From 7736f67cfb4fa847bf53dbba2ce4bf73aace3adb Mon Sep 17 00:00:00 2001
From: Surya <116063290+SuryanarayanaY@users.noreply.github.com>
Date: Wed, 31 Jan 2024 11:15:44 +0530
Subject: [PATCH] Fix bug in broadcast_to Op with Jax backend when keras
 variable as input (#19118)

* Fix bug in broadcast_to Op with Jax backend when keras variable as input

* Fix format error

* convert_to_numpy chnaged to convert_to_tensor

* Removed duplicate import of convert_to_tensor
---
 keras/backend/jax/numpy.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/keras/backend/jax/numpy.py b/keras/backend/jax/numpy.py
index ac9d10ae367..0b062489065 100644
--- a/keras/backend/jax/numpy.py
+++ b/keras/backend/jax/numpy.py
@@ -322,6 +322,7 @@ def average(x, axis=None, weights=None):
 
 
 def broadcast_to(x, shape):
+    x = convert_to_tensor(x)
     return jnp.broadcast_to(x, shape)