mirror of
https://github.com/ROCm/jax.git
synced 2025-04-16 11:56:07 +00:00
1885c4933c
This change prepares for allowing more flexible tag matching. For example, we may want to write "gpu" in a test and have it match both "cuda" and "rocm" devices, which we cannot do under the current API but can easily do under this design. Replace uses of device_under_test() in a context that performs an equality test with a call to test_device_matches(). Replace uses of if_device_under_test() with test_device_matches() and delete if_device_under_test(). PiperOrigin-RevId: 568923117
144 lines
4.4 KiB
Python
144 lines
4.4 KiB
Python
# Copyright 2021 The JAX Authors.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# https://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
|
|
import concurrent.futures
|
|
from functools import partial
|
|
import time
|
|
import unittest
|
|
|
|
from absl.testing import absltest
|
|
import numpy as np
|
|
|
|
import jax
|
|
from jax import lax
|
|
from jax import random
|
|
from jax import config
|
|
from jax.experimental import enable_x64, disable_x64
|
|
import jax.numpy as jnp
|
|
import jax._src.test_util as jtu
|
|
|
|
config.parse_flags_with_absl()
|
|
|
|
|
|
class X64ContextTests(jtu.JaxTestCase):
|
|
@jtu.sample_product(jit=jtu.JIT_IMPLEMENTATION)
|
|
def test_make_array(self, jit):
|
|
func = jit(lambda: jnp.array(np.float64(0)))
|
|
dtype_start = func().dtype
|
|
with enable_x64():
|
|
self.assertEqual(func().dtype, "float64")
|
|
with disable_x64():
|
|
self.assertEqual(func().dtype, "float32")
|
|
self.assertEqual(func().dtype, dtype_start)
|
|
|
|
@jtu.sample_product(
|
|
jit=jtu.JIT_IMPLEMENTATION,
|
|
enable_or_disable=[enable_x64, disable_x64],
|
|
)
|
|
def test_correctly_capture_default(self, jit, enable_or_disable):
|
|
# The fact we defined a jitted function with a block with a different value
|
|
# of `config.enable_x64` has no impact on the output.
|
|
with enable_or_disable():
|
|
func = jit(lambda: jnp.array(np.float64(0)))
|
|
func()
|
|
|
|
expected_dtype = "float64" if config._read("jax_enable_x64") else "float32"
|
|
self.assertEqual(func().dtype, expected_dtype)
|
|
|
|
with enable_x64():
|
|
self.assertEqual(func().dtype, "float64")
|
|
with disable_x64():
|
|
self.assertEqual(func().dtype, "float32")
|
|
|
|
@jtu.sample_product(jit=jtu.JIT_IMPLEMENTATION)
|
|
@jtu.run_on_devices("cpu") # Test presumes CPU precision
|
|
def test_near_singular_inverse(self, jit):
|
|
rng = jtu.rand_default(self.rng())
|
|
|
|
@partial(jit, static_argnums=1)
|
|
def near_singular_inverse(N=5, eps=1E-40):
|
|
X = rng((N, N), dtype='float64')
|
|
X = jnp.asarray(X)
|
|
X = X.at[-1].mul(eps)
|
|
return jnp.linalg.inv(X)
|
|
|
|
with enable_x64():
|
|
result_64 = near_singular_inverse()
|
|
self.assertTrue(jnp.all(jnp.isfinite(result_64)))
|
|
|
|
with disable_x64():
|
|
result_32 = near_singular_inverse()
|
|
self.assertTrue(jnp.all(~jnp.isfinite(result_32)))
|
|
|
|
@jtu.sample_product(jit=jtu.JIT_IMPLEMENTATION)
|
|
def test_while_loop(self, jit):
|
|
@jit
|
|
def count_to(N):
|
|
return lax.while_loop(lambda x: x < N, lambda x: x + 1.0, 0.0)
|
|
|
|
with enable_x64():
|
|
self.assertArraysEqual(count_to(10), jnp.float64(10), check_dtypes=True)
|
|
|
|
with disable_x64():
|
|
self.assertArraysEqual(count_to(10), jnp.float32(10), check_dtypes=True)
|
|
|
|
def test_thread_safety(self):
|
|
def func_x32():
|
|
with disable_x64():
|
|
time.sleep(0.1)
|
|
return jnp.array(np.int64(0)).dtype
|
|
|
|
def func_x64():
|
|
with enable_x64():
|
|
time.sleep(0.1)
|
|
return jnp.array(np.int64(0)).dtype
|
|
|
|
with concurrent.futures.ThreadPoolExecutor() as executor:
|
|
x32 = executor.submit(func_x32)
|
|
x64 = executor.submit(func_x64)
|
|
self.assertEqual(x64.result(), jnp.int64)
|
|
self.assertEqual(x32.result(), jnp.int32)
|
|
|
|
@jax.legacy_prng_key('allow')
|
|
@jtu.ignore_warning(category=UserWarning,
|
|
message="Explicitly requested dtype float64 is not available")
|
|
def test_jit_cache(self):
|
|
if jtu.test_device_matches(["tpu"]):
|
|
self.skipTest("64-bit random not available on TPU")
|
|
|
|
f = partial(random.uniform, random.PRNGKey(0), (1,), 'float64', -1, 1)
|
|
with disable_x64():
|
|
for _ in range(2):
|
|
f()
|
|
with enable_x64():
|
|
for _ in range(2):
|
|
f()
|
|
|
|
@unittest.skip("test fails, see #8552")
|
|
def test_convert_element_type(self):
|
|
# Regression test for part of https://github.com/google/jax/issues/5982
|
|
with enable_x64():
|
|
x = jnp.int64(1)
|
|
self.assertEqual(x.dtype, jnp.int64)
|
|
|
|
y = x.astype(jnp.int32)
|
|
self.assertEqual(y.dtype, jnp.int32)
|
|
|
|
z = jax.jit(lambda x: x.astype(jnp.int32))(x)
|
|
self.assertEqual(z.dtype, jnp.int32)
|
|
|
|
if __name__ == "__main__":
|
|
absltest.main(testLoader=jtu.JaxTestLoader())
|