-09-05 08:52:19.682208: W tensorflow/compiler/tf2tensorrt/utils/trt_logger.cc:83] TF-TRT Warning: DefaultLogger Using PreviewFeature::kFASTER_DYNAMIC_SHAPES_0805 can help improve performance and resolve potential functional issues.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions