Aborted (core dumped) in QuantizedInstanceNorm

[LongZE666]

Issue type

Bug

Have you reproduced the bug with TensorFlow Nightly?

Yes

Source

source

TensorFlow version

tf 2.16.1

Custom code

Yes

OS platform and distribution

Ubuntu 20.04

Mobile device

No response

Python version

No response

Bazel version

No response

GCC/compiler version

No response

CUDA/cuDNN version

No response

GPU model and memory

No response

Current behavior?

Abort is triggered when the input x does not meet the condition of dimension 4.

Standalone code to reproduce the issue

import tensorflow as tf

output_range_given = False
given_y_min = 0
given_y_max = 0
variance_epsilon = 1e-05
min_separation = 0.001
x = tf.constant([], shape=[0,0,0,1,1,0,0,0,11,9], dtype=tf.quint8)
x_min = tf.constant(-3.5e+35, shape=[], dtype=tf.float32)
x_max = tf.constant(0, shape=[], dtype=tf.float32)
tf.raw_ops.QuantizedInstanceNorm(x=x, x_min=x_min, x_max=x_max, output_range_given=output_range_given, given_y_min=given_y_min, given_y_max=given_y_max, variance_epsilon=variance_epsilon, min_separation=min_separation)

Relevant log output

2024-05-14 08:25:37.034690: F tensorflow/core/framework/tensor_shape.cc:45] Check failed: NDIMS == dims() (4 vs. 10)Asking for tensor of 4 dimensions from a tensor of 10 dimensions
Aborted (core dumped)

[SuryanarayanaY]

Hi @LongZE666 ,

I have replicated the behaviour with tf-nightly. Attached snapshot below. Thanks!