[Python] Confused by keras/tensorflow error: ValueError: None values not supported

I'm trying to build a neural net in tensorflow/keras but I'm stuck on this error:

ValueError: None values not supported.


I've been able to reduce the code to reproduce this to the following mock example:

from tensorflow.keras import layers, Model, Input
from tensorflow.keras.models import Sequential
import pandas as pd
import numpy as np
import tensorflow as tf

input_data = pd.DataFrame(np.random.rand(1000, 10))

data = tf.data.Dataset.zip({'input_values' : tf.data.Dataset.from_tensor_slices(input_data.values)})

batch_size = 100
train_split = 0.8

train_rows = int(train_split * input_data.shape[0])
train_dataset = data.take(train_rows)
validation_dataset = data.skip(train_rows)

train_data_batched = train_dataset.batch(batch_size).prefetch(tf.data.AUTOTUNE)
validation_data_batched = validation_dataset.batch(batch_size).prefetch(tf.data.AUTOTUNE)

num_outputs = 10

input_layer = Input(shape=(num_outputs,), name=f'input_values')
output = layers.Dense(num_outputs, activation='sigmoid', name='output')(input_layer)

# Define the model
model = Model(
inputs=[input_layer],
outputs=output,
)

max_epochs = 10

def loss(y_true, y_pred):
return 2.0

model.compile(
loss=loss,
optimizer='adam',
)

history = model.fit(
train_data_batched,
epochs=max_epochs,
validation_data=validation_data_batched
)


Here is the full error:

File /tmp/virtualenvs/python3.11/lib/python3.11/site-packages/keras/src/utils/traceback_utils.py:122, in filter_traceback.<locals>.error_handler(*args, **kwargs)
119 filtered_tb = _process_traceback_frames(e.__traceback__)
120 # To get the full stack trace, call:
121 # `keras.config.disable_traceback_filtering()`
--> 122 raise e.with_traceback(filtered_tb) from None
123 finally:
124 del filtered_tb

File /tmp/virtualenvs/python3.11/lib/python3.11/site-packages/optree/ops.py:752, in tree_map(func, tree, is_leaf, none_is_leaf, namespace, *rests)
750 leaves, treespec = _C.flatten(tree, is_leaf, none_is_leaf, namespace)
751 flat_args = [leaves] + [treespec.flatten_up_to(r) for r in rests]
--> 752 return treespec.unflatten(map(func, *flat_args))

ValueError: None values not supported.


But I'm stuck as to what to do, mainly because I don't understand what the error is referring to (None values where? In a tensor? In a shape? Somewhere else?)

I'm running tensorflow[and-cuda]==2.18.0

Continue reading...