This template is designed to help you create a detailed and informative bug report for TensorFlow. Please follow the steps below to ensure your report is as helpful as possible.

Steps to Follow

  1. Describe the Bug

    • What exactly is the bug? What unexpected behavior are you observing?
    • Provide a clear and concise description of the issue.

  2. Environment Details

    • TensorFlow version: pip show tensorflow
    • Python version: python --version
    • Operating system: uname -a
    • Any other relevant information (e.g., hardware, GPU, etc.)

  3. Steps to Reproduce

    • A minimal, self-contained example that demonstrates the bug.
    • Include the code that triggers the issue.

  4. Expected vs. Actual Behavior

    • Describe what you expected to happen, and what actually happened.

  5. Screenshots or Logs

    • Include screenshots or logs that illustrate the issue.

Example

Here's an example of a bug report:

  • Title: TensorFlow crashes when using GPU
  • Description: TensorFlow crashes when I try to run a simple training script on my GPU.
  • Environment Details:
    • TensorFlow version: 2.4.0
    • Python version: 3.7.5
    • Operating system: Linux x86_64
    • GPU: NVIDIA GeForce RTX 3080
  • Steps to Reproduce:
    import tensorflow as tf

model = tf.keras.models.Sequential([tf.keras.layers.Dense(10, activation='relu')])
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy')
model.fit(np.random.random((100, 10)), np.random.randint(0, 10, (100, 1)), epochs=1)
  • Expected vs. Actual Behavior:
    • Expected: The model should train without crashing.
    • Actual: TensorFlow crashes with the following error message: RuntimeError: CUDA out of memory.
  • Screenshots or Logs:
    • Bug Screenshot

For more information on creating effective bug reports, please visit our Bug Reporting Guide.

Please note that all bug reports should be respectful and non-offensive. Reports containing inappropriate content will be removed.