Fix gradient tape issue in test_gradient_tape_issue.py

[Shivam909058]

import os
import numpy as np
import tensorflow as tf
from tensorflow import keras

os.environ["KERAS_BACKEND"] = "tensorflow"
tf.random.set_seed(1234)

def create_toy_model():
inputs = keras.Input(shape=(1,))
x = keras.layers.Dense(100, activation="tanh", use_bias=True)(inputs)
x = keras.layers.Dense(1000, activation="tanh", use_bias=True)(x)
x = keras.layers.Dense(10, activation="tanh", use_bias=True)(x)
outputs = keras.layers.Dense(1, activation=None, use_bias=False)(x)
return keras.Model(inputs=inputs, outputs=outputs)

def test_gradient_tape_issue():
model = create_toy_model()

x = np.expand_dims(np.linspace(0, 10, num=20), axis=1)
x = tf.convert_to_tensor(x, dtype=tf.float32)

print(f"Number of layers: {len(model.layers)}")
for i, layer in enumerate(model.layers):
    print(f"Layer {i}: {layer.__class__.__name__}, Weights: {len(layer.weights)}")

with tf.GradientTape(watch_accessed_variables=True) as tape:
    last_layer_weights = model.layers[-1].weights[0]
    #tape.watch(last_layer_weights)
    out = model(x)

dout = tape.gradient(out, last_layer_weights)

assert dout is not None, "Gradient should not be None"
print("Gradient successfully computed!")

if name == "main":
test_gradient_tape_issue()

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 78.75%. Comparing base (458ece9) to head (a3e2089).

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #20277   +/-   ##
=======================================
  Coverage   78.75%   78.75%           
=======================================
  Files         510      510           
  Lines       48385    48385           
  Branches     8901     8901           
=======================================
  Hits        38106    38106           
  Misses       8446     8446           
  Partials     1833     1833           

Flag	Coverage Δ
keras	78.61% <ø> (ø)
keras-jax	62.15% <ø> (ø)
keras-numpy	57.31% <ø> (+<0.01%)	⬆️
keras-tensorflow	63.50% <ø> (ø)
keras-torch	62.20% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[gbaned]

Hi @Shivam909058 Can you please sign CLA? Thank you!

[gbaned]

Hi @Shivam909058 Any update on this PR? Please. Thank you!