|---| | [ View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v2.15.1/tensorflow/python/ops/numpy_ops/np_math_ops.py#L1405-L1408) | TensorFlow variant of NumPy's `argmax`. ```python tf.experimental.numpy.argmax( a, axis=None ) ``` Unsupported arguments: `out`, `keepdims`. See the NumPy documentation for [`numpy.argmax`](https://numpy.org/doc/stable/reference/generated/numpy.argmax.html).