Update references to JAX's GitHub repo

JAX has moved from https://github.com/google/jax to https://github.com/jax-ml/jax

PiperOrigin-RevId: 702886640

README.md

@@ -325,7 +325,7 @@ performance.
 
 ### Platform Specificity
 
-Natively serialized JAX programs are platform specific ([link](https://github.com/google/jax/blob/main/jax/experimental/jax2tf/README.md#natively-serialized-jax-modules-are-platform-specific)). Executing a natively
+Natively serialized JAX programs are platform specific ([link](https://github.com/jax-ml/jax/blob/main/jax/experimental/jax2tf/README.md#natively-serialized-jax-modules-are-platform-specific)). Executing a natively
 serialized program on platforms other than the one for which it was lowered,
 would raise a ValueError, e.g.:
 
@@ -399,8 +399,8 @@ ops.
 
 [DeepMind JAX Ecosystem]: https://deepmind.com/blog/article/using-jax-to-accelerate-our-research "DeepMind JAX Ecosystem"
 [DeepMind JAX Ecosystem citation]: https://github.com/google-deepmind/jax/blob/main/deepmind2020jax.txt "Citation"
-[JAX]: https://github.com/google/jax "JAX on GitHub"
+[JAX]: https://github.com/jax-ml/jax "JAX on GitHub"
 [TensorFlow]: https://github.com/tensorflow/tensorflow "TensorFlow on GitHub"
-[jax2tf documentation]: https://github.com/google/jax/blob/master/jax/experimental/jax2tf/README.md#calling-tensorflow-functions-from-jax "jax2tf documentation"
-[jax2tf_cumulative_reduction]: https://github.com/google/jax/blob/main/jax/experimental/jax2tf/jax2tf.py#L2172
+[jax2tf documentation]: https://github.com/jax-ml/jax/blob/master/jax/experimental/jax2tf/README.md#calling-tensorflow-functions-from-jax "jax2tf documentation"
+[jax2tf_cumulative_reduction]: https://github.com/jax-ml/jax/blob/main/jax/experimental/jax2tf/jax2tf.py#L2172
 [StableHLO]: https://github.com/openxla/stablehlo

test.sh

@@ -70,7 +70,7 @@ CHECK_CUSTOM_CALLS_TEST=0 pytest -n "${N_JOBS}" --pyargs tf2jax
 # Native lowering is in active development so we test against nightly and github head.
 pip uninstall --yes tensorflow
 pip install tf-nightly
-pip install git+https://github.com/google/jax.git
+pip install git+https://github.com/jax-ml/jax.git
 pip install -U --pre jaxlib -f https://storage.googleapis.com/jax-releases/jaxlib_nightly_releases.html
 CHECK_CUSTOM_CALLS_TEST=0  pytest -n "${N_JOBS}" --pyargs tf2jax._src.roundtrip_test
 cd ..

tf2jax/_src/numpy_compat.py

@@ -76,7 +76,7 @@ def is_poly_dim(x) -> bool:
     return export.is_symbolic_dim(x)
 
   # This should reflect is_poly_dim() at
-  # https://github.com/google/jax/blob/main/jax/experimental/jax2tf/shape_poly.py#L676
+  # https://github.com/jax-ml/jax/blob/main/jax/experimental/jax2tf/shape_poly.py#L676
   # Array types.
   if isinstance(x, (np.ndarray, jax.core.Tracer, xc.ArrayImpl)):  # pylint: disable=isinstance-second-argument-not-valid-type
     return False

tf2jax/_src/ops.py

@@ -844,7 +844,7 @@ def _func(x: jnp.ndarray) -> Tuple[jnp.ndarray, jnp.ndarray]:
       evals, evecs = jnp.linalg.eigh(x, symmetrize_input=False)
     else:
       # symmetrize_input does not exist for eigvalsh.
-      # See https://github.com/google/jax/issues/9473
+      # See https://github.com/jax-ml/jax/issues/9473
       evals, evecs = jnp.linalg.eigvalsh(symmetrize(x)), None
 
     # Sorting by eigenvalues to tf.raw_ops.Eig better.
@@ -2655,7 +2655,7 @@ def _xla_variadic_sort(proto):
   return _XlaVariadicSort(dict(comparator=comparator), is_stable=is_stable)
 
 
-# Taken from https://github.com/google/jax/blob/main/jax/_src/lax/lax.py#L1056
+# Taken from https://github.com/jax-ml/jax/blob/main/jax/_src/lax/lax.py#L1056
 def _get_max_identity(dtype):
   if jax.dtypes.issubdtype(dtype, np.inexact):
     return np.array(-np.inf, dtype)
@@ -2665,7 +2665,7 @@ def _get_max_identity(dtype):
     return np.array(False, np.bool_)
 
 
-# Taken from https://github.com/google/jax/blob/main/jax/_src/lax/lax.py#L1064
+# Taken from https://github.com/jax-ml/jax/blob/main/jax/_src/lax/lax.py#L1064
 def _get_min_identity(dtype):
   if jax.dtypes.issubdtype(dtype, np.inexact):
     return np.array(np.inf, dtype)

tf2jax/experimental/mhlo.py

@@ -61,7 +61,7 @@ def mhlo_apply_impl(*args, module: MhloModule):
 mhlo_apply_p.def_impl(mhlo_apply_impl)
 
 
-# See https://github.com/google/jax/blob/main/jax/_src/interpreters/mlir.py#L115
+# See https://github.com/jax-ml/jax/blob/main/jax/_src/interpreters/mlir.py#L115
 # for reference
 def ir_type_to_dtype(ir_type: ir.Type) -> jnp.dtype:
   """Converts MLIR type to JAX dtype."""
@@ -154,7 +154,7 @@ def mhlo_apply_abstract_eval(
 
 
 # Taken from
-# github.com/google/jax/blob/main/jax/experimental/jax2tf/jax_export.py#L859
+# github.com/jax-ml/jax/blob/main/jax/experimental/jax2tf/jax_export.py#L859
 def refine_polymorphic_shapes(
     module: ir.Module, validate_static_shapes: bool
 ) -> ir.Module:

tf2jax/experimental/ops.py

@@ -31,7 +31,7 @@
 
 
 # See canonicalize_platform for reference
-# https://github.com/google/jax/blob/main/jax/_src/xla_bridge.py#L344
+# https://github.com/jax-ml/jax/blob/main/jax/_src/xla_bridge.py#L344
 def _platform_to_alias(platform: str) -> str:
   aliases = {
       "cuda": "gpu",
@@ -41,7 +41,7 @@ def _platform_to_alias(platform: str) -> str:
 
 
 # Adapted from
-# https://github.com/google/jax/commit/ec8b855fa16962b1394716622c8cbc006ce76b1c
+# https://github.com/jax-ml/jax/commit/ec8b855fa16962b1394716622c8cbc006ce76b1c
 @functools.lru_cache(None)
 def _refine_with_static_input_shapes(
     module_text: str, operands: Tuple[jax.core.ShapedArray, ...]