Styjun

How can I get rid of an unhelpful parallel branch when unpivoting a single row?

All ASCII characters with a given bit count

Multiple options vs single option UI

Philosophical question on logistic regression: why isn't the optimal threshold value trained?

How do I produce this Greek letter koppa: Ϟ in pdfLaTeX?

How do I deal with a coworker that keeps asking to make small superficial changes to a report, and it is seriously triggering my anxiety?

Why do distances seem to matter in the Foundation world?

Is there metaphorical meaning of "aus der Haft entlassen"?

How can I practically buy stocks?

"The cow" OR "a cow" OR "cows" in this context

How much of a wave function must reside inside event horizon for it to be consumed by the black hole?

Find a stone which is not the lightest one

What is the best way to deal with NPC-NPC combat?

Do I need to watch Ant-Man and the Wasp and Captain Marvel before watching Avengers: Endgame?

Multiple fireplaces in an apartment building?

Can a Bard use the Spell Glyph option of the Glyph of Warding spell and cast a known spell into the glyph?

As an international instructor, should I openly talk about my accent?

How to not starve gigantic beasts

Unknown code in script

A faster way to compute the largest prime factor

Why did C use the -> operator instead of reusing the . operator?

Would the change in enthalpy (ΔH) for the dissolution of urea in water be positive or negative?

"Whatever a Russian does, they end up making the Kalashnikov gun"? Are there any similar proverbs in English?

Is Electric Central Heating worth it if using Solar Panels?

Can tensorflowjs_converter work with Keras models made with functional API?

Can I run Keras model on gpu?Keras Visualization of Model Built from Functional APIMerging two models in Keras Functional APIHow to call a multidimensional prediction on a keras model with a javascript apiInconsistency in Keras Sequential model vs Functional APIUnable to save the model made with Keras functional APIIs it possible to train a CNN starting at an intermediate layer (in general and in Keras)?Training a tf.keras model with a basic low-level TensorFlow training loop doesn't workHow can I freeze last layer of my own model?Problem converting Keras Models into Layers API format models to use with tensorflow.js

.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty height:90px;width:728px;box-sizing:border-box;

0

I'm trying to convert models made with tf.keras in Python to tensorflow.js format for use in Node.js. Here are my package versions:

tensorflowjs: 1.0.1
Keras: 2.2.4
tf-nightly-2.0-preview: 2.0.0.dev20190321 (from pip install tensorflowjs)

Here is my Sequential model also remade with the functional API:

# Sequential API
model = tf.keras.Sequential()
model.add(layers.Dense(128, activation='relu', input_shape=(22050,))
model.add(layers.Dense(9, activation='softmax'))

# Functional API
inputs = tf.keras.Input(shape=(22050,))
x = layers.Dense(128, activation='relu')(inputs)
logits = layers.Dense(9, activation='softmax')(x)

When I convert the Sequential model to tfjs_layers_model using tensorflowjs_converter, it loads fine with tensorflowjs. When I do the same thing with the functional model, I get improperly formatted model config error:

Error: Improperly formatted model config for layer "_callHook":null,"_addedWeightNames":[],"_stateful":false,"id":1,"activityRegularizer":null,"inputSpec":["minNDim":2],"supportsMasking":true,"_trainableWeights":[],"_nonTrainableWeights":[],"_losses":[],"_updates":[],"_built":false,"inboundNodes":[],"outboundNodes":[],"name":"dense_38","trainable_":true,"updatable":true,"initialWeights":null,"_refCount":null,"fastWeightInitDuringBuild":true,"activation":,"useBias":true,"kernel":null,"bias":null,"DEFAULT_KERNEL_INITIALIZER":"glorotNormal","DEFAULT_BIAS_INITIALIZER":"zeros","units":128,"kernelInitializer":"scale":1,"mode":"fanAvg","distribution":"uniform","seed":null,"biasInitializer":,"kernelConstraint":null,"biasConstraint":null,"kernelRegularizer":null,"biasRegularizer":null: "input_26"

I also tried exporting as tfjs_graph_model, but tensorflowjs_converter didn't allow that. I would like the model to eventually have multiple outputs, which is why I'd like to use the functional API rather than Sequential.

python tensorflow keras tensorflow.js

share|improve this question

edited Mar 22 at 17:00

Tomka Koliada

1,2702828

asked Mar 22 at 16:28

justinswaneyjustinswaney

12

• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  This may be due to a breakage in serialization format that was recently fixed, but may not have made it to nightly yet. Can you check if this works in the latest stable tf release (1.13.1) ?
  
  – BlessedKey
  Mar 22 at 17:55
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  I uninstalled tf-nightly-2.0-preview and installed tensorflow 1.13.1 but would get errors when importing tensorflowjs. I got ImportError: cannot import name 'convert_to_constants'
  
  – justinswaney
  Mar 23 at 2:14
  

  
  
  
  

add a comment | 

0

I'm trying to convert models made with tf.keras in Python to tensorflow.js format for use in Node.js. Here are my package versions:

tensorflowjs: 1.0.1
Keras: 2.2.4
tf-nightly-2.0-preview: 2.0.0.dev20190321 (from pip install tensorflowjs)

Here is my Sequential model also remade with the functional API:

# Sequential API
model = tf.keras.Sequential()
model.add(layers.Dense(128, activation='relu', input_shape=(22050,))
model.add(layers.Dense(9, activation='softmax'))

# Functional API
inputs = tf.keras.Input(shape=(22050,))
x = layers.Dense(128, activation='relu')(inputs)
logits = layers.Dense(9, activation='softmax')(x)

When I convert the Sequential model to tfjs_layers_model using tensorflowjs_converter, it loads fine with tensorflowjs. When I do the same thing with the functional model, I get improperly formatted model config error:

Error: Improperly formatted model config for layer "_callHook":null,"_addedWeightNames":[],"_stateful":false,"id":1,"activityRegularizer":null,"inputSpec":["minNDim":2],"supportsMasking":true,"_trainableWeights":[],"_nonTrainableWeights":[],"_losses":[],"_updates":[],"_built":false,"inboundNodes":[],"outboundNodes":[],"name":"dense_38","trainable_":true,"updatable":true,"initialWeights":null,"_refCount":null,"fastWeightInitDuringBuild":true,"activation":,"useBias":true,"kernel":null,"bias":null,"DEFAULT_KERNEL_INITIALIZER":"glorotNormal","DEFAULT_BIAS_INITIALIZER":"zeros","units":128,"kernelInitializer":"scale":1,"mode":"fanAvg","distribution":"uniform","seed":null,"biasInitializer":,"kernelConstraint":null,"biasConstraint":null,"kernelRegularizer":null,"biasRegularizer":null: "input_26"

I also tried exporting as tfjs_graph_model, but tensorflowjs_converter didn't allow that. I would like the model to eventually have multiple outputs, which is why I'd like to use the functional API rather than Sequential.

python tensorflow keras tensorflow.js

share|improve this question

edited Mar 22 at 17:00

Tomka Koliada

1,2702828

asked Mar 22 at 16:28

justinswaneyjustinswaney

12

• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  This may be due to a breakage in serialization format that was recently fixed, but may not have made it to nightly yet. Can you check if this works in the latest stable tf release (1.13.1) ?
  
  – BlessedKey
  Mar 22 at 17:55
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  I uninstalled tf-nightly-2.0-preview and installed tensorflow 1.13.1 but would get errors when importing tensorflowjs. I got ImportError: cannot import name 'convert_to_constants'
  
  – justinswaney
  Mar 23 at 2:14
  

  
  
  
  

add a comment | 

I'm trying to convert models made with tf.keras in Python to tensorflow.js format for use in Node.js. Here are my package versions:

tensorflowjs: 1.0.1
Keras: 2.2.4
tf-nightly-2.0-preview: 2.0.0.dev20190321 (from pip install tensorflowjs)

Here is my Sequential model also remade with the functional API:

# Sequential API
model = tf.keras.Sequential()
model.add(layers.Dense(128, activation='relu', input_shape=(22050,))
model.add(layers.Dense(9, activation='softmax'))

# Functional API
inputs = tf.keras.Input(shape=(22050,))
x = layers.Dense(128, activation='relu')(inputs)
logits = layers.Dense(9, activation='softmax')(x)

When I convert the Sequential model to tfjs_layers_model using tensorflowjs_converter, it loads fine with tensorflowjs. When I do the same thing with the functional model, I get improperly formatted model config error:

Error: Improperly formatted model config for layer "_callHook":null,"_addedWeightNames":[],"_stateful":false,"id":1,"activityRegularizer":null,"inputSpec":["minNDim":2],"supportsMasking":true,"_trainableWeights":[],"_nonTrainableWeights":[],"_losses":[],"_updates":[],"_built":false,"inboundNodes":[],"outboundNodes":[],"name":"dense_38","trainable_":true,"updatable":true,"initialWeights":null,"_refCount":null,"fastWeightInitDuringBuild":true,"activation":,"useBias":true,"kernel":null,"bias":null,"DEFAULT_KERNEL_INITIALIZER":"glorotNormal","DEFAULT_BIAS_INITIALIZER":"zeros","units":128,"kernelInitializer":"scale":1,"mode":"fanAvg","distribution":"uniform","seed":null,"biasInitializer":,"kernelConstraint":null,"biasConstraint":null,"kernelRegularizer":null,"biasRegularizer":null: "input_26"

I also tried exporting as tfjs_graph_model, but tensorflowjs_converter didn't allow that. I would like the model to eventually have multiple outputs, which is why I'd like to use the functional API rather than Sequential.

python tensorflow keras tensorflow.js

share|improve this question

edited Mar 22 at 17:00

Tomka Koliada

1,2702828

asked Mar 22 at 16:28

justinswaneyjustinswaney

12

tensorflowjs: 1.0.1
Keras: 2.2.4
tf-nightly-2.0-preview: 2.0.0.dev20190321 (from pip install tensorflowjs)

# Sequential API
model = tf.keras.Sequential()
model.add(layers.Dense(128, activation='relu', input_shape=(22050,))
model.add(layers.Dense(9, activation='softmax'))

# Functional API
inputs = tf.keras.Input(shape=(22050,))
x = layers.Dense(128, activation='relu')(inputs)
logits = layers.Dense(9, activation='softmax')(x)

Error: Improperly formatted model config for layer "_callHook":null,"_addedWeightNames":[],"_stateful":false,"id":1,"activityRegularizer":null,"inputSpec":["minNDim":2],"supportsMasking":true,"_trainableWeights":[],"_nonTrainableWeights":[],"_losses":[],"_updates":[],"_built":false,"inboundNodes":[],"outboundNodes":[],"name":"dense_38","trainable_":true,"updatable":true,"initialWeights":null,"_refCount":null,"fastWeightInitDuringBuild":true,"activation":,"useBias":true,"kernel":null,"bias":null,"DEFAULT_KERNEL_INITIALIZER":"glorotNormal","DEFAULT_BIAS_INITIALIZER":"zeros","units":128,"kernelInitializer":"scale":1,"mode":"fanAvg","distribution":"uniform","seed":null,"biasInitializer":,"kernelConstraint":null,"biasConstraint":null,"kernelRegularizer":null,"biasRegularizer":null: "input_26"

share|improve this question

edited Mar 22 at 17:00

Tomka Koliada

1,2702828

asked Mar 22 at 16:28

justinswaneyjustinswaney

12

share|improve this question

edited Mar 22 at 17:00

Tomka Koliada

1,2702828

asked Mar 22 at 16:28

justinswaneyjustinswaney

12

edited Mar 22 at 17:00

Tomka Koliada

1,2702828

edited Mar 22 at 17:00

Tomka Koliada

1,2702828

asked Mar 22 at 16:28

justinswaneyjustinswaney

12

asked Mar 22 at 16:28

justinswaneyjustinswaney

12