TensorFlow.js conversion - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to load a TensorFlow.js model from a URL.

TensorFlow

const model = await tf.loadLayersModel('[1]');

Drag options to blanks, or click blank then click option'

Amodel.h5

Bhttps://example.com/model.json

C./model.pb

Dmodel.pkl

Attempts:

3 left

💡 Hint

Common Mistakes

Using a Keras H5 file instead of a TensorFlow.js JSON model file.

Using a local file path without proper URL format.

The tf.loadLayersModel function loads a TensorFlow.js model from a URL or local path to a JSON file. The URL must point to a model.json file.

2fill in blank

medium

Complete the code to convert a Keras model to TensorFlow.js format using the CLI.

TensorFlow

tensorflowjs_converter --input_format=keras [1] [2]

Drag options to blanks, or click blank then click option'

Amodel.h5

Bmodel.json

Ctfjs_model

Dsaved_model

Attempts:

3 left

💡 Hint

Common Mistakes

Using model.json as input instead of output.

Using saved_model as output folder for Keras input.

The tensorflowjs_converter command converts a Keras H5 model file (model.h5) to a TensorFlow.js format saved in a directory (tfjs_model).

3fill in blank

hard

Fix the error in the code to correctly save a TensorFlow SavedModel for conversion.

TensorFlow

model.save('[1]');

Drag options to blanks, or click blank then click option'

A./saved_model

Bmodel.h5

Cmodel.json

Dtfjs_model

Attempts:

3 left

💡 Hint

Common Mistakes

Saving with a filename ending in .h5 instead of a folder path.

Using a JSON filename which is for TensorFlow.js, not TensorFlow SavedModel.

To save a TensorFlow SavedModel format, you must provide a directory path (e.g., './saved_model'). Saving as .h5 saves a Keras H5 file, not a SavedModel.

4fill in blank

hard

Fill both blanks to create a TensorFlow.js LayersModel from a Keras model and save it.

TensorFlow

const model = await tf.[1]('https://example.com/model.json');
await model.[2]('localstorage://my-model');

Drag options to blanks, or click blank then click option'

AloadLayersModel

Bsave

Cfit

Dcompile

Attempts:

3 left

💡 Hint

Common Mistakes

Using fit or compile instead of loading or saving methods.

Confusing method names for loading and saving.

Use tf.loadLayersModel() to load a Keras converted model, then model.save() to save it to local storage in TensorFlow.js.

5fill in blank

hard

Fill all three blanks to convert a TensorFlow SavedModel to TensorFlow.js format specifying input format and output directory.

TensorFlow

tensorflowjs_converter --input_format=[1] [2] [3]

Drag options to blanks, or click blank then click option'

Akeras

B./saved_model

C./tfjs_model

Dtf_saved_model

Attempts:

3 left

💡 Hint

Common Mistakes

Using keras as input format for SavedModel.

Not specifying output directory path.

To convert a TensorFlow SavedModel to TensorFlow.js format, specify --input_format=tf_saved_model, the SavedModel directory ./saved_model, and the output directory ./tfjs_model.