Update Python tests to reflect JAX changes

JAX no longer accepts lists in many of its NumPy API functions, but
casting them to arrays explicitly fixes the issue.

python/tests/jax_test.py

@@ -52,12 +52,12 @@ def test_jit_scale(self):
 
   def test_vmap(self):
     add_two = primitive(dex.eval(r'\x:((Fin 2)=>Float). for i. x.i + 2.0'))
-    x = jnp.linspace([0, 3], [5, 8], num=4, dtype=jnp.float32)
+    x = jnp.linspace(jnp.array([0, 3]), jnp.array([5, 8]), num=4, dtype=jnp.float32)
     np.testing.assert_allclose(jax.vmap(add_two)(x), x + 2.0)
 
   def test_vmap_nonzero_index(self):
     add_two = primitive(dex.eval(r'\x:((Fin 4)=>Float). for i. x.i + 2.0'))
-    x = jnp.linspace([0, 3], [5, 8], num=4, dtype=jnp.float32)
+    x = jnp.linspace(jnp.array([0, 3]), jnp.array([5, 8]), num=4, dtype=jnp.float32)
     np.testing.assert_allclose(
         jax.vmap(add_two, in_axes=1, out_axes=1)(x), x + 2.0)
 
@@ -72,12 +72,12 @@ def test_vmap_unbatched_array(self):
 
   def test_vmap_jit(self):
     add_two = primitive(dex.eval(r'\x:((Fin 2)=>Float). for i. x.i + 2.0'))
-    x = jnp.linspace([0, 3], [5, 8], num=4, dtype=jnp.float32)
+    x = jnp.linspace(jnp.array([0, 3]), jnp.array([5, 8]), num=4, dtype=jnp.float32)
     np.testing.assert_allclose(jax.jit(jax.vmap(add_two))(x), x + 2.0)
 
   def test_vmap_jit_nonzero_index(self):
     add_two = primitive(dex.eval(r'\x:((Fin 4)=>Float). for i. x.i + 2.0'))
-    x = jnp.linspace([0, 3], [5, 8], num=4, dtype=jnp.float32)
+    x = jnp.linspace(jnp.array([0, 3]), jnp.array([5, 8]), num=4, dtype=jnp.float32)
     np.testing.assert_allclose(
         jax.jit(jax.vmap(add_two, in_axes=1, out_axes=1))(x), x + 2.0)