Custom layer calling in Tensorflow.js loadLayersModel()

[stevencheng1220]

tfjs/tfjs-layers/src/models.ts

Line 256 in 3688c73

return loadLayersModelFromIOHandler(pathOrIOHandler, undefined, options);

It seems like there is already functionality for custom layer calling in Tensorflow.js's loadLayersModel() method. However, it seems like behind the scenes as seen from the function LoadLayersModelFromIOHandler(pathOrIOHandler, undefined, options), the input for custom objects is set as "undefined". Is this an intentional design?

Just linking the repo contributors here:
@pyu10055 @nsthorat @dsmilkov @caisq

[pyu10055]

Hi @stevencheng1220 I think that is a placeholder for a great feature to add.

If you have custom layers, I think currently there is a work around for that:

1. save the keras model as TF saved model format
2. use tfjs converter to convert saved model to tfjs_graph_model.

[gaikwadrahul8]

Hi, @stevencheng1220

Apologize for the delayed response and we have updated our official documentation w.r.t tf.loadLayersModel and This method is applicable to:

Models created with the tf.layers.*, tf.sequential(), and tf.model() APIs of TensorFlow.js and later saved with the tf.LayersModel.save() method.
Models converted from Keras or TensorFlow tf.keras using the tensorflowjs_converter.
This mode is not applicable to TensorFlow SavedModels or their converted forms. For those models, use tf.loadGraphModel().

Please refer these comments Ref-1 and Ref-2, I hope it help you to resolve your issue and if not please help me with sample code with model and error log to replicate the issue from our end ? Thank you!

[google-ml-butler]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 7 days if no further activity occurs. Thank you.

[google-ml-butler]

Closing as stale. Please @mention us if this needs more attention.