mirrors/rocm_jax
1
0
Fork 0
mirror of https://github.com/ROCm/jax.git synced 2025-04-14 10:56:06 +00:00
Code Issues Packages Projects Releases Wiki Activity
rocm_jax/tests/shape_poly_test.py
Sergei Lebedev 0ff234049b Removed trivial docstrings from JAX tests 
These docstrings do not make the tests any more clear and typically just duplicate the test module name.

PiperOrigin-RevId: 737611977
2025-03-17 07:49:37 -07:00

3691 lines
164 KiB
Python
Raw Permalink Blame History

# Copyright 2020 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.
from __future__ import annotations
import enum
from collections.abc import Callable, Sequence
import cProfile
import itertools
import math
import os
from pstats import Stats
from typing import Any
import unittest
from absl import logging
from absl.testing import absltest
import collections
import functools
from functools import partial
import operator as op
import re
import jax
from jax import export
from jax.experimental import pjit
from jax import lax
import jax.numpy as jnp
from jax import ops
from jax import random
from jax._src import config
from jax._src import core
from jax._src import test_util as jtu
from jax._src.export import shape_poly
from jax._src.export import shape_poly_decision
from jax._src.lax import lax as lax_internal
from jax._src.lax import control_flow as lax_control_flow
from jax._src.state import discharge
from jax._src.state import primitives as ref_primitives
import numpy as np
config.parse_flags_with_absl()
# Import after parsing flags
from jax._src.internal_test_util import test_harnesses
from jax._src.internal_test_util.test_harnesses import Harness, CustomArg, RandArg, StaticArg
_f32 = np.float32
_i32 = np.int32
DimSize = shape_poly.DimSize
expect_error_associative_scan = (
NotImplementedError,
"associative scan over axis of non-constant size",
)
# We want to have a complete test suite, even if the decision procedure is not
# yet as tight as we want. We will write the expected values as
# _expect(best=..., current=...) and we use the `current` value in the
# test, but we aspire to a decision procedure where we could compute `best`.
def _expect(*, current, best):
return current
def _bounds(e: shape_poly.DimSize) -> tuple[float, float]:
return shape_poly._bounds_decision(e, shape_poly.BoundsPrecision.BEST)
def _assert_equal_bounds(tst: jtu.JaxTestCase,
e: shape_poly.DimSize,
bounds: tuple[float, float]):
if isinstance(e, shape_poly._DimExpr):
scope = e.scope
else:
scope = shape_poly.SymbolicScope()
decision = shape_poly._make_decision_state(scope)
found_bounds = decision.bounds(e)
tst.assertEqual(bounds, found_bounds)
def _start_profile(tst: jtu.JaxTestCase):
tst.prof = None
if os.getenv("JAX_PROFILE_TEST", False):
tst.prof = cProfile.Profile()
tst.prof.enable()
def _stop_profile(tst: jtu.JaxTestCase):
if tst.prof is not None:
outfile = os.getenv("JAX_PROFILE_OUT")
if outfile:
tst.prof.dump_stats(outfile)
p = Stats(tst.prof)
p.strip_dirs()
p.sort_stats("cumtime").print_stats(.2)
p.print_callers(.2)
class DimExprTest(jtu.JaxTestCase):
def setUp(self):
_start_profile(self)
super().setUp()
def tearDown(self):
super().tearDown()
_stop_profile(self)
class AssertionType(enum.Enum):
EQ = 1,
GEQ = 2
def sampled_assertion(self,
e: DimSize,
fun: Callable,
*operands_sym: shape_poly.DimSize,
assertion: AssertionType = AssertionType.EQ
):
"""Checks `assertion(e, fun(*operands))` symbolically and concretely.
For the concrete check, it will sample the space of dimension variable
assignments for the dimension variables in `e`.
This is useful when `fun` can operate both with symbolic and with
concrete values, and we want to check that the behavior is sound.
"""
computed_sym = fun(*operands_sym)
assertion_fun = {
DimExprTest.AssertionType.EQ: self.assertEqual,
DimExprTest.AssertionType.GEQ: self.assertGreaterEqual}[assertion]
assertion_fun(e, computed_sym)
dim_vars: set[str] = set()
for a in (e, computed_sym, *operands_sym):
if core.is_symbolic_dim(a):
dim_vars = dim_vars.union(a._get_vars())
if not dim_vars:
return
dim_vars_tuple = tuple(dim_vars)
# All combinations of values
for dim_values in itertools.product(*([(1, 2, 5, 10)] * len(dim_vars_tuple))):
env = dict(zip(dim_vars_tuple, dim_values))
def eval(d: shape_poly.DimSize):
return d._evaluate(env) if core.is_symbolic_dim(d) else d # type: ignore
compute_concrete = fun(*map(eval, operands_sym))
expected_concrete = eval(e)
assertion_fun(
expected_concrete, compute_concrete,
f"{e=} {expected_concrete=} {compute_concrete=} {env=}")
def test_parse_shape(self):
self.assertEqual((), shape_poly.symbolic_shape("", like=()))
self.assertEqual((), shape_poly.symbolic_shape("()", like=()))
self.assertEqual((2, 3), shape_poly.symbolic_shape(None, like=(2, 3)))
self.assertEqual((2, 3), shape_poly.symbolic_shape("2, 3,", like=(2, 3)))
self.assertEqual((2, 3), shape_poly.symbolic_shape("2, _", like=(2, 3)))
self.assertEqual((2, 3), shape_poly.symbolic_shape("2, ...", like=(2, 3)))
self.assertEqual((2,), shape_poly.symbolic_shape("2, ...", like=(2,)))
self.assertEqual((2, 3), shape_poly.symbolic_shape("...", like=(2, 3)))
self.assertEqual((2, 3), shape_poly.symbolic_shape(" ( 2 , 3 ) "))
a, b = shape_poly.symbolic_shape("a, b")
self.assertEqual((a, 3), shape_poly.symbolic_shape("(a, ...) ", like=(None, 3),
scope=a.scope))
a, b = shape_poly.symbolic_shape("a, b")
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=f"_{dim_spec}",
dim_spec=dim_spec, dim_poly=dim_poly, expected_str=expected_str)
for dim_spec, dim_poly, expected_str in [
("2*b*a", 2 * a * b, "2*a*b"),
("-2 * b * a^2 + b^2", -2 * a * a * b + b * b,
"- 2*a^2*b + b^2"),
("-2 * a^2 * b + -1 *b^2*a", -2 * a * a * b - a * b * b,
"- a*b^2 - 2*a^2*b"),
("3 * a * b * a - 2", 3 * a * b * a - 2,
"3*a^2*b - 2"),
("a - b", a - b, "- b + a"),
("b - a", b - a, "b - a"),
("-1 * a", -a, "- a"),
("-2*a", -2 * a, "- 2*a"),
("a + -1", a - 1, "a - 1"),
("a + 1 ,", a + 1, "a + 1"),
("a - 1", a - 1, "a - 1"),
("+ a - b", a - b, "- b + a"),
("+ 2", 2, "2"),
("3 * a * mod(a + 2, b + 2)", 3 * a * ((a + 2) % (b + 2)),
"3*a*mod(a + 2, b + 2)"),
("3 * floordiv(a + 2, b + 2) * 2", 3 * ((a + 2) // (b + 2)) * 2,
"6*floordiv(a + 2, b + 2)"),
("max(a, b)", "build_inside", "max(a, b)"),
("min(a, b)", "build_inside", "min(a, b)"),
]])
def test_parse_dim(self, dim_spec, dim_poly, expected_str):
if dim_spec == "max(a, b)":
dim_poly = core.max_dim(DimExprTest.a, DimExprTest.b)
elif dim_spec == "min(a, b)":
dim_poly = core.min_dim(DimExprTest.a, DimExprTest.b)
self.assertEqual((dim_poly,), shape_poly.symbolic_shape(dim_spec,
scope=DimExprTest.a.scope))
self.assertEqual(expected_str, str(dim_poly))
self.assertEqual((dim_poly,), shape_poly.symbolic_shape(expected_str,
scope=DimExprTest.a.scope))
@jtu.parameterized_filterable(
kwargs=[
dict(dim_spec=dim_spec)
for dim_spec in [
"b + a",
"b - a",
"b + 3*a",
"a*b + a^2 + b + a",
"mod(a, 4) + floordiv(a, 4) + a",
"2*a^2 - 3*a - 1",
"a^2 + 3*a - 1",
"- a + 3",
"- mod(a, 4) + 3",
"- 2*a + 3",
"a*floordiv(b, 8)*mod(b, 4)",
]
]
)
def test_print_dim(self, *, dim_spec: str):
e, = shape_poly.symbolic_shape(dim_spec)
self.assertEqual(str(e), dim_spec)
@jtu.parameterized_filterable(
kwargs=[
# sanitized shape_spec sometimes collide
dict(testcase_name=(
f"_shape_spec={shape_spec}" +
{"...": "name=ellipsis",
")(": "name=bad_parens",
"a ;": "name=a_semicolon",
"'a'": "name=a_quotes"}.get(shape_spec, "")),
shape_spec=shape_spec)
for shape_spec in [
"2.5", "a + a a", "a ^ a", "a; a",
"_", "...", "a ;", ")(", "2a", "a@", "'a'", "('a', ...)",
"mod(a)", "floordiv(a, b, c)", "..., 3"
]])
def test_parse_error(self,
shape_spec="a + a a"):
with self.assertRaisesRegex(ValueError,
"syntax error in symbolic shape"):
shape_poly.symbolic_shape(shape_spec)
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=f"_{shape_spec=}",
shape_spec=shape_spec, arg_shape=arg_shape, error_msg=error_msg)
for shape_spec, arg_shape, error_msg in [
("3", (4,), "different size"),
("b, 3", (None, 4), "different size"),
("b, ...", None, "no 'like' shape was given"),
("b, 5, _", None, "no 'like' shape was given"),
("b, 6, c, _", (4, 6, 7),
"because we parsed 3 already and 'like' shape has only 3 dimensions"),
("b, 7, c, ...", (4, 7),
"because we parsed 3 already and 'like' shape has only 2 dimensions"),
]])
def test_parse_mismatch_error(self, *,
shape_spec, arg_shape, error_msg):
with self.assertRaisesRegex(ValueError,
"syntax error in symbolic shape .*" + error_msg):
shape_poly.symbolic_shape(shape_spec, like=arg_shape)
def test_dim_vars(self):
a, b, a1 = shape_poly.symbolic_shape("a, b, a")
self.assertEqual(True, a == a)
self.assertEqual(True, a == a1)
self.assertEqual(False, a != a)
self.assertNotEqual(a, b)
self.assertLen({a, a}, 1)
self.assertLen({a, b}, 2)
self.assertIn(a, {a, b})
self.assertIn(b, {a, b})
self.assertIn(a, [a, b])
self.assertIn(b, [a, b])
def test_sizes(self):
a, b = shape_poly.symbolic_shape("a, b")
self.assertEqual(a._size, b._size)
self.assertGreater(a._size, shape_poly._ensure_poly(1, "", a.scope)._size)
self.assertGreater((a + 1)._size, a._size)
self.assertGreater((a + b)._size, a._size)
self.assertGreater((a + b)._size, (a + 1)._size)
self.assertGreater((2 * a)._size, a._size)
self.assertGreater((2 * a)._size, (a + 1)._size)
self.assertGreater((a * a)._size, a._size)
self.assertGreater((a * a * a)._size, (a * a)._size)
self.assertGreater((a * b)._size, a._size)
def test_get_vars(self):
a, b = shape_poly.symbolic_shape("a, b")
self.assertEqual({"a"}, a._get_vars())
self.assertEqual({"a", "b"}, (a * b * a)._get_vars())
def test_evaluate(self):
a, b = shape_poly.symbolic_shape("a, b")
self.assertEqual(1, (a * a - b)._evaluate(dict(a=2, b=3)))
self.assertEqual(1, ((a * a) // b)._evaluate(dict(a=2, b=3)))
self.assertEqual(4, ((a * a) % b)._evaluate(dict(a=5, b=7)))
def test_dim_vars_definitely_equal(self):
a, b = shape_poly.symbolic_shape("a, b")
self.assertTrue(core.definitely_equal(a, a))
self.assertFalse(core.definitely_equal(a, 1))
self.assertFalse(core.definitely_equal(a, b))
self.assertTrue(core.definitely_equal_one_of_dim(a, [2, a]))
self.assertFalse(core.definitely_equal_one_of_dim(a, [2, b]))
self.assertFalse(core.definitely_equal_one_of_dim(a, []))
self.assertTrue(core.definitely_equal_one_of_dim(2, [a, 3, 2]))
self.assertFalse(core.definitely_equal_one_of_dim(1, [2, b]))
self.assertFalse(core.definitely_equal_one_of_dim(3, []))
self.assertTrue(core.definitely_equal(1, jnp.add(0, 1))) # An Array
self.assertFalse(core.definitely_equal(1, "a"))
def test_factors_ordering(self):
a, b = shape_poly.symbolic_shape("a, b")
self.assertTrue(a._to_factor() < b._to_factor())
self.assertFalse(a._to_factor() >= b._to_factor())
self.assertTrue(a._to_factor() <= b._to_factor())
self.assertTrue(a._to_factor() != b._to_factor())
self.assertTrue(a._to_factor() < (a % 4)._to_factor())
self.assertFalse(a._to_factor() > (a % 4)._to_factor())
# FLOORDIV comes before MON because we compare operations alphabetically
self.assertTrue((a // 4)._to_factor() < (a % 4)._to_factor())
self.assertEqual(hash((a // 4)._to_factor()), hash((a // 4)._to_factor()))
def test_term_ordering(self):
a, b = shape_poly.symbolic_shape("a, b")
self.assertTrue(a._to_term() < b._to_term())
self.assertTrue(a._to_term() <= b._to_term())
self.assertTrue(b._to_term() >= a._to_term())
self.assertTrue(b._to_term() > a._to_term())
self.assertTrue(((3 * b) // a)._to_term() >= ((2 * b) // a)._to_term())
self.assertTrue(((3 * b) // a)._to_term() <= ((4 * a) // b)._to_term())
self.assertTrue(a._to_term() < (a * a)._to_term())
self.assertTrue(b._to_term() < (a * a)._to_term())
self.assertTrue((a * a * b)._to_term() < (a * b * b)._to_term())
e1 = 2 + a * a * b + a * b * b + a * b + a * a + a + b
sorted_e1 = [shape_poly._DimExpr._from_term(m, m_count, a.scope)
for m, m_count in e1._sorted_terms]
self.assertSequenceEqual(sorted_e1,
[a * b * b, a * a * b, a * b, a * a, b, a, 2])
e2 = a * (a // 4) + (a // 4) + b * (a // 4) + b * (a % 4) + a * a + b + 15
sorted_e2 = [shape_poly._DimExpr._from_term(m, m_count, a.scope)
for m, m_count in e2._sorted_terms]
self.assertSequenceEqual(sorted_e2,
[b * (a % 4), b * (a // 4), a * (a // 4), a // 4,
a * a, b, 15])
def test_bounds_arithmetic(self):
a, b, c = shape_poly.symbolic_shape("a, b, c")
bounded_le4 = 5 - a
bounded_ge2 = b + 1
bounded_ge0_le4 = a % 5
self.assertEqual(_bounds(a), (1, np.inf))
self.assertEqual(_bounds(- a), (-np.inf, -1))
self.assertEqual(_bounds(bounded_le4), (-np.inf, 4))
self.assertEqual(_bounds(bounded_ge2), (2, np.inf))
self.assertEqual(_bounds(bounded_ge0_le4), (0, 4))
self.assertEqual(_bounds(b + 1 - b), (1, 1))
self.assertEqual(_bounds(a), (1, np.inf))
self.assertEqual(_bounds(a - 1), (0, np.inf))
self.assertEqual(_bounds(5*a - 2), (3, np.inf))
self.assertEqual(_bounds(-5*a - 2), (-np.inf, -7))
self.assertEqual(_bounds(2*a + 3*b + 5*c - 1), (9, np.inf))
self.assertEqual(_bounds(2*a - 3*b + 5*c - 1), (-np.inf, np.inf))
self.assertEqual(_bounds(-2*a + -3*b + -5*c + 20), (-np.inf, 10))
# Additions
self.assertEqual(_bounds(bounded_ge0_le4 + bounded_le4), (-np.inf, 8))
self.assertEqual(_bounds(bounded_ge0_le4 + bounded_ge2), (2, np.inf))
self.assertEqual(_bounds(bounded_le4 + bounded_ge2), (-np.inf, np.inf))
# Subtractions
self.assertEqual(_bounds(bounded_ge0_le4 - bounded_le4), (-4, np.inf))
self.assertEqual(_bounds(- bounded_ge0_le4 + bounded_le4), (-np.inf, 4))
self.assertEqual(_bounds(bounded_ge0_le4 - bounded_ge2), (-np.inf, 2))
self.assertEqual(_bounds(- bounded_ge0_le4 + bounded_ge2), (-2, np.inf))
self.assertEqual(_bounds(bounded_le4 - bounded_ge2), (-np.inf, 2))
self.assertEqual(_bounds(- bounded_le4 + bounded_ge2), (-2, np.inf))
# Multiplications
self.assertEqual(_bounds(2 * a - 3), (-1, np.inf))
self.assertEqual(_bounds(-2 * a - 3), (-np.inf, -5))
self.assertEqual(_bounds(3 * a * b * b + 5 * a - 7), (1, np.inf))
self.assertEqual(_bounds(3 * a * b * b - 5 * a - 7), (-np.inf, np.inf))
self.assertEqual(_bounds(a + b - a * b + a * b * a), (-np.inf, np.inf))
self.assertEqual(_bounds(a + 2 * b - a), (2, np.inf))
self.assertEqual(_bounds(a + 2 * b - a), (2, np.inf))
# Higher order polynomial
self.assertEqual(_bounds(a*a + b - 2), (0, np.inf))
self.assertEqual(_bounds(-2*a*b - b - 2), (-np.inf, -5))
def test_bounds_mod(self):
a, b = shape_poly.symbolic_shape("a, b")
# self.assertEqual(_bounds(5 - a % 5), (1, 5))
self.assertEqual(_bounds(-5 - a % (-5)), (-5, -1))
self.assertEqual(_bounds(a - 5 % a), (1, np.inf))
self.assertEqual(_bounds(a - 5 % a), (1, np.inf))
self.assertEqual(_bounds(3 * (a + b) - 5 % (3 * (a + b))), (1, np.inf))
self.assertEqual(_bounds(- a + (b - 5) % a), (-np.inf, -1))
# mod
self.assertEqual(_bounds((b + 1) % 2), (0, 1))
self.assertEqual(_bounds((b + 1) % -2), (-1, 0))
self.assertEqual(_bounds((b - 4) % 2), (0, 1))
self.assertEqual(_bounds((b + 1) % a), (0, np.inf))
self.assertEqual(_bounds(11 % (a + 1)), (0, np.inf))
self.assertEqual(_bounds(-11 % (a + 1)), (0, np.inf))
self.assertEqual(_bounds(b % (a - 2)), (-np.inf, np.inf))
# This arises in convolutions, because we use "-2 * div(-b, 2)" to get
# the "2*ceil(b / 2)".
self.assertGreaterEqual(-2 * ((- b) // 2), b)
def test_bounds_floordiv(self):
a, b = shape_poly.symbolic_shape("a, b")
self.assertEqual(_bounds((a + 4) // 2), (2, np.inf))
self.assertEqual(_bounds((a + 4) // -2), (-np.inf, -3))
self.assertEqual(_bounds((a + 5) // 2), (3, np.inf))
self.assertEqual(_bounds((a + 5) // -2), (-np.inf, -3))
self.assertEqual(_bounds(11 // (a + 1)), (0, 5))
self.assertEqual(_bounds(-11 // (a + 1)), (-6, -1))
self.assertEqual(_bounds(-11 // (- a)), (0, 11)) # finite negative dividend, infinite divisor
self.assertEqual(_bounds((b + 1) // (a + 1)), (0, np.inf))
self.assertEqual(_bounds(-b // (a + 1)), (-np.inf, -1))
self.assertEqual(_bounds(a - a // 2), (1, np.inf))
self.assertEqual(_bounds((a + 3) - (a + 3) // 2), (2, np.inf))
self.assertEqual(_bounds((a + 6) - 1 * (a + 6) // 4), (6, np.inf))
self.assertEqual(_bounds((a + 6) - 2 * ((a + 6) // 4)), (4, np.inf))
self.assertEqual(_bounds((a + 6) - 3 * ((a + 6) // 4)), (2, np.inf))
self.assertEqual(_bounds(a - 2 * (a // 2)), (0, 1))
with self.assertRaisesRegex(core.InconclusiveDimensionOperation,
"Possible division by 0"):
_bounds(a // (a - 3))
def test_bounds_floordiv_against_concrete_evaluation(self):
a, b = shape_poly.symbolic_shape("a, b")
# Generate test cases for floordiv and mod: (a + N) // +-2, (N - a) // +-2
# and then evaluate them for a = 1, 5, 10000
div_mod_factors = [
operation(op1 + n, div)
for op1 in (a, a + 10, a + 11, -a, -a + 10, -a + 11)
for n in (-3, -1, 0, 1, 3)
for div in (-2, 2, a + 4, -4 - a) # Either negative, or positive
for operation in (op.floordiv, op.mod)
]
for fact in div_mod_factors:
lb, ub = _bounds(fact)
self.assertLessEqual(lb, ub)
for a_val in (1, 5, 10000):
fact_val = fact._evaluate(dict(a=a_val))
self.assertGreaterEqual(fact_val, lb)
self.assertLessEqual(fact_val, ub)
def test_max_dim(self):
a, b, c, d = shape_poly.symbolic_shape("a, b, c, d")
self.assertEqual(_bounds(core.max_dim(a, b)), (1, np.inf))
self.assertEqual(_bounds(core.max_dim(2, b)), (2, np.inf))
self.assertEqual(_bounds(core.max_dim(a, 2)), (2, np.inf))
self.assertEqual(_bounds(core.max_dim(a - 5, 1)), (1, np.inf))
self.assertEqual(_bounds(core.max_dim(15 - a, 0)), (0, 14))
self.assertEqual(_bounds(core.max_dim(15 - a, 0) // 3), (0, 4))
self.assertEqual(a + 2, core.max_dim(a, a + 2))
self.assertEqual(a , core.max_dim(a, a - 2))
self.assertEqual(core.max_dim(a % 2 - 1, 0), 0)
self.assertEqual(core.max_dim(0, a % 2 - 1), 0)
self.assertGreaterEqual(core.max_dim(a, b), a)
self.assertGreaterEqual(core.max_dim(a, b) + c - 1, a)
self.assertGreaterEqual(core.max_dim(a, b), b)
self.assertGreaterEqual(core.max_dim(a, b) + c - 1, b)
self.assertGreaterEqual(core.max_dim(a, b), core.min_dim(a, b))
self.assertEqual(_bounds(b - core.max_dim(b - a, 0)), (1, np.inf))
self.assertEqual(_bounds(a - core.min_dim(a, b)), (0, np.inf))
self.assertEqual(_bounds(b - core.min_dim(a, b)), (0, np.inf))
self.assertEqual(_bounds(core.max_dim(a, b) - a), (0, np.inf))
self.assertEqual(_bounds(core.max_dim(a, b) - b), (0, np.inf))
self.assertEqual((0, 0), _bounds(core.max_dim(1 - b, 0)))
self.assertEqual(_bounds(core.max_dim(a, b) - a - core.max_dim(b - a, 0)),
_expect(best=(0, 0), current=(0, np.inf)))
self.assertEqual(_bounds(core.max_dim(a, b) + core.max_dim(c, d) -
core.max_dim(a + core.max_dim(c, d),
b + core.max_dim(c, d))),
_expect(best=(0, 0), current=(0, np.inf)))
self.assertEqual(_bounds(core.max_dim(a, b) + core.min_dim(c, d) -
core.max_dim(a + core.min_dim(c, d),
b + core.min_dim(c, d))),
_expect(best=(0, 0), current=(0, np.inf)))
self.sampled_assertion(core.max_dim(a, 5), core.max_dim, a, 5)
self.sampled_assertion(core.max_dim(5, a), core.max_dim, 5, a)
def test_min_dim(self):
a, b, c = shape_poly.symbolic_shape("a, b, c")
self.assertEqual(_bounds(core.min_dim(a, b)), (1, np.inf))
self.assertEqual(_bounds(core.min_dim(2, b)), (1, 2))
self.assertEqual(core.min_dim(a, -2), -2)
self.assertEqual(_bounds(core.min_dim(a - 5, 1)), (-4, 1))
self.assertEqual(_bounds(core.min_dim(15 - a, 10)), (-np.inf, 10))
self.assertEqual(_bounds(core.min_dim(15 - a, 20)), (-np.inf, 14))
# Test simplification during construction
self.assertEqual(a, core.min_dim(a, a + 2))
self.assertEqual(a - 2, core.min_dim(a, a - 2))
self.assertEqual(core.min_dim(a % 2 - 1, -1), -1)
self.assertEqual(core.min_dim(-1, a % 2 - 1), -1)
self.assertGreaterEqual(a, core.min_dim(a, b))
self.assertGreaterEqual(a + c - 1, core.min_dim(a, b))
self.assertGreaterEqual(b, core.min_dim(a, b))
self.assertGreaterEqual(b + c - 1, core.min_dim(a, b))
self.sampled_assertion(core.min_dim(a, 5), core.min_dim, a, 5)
self.sampled_assertion(core.min_dim(5, a), core.min_dim, 5, a)
def test_min_max_type_check(self):
a, = shape_poly.symbolic_shape("a")
for i, f in enumerate([lambda x: core.max_dim(x, a),
lambda x: core.max_dim(a, x),
lambda x: core.min_dim(x, a),
lambda x: core.min_dim(a, x)]):
with self.subTest(f"jit_{i}"):
with self.assertRaisesRegex(core.ConcretizationTypeError, ""):
jax.jit(f)(1)
arr = jnp.array([1], dtype=np.int32)
for i, f in enumerate([lambda: core.max_dim(arr, a),
lambda: core.max_dim(a, arr),
lambda: core.min_dim(arr, a),
lambda: core.min_dim(a, arr)]):
with self.subTest(f"array_{i}"):
with self.assertRaisesRegex(
TypeError,
"Only integer scalar arrays can be converted to a scalar index"):
f()
def test_clamp_dim(self):
a, b = shape_poly.symbolic_shape("a, b")
# Clamping b <= a <= b + 10
clamp = core.max_dim(core.min_dim(a, b + 10), b)
self.assertLessEqual(b, clamp)
self.assertLessEqual(clamp, b + 10)
def test_bounds_complex(self):
a, b = shape_poly.symbolic_shape("a, b")
min_a_b = b - core.max_dim(0, b - a)
# This comes up in slicing with stride
self.assertGreaterEqual(min_a_b // 2, 0)
# Comes up in slice[-2:-2b-2:-2]
self.assertGreaterEqual((-1 * core.max_dim(-1, a - 2 * b - 2) + core.max_dim(-1, a - 2) + 1),
0)
self.assertGreaterEqual((-1 * core.max_dim(-1, a - 2 * b - 2) + core.max_dim(-1, a - 2) + 1) // 2,
0)
self.assertEqual((0, 0),
_bounds(core.max_dim(-1*b + 1, 0) // 2))
def test_equal(self):
a, b = shape_poly.symbolic_shape("a, b")
poly3 = a + 3 - a
self.assertEqual(poly3, 3)
self.assertEqual(poly3, np.array(3, np.int64))
self.assertEqual(poly3, np.array(3, np.int64)[()])
self.assertNotEqual(poly3 + 1, 3)
self.assertNotEqual(poly3, poly3 + 1)
self.assertTrue((2 * a * b * a + 3)._eq(1 + b * a * a + a * a * b + 2))
self.assertFalse((2 * a * b * a + 3)._eq(a * b * a + 3))
self.assertFalse((a * b * a + 3)._eq(a * b * a + 4))
self.assertFalse((2 * a * b * a)._eq(a * b * a))
self.assertFalse((2 * a * b * a + 1)._eq(a * b * a))
self.assertFalse((3 * a * b * a - 1)._eq(a * b * a))
self.assertFalse((3 * a * b * a - 2)._eq(a * b * a))
self.sampled_assertion(a % b,
lambda x: x, a % b)
self.sampled_assertion(a % b - a % b,
lambda x: x, 0)
self.sampled_assertion(a // b,
lambda x: x, a // b)
self.sampled_assertion(a // b - a // b,
lambda x: x, 0)
self.sampled_assertion(a % b,
lambda x: x, (2 * a // 2) % (a + b - a))
self.sampled_assertion(a // b,
lambda x: x, (2 * a // 2) // (a + b - a))
self.sampled_assertion(a, lambda x: x,
a + (a + b) // b - (b + a) // b)
# Test the normalization (a // b) * b == a - a % b
# We sacrifice this with a faster implementation of equality.
# We could fix this by adding: `core.expensive_eq()`
# self.sampled_assertion((a // 2) * 2,
# lambda x: x, a - a % 2)
# self.sampled_assertion((a // 2) + (a // 2),
# lambda x: x, a - a % 2)
# self.sampled_assertion((a // 2) * 6,
# lambda x: x, 3 * a - 3 * (a % 2))
def test_compare_ge(self):
a, b = shape_poly.symbolic_shape("a, b")
poly = 4 * a + b + 3
self.assertGreaterEqual(a, a)
self.assertGreaterEqual(a, 0)
self.assertGreaterEqual(a, 1)
with self.assertRaisesRegex(core.InconclusiveDimensionOperation, "inconclusive"):
a >= 2
self.assertGreaterEqual(poly, 0)
self.assertGreaterEqual(poly, 8)
self.assertGreaterEqual(poly, poly)
self.assertGreaterEqual(poly, poly - 1)
with self.assertRaisesRegex(
core.InconclusiveDimensionOperation,
re.escape("comparison 'b + 4*a + 3' >= '9' is inconclusive")):
poly >= 9
with self.assertRaisesRegex(
core.InconclusiveDimensionOperation,
re.escape("comparison 'b + 4*a + 3' > '9' is inconclusive")):
poly > 9
with self.assertRaisesRegex(core.InconclusiveDimensionOperation,
"inconclusive"):
(4 * a - b) >= 0
# LHS is an integer
self.assertLessEqual(8, poly)
self.assertLess(7, poly)
self.assertGreaterEqual(-8, -poly)
self.assertGreater(-7, -poly)
def test_int_results(self):
# Whenever the result is an integer, it should be represented as a
# Python integer, not a symbolic dimension.
a, b = shape_poly.symbolic_shape("a, b")
self.assertEqual(a + 2 - a, 2)
self.assertIsInstance(a + 2 - a, int)
self.assertEqual(a + (2 - a), 2)
self.assertIsInstance(a + (2 - a), int)
self.assertEqual(a * 2 // a, 2)
self.assertIsInstance(a * 2 // a, int)
self.assertEqual((4*a)*0, 0)
self.assertEqual(0*(4*a), 0)
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=f"_D={dividend}_d={divisor}_q={quotient}_r={remainder}",
dividend=dividend, divisor=divisor, quotient=quotient,
remainder=remainder)
for dividend, divisor, quotient, remainder in [
(a, 1, a, 0),
(3 * a, 3, a, 0),
(3 * a + 3, 3, a + 1, 0),
(3 * a + 2, 3, a, 2),
(3 * a + 5, 3, a + 1, 2),
(3 * a - 2, 3, a - 1, 1),
(3 * a * a * b + 2 * b * b * a, a * b, 3 * a + 2 * b, 0),
(a * a - b * b, a + b, a - b, 0),
(256 * a * b, 32, 8 * a * b, 0),
(0, b, 0, 0),
(a, b, "floordiv(a, b)", "mod(a, b)"),
(3 * a, 2, "floordiv(3*a, 2)", "mod(3*a, 2)"),
(2 * a * b + b * b, a + b, "floordiv(2*a*b + b^2, b + a)", "mod(2*a*b + b^2, b + a)"),
(3, a, "floordiv(3, a)", "mod(3, a)"),
]])
def test_divmod(self, *, dividend, quotient, divisor, remainder):
if isinstance(quotient, str):
d1, d2 = divmod(dividend, divisor)
self.assertEqual((quotient, remainder), (str(d1), str(d2)))
else:
self.sampled_assertion(quotient, lambda *args: divmod(*args)[0],
dividend, divisor)
self.sampled_assertion(remainder, lambda *args: divmod(*args)[1],
dividend, divisor)
def test_unit_combine_term_with_constraints(self):
a, b, c, d, e = shape_poly.symbolic_shape("a, b, c, d, e",
constraints=[
"a >= 4",
"b <= 5",
"c + 3*d >= 10", "c - 2*d >= -4", # -> 5c >= 8 -> c >= 2
])
scope = a.scope
def _m(e: shape_poly._DimExpr) -> shape_poly._DimTerm:
return e._to_term()
Comparator = shape_poly.Comparator
decision = shape_poly_decision._DecisionByElimination(scope).initialize()
self.assertSetEqual(
set(),
set(decision.combine_term_with_existing(_m(e), 2, scope=scope)))
self.assertSetEqual(
{(Comparator.GEQ, a - 4, 2, -1)},
set(decision.combine_term_with_existing(_m(a), 2, scope=scope)))
self.assertSetEqual(
{(Comparator.GEQ, a - 4, 2, 1)},
set(decision.combine_term_with_existing(_m(a), -2, scope=scope)))
self.assertSetEqual(
{(Comparator.GEQ, b - 1, 2, -1),
(Comparator.GEQ, -b + 5, 2, 1)},
set(decision.combine_term_with_existing(_m(b), 2, scope=scope)))
self.assertSetEqual(
{(Comparator.GEQ, b - 1, 2, 1),
(Comparator.GEQ, -b + 5, 2, -1)},
set(decision.combine_term_with_existing(_m(b), -2, scope=scope)))
self.assertSetEqual(
{(Comparator.GEQ, c - 2, 2, -1),
(Comparator.GEQ, -2*d + c + 4, 2, -1),
(Comparator.GEQ, 3*d + c - 10, 2, -1)},
set(decision.combine_term_with_existing(_m(c), 2, scope=scope,
only_smaller_than_t=False)))
self.assertSetEqual(
{(Comparator.GEQ, c - 2, 2, 1),
(Comparator.GEQ, -2*d + c + 4, 2, 1),
(Comparator.GEQ, 3*d + c - 10, 2, 1)},
set(decision.combine_term_with_existing(_m(c), -2, scope=scope,
only_smaller_than_t=False)))
self.assertSetEqual(
{(Comparator.GEQ, c - 2, 2, -1)},
set(decision.combine_term_with_existing(_m(c), 2, scope=scope,
only_smaller_than_t=True)))
self.assertSetEqual(
{(Comparator.GEQ, d - 1, 2, -1),
(Comparator.GEQ, -2*d + c + 4, 2, 2),
(Comparator.GEQ, 3*d + c - 10, 2, -3)},
set(decision.combine_term_with_existing(_m(d), 2, scope=scope,
only_smaller_than_t=False)))
self.assertSetEqual(
{(Comparator.GEQ, d - 1, 2, -1),
(Comparator.GEQ, -2*d + c + 4, 2, 2),
(Comparator.GEQ, 3*d + c - 10, 2, -3)},
set(decision.combine_term_with_existing(_m(d), 2, scope=scope,
only_smaller_than_t=True)))
def test_dilate_dim(self):
"""0 if d == 0 else 1 + dilation * (d - 1))"""
a, = shape_poly.symbolic_shape("a,")
self.sampled_assertion(4, core.dilate_dim, 2, 3)
self.sampled_assertion(7, core.dilate_dim, 3, 3)
self.sampled_assertion(0, core.dilate_dim, 0, 3)
self.sampled_assertion(a, core.dilate_dim, a, 1)
self.sampled_assertion(2 * a - 1, core.dilate_dim, a, 2)
self.sampled_assertion(core.max_dim(2 * a - 3, 0),
core.dilate_dim, a - 1, 2)
def test_stride_dim(self):
"""(d - window_size) // window_stride + 1
If d < window_size, returns 0.
"""
a, stride = shape_poly.symbolic_shape("a, s")
self.sampled_assertion(8, core.stride_dim, 10, 3, 1)
self.sampled_assertion(9, core.stride_dim, 20, 3, 2)
self.sampled_assertion(9, core.stride_dim, 20, 4, 2)
self.sampled_assertion(a, core.stride_dim, a, 1, 1)
self.sampled_assertion(a - 1, core.stride_dim, a, 2, 1)
self.sampled_assertion(a + 1, core.stride_dim, a * stride + 2, 2, stride)
self.sampled_assertion((a - 1) // 2 + 1, core.stride_dim, a, 1, 2)
self.sampled_assertion(core.max_dim((a - 4) // 2 + 1, 0),
core.stride_dim, a, 4, 2)
def test_constraints_ge_basic(self):
a, b = shape_poly.symbolic_shape("a, b",
constraints=("a >= 5", "b <= 16"))
self.assertEqual(_bounds(a), (5, np.inf))
self.assertEqual(_bounds(b), (1, 16))
def test_constraints_ge_trivial(self):
a, = shape_poly.symbolic_shape("a",
# Trivial, is dropped
constraints=("a <= a + 1",))
self.assertSequenceEqual(a.scope._explicit_constraints, ())
with self.assertRaisesRegex(ValueError,
"Unsatisfiable.*a <= a - 1"):
_ = shape_poly.symbolic_shape("a, b",
constraints=("a <= a - 1",))
with self.assertRaisesRegex(ValueError,
"Unsatisfiable.*a <= 0"):
a, b = shape_poly.symbolic_shape("a, b",
# Contradicts the default a >= 1
constraints=("a <= 0",))
a >= b
with self.assertRaisesRegex(ValueError,
"Unsatisfiable.*a <= 0"):
a, b = shape_poly.symbolic_shape("a, b",
# Contradicts the default a >= 1
constraints=("a <= 0",))
a >= b
def test_constraints_ge_smallest_first(self):
# Since we process the smaller constraints first, they take effect for the
# processing of the remaining constraints.
a, = shape_poly.symbolic_shape("a",
# max(a, 2) = a
constraints=("a >= 4", "max(a, 2) >= 6"))
self.assertEqual(core.max_dim(a, 2), a)
self.assertEqual(_bounds(a), (6, np.inf))
def test_constraints_ge_smallest_first_caching(self):
# When we process the constraint with max(a, 5) we will compute the bounds
# of (a, 5) and at that time we will cache that "-1 <= a - 5". We should
# not cache that, because later we learn that a >= 5
a, = shape_poly.symbolic_shape("a",
# max(a, 2) = a
constraints=("a >= 4", "max(a, 5) >= 5",
"max(a, 2) >= 5"))
self.assertEqual(core.max_dim(a, 2), a)
self.assertEqual(core.max_dim(a, 5), a)
def test_constraints_ge_a_minus_4d(self):
# simulates d = div(a, 4) and m = mod(a, 4)
assumptions = ["a >= 4*d + m ",
"a <= 4*d + m",
"m >= 0", "m <= 3"]
scope = shape_poly.SymbolicScope(assumptions)
a, d = shape_poly.symbolic_shape("a, d", scope=scope)
self.assertEqual(_bounds(a - 4*d), (1, 3)) # a - 4d = m >= 1
self.assertEqual(_bounds(a - 2*d), (3, np.inf)) # a - 2d = m + 2d >= 3
# TODO: The incompleteness is due to the way we combine external constraints
self.assertEqual(_bounds(a),
_expect(best=(5, np.inf), current=(4, np.inf))) # a >= 4d + m >= 5
def test_constraints_errors(self):
with self.assertRaisesRegex(ValueError,
"The symbolic constraints should be a sequence of strings"):
_ = shape_poly.symbolic_shape("a",
constraints="a <= a - 1")
with self.assertRaisesRegex(ValueError,
"Constraint parsing error: must contain one of '==' or '>=' or '<='"):
_ = shape_poly.symbolic_shape("a",
constraints=("a != 0",))
with self.assertRaisesRegex(ValueError,
re.escape("Unsatisfiable explicit constraint: a == a + 1")):
_ = shape_poly.symbolic_shape("a",
# Contradicts the default a >= 1
constraints=("a == a + 1",))
def test_constraints_ge_term(self):
a, b = shape_poly.symbolic_shape(
"a, b",
constraints=("max(a, b) >= 8", "min(a, b) <= 2",
"floordiv(a, 4) >= 2",
"mod(a, b) >= 2",
"mod(max(a, b), 4) >= 2"))
self.assertGreaterEqual(core.max_dim(a, b), 6)
self.assertLessEqual(core.min_dim(a, b), 6)
self.assertGreaterEqual(a // 4, 1)
self.assertGreaterEqual(a % b, 1)
self.assertGreaterEqual(core.max_dim(a, b) % 4, 2)
def test_constraints_ge_term_derived(self):
a, b = shape_poly.symbolic_shape(
"a, b",
constraints=("a >= 4", "3 >= b"))
self.assertGreaterEqual(a, b)
self.assertEqual(core.max_dim(a, b), a)
self.assertEqual(core.min_dim(a, b), b)
self.assertGreaterEqual(a // 3, 1)
self.assertEqual(_bounds(a + b), (5, np.inf))
self.assertEqual(_bounds(a - b), (1, np.inf))
def test_constraints_ge_not_term(self):
a, b = shape_poly.symbolic_shape("a, b",
constraints=("a >= b",))
self.assertGreaterEqual(a, b)
def test_constraints_ge_complex(self):
a, b, c = shape_poly.symbolic_shape(
"a, b, c",
constraints=("a + 2 <= b", "b <= a + 5", "a + b >= c"))
self.assertEqual(_bounds(b), (3, np.inf))
self.assertEqual(_bounds(b - a), (2, 5))
self.assertEqual(_bounds(b - a - 7), (-5, -2))
self.assertEqual(_bounds(c - 2*a - 5), (-np.inf, 0))
def test_constraints_ge_fractional(self):
a, = shape_poly.symbolic_shape("a",
constraints=("2 * a >= 5", "3 * a <= 10",))
self.assertEqual(_bounds(5*a - 2), (13, 13))
@jtu.parameterized_filterable(
kwargs=[
dict(constraint="2*a + 3*b >= 10", exp="a + b - 2",
bounds=(2, np.inf)),
dict(constraint="-2*a + 3*b >= 10", exp="a + 2*b",
bounds=(9, np.inf)),
dict(constraint="-2*a + -3*b >= -10", exp="-1*a + 2*b",
bounds=(-1, 3)),
dict(constraint="2*a + -3*b >= 10", exp="-1*a + 2*b", bounds=(-np.inf, np.inf)),
]
)
def test_constraints_ge_complex_gen(self,
constraint: str, exp: str,
bounds: tuple[float, float]):
a, b, exp = shape_poly.symbolic_shape(
"a, b, " + exp,
constraints=(constraint,))
self.assertEqual(bounds, _bounds(exp))
def test_constraints_ge_override(self):
# Some constaints override other
a, b = shape_poly.symbolic_shape("a, b",
constraints=("a >= 5", "b <= 16",
"a >= 10", "b <= 10"))
self.assertEqual(_bounds(a), (10, np.inf))
self.assertEqual(_bounds(b), (1, 10))
def test_constraint_eq_0(self):
a, b, c, d = shape_poly.symbolic_shape(
"a, b, c, d",
constraints=("b == a", "c == a + b", "d == 5"))
# Check that we have already applied the normalizaton rules
self.assertEqual(a._to_var(), "a")
self.assertEqual(b._to_var(), "a")
self.assertEqual(c._to_single_term(), (0, 2, a._to_term()))
self.assertIs(d, 5)
def test_constraints_eq_1(self):
# Some constaints override other
a, b, c = shape_poly.symbolic_shape("a, b, c",
constraints=("max(a, b) == c",))
self.assertEqual(_bounds(core.max_dim(a, b) - c + 3), (3, 3))
def test_constraints_eq_2(self):
a, b = shape_poly.symbolic_shape("a, b",
constraints=("max(a, b) == 5",
"min(a, b) == 3"))
self.assertEqual(_bounds(core.max_dim(a, b)), (5, 5))
self.assertEqual(_bounds(core.min_dim(a, b)), (3, 3))
self.assertEqual(_bounds(a), (3, 5))
def test_constraints_eq_3(self):
a, b = shape_poly.symbolic_shape(
"a, b",
# The constraint implies `b >= 2` thus `min(b, 2)` gets normalized
# to `2`
constraints=("min(a, b) == b - min(b, 2)",))
self.assertEqual(core.min_dim(b, 2), 2)
self.assertEqual(_bounds(b), (2, np.inf))
# TODO: the following ought to work, but the way we wrote the equality
# constraint, `min(b, 2)` gets rewritten to `2`.
#self.assertEqual(core.min_dim(a, b), b - core.min_dim(b, 2))
def test_constraints_eq_4(self):
# Equalities of a variable with an expression
a, b, c, d = shape_poly.symbolic_shape(
"a, b, c, d",
constraints=("floordiv(d, 2) == a + 5",
"floordiv(c, 3) == a + 3",
"b == a + 1"))
self.assertEqual(b, a + 1)
self.assertEqual(d // 2, a + 5)
self.assertEqual(c // 3, b + 2)
def test_constraints_eq_5(self):
# Constraints involving both inequalities and equalities
a, b, c, d = shape_poly.symbolic_shape(
"a, b, c, d",
constraints=(# max(c, a) == c because c >= b + 3 = a + 3
"b == a",
"c >= b + 3",
"floordiv(d, 2) == floordiv(max(c, a), 3) + 2",
))
self.assertEqual(core.max_dim(c, a), c)
# d // 2 = c // 3 + 2 = b + 5
self.assertEqual(_bounds(d // 2 - c // 3), (2, 2))
def test_constraints_eq_6(self):
# Constraints involving both inequalities and equalities
a, b, c, d = shape_poly.symbolic_shape(
"a, b, c, d",
constraints=("b == a",
"c >= b + 3",
"max(d, a) == b + 3",))
self.assertEqual(core.max_dim(c, a), c)
# d <= max(d, a) == a + 3, thus d - a <= 3
self.assertEqual(_bounds(d - b), (-np.inf, 3))
def test_constraints_eq_7(self):
# This came up in b/324989597
b, t = shape_poly.symbolic_shape(
"b, t",
constraints=(
"128 * b * floordiv(t + mod(-1*t, 128), 128) == b * t + b * mod(-1*t, 128)",))
t_ceil = t + (-t) % 128 # A multiple of 128, but hard to tell
self.assertEqual(128 * b * (t_ceil // 128), b * t_ceil)
b1, t1 = shape_poly.symbolic_shape(
"b1, t1",
constraints=(
"128 * floordiv(t1 + mod(-1*t1, 128), 128) == t1 + mod(-1*t1, 128)",))
t1_ceil = t1 + (-t1) % 128 # A multiple of 128, but hard to tell
self.assertEqual(128 * (t1_ceil // 128), t1_ceil)
self.assertEqual(128 * b1 * (t1_ceil // 128), b1 * t1_ceil)
def test_constraints_eq_bug_23456(self):
b, = jax.export.symbolic_shape('b', constraints=['b==5'])
jax.eval_shape(lambda k: jnp.tile(k, 3), jax.ShapeDtypeStruct((b,), jnp.float32))
def test_constraints_eq_bug_23437(self):
def f1(x, y):
return x + y
x = jnp.ones((4,), dtype=jnp.int32)
y = jnp.ones((4,), dtype=jnp.int32)
args_specs = jax.export.symbolic_args_specs((x, y), ("a*2", "b*2"), constraints=("a==b",))
exp = jax.export.export(jax.jit(f1))(*args_specs)
self.assertEqual(exp.in_avals[0], exp.in_avals[1])
def test_constraints_eq_threefry(self):
# Test equalities that arise out of the threefry lowering
# x : i32[a] # a may be even or odd
# x_padded: i32[a + a % 2] = jnp.concat([x, jnp.zeros((a % 2,))])
# x_reshaped: i32[(a + a % 2) // 2, 2] = x_padded.reshape((-1, 2))
# x_1: i32[a + a % 2] = x_reshaped.reshape((-1,))
a, = shape_poly.symbolic_shape(
"a",
constraints=("mod(a + mod(a, 2), -2) == 0",))
x_reshaped, r = divmod(a + a % 2, -2)
self.assertEqual(r, 0)
self.assertEqual(- x_reshaped, -1 * ((a + a % 2) // -2))
self.assertEqual(-2 * x_reshaped, a + a % 2)
def test_constraints_eq_mod_0(self):
# mod(b, N) == 0 is a common constraint, we need to ensure we can use it
# to infer things like: N * floordiv(b, N) == b, b >= N.
b, c, d = shape_poly.symbolic_shape(
"b, c, d",
constraints=("mod(b, 4) == 0",))
# Inequalities work, because we use more expensive reasoning
self.assertGreaterEqual(b, 4 * (b // 4))
self.assertGreaterEqual(4 * (b // 4), b)
# Equalities used to fail
self.assertEqual(b, 4 * (b // 4))
# And an equality that may come up in a reshape
self.assertEqual(math.prod([b, c, d]), math.prod([b // 4, c, d, 2, 2]))
self.assertGreaterEqual(b, b // 4)
self.assertGreaterEqual(b, 3 * (b // 4))
def test_constraints_eq_a_minus_4d(self):
# simulates d = div(a, 4) and m = mod(a, 4)
constraints = ["4*d == a - m", "m >= 0", "m <= 3"]
scope = shape_poly.SymbolicScope(constraints)
a, d = shape_poly.symbolic_shape("a, d", scope=scope)
self.assertEqual(_bounds(a - 4*d), (1, 3)) # a - 4d = m >= 1
# TODO: The incompleteness is due to the way we combine external constraints
self.assertEqual(_bounds(a - 2*d),
_expect(best=(3, np.inf), current=(-np.inf, np.inf))) # a - 2d = m + 2d >= 3
# TODO: The incompleteness is due to the way we combine external constraints
self.assertEqual(_bounds(a),
_expect(best=(5, np.inf), current=(4, np.inf))) # a >= 4d + m >= 5
# Now with a different order of constraints
constraints1 = ["m1 >= 0", "m1 <= 3", "a1 == 4*d1 + m1"]
scope1 = shape_poly.SymbolicScope(constraints1)
a1, d1, m1 = shape_poly.symbolic_shape("a1, d1, m1", scope=scope1)
self.assertEqual(_bounds(a1 - 4*d1), (1, 3)) # a - 4d = m >= 1
self.assertEqual(_bounds(a1 - 2*d1), (3, np.inf)) # a - 2d = m + 2d >= 3
self.assertEqual(_bounds(a1), (5, np.inf)) # a >= 4d + m >= 5
def test_constraints_eq_geq(self):
# We ensure that an equality constraint it is usable not just for
# normalization but also for inequality reasoning.
a, b = export.symbolic_shape(
"a, b", constraints=["4 * a == b"])
self.assertGreaterEqual(b, a)
self.assertGreaterEqual(b, 3*a)
self.assertGreaterEqual(b, 4 * a)
self.assertGreaterEqual(5 * a, b)
self.assertGreaterEqual(9 * a, 2*b)
def test_constraints_error_msg(self):
a, b = shape_poly.symbolic_shape("a, b",
constraints=("a >= 5",))
with self.assertRaisesRegex(
core.InconclusiveDimensionOperation,
re.compile("with constraints.*"
r"a >= 5", re.DOTALL)):
self.assertGreaterEqual(a, 6)
def test_scope_sharing(self):
a, b = shape_poly.symbolic_shape("a, b",
constraints=("a >= 5", "a <= 10"))
self.assertIs(a.scope, b.scope)
self.assertEqual(_bounds(a), (5, 10))
def test_constraints_seq(self):
a1, a2, a3, a4, a5 = shape_poly.symbolic_shape(
"a1, a2, a3, a4, a5",
constraints=(
"a1 >= a2",
"a2 >= a3",
"a3 >= a4",
"a4 >= a5",
)
)
self.assertEqual(_bounds(a1 - a5), (0, np.inf))
def test_constraints_rounding_terms(self):
a1, a2 = shape_poly.symbolic_shape(
"a1, a2",
constraints=(
# a1 >= 2 and a1 <= 2
"2 * a1 >= 3", "-2 * a1 >= -5"
)
)
self.assertEqual(_bounds(a1), (2, 2))
def test_constraints_rounding_not_terms(self):
a1, a2 = shape_poly.symbolic_shape(
"a1, a2",
constraints=(
# a1 >= a2 + 2 and a1 <= a2 + 2
"2*a1 >= 2*a2 + 3", "-2*a1 + 2*a2 >= -5"
)
)
self.assertEqual(_bounds(a1 - a2 - 2), (0, 0))
self.assertEqual(_bounds(a2 - a1 + 2), (0, 0))
def test_constraints_unsat_trivial(self):
with self.assertRaisesRegex(ValueError,
r"Unsatisfiable explicit constraint: a1 >= a1 \+ 1"):
_ = shape_poly.symbolic_shape(
"a1", constraints=("a1 >= a1 + 1",))
def test_constraints_unsat_terms(self):
with self.assertRaisesRegex(ValueError,
"Unsatisfiable constraint"):
a1, a2, *_ = shape_poly.symbolic_shape(
"a1, a2, a3, a4",
constraints=(
# The following -> a1 >= 5
"a1 >= a3 + 1", "a3 >= 4",
# The following -> a1 <= 4
"a1 <= a4 + 2", "a4 <= 2"))
a1 >= a2
def test_constraints_unsat_not_terms(self):
with self.assertRaisesRegex(ValueError,
"Unsatisfiable constraint"):
a1, a2, a3, a4 = shape_poly.symbolic_shape(
"a1, a2, a3, a4",
constraints=(
# The following -> a1 >= a2 + 5
"a1 >= a3 + 1", "a3 >= a2 + 4",
# The following -> a1 <= a2 + 4
"a1 <= a4 + 2", "a4 <= a2 + 2"))
self.assertGreaterEqual(a1, a2)
def test_constraints_sharing(self):
a, = shape_poly.symbolic_shape("a",
constraints=("a >= 5", "a <= 10"))
self.assertEqual(_bounds(a), (5, 10))
# The constraints order does not matter, and they are canonicalized
a1, = shape_poly.symbolic_shape("a",
constraints=("2*a + 5 <= a + 15", "a >= 5"))
self.assertNotEqual(a, a1) # Two separate scopes are not equal
self.assertNotEqual(hash(a), hash(a1)) # Different hash because of different scopes
self.assertIsNot(a.scope, a1.scope)
def test_constraints_different_scope(self):
a, = shape_poly.symbolic_shape("a",
constraints=("a >= 5", "a <= 10"))
self.assertEqual(_bounds(a), (5, 10))
a1, = shape_poly.symbolic_shape("a",
constraints=("a <= 10",))
self.assertNotEqual(a, a1)
self.assertNotEqual(hash(a), hash(a1)) # Hope there is no collision
self.assertIsNot(a.scope, a1.scope)
# Test the error message
with self.subTest(msg="error_message"):
with self.assertRaisesRegex(ValueError,
re.compile(
"Invalid mixing of symbolic scopes.*"
"with constraints.*"
"a >= 5.*a <= 10.*"
"and found.*with constraints.*"
"a <= 10",
re.DOTALL)):
a + a1
for o in (op.add, op.sub, op.mul, op.mod, op.floordiv,
op.ge, op.gt, op.le, op.lt, core.max_dim, core.min_dim):
with self.subTest(o=str(o)):
with self.assertRaisesRegex(ValueError,
"Invalid mixing of symbolic scopes"):
o(a, a1)
class PolyHarness(Harness):
"""Tests a function with shape polymorphism.
Exports `fun` with shape polymorphism, then checks that the JAX native and
the exported function produce the same results.
"""
def __init__(self,
group_name: str, name: str,
fun: Callable[..., Any],
*,
arg_descriptors: Sequence[test_harnesses.ArgDescriptor] = (),
polymorphic_shapes: Sequence[str | None] = (),
symbolic_constraints: Sequence[str] = (),
expect_error: tuple[Any, str] | None = None,
check_result: bool = True,
tol: float | None = None,
limitations: Sequence[test_harnesses.Limitation] = (),
override_jax_config_flags: dict[str, Any] = {}):
"""Args:
group_name, name: The name for the harness. See `Harness.__init__`.
fun: the function to be converted. See `Harness.__init__`.
arg_descriptors: The argument descriptors. See `Harness.__init__`.
polymorphic_shapes: For `export.args_specs`.
symbolic_constraints: For `export.args_specs`.
expect_error: an optional pair of an Exception type and a regular
expression to match the expected exception string.
We expect this error during tracing and exporting with shape
polymorphism.
check_result: specifies if we want to check that the result of invoking
the shape polymorphic export produces the same result as the
native JAX function.
tol: the tolerance to use for checking results.
limitations: a sequence of Limitation(s), used for obtaining the default
tolerance (if `tol` is not specified).
override_jax_config_flags: jax.config flags to override for the duration
of the test.
"""
super().__init__(group_name, name, fun, arg_descriptors,
dtype=np.float32)
self.polymorphic_shapes = polymorphic_shapes
self.symbolic_constraints = symbolic_constraints
self.expect_error = expect_error
self.tol = tol
self.check_result = check_result
self.limitations = limitations
self.override_jax_config_flags = override_jax_config_flags
def run_test(self, tst: jtu.JaxTestCase) -> jax.Array | None:
def log_message(extra: str):
return f"[{tst._testMethodName}]: {extra}"
# Check that we have overridden the jax.config flags
for fname, fvalue in self.override_jax_config_flags.items():
tst.assertEqual(getattr(jax.config, fname), fvalue, (
f"Flag {fname} current value {getattr(jax.config, fname)} != {fvalue}"))
f_jax = jax.jit(self.dyn_fun)
args = self.dyn_args_maker(tst.rng())
args = jax.tree.map(jnp.array, args)
args_specs = export.symbolic_args_specs(args, self.polymorphic_shapes,
constraints=self.symbolic_constraints)
if self.expect_error is not None:
with tst.assertRaisesRegex(self.expect_error[0], self.expect_error[1]):
export.export(f_jax)(*args_specs)
return None
exp = export.export(f_jax)(*args_specs)
if not self.check_result:
return None
# Run the JAX natively and then the exported function and compare
res_jax_native = f_jax(*args)
res_jax_exported = exp.call(*args)
custom_assert_lims = [
l for l in self.limitations if l.custom_assert is not None]
assert len(custom_assert_lims) <= 1, custom_assert_lims
tol = None
if self.tol is not None:
tol = self.tol
elif self.limitations:
max_lim = self.limitations[0].get_max_tolerance_limitation(
self.limitations)
if max_lim is not None:
tol = max_lim.tol
if not custom_assert_lims:
tst.assertAllClose(res_jax_native, res_jax_exported,
atol=tol, rtol=tol)
else:
logging.info(log_message(
f"Running custom_assert with tol={tol} due "
f"to {custom_assert_lims[0]}"))
custom_assert_lims[0].custom_assert(tst, res_jax_native,
res_jax_exported, args=args, # type: ignore
tol=tol, err_msg=None)
return res_jax_exported
def check_shape_poly(tst, f_jax: Callable, *,
arg_descriptors: Sequence[test_harnesses.ArgDescriptor] = (),
polymorphic_shapes: Sequence[str | None] = (),
symbolic_constraints: Sequence[str] = (),
expect_error=None) -> jax.Array | None:
# Builds a PolyHarness and runs the test. See PolyHarness documentation.
h = PolyHarness("", "", jax.jit(f_jax),
arg_descriptors=arg_descriptors,
polymorphic_shapes=polymorphic_shapes,
symbolic_constraints=symbolic_constraints,
expect_error=expect_error)
return h.run_test(tst)
class ShapePolyTest(jtu.JaxTestCase):
def setUp(self):
_start_profile(self)
super().setUp()
def tearDown(self):
super().tearDown()
_stop_profile(self)
def test_simple_unary(self):
"""Test shape polymorphism for a simple case, unary function."""
def f_jax(x):
return x + jnp.sin(x)
for polymorphic_shapes in [None, "_, h", "h, h"]:
with self.subTest(shapes=polymorphic_shapes):
check_shape_poly(self,
f_jax,
arg_descriptors=[RandArg((3, 3), _f32)],
polymorphic_shapes=polymorphic_shapes)
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=f"expr={name}", expr=expr)
for name, expr in [
("d + 2", lambda d: d + 2),
("2 - d", lambda d: 2 - d),
("d * 2", lambda d: d * 2),
("d * d", lambda d: d * d),
("(- d) * d", lambda d: (- d) * d),
("d * d - d", lambda d: d * d - d),
# Division
("d // 2", lambda d: d // 2),
("(d + 1) // 2", lambda d: (d + 1) // 2),
("d // -2", lambda d: d // -2),
("(d + 1) // -2", lambda d: (d + 1) // -2),
("(-d) // 2", lambda d: (-d) // 2),
("(-d - 1) // 2", lambda d: (-d - 1) // 2),
("(-d) // -2", lambda d: (-d) // -2),
("(-d - 1) // -2", lambda d: (-d - 1) // -2),
# Remainder
("d % 2", lambda d: d % 2),
("(d + 1) % 2", lambda d: (d + 1) % 2),
("d % -2", lambda d: d % -2),
("(d + 1) % -2", lambda d: (d + 1) % -2),
("(-d) % 2", lambda d: (-d) % 2),
("(-d - 1) % 2", lambda d: (-d - 1) % 2),
("(-d) % -2", lambda d: (-d) % -2),
("(-d - 1) % -2", lambda d: (-d - 1) % -2),
]
])
def test_non_trivial_dim_expr(self, expr=lambda d: d % -2):
# Check the lowering for shape expressions
check_shape_poly(
self,
lambda x: x[0] * 0 + expr(x.shape[0]),
arg_descriptors=[RandArg((3,), np.int64)],
polymorphic_shapes=["b"])
@jtu.parameterized_filterable(
# The function `f` will be called with x: f32[b]
kwargs=[
dict(testcase_name="cube", f=lambda x: x.shape[0] ** 3),
dict(testcase_name="zero", f=lambda x: x.shape[0] ** 0),
dict(testcase_name="rpow", f=lambda x: 2 ** x.shape[0]),
dict(testcase_name="negative",
f=lambda x: x.shape[0] ** -2,
expect_error=(ValueError, "cannot be raised to negative powers")),
dict(testcase_name="non_integer",
f=lambda x: x.shape[0] ** 1.5,
expect_error=(ValueError, "cannot be raised to non-integer powers")),
dict(testcase_name="sym_pow",
f=lambda x: x.shape[0] ** x.shape[0]),
]
)
def test_pow(self, f, expect_error: tuple[Exception, str] | None = None):
check_shape_poly(self,
f,
arg_descriptors=[RandArg((3,), np.float32)],
polymorphic_shapes=["b"],
expect_error=expect_error)
def test_static_shape_result(self):
"""The result has static shape."""
def f_jax(x):
return jnp.sum(x + jnp.sin(x), axis=0)
check_shape_poly(self,
f_jax,
arg_descriptors=[RandArg((2, 3), _f32)],
polymorphic_shapes=[None])
check_shape_poly(self,
f_jax,
arg_descriptors=[RandArg((2, 3), _f32)],
polymorphic_shapes=["b, _"])
def test_kwargs(self):
"""Test shape polymorphism for a function with kwargs."""
x = np.ones(3, dtype=np.float32)
y = np.ones(1, dtype=np.float32)
def f_jax(x, *, y):
return x + jnp.sin(y)
exp = export.export(jax.jit(f_jax))(
jax.ShapeDtypeStruct(export.symbolic_shape("b"), x.dtype),
y=jax.ShapeDtypeStruct(y.shape, y.dtype))
self.assertAllClose(f_jax(x, y=y), exp.call(x, y=y))
def test_arg_avals_errors(self):
"""Test error reporting for shape polymorphism."""
def conv_and_run(*, arg_shape: core.Shape,
polymorphic_shape: str):
arg = np.arange(math.prod(arg_shape), dtype=np.float32).reshape(arg_shape)
check_shape_poly(self, lambda x: x,
arg_descriptors=[arg],
polymorphic_shapes=[polymorphic_shape])
with self.assertRaisesRegex(ValueError,
re.escape("polymorphic shape spec should be")):
conv_and_run(arg_shape=(2,), polymorphic_shape=5.)
with self.assertRaisesRegex(ValueError,
re.escape("pytree structure error: different types")):
conv_and_run(arg_shape=(2,), polymorphic_shape=["a list"])
with self.assertRaisesRegex(ValueError,
re.escape("pytree structure error: different types")):
conv_and_run(arg_shape=(2,), polymorphic_shape=("a tuple",))
with self.assertRaisesRegex(ValueError,
"Cannot solve for values of dimension variables {'b'}"):
conv_and_run(arg_shape=(4, 36, 3), polymorphic_shape="b * b, b * d * d, d")
with self.assertRaisesRegex(ValueError,
"Division had remainder 2 when computing the value of 'b'"):
conv_and_run(arg_shape=(5, 36), polymorphic_shape="3 * b, ...")
with self.assertRaisesRegex(ValueError,
"Expected value >= 1 for dimension variable 'b'"):
conv_and_run(arg_shape=(10, 3), polymorphic_shape="3 * b + 10, ...")
with self.assertRaisesRegex(ValueError,
"Expected value >= 1 for dimension variable 'b'"):
conv_and_run(arg_shape=(7, 3), polymorphic_shape="3 * b + 10, ...")
with self.assertRaisesRegex(
ValueError,
re.escape(
"Found inconsistency between dimension size "
"args[0].shape[1] (= 3) and the specification 'a' (= 2)")):
conv_and_run(arg_shape=(2, 3), polymorphic_shape="(a, a)")
def test_pytree(self):
"""Arguments and polymorphic_shapes are pytrees."""
# Arguments are of the form [([x00, x01], [x10]), dict(a=ya, b=yb)]
def add_all_jax(x_pair_of_list, y_dict):
x_list_0, x_list_1 = x_pair_of_list
return functools.reduce(op.add,
x_list_0 + x_list_1 + [y_dict["a"], y_dict["b"]])
x = np.arange(4, dtype=_f32)
args = (([x, x], [x]), dict(a=x, b=x))
check_shape_poly(self,
add_all_jax,
arg_descriptors=args,
polymorphic_shapes=[(["v", "v"], ["v"]),
dict(a="v", b="v")])
# Prefix polymorphic shapes
check_shape_poly(self,
add_all_jax,
arg_descriptors=args,
polymorphic_shapes="v")
check_shape_poly(self,
add_all_jax,
arg_descriptors=args,
polymorphic_shapes=["v", "v"])
check_shape_poly(self,
add_all_jax,
arg_descriptors=args,
polymorphic_shapes=[("v", "v"), "v"])
# Now partial polymorphic_shapes.
check_shape_poly(self,
add_all_jax,
arg_descriptors=args,
polymorphic_shapes=[(["(4,)", "(_,)"], [("4,")]),
dict(a="(_,)", b="(4,)")])
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=name, polymorphic_shapes=polymorphic_shapes)
for name, polymorphic_shapes in [
("1", ("b", "b", "b")),
("2", dict(a="b")),
("3", (dict(a="b"), "b")),
]]
)
def test_pytree_errors(self, polymorphic_shapes=("b", "b", "b")):
"""Arguments and polymorphic_shapes are not-matching pytrees."""
# Arguments are of the form [([x00, x01], [x10]), dict(a=ya, b=yb)]
x = np.arange(4, dtype=_f32)
args = (([x, x], [x]), dict(a=x, b=x))
def add_all_jax(x_pair_of_list, y_dict):
x_list_0, x_list_1 = x_pair_of_list
return functools.reduce(op.add,
x_list_0 + x_list_1 + [y_dict["a"], y_dict["b"]])
with self.assertRaisesRegex(ValueError, "pytree structure error"):
check_shape_poly(self,
add_all_jax,
arg_descriptors=args,
polymorphic_shapes=polymorphic_shapes)
def test_with_nested_jit(self):
def f_jax(x): # x: f32[w, h]
return jnp.sin(x) + jnp.arange(x.shape[0] * x.shape[1],
dtype=x.dtype).reshape(x.shape)
check_shape_poly(self,
lambda x: x + jax.jit(f_jax)(x),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["a, b"])
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=str(polymorphic_shapes), polymorphic_shapes=polymorphic_shapes)
# The polymorphic_shapes should have three comma-separated DimExpr matching
# 16, 24, 32
for polymorphic_shapes in [
"b1+6,b1+14,b2", # b1=10, b2=32
"2*b1,4*b2,b1+b2+18", # b1=8,b2=6
"b1+2*b2,4*b2,b1*b1+16", # b1=4,b2=6
]
])
def test_non_trivial_polynomials_spec(self,
polymorphic_shapes="2*b1,4*b2,b1+b2+18"):
# We can handle non-trivial polynomials in the input shape,
# as long as all variables also occur in trivial expressions
check_shape_poly(self,
lambda x: 2 * x.shape[0] + 3 * x.shape[1] + 4 * x.shape[2],
arg_descriptors=[RandArg((16, 24, 32), _f32)],
polymorphic_shapes=[polymorphic_shapes])
def test_constraints_basic(self):
def f(x): # x: i32[a], with a >= 8
assert x.shape[0] >= 6
check_shape_poly(self, f,
arg_descriptors=[RandArg((16,), _i32)],
polymorphic_shapes=["a"],
symbolic_constraints=["a >= 8"])
def test_constraints_slice_in_dim(self):
def f(x): # x: i32[a], with a >= 8
return lax.dynamic_slice_in_dim(x, 0, 8, 0)
check_shape_poly(self, f,
arg_descriptors=[RandArg((16,), _i32)],
polymorphic_shapes=["a"],
symbolic_constraints=["a >= 8"])
def test_constraints_slice_in_dim_eq(self):
def f(x, y): # x: i32[a], y: i32[b]
v1 = x[:y.shape[0]] # i32[min(a, b)]
v2 = y[2:] # i32[b - min(b, 2)]
return v1 + v2
check_shape_poly(self, f,
arg_descriptors=[RandArg((16,), _i32), # a == 16
RandArg((18,), _i32)], # b == 18
polymorphic_shapes=["a", "b"],
# This is the exact failure we are getting.
symbolic_constraints=[
"b >= 2",
"min(a, b) == b - 2"
])
def test_constraints_eq_threefry(self):
# This pattern arises in the lowering of threefry
def f(x): # x: i32[a]
a, = x.shape
x_padded = jnp.concatenate([x, jnp.zeros((a % 2,), dtype=x.dtype)])
x_reshaped = x_padded.reshape((2, -1))
x_reshaped += 2
x_padded_1 = x_reshaped.reshape((-1,))
x_1 = x_padded_1[:a]
# Ensure x and x_1 have the same shape
return x + x_1
check_shape_poly(self, f,
arg_descriptors=[RandArg((16,), _i32)], # a == 16
polymorphic_shapes=["a"],
# These are the exact failure we are getting.
symbolic_constraints=[
"mod(a + mod(a, 2), -2) == 0",
"-2*floordiv(a + mod(a, 2), -2) == a + mod(a, 2)"])
def test_constraints_eq_mod_0(self):
# mod(b, N) == 0 is a common constraint, we need to ensure we can use it
# to infer things like: N * floordiv(b, N) == b, b >= N.
def f(x): # x: f32[b] and b % 4 == 0
b = x.shape[0]
y1 = jnp.ones((1, 3, 4, b // 4), dtype=x.dtype)
y2 = y1.reshape((1, 3, -1)) # : f32[1, 3, b]
y3 = x.reshape((1, 1, b)) + y2 # : f32[1, 3, b]
slice0 = lax.slice(x, (0,), (b // 4,)) # Requires b >= b // 4
slice1 = lax.slice(x, (0,), (2 * (b // 4),)) # Requires b >= 2 * (b // 4)
slice2 = lax.slice(x, (0,), (3 * (b // 4),)) # Requires b >= 2 * (b // 4)
slice3 = lax.slice(x, (0,), (4 * (b // 4),)) # Requires b >= 2 * (b // 4)
return (jnp.sum(y3) +
jnp.sum(slice0) + jnp.sum(slice1) +
jnp.sum(slice2) + jnp.sum(slice3))
check_shape_poly(self, f,
arg_descriptors=[RandArg((16,), _i32)],
polymorphic_shapes=["b"],
symbolic_constraints=["mod(b, 4) == 0"])
def test_constraints_for_profile(self):
# A somewhat more involved tests to stress test the correctness and
# performance
def f(x): # x: i32[a, b]
acc = 0
for start in range(0, 50):
slice = x[start::2] # exercises floordiv and min
acc += jnp.sum(slice, axis=0)
slice = x[start:(x.shape[0] - x.shape[0] % 2):2] # exercises max and min
acc += jnp.sum(slice, axis=0)
return acc
_ = export.export(jax.jit(f))(
jax.ShapeDtypeStruct(export.symbolic_shape("a, b"), np.int32))
def test_constraints_ge_compile_time_check(self):
def f(x): # x: i32[a]
a = x.shape[0]
assert _bounds(a) == (2, 4)
return lax.dynamic_slice_in_dim(x, 1, 2, 0)
x_spec = jax.ShapeDtypeStruct(
export.symbolic_shape("a",
constraints=["a >= 2", "a <= 4"]), np.int32)
exp = export.export(jax.jit(f))(x_spec)
x_2 = np.arange(2, dtype=np.int32)
res_2 = exp.call(x_2)
self.assertAllClose(x_2[0:2], res_2)
x_4 = np.arange(4, dtype=np.int32)
res_4 = exp.call(x_4)
self.assertAllClose(x_4[1:3], res_4)
with self.assertRaisesRegex(
ValueError,
re.escape("Expected 'a - 2' to be greater or equal to 0, but found -1")):
exp.call(np.arange(1, dtype=np.int32))
with self.assertRaisesRegex(
ValueError,
re.escape("Expected '- a + 4' to be greater or equal to 0, but found -1")):
exp.call(np.arange(5, dtype=np.int32))
def test_constraints_eq_0_compile_time_check(self):
def f(x): # x: i32[a, b]
return x
x_spec = jax.ShapeDtypeStruct(
export.symbolic_shape("a, b",
constraints=["max(a, b) == b"]), np.int32)
exp = export.export(jax.jit(f))(x_spec)
with self.assertRaisesRegex(
ValueError,
re.escape("Expected 'max(a, b) - b' to be equal to 0, but found 1")):
exp.call(np.ones((3, 2), dtype=np.int32))
def test_constraints_eq_1_compile_time_check(self):
def f(x): # x: i32[a, b]
return x
x_spec = jax.ShapeDtypeStruct(
export.symbolic_shape("a, b",
constraints=["a == b"]), np.int32)
exp = export.export(jax.jit(f))(x_spec)
exp.call(np.ones((3, 3), dtype=np.int32))
def test_constraints_eq_2_compile_time_check(self):
def f(x): # x: i32[a, b]
return x
x_spec = jax.ShapeDtypeStruct(
export.symbolic_shape("a, b",
constraints=["max(a, b) == 4", "a == b"]), np.int32)
exp = export.export(jax.jit(f))(x_spec)
with self.assertRaisesRegex(
ValueError,
re.escape("Expected 'max(a, b) - 4' to be equal to 0, but found -1")):
exp.call(np.ones((3, 3), dtype=np.int32))
def test_caching_with_scopes(self):
f_tracing_count = 0
expected_a_bounds = (1, np.inf)
@jax.jit
def f(x): # x: i32[a]
nonlocal f_tracing_count
f_tracing_count += 1
a = x.shape[0]
assert _bounds(a) == expected_a_bounds
x_spec = jax.ShapeDtypeStruct(export.symbolic_shape("a"), np.int32)
_ = export.export(f)(x_spec)
self.assertEqual(1, f_tracing_count)
_ = export.export(f)(x_spec) # This hits the tracing cache
self.assertEqual(1, f_tracing_count)
# Recreate x_spec, with the same symbolic expressions, and the same scope
x_spec = jax.ShapeDtypeStruct(export.symbolic_shape("a",
scope=x_spec.shape[0].scope), np.int32)
_ = export.export(f)(x_spec) # This hits the tracing cache
self.assertEqual(1, f_tracing_count)
# Recreate x_spec, with the same symbolic expressisions with a fresh scope
x_spec = jax.ShapeDtypeStruct(export.symbolic_shape("a"), np.int32)
_ = export.export(f)(x_spec) # This is a cache miss
self.assertEqual(2, f_tracing_count)
# Now with constraints
x_spec = jax.ShapeDtypeStruct(export.symbolic_shape("a",
constraints=["a >= 2"]),
np.int32)
expected_a_bounds = (2, np.inf)
_ = export.export(f)(x_spec) # Cache miss
self.assertEqual(3, f_tracing_count)
def test_unused_args(self):
# Tests with functions that do not use their inputs.
# First arg unused, not polymorphic
check_shape_poly(self,
lambda x_unused, y: y * 2.0,
arg_descriptors=[RandArg((2, 3), _f32), RandArg((3,), _f32)],
polymorphic_shapes=[None, "b"])
# Some args unused, not polymorphic
check_shape_poly(self,
lambda x_unused, y, z_unused, w: jnp.concatenate([y, w]),
arg_descriptors=[RandArg((3,), _f32), RandArg((4,), _f32),
RandArg((5,), _f32), RandArg((6,), _f32)],
polymorphic_shapes=[None, "b1", None, "b2"])
# A polymorphic arg is not used, but the dimension var appears
# in a used arg also
check_shape_poly(self,
lambda x_unused, y: y * 2.0,
arg_descriptors=[RandArg((3,), _f32), RandArg((3,), _f32)],
polymorphic_shapes=["b", "b"])
# A polymorphic arg is not used, and the dimension var does not appear
# elsewhere.
check_shape_poly(self,
lambda x_unused, y: y * 2.0,
arg_descriptors=[RandArg((4,), _f32), RandArg((3,), _f32)],
polymorphic_shapes=["b1", "b2"])
# A polymorphic arg is not used, and the dimension var does appear
# elsewhere but not as a trivial term.
check_shape_poly(self,
lambda x_unused, y: y * 2.0,
arg_descriptors=[RandArg((3,), _f32), RandArg((9,), _f32)],
polymorphic_shapes=["b1", "b1 * b1"])
# It is not sufficient to just use the shape of an input; it is still unused
check_shape_poly(self,
lambda x_unused, y: y + x_unused.shape[0],
arg_descriptors=[RandArg((3,), _f32), RandArg((9,), _f32)],
polymorphic_shapes=["b1", "b2"])
def test_cond(self):
# Test the primitive under conditional
def f(x, y):
# x: f32[B, H], y : f32[H]
return lax.cond(
jnp.sum(x) > 0.,
lambda _: x + jnp.reshape(y, (1, y.shape[0])),
lambda _: jnp.zeros_like(x),
operand=None)
x = np.ones((2, 3))
y = np.ones((3,))
res_jax = f(x, y)
self.assertAllClose(
res_jax,
check_shape_poly(self, f, arg_descriptors=[x, y],
polymorphic_shapes=["(b, h)", "h"]))
def test_while(self):
def f(x):
# x: f32[B], iter: i32
return lax.while_loop(lambda x_iter: x_iter[1] < 5,
lambda x_iter: (x_iter[0] + jnp.arange(x_iter[0].shape[0], dtype=np.float32), x_iter[1] + 1),
(x, 0))
x = np.ones((3,), dtype=np.float32)
res_tf = check_shape_poly(self, f, arg_descriptors=[x],
polymorphic_shapes=["(b,)"])
self.assertAllClose(f(x), res_tf)
@jax.debug_key_reuse(False)
def test_prng(self):
# The PRNG implementation uses opaque types, test shape polymorphism
with config.enable_custom_prng(True):
def f_jax(x): # x: f32[b1, b2]
key = random.PRNGKey(123) # key: key<fry>[]
# Exercise key operations that have custom lowering rules
broadcast_keys = lax.broadcast_in_dim(key, x.shape, ()) # key<fry>[b1, b2]
gather_keys = lax.broadcast_in_dim(broadcast_keys[0], (1, x.shape[1]), (1,)) # : key[1, b2]
slice_keys1 = lax.slice(broadcast_keys, (0, 0), (1, x.shape[1]), (1, 1)) # key[1, b2]
slice_keys2 = lax.dynamic_slice(broadcast_keys, (0, 0), slice_sizes=(1, x.shape[1])) # key[1, b2]
upd1 = lax.dynamic_update_slice(slice_keys2, slice_keys1, start_indices=(0, 0)) # key[1, b2]
_ = lax.dynamic_update_slice(upd1, gather_keys, start_indices=(0, 0))
# We need to test the special case for vmap(while)
xs = broadcast_keys
counts = jnp.arange(broadcast_keys.shape[0], dtype=np.int32)
def f_vmap_jax(counts, xs): # counts: i32[b1], xs: key<fry>[b1, b2]
def inner(count, x): # count i32, x: key<fry>[b2]
return lax.fori_loop(0, count, lambda _, acc: acc, x)
return jax.vmap(inner)(counts, xs)
_ = f_vmap_jax(counts, xs)
return x
check_shape_poly(self, f_jax,
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b1, b2"])
def test_dynamic_shapes(self):
# Test dim_as_value with dynamic shapes.
def f(x):
return jnp.sum(x, axis=0) * x.shape[0]
x = np.arange(3.)
self.assertAllClose(9.,
check_shape_poly(self, f,
arg_descriptors=[x],
polymorphic_shapes=["(b,)"]))
self.assertAllClose(
9.,
check_shape_poly(self, jax.jit(f),
arg_descriptors=[x], polymorphic_shapes=["(b,)"]))
res_primal, res_tangent = check_shape_poly(self,
lambda x, xt: jax.jvp(f, (x,), (xt,)),
arg_descriptors=[x, np.array([0.1, 0.2, 0.3])],
polymorphic_shapes=["b", "b"])
self.assertAllClose((9., 1.8), (res_primal, res_tangent))
self.assertAllClose(
np.array([3., 3., 3.]),
check_shape_poly(self, jax.grad(f),
arg_descriptors=[x],
polymorphic_shapes=["b"]))
xv = np.arange(24.).reshape((2, 3, 4))
res_vmap = jax.vmap(f, in_axes=1)(xv)
# Implement by iteration
res_iter = jnp.stack([f(xv[:, i, :]) for i in range(xv.shape[1])])
self.assertAllClose(res_iter, res_vmap)
res_vmap_tf = check_shape_poly(self, jax.vmap(f, in_axes=1),
arg_descriptors=[xv],
polymorphic_shapes=["b1, b2, ..."])
self.assertAllClose(res_iter, res_vmap_tf)
def test_with_hash_collision_vmap(self):
# Batching caches based on Jaxpr, and Jaxpr include _DimExpr. If we have
# a collision for the hashing of a _DimExpr, then Python will call the
# equality, which will raise InconclusiveDimensionOperation.
def f_jax(x):
return jnp.reshape(x, (2, -1,))
orig_hash = None
try:
# Override the hashing to create collisions
orig_hash = getattr(shape_poly._DimExpr, "__hash__")
def collision_hash(obj):
return hash(5)
setattr(shape_poly._DimExpr, "__hash__", collision_hash)
xs = [np.ones((3, 5, 6), dtype=np.float32)]
f_toconvert = jax.vmap(pjit.pjit(f_jax))
res_1 = check_shape_poly(self, f_toconvert, arg_descriptors=xs,
polymorphic_shapes=["..."])
res_2 = check_shape_poly(self, f_toconvert, arg_descriptors=xs,
polymorphic_shapes=["b1, b2, ..."])
self.assertAllClose(res_1, res_2)
finally:
setattr(shape_poly._DimExpr, "__hash__", orig_hash)
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=op_name, op=op)
for op, op_name in [
(jnp.array, "array"),
(jnp.sin, "sin"),
(lambda x: x, "id"),
(core.dimension_as_value, "dimension_as_value"),
]])
def test_poly_unary_op(self, *, op=jnp.array):
def f_jax(x): # x: f32[b]
poly = 2 * x.shape[0]
return (op(poly), x) # Make sure we are using x
check_shape_poly(self,
f_jax,
arg_descriptors=[RandArg((3,), _f32)],
polymorphic_shapes=["b"])
@jtu.parameterized_filterable(
kwargs=[
dict(testcase_name=f"_{op.__name__}_other={other}:{type(other)}{'_other_jnp_array' if other_jnp_array else ''}{'_swap' if swap else ''}",
op=op, other=other,
other_jnp_array=other_jnp_array, swap=swap)
for op in [op.add, op.mul, op.sub,
op.mod, op.floordiv, op.truediv]
for other in [
2, np.int32(2), 2., np.float32(2),
np.array(2, dtype=np.int32), np.arange(1, 5, dtype=np.int32),
np.array(2., dtype=np.float32), np.arange(1., 7., dtype=np.float32)
]
for other_jnp_array in (
[True, False] if np.shape(other) == (7,) else [False]) # type: ignore
for swap in [False, True] # The poly is the left op by default
])
def test_poly_binary_op(self, *, op=op.add,
other=np.arange(2, dtype=np.int32),
other_jnp_array=False,
swap=True):
# Test arithmetic operations with poly and a variety of other operand types
def f_jax(x): # x: f32[b]
poly = 2 * x.shape[0] # This will allow divisions with 2
other_wrapped = jnp.array(other) if other_jnp_array else other
ops = (poly, other_wrapped) if not swap else (other_wrapped, poly)
res = op(*ops)
# If the other op is an integer then the result is a symbolic dim
try:
op.index(other)
other_isint = True
except Exception:
other_isint = False
if (hasattr(poly, "dimension_as_value") and
other_isint and
op.__name__ != "truediv"):
# If we running under jax2tf and "other" is an integer the result
# should be a symbolic dimension
self.assertTrue(isinstance(res, int) or hasattr(res, "dimension_as_value"))
if config.enable_x64.value:
# Outside jax2tf, x.shape[0] is a Python (64-bit) integer and for most
# operations here JAX is not involved at all because the other operand
# is a Python or NumPy constant. So the result will be 64-bits. But under
# jax2tf, x.shape[0] is rewritten to jnp.array(x.shape[0]) which when
# used with int32 or float32 values will produce 32-bit values.
return (lax.convert_element_type(res, np.float32), x)
return (res, x) # Make sure we are using x
check_shape_poly(self,
f_jax,
arg_descriptors=[RandArg((3,), np.int32)],
polymorphic_shapes=["b"])
def test_shape_as_array(self):
def f_jax(x):
# The entire x.shape is passed to jnp.array
return x + jnp.sum(jnp.array(x.shape)).astype(np.int32)
check_shape_poly(self,
f_jax,
arg_descriptors=[RandArg((3, 4), _i32)],
polymorphic_shapes=["b, _"])
def test_dim_as_value_weak_type(self):
def f_jax(x): # x: f32[b]
d0 = jnp.array(x.shape[0]) # in JAX should have weak_type=True
if isinstance(d0, core.Tracer):
self.assertTrue(d0.aval.weak_type)
# And an implicit conversion to array
d1 = x.shape[0] + jnp.array(4)
if isinstance(d1, core.Tracer):
self.assertTrue(d1.aval.weak_type)
return d0 + np.array(5., dtype=np.float32) + d1 + x[0]
with config.numpy_dtype_promotion("strict"):
# strict type promotion is sensitive to weak_types
check_shape_poly(self,
f_jax,
arg_descriptors=[RandArg((3,), _f32)],
polymorphic_shapes=["b"])
def test_vmap_while(self):
def cond_func(x): # x: f32[3]
return jnp.sum(x) >= 0.
def body_func(x): # x: f32[3]
return x - 1.
def f_jax(x):
return lax.while_loop(cond_func, body_func, x)
check_shape_poly(self,
jax.vmap(f_jax),
arg_descriptors=[RandArg((5, 3), _f32)],
polymorphic_shapes=["b, ..."])
def test_vmap_error(self):
# vmap is careful to give nice error messages when mapped axes have
# different sizes, but this can be foiled by InconsistentDimensionOperation
x = y = np.ones((3, 5), dtype=np.float32)
with self.assertRaisesRegex(ValueError,
"vmap got inconsistent sizes for array axes to be mapped"):
check_shape_poly(self, jax.vmap(lambda x, y: x + y),
arg_descriptors=[x, y],
polymorphic_shapes=["b, ...", None])
z = x
with self.assertRaisesRegex(ValueError,
"vmap got inconsistent sizes for array axes to be mapped"):
check_shape_poly(self, jax.vmap(lambda x, y, z: x + y + z),
arg_descriptors=[x, y, z],
polymorphic_shapes=["b, ...", "c, ...", None])
@jtu.parameterized_filterable(
kwargs=[
dict(slc=slc)
for slc in [
slice(None, None, None),
slice(2, 5),
]
])
def test_stateful(self, slc: slice):
w, = export.symbolic_shape("w", constraints=["w >= 3"])
def f(x_ref):
ones = jnp.ones_like(x_ref)[slc]
ref_primitives.ref_addupdate(x_ref, slc, ones)
x1 = ref_primitives.ref_get(x_ref, slc)
x2 = x1 + ones
ref_primitives.ref_set(x_ref, slc, x2)
exp = export.export(jax.jit(discharge.run_state(f)))(
jax.ShapeDtypeStruct((w,), dtype=_f32))
x = np.ones((32,), dtype=_f32)
expected = np.copy(x)
expected[slc] = 3.
self.assertAllClose(exp.call(x), expected)
# List containing either harnesses, or lists of harnesses
_POLY_SHAPE_TEST_HARNESSES = [
PolyHarness("add", "",
lambda x, y: jnp.add(jnp.expand_dims(x, axis=0), y),
arg_descriptors=[RandArg((3, 4), _f32), RandArg((2, 3, 4), _f32)],
polymorphic_shapes=["b, _", "_, b, _"]),
PolyHarness("add_transpose", "",
jax.grad(lambda x: jnp.sum(
jnp.expand_dims(jnp.sum(x, axis=0, keepdims=False), axis=0)
+ jnp.sin(x))),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
[ # approx_max_k
# x: f32[b, {n}, 32] with n being either 8 or the symbol "n"
# we reduce on dim=1, with size n
# k is either the constant 4 or the symbol "k"
PolyHarness("approx_max_k", f"n_{n}_k_{k}_agg={agg}",
lambda x, x_k, agg: lax.approx_max_k(
x, k=x_k.shape[0], reduction_dimension=1,
aggregate_to_topk=agg),
arg_descriptors=[RandArg((3, 8, 32), _f32),
RandArg((4,), _f32),
StaticArg(agg)],
polymorphic_shapes=[f"b, {n}, 32", f"{k},"],
# k must be at most the reduction dimension size
symbolic_constraints=[f"{k} <= {n}"],
expect_error=(
(NotImplementedError, "aggregate_to_topk=False") if (
not agg and (isinstance(k, str) or
isinstance(n, str))) else
None
))
for n in [8, "n"]
for k in [4, "k"]
for agg in [True, False]
],
[ # arange
PolyHarness("arange", name,
f_jax,
arg_descriptors=[RandArg((6,), np.float32)],
polymorphic_shapes=["b"])
for name, f_jax in [
# Positive step
("b", lambda x: jnp.arange(x.shape[0], None, None, None)),
("0_b+1", lambda x: jnp.arange(0, x.shape[0] + 1, None, None)),
("0_5b_2", lambda x: jnp.arange(0, 5 * x.shape[0], 2, None)),
("0_5b+1_2", lambda x: jnp.arange(0, 5 * x.shape[0] + 1, 2, None)),
("b_5b+2_2", lambda x: jnp.arange(x.shape[0], 5 * x.shape[0] + 2, 2, None)),
("0_b-1_2", lambda x: jnp.arange(0, x.shape[0] - 1, 2, None)),
("0_b-2_2", lambda x: jnp.arange(0, x.shape[0] - 2, 2, None)),
("0_-b_2", lambda x: jnp.arange(0, -x.shape[0], 2, None)),
("0_1-b_2", lambda x: jnp.arange(0, 1 - x.shape[0], 2, None)),
("0_b-3_2", lambda x: jnp.arange(0, x.shape[0] - 3, 2, None)),
# Cannot tell if size >= 0
# Negative step
("b_0_-1", lambda x: jnp.arange(x.shape[0], 0, -1, None)),
("b_1_-2", lambda x: jnp.arange(x.shape[0], 1, -2, None)),
("b_-1_-1", lambda x: jnp.arange(x.shape[0], -1, -1, None)),
("5b+1_0_-2", lambda x: jnp.arange(5 * x.shape[0] + 1, 0, -2, None)),
("5b+2_0_-2", lambda x: jnp.arange(5 * x.shape[0] + 2, 0, -2, None)),
("b-3_0_-2", lambda x: jnp.arange(x.shape[0] - 3, 0, -2, None)),
# Cannot tell if size >= 0
# Symbolic step
("0_10_b", lambda x: jnp.arange(0, 10, x.shape[0])),
("0_0_b", lambda x: jnp.arange(0, 0, x.shape[0])),
("10_0_-b", lambda x: jnp.arange(10, 0, -x.shape[0])),
("b_1_-b", lambda x: jnp.arange(x.shape[0], 1, -x.shape[0])),
# Float return type
("0_b_1_f32", lambda x: jnp.arange(0, x.shape[0], 1, np.float32))
]
],
[ # arange errors
PolyHarness("arange_error", name,
# x: i32[b]
f_jax,
arg_descriptors=[RandArg((3,), dtype=np.int32)],
polymorphic_shapes=["b"],
expect_error=(expect_error, expect_msg))
for name, f_jax, expect_error, expect_msg in [
# make_args invoked with op.shape[0]: start, stop, step
("float_start", lambda x: x[0] + jnp.arange(0., x.shape[0], None),
ValueError, "must be either dimension expressions or integers"),
("float_step", lambda x: x[0] + jnp.arange(0, x.shape[0], 0.5),
ValueError, "must be either dimension expressions or integers"),
("step_0", lambda x: x[0] + jnp.arange(0, x.shape[0], 0),
ValueError, "has step == 0"),
("inconclusive_step_sign", lambda x: x[0] + jnp.arange(0, x.shape[0],
x.shape[0] - 2),
core.InconclusiveDimensionOperation,
"must be resolved statically if it is > 0 or < 0"),
]
],
[ # indexing
# operand is non-poly, index is poly
PolyHarness("indexing", "op=static_idx=poly",
lambda a, i: a[i],
arg_descriptors=[RandArg((3, 4), _f32),
np.array([2, 2], np.int32)],
polymorphic_shapes=[None, "b0, ..."]),
# operand is poly, index is integer
PolyHarness("indexing", "op=poly_idx=const",
lambda a: a[1],
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
# Both the operand and the index are poly
PolyHarness("indexing", "op=poly_idx=poly",
lambda a, i: a[i],
arg_descriptors=[RandArg((3, 4), _f32),
np.array([1, 2, 0], np.int32)],
polymorphic_shapes=["b, ...", "b, ..."]),
],
[ # indexing with slices
PolyHarness("indexing", f"start_{start_name}_stop_{stop_name}_step_{step_name}",
partial(lambda start, stop, step, x: x[slice(start(x.shape[1]),
stop(x.shape[1]),
step(x.shape[1]))],
start, stop, step),
arg_descriptors=[RandArg((16, 8), np.float32)],
polymorphic_shapes=["c, b"])
# start, stop, step are functions that take the argument "b"
# "b" actual value is 8 and "c" is 16
for start_name, start in [
("None", lambda b: None),
("0", lambda b: 0),
("2", lambda b: 2),
("b", lambda b: b),
("2b2", lambda b: 2*b + 2),
("-2", lambda b: -2),
("-b", lambda b: -b),
("-2b2", lambda b: -2*b - 2),
]
for stop_name, stop in [
("None", lambda b: None),
("0", lambda b: 0),
("b", lambda b: b),
("2b2", lambda b: 2*b + 2),
("4", lambda b: 4),
("-4", lambda b: -4),
("-b", lambda b: -b),
("-2b2", lambda b: -2*b - 2),
]
for step_name, step in [
("None", lambda b: None),
("1", lambda b: 1),
("2", lambda b: 2),
("b", lambda b: b),
("-1", lambda b: -1),
("-2", lambda b: -2),
("-b", lambda b: -b),
]
],
# Reduce the poly dimension
PolyHarness("argmax", "0",
lambda op: lax.argmax(op, axis=0, index_dtype=np.int32),
arg_descriptors=[RandArg((3, 4, 5), _f32)],
polymorphic_shapes=["b, ..."]),
# Reduce the non-poly dimension
PolyHarness("argmax", "1",
lambda op: lax.argmax(op, axis=1, index_dtype=np.int32),
arg_descriptors=[RandArg((3, 4, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.argsort", "",
jnp.argsort,
arg_descriptors=[RandArg((3, 4, 5), _f32)],
polymorphic_shapes=["b, ..."]),
[
PolyHarness("average",
f"{axis=}_weights=None",
lambda x, axis: jnp.average(x, axis=axis, returned=False, weights=None),
arg_descriptors=[RandArg((7, 8, 4), _f32), StaticArg(axis)],
polymorphic_shapes=["b, ..."])
for axis in [None, 0, 1]
],
[
PolyHarness("average",
f"{axis=}_weights=Some",
lambda x, weights, axis: jnp.average(x, axis=axis, returned=False, weights=weights),
arg_descriptors=[RandArg((7, 8, 4), _f32), RandArg((7, 8, 4), _f32), StaticArg(axis)],
polymorphic_shapes=["b, ...", "b, ..."])
for axis in [None, 0, 1]
],
PolyHarness("jnp.bincount", "length=constant",
lambda x: jnp.bincount(x % 2, length=4),
arg_descriptors=[RandArg((12,), np.int32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.bincount", "length=poly",
lambda x: jnp.bincount(x % 4, length=x.shape[0]),
arg_descriptors=[RandArg((12,), np.int32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("broadcast_to", "",
lambda x: jnp.broadcast_to(x, [x.shape[0], x.shape[0], 4]),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("broadcast_in_dim", "0",
lambda x: lax.broadcast_in_dim(x, [x.shape[0], 4, 5, 6],
broadcast_dimensions=(0, 2, 3)),
arg_descriptors=[RandArg((3, 1, 6), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("broadcast_in_dim", "poly",
lambda x: lax.broadcast_in_dim(x, [x.shape[0], x.shape[0] + x.shape[0], 4],
broadcast_dimensions=(0, 1, 2)),
arg_descriptors=[RandArg((3, 1, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("broadcast_in_dim", "poly2",
lambda x: lax.broadcast_in_dim(x, [x.shape[0], 5, 6, x.shape[2], 4],
broadcast_dimensions=(0, 2, 3)),
arg_descriptors=[RandArg((3, 1, 4), _f32)],
polymorphic_shapes=["b1, _, b2"]),
PolyHarness("broadcast_in_dim", "transpose",
jax.grad(lambda x: jnp.sum(
lax.broadcast_in_dim(jnp.sin(x), [2, x.shape[0], 5, x.shape[2], 4],
broadcast_dimensions=(1, 2, 3)))),
arg_descriptors=[RandArg((3, 1, 4), _f32)],
polymorphic_shapes=["b1, _, b2"]),
PolyHarness("clamp", "",
lax.clamp,
arg_descriptors=[RandArg((3, 4, 5), _f32), RandArg((3, 4, 5), _f32),
RandArg((3, 4, 5), _f32)],
polymorphic_shapes=["b, ...", "b, ...", "b, ..."]),
PolyHarness("collapse", "",
lambda x: lax.collapse(x, 1, 4),
arg_descriptors=[RandArg((3, 4, 5, 6, 7), _f32)],
polymorphic_shapes=["b0, b1, _, b3, ..."]),
PolyHarness("concatenate", "",
lambda x: jnp.concatenate([x, x], axis=0),
arg_descriptors=[RandArg((3, 4, 5), _f32)],
polymorphic_shapes=["b0, b1, _"]),
PolyHarness("concatenate", "grad",
jax.grad(lambda x: jnp.sum(jnp.concatenate([x, jnp.sin(x)], axis=0))),
arg_descriptors=[RandArg((3, 4, 5), _f32)],
polymorphic_shapes=["b0, b1, _"]),
PolyHarness("conv_general_dilated", "1d_stride=1",
lambda lhs, rhs: lax.conv_general_dilated(
lhs, rhs,
window_strides=(1,),
padding="SAME",
rhs_dilation=None,
dimension_numbers=lax.ConvDimensionNumbers(lhs_spec=(0, 2, 1),
rhs_spec=(2, 1, 0),
out_spec=(0, 2, 1))),
arg_descriptors=[RandArg((1, 12, 16), _f32), RandArg((4, 16, 16), _f32)],
polymorphic_shapes=["_, b, _", None]),
# The same example from above, but with stride=2.
PolyHarness("conv_general_dilated", "1d_stride=2_even",
lambda lhs, rhs: lax.conv_general_dilated(
lhs, rhs,
window_strides=(2,),
padding="SAME",
rhs_dilation=None,
dimension_numbers=lax.ConvDimensionNumbers(lhs_spec=(0, 2, 1),
rhs_spec=(2, 1, 0),
out_spec=(0, 2, 1))),
arg_descriptors=[RandArg((1, 12, 16), _f32), RandArg((4, 16, 16), _f32)],
polymorphic_shapes=["_, b, _", None]),
# The same example from above, but with stride=2 and odd input size.
PolyHarness("conv_general_dilated", "1d_stride=2_odd",
lambda lhs, rhs: lax.conv_general_dilated(
lhs, rhs,
window_strides=(2,),
padding="SAME",
rhs_dilation=None,
dimension_numbers=lax.ConvDimensionNumbers(lhs_spec=(0, 2, 1),
rhs_spec=(2, 1, 0),
out_spec=(0, 2, 1))),
arg_descriptors=[RandArg((1, 13, 16), _f32), RandArg((4, 16, 16), _f32)],
polymorphic_shapes=["_, b, _", None]),
PolyHarness("conv_general_dilated", "1d_stride=2_zero_output",
lambda lhs, rhs: lax.conv_general_dilated(
lhs, rhs,
window_strides=(2,),
padding="VALID",
rhs_dilation=None,
dimension_numbers=lax.ConvDimensionNumbers(lhs_spec=(0, 2, 1),
rhs_spec=(2, 1, 0),
out_spec=(0, 2, 1))
).shape[1], # should be 0 in JAX native
arg_descriptors=[RandArg((1, 4, 16), _f32),
RandArg((8, 16, 16), _f32)],
polymorphic_shapes=["_, b, _",
None]),
# Issue #11402
PolyHarness("conv_general_dilated", "1d_2",
lambda lhs, rhs: lax.conv_transpose(lhs, rhs,
strides=(2,),
padding="SAME",
rhs_dilation=None,
transpose_kernel=False),
arg_descriptors=[RandArg((5, 12, 16), _f32), RandArg((4, 16, 16), _f32)],
polymorphic_shapes=["b, _, _", None],
tol=5e-5),
# Issue #11402
PolyHarness("conv_general_dilated", "1d_3",
lambda lhs, rhs: lax.conv_transpose(lhs, rhs,
strides=(2,),
padding="SAME",
rhs_dilation=None,
transpose_kernel=False),
arg_descriptors=[RandArg((5, 12, 16), _f32), RandArg((4, 16, 16), _f32)],
polymorphic_shapes=["_, b, _", None],
tol=5e-5),
PolyHarness("conv_general_dilated", "",
lambda lhs, rhs: lax.conv_general_dilated(
lhs, rhs,
window_strides=(2, 3),
padding=((0, 0), (0, 0)),
lhs_dilation=(1, 1),
rhs_dilation=(1, 2),
dimension_numbers=("NCHW", "OIHW", "NCHW"),
feature_group_count=1,
batch_group_count=1,
precision=None),
arg_descriptors=[RandArg((7, 3, 9, 10), _f32), RandArg((3, 3, 4, 5), _f32)],
polymorphic_shapes=["b, ...", None]),
[
[
PolyHarness(cum_name, "reduce_axis_poly",
lambda x: cum_func(x, axis=0),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness(cum_name, "reduce_axis_static",
lambda x: cum_func(x, axis=1),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."])
]
for cum_name, cum_func in [
("cumlogsumexp", lax_control_flow.cumlogsumexp),
("cummax", lax_control_flow.cummax),
("cummin", lax_control_flow.cummin),
("cumsum", lax_control_flow.cumsum),
("cumprod", lax_control_flow.cumprod)
]
],
PolyHarness("delta", "0",
lambda x: lax_internal._delta(_f32, x.shape, axes=(0, 1)) + x,
arg_descriptors=[RandArg((3, 1), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("dot_general", "",
lambda lhs, rhs: lax.dot_general(lhs, rhs,
dimension_numbers=(((2,), (1,)), ((0,), (0,)))),
arg_descriptors=[RandArg((3, 4, 4), _f32), RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ...", "b, ..."]),
PolyHarness("dynamic_slice", "idx=tuple_int",
# x:shape: (b, 4)
lambda x: lax.dynamic_slice(x, (0, 1), (x.shape[0], 2)),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("dynamic_slice", "idx=tuple_arg",
# x:shape: (b, 4)
lambda x, i0: lax.dynamic_slice(x, (i0, np.int32(1)), (x.shape[0], 2)),
arg_descriptors=[RandArg((3, 4), _f32), np.array(-2, dtype=np.int32)],
polymorphic_shapes=["b, ...", None]),
PolyHarness("dynamic_slice", "idx=array",
# x:shape: (b, 4)
lambda x, idx: lax.dynamic_slice(x, idx, (x.shape[0], 2)),
arg_descriptors=[RandArg((3, 4), _f32), np.array([-2, -1], dtype=np.int32)],
polymorphic_shapes=["b, ...", None]),
PolyHarness("dynamic_slice", "idx=tuple_int_start_oob_large",
# x:shape: (b, 4)
lambda x: lax.dynamic_slice(x, (1, 1), (x.shape[0], 2)),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("dynamic_slice", "idx=tuple_int_start_oob_small",
# x:shape: (b, 4)
lambda x: lax.dynamic_slice(x, (-1, 1), (x.shape[0] - 1, 2)),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("dynamic_slice_in_dim", "idx=0",
# x:shape: (b, 4)
lambda x: lax.dynamic_slice_in_dim(x, 0, x.shape[0], axis=0),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("dynamic_update_slice", "idx=tuple_int",
# x:shape: (b, 4)
lambda x: lax.dynamic_update_slice(x, x, (0, 0)),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("dynamic_update_slice", "idx=tuple_arg",
# x:shape: (b, 4)
lambda x, i0: lax.dynamic_update_slice(x, x, (i0, np.int32(0))),
arg_descriptors=[RandArg((3, 4), _f32), np.array(-2, dtype=np.int32)],
polymorphic_shapes=["b, ...", None]),
PolyHarness("dynamic_update_slice", "idx=array",
# x:shape: (b, 4)
lambda x, idx: lax.dynamic_update_slice(x, x, idx),
arg_descriptors=[RandArg((3, 4), _f32), np.array([-2, -1], dtype=np.int32)],
polymorphic_shapes=["b, _", None]),
[
PolyHarness("eig", f"shape={jtu.format_shape_dtype_string((3, 5, 5), dtype)}_poly={poly}_{left=}_{right=}",
lambda x, left, right: lax.linalg.eig(x, compute_left_eigenvectors=left, compute_right_eigenvectors=right),
arg_descriptors=[RandArg((3, 5, 5), dtype),
StaticArg(left), StaticArg(right)],
polymorphic_shapes=[poly])
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for poly in ["b, ...", "b, w, w"]
for left in ([True, False] if dtype == np.float32 else [True])
for right in ([True, False] if dtype == np.float32 else [False])
],
PolyHarness("einsum", "0",
lambda x: jnp.einsum("...i->...", x),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("einsum", "0_alt",
lambda x: jnp.einsum(x, (..., 1), [...]),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("einsum", "1",
lambda x, y: jnp.einsum("...ij,...jk->...ik", x, y),
arg_descriptors=[RandArg((3, 4, 5), _f32), RandArg((3, 5, 6), _f32)],
polymorphic_shapes=["b, ...", "b, ..."]),
PolyHarness("einsum", "1_alt",
lambda x, y: jnp.einsum(x, [..., 0, 1], y, (..., 1, 2), [..., 0, 2]),
arg_descriptors=[RandArg((3, 4, 5), _f32), RandArg((3, 5, 6), _f32)],
polymorphic_shapes=["b, ...", "b, ..."]),
PolyHarness("einsum", "2",
lambda x, y: jnp.einsum("...ij,jk->...ik", x, y),
arg_descriptors=[RandArg((3, 4, 5), _f32), RandArg((5, 6), _f32)],
polymorphic_shapes=["b, ...", None]),
PolyHarness("einsum", "2_alt",
lambda x, y: jnp.einsum(x, [..., 0, 1], y, [1, 2], [..., 0, 2]),
arg_descriptors=[RandArg((3, 4, 5), _f32), RandArg((5, 6), _f32)],
polymorphic_shapes=["b, ...", None]),
PolyHarness("einsum", "3",
# Reduced dimension is polymorphic
lambda x, y: jnp.einsum("ij,jk->ik", x, y),
arg_descriptors=[RandArg((3, 4), _f32), RandArg((4, 5), _f32)],
polymorphic_shapes=["_, b", "b, ..."]),
PolyHarness("einsum", "3_alt",
# Reduced dimension is polymorphic
lambda x, y: jnp.einsum(x, [0, 1], y, [1, 2], [0, 2]),
arg_descriptors=[RandArg((3, 4), _f32), RandArg((4, 5), _f32)],
polymorphic_shapes=["_, b", "b, ..."]),
PolyHarness("einsum", "4",
# Reduced dimension is polymorphic, and is 2*b
lambda x, y: jnp.einsum("ij,jk->ik",
jnp.concatenate([x, x], axis=1),
jnp.concatenate([y, y], axis=0)),
arg_descriptors=[RandArg((3, 4), _f32), RandArg((4, 5), _f32)],
polymorphic_shapes=["_, b", "b, ..."]),
PolyHarness("einsum", "4_alt",
# Reduced dimension is polymorphic, and is 2*b
lambda x, y: jnp.einsum(jnp.concatenate([x, x], axis=1), [0, 1],
jnp.concatenate([y, y], axis=0), [1, 2],
[0, 2]),
arg_descriptors=[RandArg((3, 4), _f32), RandArg((4, 5), _f32)],
polymorphic_shapes=["_, b", "b, ..."]),
PolyHarness("einsum", "multiple_contractions",
lambda x, y, z: jnp.einsum("ab,bc,cd->ad", x, y, z),
arg_descriptors=[RandArg((3, 2), _f32), RandArg((2, 3), _f32), RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ...", None, None]),
PolyHarness("einsum", "incompatible_contractions_error",
lambda x, y: jnp.einsum("ab,cb->ac", x, y),
arg_descriptors=[RandArg((2, 3), _f32), RandArg((2, 3), _f32)],
polymorphic_shapes=["(2, b0)", "(2, b1)"],
expect_error=(AssertionError,
"Incompatible reduction dimensions")),
PolyHarness("eye", "N=poly_M=None",
lambda x: jnp.eye(x.shape[0], dtype=_f32) + x,
arg_descriptors=[RandArg((3, 1), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("eye", "N=poly_M=poly",
lambda x: jnp.eye(x.shape[0], M=x.shape[0] + 2, dtype=_f32) + x,
arg_descriptors=[RandArg((3, 1), _f32)],
polymorphic_shapes=["b, ..."]),
[
PolyHarness("fft", f"{fft_type=}_{nr_fft_lengths=}",
lambda x, fft_type, nr_fft_lengths: lax.fft_p.bind(
x, fft_type=fft_type,
fft_lengths=tuple(
x.shape[-nr_fft_lengths:] if fft_type != lax.FftType.IRFFT else
[(x.shape[-1] - 1) * 2])),
arg_descriptors=[
RandArg((3, 4, 5, 6),
np.float32 if fft_type == lax.FftType.RFFT else np.complex64),
StaticArg(fft_type),
StaticArg(nr_fft_lengths)],
# All axes but the last one are dynamic. This means that the test
# with nr_fft_lengths==1 will not have dynamic fft_lengths.
polymorphic_shapes=["b0, b1, b2, ..."],
tol=1e-4)
for fft_type in (lax.FftType.FFT, lax.FftType.IFFT,
lax.FftType.RFFT, lax.FftType.IRFFT)
for nr_fft_lengths in (1, 2)
],
PolyHarness("full", "",
lambda x: lax.full((x.shape[0], 2), 3.) + x,
arg_descriptors=[RandArg((3, 1), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("gather", "1d",
lambda operand, start_indices, x: lax.gather(
operand,
start_indices,
dimension_numbers=lax.GatherDimensionNumbers(
offset_dims=(1,),
collapsed_slice_dims=(),
start_index_map=(0,)),
slice_sizes=x.shape,
mode="promise_in_bounds"),
arg_descriptors=[
RandArg((10,), np.float32),
np.random.randint(0, high=10, size=(3, 1),
dtype=np.int32),
np.zeros((10,), dtype=jnp.int32),
],
polymorphic_shapes=["(t, )", "(3, 1)", "(t)"]),
PolyHarness("image_resize", "linear_0",
lambda x: jax.image.resize(x, (x.shape[0], 2 * x.shape[1], 2 * x.shape[2], x.shape[3]),
method="linear"),
arg_descriptors=[RandArg((3, 16, 32, 3), _f32)],
polymorphic_shapes=["_, b1, b2, ..."]),
PolyHarness("image_resize", "linear_to_fixed_dim",
lambda x: jax.image.resize(x, (x.shape[0], 64, 64, x.shape[3]),
method="linear"),
arg_descriptors=[RandArg((3, 16, 32, 3), _f32)],
polymorphic_shapes=["_, b1, b2, ..."]),
PolyHarness("image_resize", "nearest_0",
lambda x: jax.image.resize(x, (x.shape[0], 2 * x.shape[1], 2 * x.shape[2], x.shape[3]),
method="nearest"),
arg_descriptors=[RandArg((3, 5, 7, 3), _f32)],
polymorphic_shapes=["_, b1, b2, ..."]),
PolyHarness("index_in_dim", "0",
lambda x: lax.index_in_dim(x, -1, axis=0, keepdims=False),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("index_in_dim", "idx=neg",
lambda x: lax.index_in_dim(x, -1, axis=0, keepdims=False),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("index_in_dim", "idx=last",
lambda x: lax.index_in_dim(x, x.shape[0] - 1, axis=0, keepdims=False),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.insert", "insert=constant",
lambda x: jnp.insert(x, jnp.arange(3, dtype=_i32), np.array([3, 4, 5], dtype=_i32)),
arg_descriptors=[RandArg((12,), _i32)],
polymorphic_shapes=["b, ..."],
expect_error=expect_error_associative_scan),
PolyHarness("jnp.insert", "insert=poly",
lambda x: jnp.insert(x, jnp.arange(x.shape[0], dtype=_i32), x, axis=0),
arg_descriptors=[RandArg((12, 3), _i32)],
polymorphic_shapes=["b0, b1, ..."],
expect_error=expect_error_associative_scan),
PolyHarness("iota", "",
lambda x: x + lax.iota(_f32, x.shape[0]),
arg_descriptors=[RandArg((3,), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("linspace", "",
lambda x: jnp.linspace(0, x.shape[0], 4),
arg_descriptors=[RandArg((30,), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("linspace", "num_poly",
lambda x: jnp.linspace(0, 100, x.shape[0]),
arg_descriptors=[RandArg((30,), _f32)],
polymorphic_shapes=["b, ..."],
symbolic_constraints=["b >= 2"]),
PolyHarness("matmul", "0",
jnp.matmul,
arg_descriptors=[RandArg((7, 8, 4), _f32), RandArg((7, 4, 5), _f32)],
polymorphic_shapes=["b, ...", "b, ..."],
tol=1e-5),
PolyHarness("matmul", "1",
jnp.matmul,
arg_descriptors=[RandArg((7, 8, 4), _f32), RandArg((4, 5), _f32)],
polymorphic_shapes=["b, ...", None],
tol=1e-5),
[
PolyHarness("mean",
f"{axis=}_{keepdims=}_where=None",
lambda x, axis, keepdims: jnp.mean(x, axis=axis, keepdims=keepdims, where=None),
arg_descriptors=[RandArg((7, 8, 4), _f32), StaticArg(axis), StaticArg(keepdims)],
polymorphic_shapes=["b, ..."])
for keepdims in [False, True]
for axis in [None, (0,), (0, 1), (1,)]
],
[
PolyHarness("mean",
f"{axis=}_{keepdims=}_where=Some",
lambda x, where, axis, keepdims: jnp.mean(x, axis=axis, keepdims=keepdims, where=where),
arg_descriptors=[RandArg((7, 8, 4), _f32), RandArg((7, 8, 4), np.bool_),
StaticArg(axis), StaticArg(keepdims)],
polymorphic_shapes=["b, ...", "b, ..."])
for keepdims in [False, True]
for axis in [None, (0,), (0, 1), (1,)]
],
PolyHarness("jnp.nonzero", "size=constant",
lambda x: jnp.nonzero(x % 3, size=10, fill_value=100),
arg_descriptors=[RandArg((3, 2, 4), _i32)],
polymorphic_shapes=["b, ..."],
expect_error=expect_error_associative_scan),
PolyHarness("jnp.nonzero", "size=poly",
lambda x: jnp.nonzero(x % 3, size=x.shape[0] * 2, fill_value=100),
arg_descriptors=[RandArg((3, 2, 4), _i32)],
polymorphic_shapes=["b, ..."],
expect_error=expect_error_associative_scan),
PolyHarness("one_hot", "poly_num_classes",
lambda x, y: jax.nn.one_hot(x, y.shape[0]),
arg_descriptors=[np.arange(16, dtype=_i32), RandArg((16,), _f32)],
polymorphic_shapes=[None, "b0, ..."]),
PolyHarness("one_hot", "all_poly",
lambda x, y: jax.nn.one_hot(x, y.shape[0]),
arg_descriptors=[np.arange(16, dtype=_i32), RandArg((16,), _f32)],
polymorphic_shapes=["b, ...", "b, ..."]),
PolyHarness("ones", "",
lambda x: jnp.ones(x.shape, dtype=_f32) + x,
arg_descriptors=[RandArg((3, 2, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("pad", "",
lax.pad,
arg_descriptors=[RandArg((3, 2, 5), _f32), np.float32(5.),
StaticArg(((0, 0, 0), (0, 0, 0), (1, 1, 1)))],
polymorphic_shapes=["b, ...", None]),
PolyHarness("pad", "poly_padding_config",
lambda x: lax.pad(x, _f32(0.),
((x.shape[0], x.shape[1], x.shape[0]),
(0, 0, 0))),
arg_descriptors=[RandArg((3, 2), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.pad", "mode=constant",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="constant"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.pad", "mode=constant_bminus1",
# We slice first the unknown dimension to make it of size b - 1
# which may be 0.
lambda x: jnp.pad(lax.dynamic_slice_in_dim(x, 1, x.shape[0] - 1,
axis=0),
[[x.shape[0], 0], [x.shape[1], 1]],
mode="constant"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.pad", "mode=edge",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="edge"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.pad", "mode=maximum",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="maximum"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("jnp.pad", "mode=maximum_stat_length=b",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="maximum", stat_length=((x.shape[0] // 2, 2), (2, 2))),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."],
symbolic_constraints=["b >= 2"]),
PolyHarness("jnp.pad", "mode=linear_ramp",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="linear_ramp"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."],
symbolic_constraints=["b >= 2"]),
PolyHarness("jnp.pad", "mode=reflect_odd",
lambda x: jnp.pad(x, [[x.shape[0] - 1, 0], [x.shape[1], 1]],
mode="reflect", reflect_type="odd"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."],
symbolic_constraints=["b >= 2"]),
PolyHarness("jnp.pad", "mode=reflect_odd_error",
lambda x: jnp.pad(x, [[x.shape[0] - 1, 0], [x.shape[1], 1]],
mode="reflect", reflect_type="odd"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."],
expect_error=(ValueError, "Shape polymorphism is supported for jnp.pad")),
PolyHarness("jnp.pad", "mode=reflect_even",
lambda x: jnp.pad(x, [[x.shape[0] - 1, 0], [x.shape[1], 1]],
mode="reflect", reflect_type="even"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."],
symbolic_constraints=["b >= 2"]),
PolyHarness("jnp.pad", "mode=symmetric_odd",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="symmetric", reflect_type="odd"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."],
symbolic_constraints=["b >= 2"]),
PolyHarness("jnp.pad", "mode=symmetric_even",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="symmetric", reflect_type="even"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."],
symbolic_constraints=["b >= 2"]),
PolyHarness("jnp.pad", "mode=wrap",
lambda x: jnp.pad(x, [[x.shape[0], 0], [x.shape[1], 1]],
mode="wrap"),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("percentile", "axis=None",
lambda x: jnp.percentile(x, 50, axis=None),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("nanquantile", "axis=None",
lambda x: jnp.nanquantile(x, .5, axis=None),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("percentile", "axis=0",
lambda x: jnp.percentile(x, 50, axis=0),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("nanquantile", "axis=0",
lambda x: jnp.nanquantile(x, .5, axis=0),
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("inv", "",
lambda x: jnp.linalg.inv(jnp.eye(x.shape[0])),
arg_descriptors=[RandArg((3, 3), _f32)],
polymorphic_shapes=["b, b, ..."],
override_jax_config_flags={"jax_export_ignore_forward_compatibility": True}),
[
PolyHarness( # pylint: disable=g-complex-comprehension
"qr", f"shape={jtu.format_shape_dtype_string(shape, dtype)}_poly={poly}_{full_matrices=}",
lambda x, full_matrices: lax.linalg.qr(x, full_matrices=full_matrices),
arg_descriptors=[RandArg(shape, dtype), StaticArg(full_matrices)],
polymorphic_shapes=[poly],
symbolic_constraints=constraints)
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for shape, poly, full_matrices, constraints in [
((2, 0, 4), "b, ...", False, ()), # m = 0
((2, 4, 0), "b, ...", False, ()), # n = 0
((2, 3, 4, 4), "b1, b2, ...", False, ()), # m == n
((2, 3, 4, 4), "b1, b2, ...", True, ()),
((2, 3, 4, 5), "b1, b2, ...", False, ()), # m < n
((2, 3, 4, 5), "b1, b2, ...", True, ()),
((2, 3, 8, 4), "b1, b2, ...", False, ()), # m > n
((2, 3, 8, 4), "b1, b2, ...", True, ()),
# Dynamic shapes are also supported for non-batch dimensions with
# some constraints.
((2, 3, 4, 4), "b1, b2, m, m", False, ()), # m == n
((2, 3, 4, 4), "b1, b2, m, m", True, ()),
((2, 3, 4, 5), "b1, b2, m, n", False, ["m + 1 <= n"]), # m < n
((2, 3, 4, 5), "b1, b2, m, n", True, ["m + 1 <= n"]),
((2, 3, 8, 4), "b1, b2, m, n", False, ["n <= m"]), # m > n
((2, 3, 8, 4), "b1, b2, m, n", True, ["n <= m"]),
]
],
[
PolyHarness(
"lu_pivots_to_permutation",
f"shape={jtu.format_shape_dtype_string(shape, np.int32)}_poly={poly}_{permutation_size=}",
lax.linalg.lu_pivots_to_permutation,
arg_descriptors=[RandArg(shape, np.int32), StaticArg(permutation_size)],
polymorphic_shapes=[poly],
symbolic_constraints=constraints,
)
for shape, poly, permutation_size, constraints in [
((4,), None, 8, ()),
((2, 3, 4), "b1, b2, ...", 8, ()),
((4,), "b", 8, ["b <= 8"]),
((2, 3, 4), "b1, b2, b3", 8, ["b3 <= 8"]),
]
],
[
# Tracing errors are only thrown when the trailing dimension of pivots
# is static. Otherwise, the error is thrown at runtime.
PolyHarness(
"lu_pivots_to_permutation_error",
f"shape={jtu.format_shape_dtype_string(shape, np.int32)}_poly={poly}_{permutation_size=}",
lax.linalg.lu_pivots_to_permutation,
arg_descriptors=[RandArg(shape, np.int32), StaticArg(permutation_size)],
polymorphic_shapes=[poly],
symbolic_constraints=constraints,
expect_error=(ValueError, "Output permutation size"),
)
for shape, poly, permutation_size, constraints in [
((4,), None, 3, ()),
((2, 3, 4), "b1, b2, ...", 3, ()),
((4,), "b", 8, ["b >= 9"]),
((2, 3, 4), "b1, b2, b3", 8, ["b3 >= 9"]),
]
],
[
PolyHarness( # pylint: disable=g-complex-comprehension
"lu", f"shape={jtu.format_shape_dtype_string(shape, dtype)}_poly={poly}",
lax.linalg.lu,
arg_descriptors=[RandArg(shape, dtype)],
polymorphic_shapes=[poly])
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for shape, poly in [
((5, 4), "m, n"),
((2, 0, 4), "b, ..."),
((2, 4, 0), "b, ..."),
((2, 3, 4, 4), "b1, b2, ..."),
((2, 3, 4, 5), "b1, b2, ..."),
((2, 3, 8, 4), "b1, b2, ..."),
((2, 3, 4, 5), "b1, b2, m, n"),
]
],
[
PolyHarness( # pylint: disable=g-complex-comprehension
"eigh", f"shape={jtu.format_shape_dtype_string(shape, dtype)}_poly={poly}_{lower=}",
lambda x, lower: lax.linalg.eigh(x, lower=lower),
arg_descriptors=[RandArg(shape, dtype), StaticArg(lower)],
polymorphic_shapes=[poly])
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for lower in [True, False]
for shape, poly in [
((4, 4), "n, n"),
((2, 3, 4, 4), "b1, b2, ..."),
((2, 3, 4, 4), "b1, b2, n, n"),
]
],
[
PolyHarness( # pylint: disable=g-complex-comprehension
"eigh_shape_error", f"shape={jtu.format_shape_dtype_string(shape, dtype)}_poly={poly}",
lambda x: lax.linalg.eigh(x, symmetrize_input=False),
arg_descriptors=[RandArg(shape, dtype)],
polymorphic_shapes=[poly],
expect_error=(ValueError, "Argument to symmetric eigendecomposition"))
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for shape, poly in [
((4, 5), "m, n"),
((2, 3, 4, 5), "b1, b2, ..."),
((2, 3, 4, 5), "b1, b2, m, n"),
]
],
[
PolyHarness( # pylint: disable=g-complex-comprehension
"svd", f"shape={jtu.format_shape_dtype_string(shape, dtype)}_poly={poly}_{full_matrices=}_{compute_uv=}",
lambda x, full_matrices, compute_uv: lax.linalg.svd(x, full_matrices=full_matrices, compute_uv=compute_uv),
arg_descriptors=[RandArg(shape, dtype), StaticArg(full_matrices), StaticArg(compute_uv)],
polymorphic_shapes=[poly],
symbolic_constraints=constraints)
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for compute_uv in [True, False]
for full_matrices in ([True, False] if compute_uv else [True])
for shape, poly, constraints in [
((2, 0, 4), "b, ...", ()),
((2, 4, 0), "b, ...", ()),
((2, 3, 4, 4), "b1, b2, ...", ()),
((2, 3, 4, 5), "b1, b2, ...", ()),
((2, 3, 8, 4), "b1, b2, ...", ()),
# The constraints listed here are only for the GPU implementation
# which selects an algorithm based on the size of the matrix.
((5, 4), "m, n", ["n <= m", "m <= 32", "n <= 32"]),
((2, 3, 4, 5), "b1, b2, m, n", ["m <= n", "m <= 32", "n <= 32"]),
]
],
[
# The random primitive tests, with threefry (both partitionable and
# non-partitionable), and unsafe_rbg.
[
PolyHarness("random_gamma", f"{flags_name}",
lambda key, a: jax.vmap(jax.random.gamma)(
jax.random.wrap_key_data(key), a),
arg_descriptors=[RandArg((3, key_size), np.uint32),
RandArg((3, 4, 5), _f32)],
polymorphic_shapes=["b, ...", "b, w, ..."], tol=1E-5,
override_jax_config_flags=override_jax_config_flags), # type: ignore
PolyHarness("random_categorical", f"axis=0_{flags_name}",
lambda key, a: jax.random.categorical(
jax.random.wrap_key_data(key), a, axis=0),
arg_descriptors=[RandArg((key_size,), np.uint32),
RandArg((3, 8), _f32)],
polymorphic_shapes=[None, "b0, b1"],
override_jax_config_flags=override_jax_config_flags), # type: ignore
PolyHarness("random_categorical", f"axis=1_{flags_name}",
lambda key, a: jax.random.categorical(
jax.random.wrap_key_data(key), a, axis=1),
arg_descriptors=[RandArg((key_size,), np.uint32),
RandArg((3, 5, 8), _f32)],
polymorphic_shapes=[None, "b0, b1, b2"],
override_jax_config_flags=override_jax_config_flags), # type: ignore
PolyHarness("random_categorical", f"axis=1_then_reshape_{flags_name}",
lambda key, a: jax.random.categorical(
jax.random.wrap_key_data(key), a, axis=1).reshape(-1),
arg_descriptors=[RandArg((key_size,), np.uint32),
RandArg((3, 5, 8), _f32)],
polymorphic_shapes=[None, "b0, b1, b2"],
override_jax_config_flags=override_jax_config_flags), # type: ignore
PolyHarness("random_categorical", f"0_dim_{flags_name}", # One axis has 0 size
lambda key, a: jax.random.categorical(
jax.random.wrap_key_data(key), a, axis=1),
arg_descriptors=[RandArg((key_size,), np.uint32),
RandArg((3, 5, 0), _f32)],
polymorphic_shapes=[None, "b0, b1, ..."],
override_jax_config_flags=override_jax_config_flags), # type: ignore
[
PolyHarness("random_choice", f"{flags_name}_arr_poly={arr_poly}_shape_poly={shape_poly}_replace={replace}_use_p={use_p}",
lambda key, a, res_shape, use_p: jax.random.choice(
jax.random.wrap_key_data(key),
a,
shape=res_shape.shape,
p=jnp.full((a.shape[1],), 0.1, dtype=_f32) if use_p else None,
axis=1,
replace=replace),
arg_descriptors=[RandArg((key_size,), np.uint32),
RandArg((64, 12, 4), _f32), # sample on axis=1
RandArg((3, 4), _f32),
StaticArg(use_p)],
polymorphic_shapes=[None,
"b0, b1, b2" if arr_poly else None,
"b3, b4" if shape_poly else None],
# The array sampled dimension must be larger than res_shape.size
symbolic_constraints=[
"b1 >= 12" if arr_poly else "1 >= 0",
"b1 >= b3*b4" if arr_poly and shape_poly else "1 >= 0",
"12 >= b3*b4" if shape_poly else "1 >= 0"
],
override_jax_config_flags=override_jax_config_flags,
expect_error=(
(NotImplementedError, "permutation")
if arr_poly and not use_p else None)) # type: ignore
# np.insert used in random.choice tries to coerce shape_poly to
# integer arrays, but only when the arr_poly is False.
for arr_poly in [True, False]
for shape_poly in [True, False]
for replace in [True, False]
for use_p in [True, False]
],
PolyHarness("random_split", f"{flags_name}",
lambda key, a: jax.random.key_data(
jax.random.split(jax.random.wrap_key_data(key),
2 * a.shape[0])),
arg_descriptors=[RandArg((key_size,), np.uint32),
RandArg((3, 4), _f32)],
polymorphic_shapes=[None, "b0, ..."],
override_jax_config_flags=override_jax_config_flags), # type: ignore
# Works when the known dimensions are known to be even or odd.
PolyHarness("random_uniform", f"even_1_{flags_name}",
lambda key, a: jax.random.uniform(jax.random.wrap_key_data(key),
a.shape, dtype=_f32),
arg_descriptors=[RandArg((key_size,), np.uint32), RandArg((3, 4, 5), _f32)],
polymorphic_shapes=[None, "b0, 4, 5"],
override_jax_config_flags=override_jax_config_flags), # type: ignore
PolyHarness("random_uniform", f"even_2_{flags_name}",
lambda key, a: jax.random.uniform(jax.random.wrap_key_data(key),
a.shape, dtype=_f32),
arg_descriptors=[RandArg((key_size,), np.uint32), RandArg((3, 4), _f32)],
polymorphic_shapes=[None, "b0, 2*b1"],
override_jax_config_flags=override_jax_config_flags), # type: ignore
PolyHarness("random_uniform", f"error_unknown_evenness_{flags_name}",
lambda key, a: jax.random.uniform(jax.random.wrap_key_data(key),
a.shape, dtype=_f32),
arg_descriptors=[RandArg((key_size,), np.uint32),
RandArg((3, 5), _f32)],
polymorphic_shapes=[None, "b0, b1"],
override_jax_config_flags=override_jax_config_flags) # type: ignore
]
for key_size, flags_name, override_jax_config_flags in [
(2, "threefry_non_partitionable",
dict(jax_default_prng_impl="threefry2x32", jax_threefry_partitionable=False)),
(2, "threefry_partitionable",
dict(jax_default_prng_impl="threefry2x32", jax_threefry_partitionable=True)),
(4, "unsafe_rbg",
dict(jax_default_prng_impl="unsafe_rbg"))
]
],
# For reduce_window we have a variant with one reduction axis of
# non-static shape, and one with additionally the dimension window
# non-static.
PolyHarness("reduce_window", "min_window_size=static",
# x: f32[b, 8]
lambda x: lax.reduce_window(x, np.array(1., _f32), lax.min,
(2, 2), (1, 1), "VALID"),
arg_descriptors=[RandArg((3, 8), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("reduce_window", "min_window_size=dynamic",
# x: f32[b, 8]
lambda x: lax.reduce_window(x, np.array(1., _f32), lax.min,
(2, x.shape[0]), (1, 1), "VALID"),
arg_descriptors=[RandArg((3, 8), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("reduce_window", "min_plus_max_window_size=static",
# x: f32[b, 8]
lambda x: (
# Test that we don't get confusion for the reducer name.
lax.reduce_window(x, np.array(1., _f32), lax.min,
(2, 2), (1, 1), "VALID") +
lax.reduce_window(x, np.array(1., _f32), lax.max,
(2, 2), (1, 1), "VALID")),
arg_descriptors=[RandArg((3, 8), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("reduce_window", "min_plus_max_window_size=dynamic",
# x: f32[b, 8]
lambda x: (
# Test that we don't get confusion for the reducer name.
lax.reduce_window(x, np.array(1., _f32), lax.min,
(2, x.shape[0]), (1, 1), "VALID") +
lax.reduce_window(x, np.array(1., _f32), lax.max,
(2, x.shape[0]), (1, 1), "VALID")),
arg_descriptors=[RandArg((3, 8), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("reduce_window", "add_monoid_base_window_size=static",
# x: f32[b, 8]
lambda x: lax.reduce_window(x, np.array(0., _f32), lax.add,
(2, 2), (1, 1), "VALID"),
arg_descriptors=[RandArg((3, 8), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("reduce_window", "add_monoid_base_window_size=dynamic",
# x: f32[b, 8]
lambda x: lax.reduce_window(x, np.array(0., _f32), lax.add,
(2, x.shape[0]), (1, 1), "VALID"),
arg_descriptors=[RandArg((3, 8), _f32)],
polymorphic_shapes=["b, ..."]),
# https://github.com/jax-ml/jax/issues/11804
# Use the reshape trick to simulate a polymorphic dimension of 16*b.
# (See test "conv_general_dilated.1d_1" above for more details.)
PolyHarness("reduce_window", "add_monoid_strides_window_size=static",
# x: f32[1, 16*b, 1]
lambda x: lax.reduce_window(
jnp.reshape(x, (1, -1, 1)),
np.array(0., _f32), lax.add, (1, 4, 1), (1, 2, 1), "SAME"),
arg_descriptors=[RandArg((1, 128, 16), _f32)],
polymorphic_shapes=["_, b1, ..."]),
PolyHarness("reduce_window", "add_generic_window_size=static",
# x: f32[1, 16*b, 1]
# Use an initial value of 1. to trigger the generic reduction path
lambda x: lax.reduce_window(
jnp.reshape(x, (1, -1, 1)),
np.array(1., _f32), lax.add, (1, 4, 1), (1, 2, 1), "SAME"),
arg_descriptors=[RandArg((1, 128, 16), _f32)],
polymorphic_shapes=["_, b1, ..."]),
PolyHarness("reduce_window", "variadic_generic_window_size=static",
# x: f32[b, 8] y: f32[b, 8]
lambda x, y: lax.reduce_window(
(x, y), (np.array(1., _f32), np.array(2, _i32)),
lambda xy0, xy1: (lax.add(xy0[0], xy1[0]),
lax.sub(xy0[1], xy1[1])),
(2, 2), (1, 1), "VALID"),
arg_descriptors=[RandArg((3, 8), _f32), RandArg((3, 8), _i32)],
polymorphic_shapes=["b, ...", "b, ..."]),
PolyHarness("reduce_window", "variadic_generic_window_size=dynamic",
# x: f32[b, 8] y: f32[b, 8]
lambda x, y: lax.reduce_window(
(x, y), (np.array(1., _f32), np.array(2, _i32)),
lambda xy0, xy1: (lax.add(xy0[0], xy1[0]),
lax.sub(xy0[1], xy1[1])),
(2, x.shape[0]), (1, 1), "VALID"),
arg_descriptors=[RandArg((3, 8), _f32), RandArg((3, 8), _i32)],
polymorphic_shapes=["b, ...", "b, ..."]),
# TODO(necula): not yet supported, but also unlikely to come up.
# PolyHarness("random_uniform", "odd",
# lambda key, a: jax.random.uniform(key, (2 * a.shape[0] + 1, a.shape[1]),
# dtype=_f32),
# [RandArg((2,), np.uint32), RandArg((3, 5), _f32)],
# polymorphic_shapes=[None, "b0, ..."]),
[
PolyHarness("reduce", reduce_op.__name__,
lambda x: reduce_op(x, axis=-1, keepdims=True), # type: ignore
arg_descriptors=[RandArg((3, 5), _f32)],
polymorphic_shapes=["b, ..."])
for reduce_op in [jnp.all, jnp.any, jnp.max, jnp.min, jnp.prod, jnp.sum]
],
# Repeat f32[b, 2] * 3
PolyHarness("repeat", "repeats=int_axis=0",
lambda x: jnp.repeat(x, repeats=3, axis=0),
arg_descriptors=[RandArg((3, 2), _f32)],
polymorphic_shapes=["b, ..."]),
# Repeat f32[b, 2] * b
PolyHarness("repeat", "repeats=poly_axis=0",
lambda x: jnp.repeat(x, repeats=x.shape[0], axis=0),
arg_descriptors=[RandArg((3, 2), _f32)],
polymorphic_shapes=["b, ..."]),
# Repeat f32[b, 2] * b
PolyHarness("repeat", "repeats=poly_axis=None",
lambda x: jnp.repeat(x, repeats=x.shape[0], axis=None),
arg_descriptors=[RandArg((3, 2), _f32)],
polymorphic_shapes=["b, ..."]),
# Repeat f32 * b
PolyHarness("repeat", "repeats=poly_axis=None_scalar",
lambda x, y: jnp.expand_dims(jnp.repeat(x, repeats=y.shape[0], axis=None), axis=1) + y,
arg_descriptors=[RandArg((), _f32), RandArg((3, 1), _f32)],
polymorphic_shapes=[None, "b0, ..."]),
PolyHarness("repeat", "repeats=poly_axis=None_total_repeat_length1",
lambda x: jnp.repeat(x, repeats=x.shape[0], axis=None, total_repeat_length=8),
arg_descriptors=[RandArg((3, 2), _f32)],
polymorphic_shapes=["b, ..."],
expect_error=(ValueError, "jnp.repeat with a non-constant `repeats` is supported only .*")),
PolyHarness("reshape", "0",
lambda x: x.reshape([x.shape[0], -1]),
arg_descriptors=[RandArg((3, 2, 3), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("reshape", "1",
lambda x: x.reshape([x.shape[0], -1]),
arg_descriptors=[RandArg((3, 2, 3), _f32)],
polymorphic_shapes=["b0, b1, ..."]),
PolyHarness("reshape", "2",
lambda x: x.reshape([x.shape[0], -1, x.shape[3], x.shape[2]]),
arg_descriptors=[RandArg((3, 4, 5, 6, 7), _f32)],
polymorphic_shapes=["b0, _, b2, b3, ..."]),
PolyHarness("reshape", "3",
lambda x: jnp.reshape(x, [2, -1]),
arg_descriptors=[RandArg((3, 4, 5, 6, 7), _f32)],
polymorphic_shapes=["b0, _, b2, ..."]),
PolyHarness("reshape", "_issue_9975",
# The newshape is a scalar
lambda x: jnp.reshape(x, x.shape[0] * x.shape[1]),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("reshape", "error",
lambda x: x.reshape([x.shape[0], -1, 3]),
arg_descriptors=[RandArg((3, 2, 4), _f32)],
polymorphic_shapes=["b, ..."],
expect_error=(core.InconclusiveDimensionOperation,
re.escape(
"Cannot divide evenly the sizes of shapes (b, 2, 4) and (b, -1, 3)"))),
PolyHarness("roll", "axis=0",
lambda x: jnp.roll(x, 2, axis=0),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("roll", "axis=None",
lambda x: jnp.roll(x, 2),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("scatter_add", "",
partial(lax.scatter_add, indices_are_sorted=False, unique_indices=True),
arg_descriptors=[RandArg((7, 4), _f32), # op: [b, 4]
np.array([[1], [2]], np.int32), # indices: [2, 1]
RandArg((7, 2), _f32), # updates: [b, 2]
StaticArg(lax.ScatterDimensionNumbers((0,), (1,), (1,)))],
polymorphic_shapes=["b, ...", None, "b, ..."]),
PolyHarness("scatter_add", "clip0",
partial(lax.scatter_add, indices_are_sorted=False, unique_indices=True, mode=lax.GatherScatterMode.CLIP),
arg_descriptors=[RandArg((7, 4), _f32), # op: [b, 4]
np.array([[1], [2]], np.int32), # indices: [2, 1]
RandArg((7, 2), _f32), # updates: [b, 2]
StaticArg(lax.ScatterDimensionNumbers((0,), (1,), (1,)))],
polymorphic_shapes=["b, ...", None, "b, ..."]),
PolyHarness("scatter_add", "clip1",
partial(lax.scatter_add, indices_are_sorted=False, unique_indices=True, mode=lax.GatherScatterMode.CLIP),
arg_descriptors=[RandArg((7, 4), _f32), # op: [b, 4]
# indices: [b, 2]
np.array([[1, 2], [-2, 0], [6, 4], [7, -1], [1, 0], [3, 0], [0, 5]], np.int32),
RandArg((7, 1), _f32), # updates: [b, 1]
StaticArg(lax.ScatterDimensionNumbers((1,), (0,), (0, 1,)))],
polymorphic_shapes=["b, ...", "b, ...", "b, ..."]),
PolyHarness("scatter_grad", "",
lambda *args: jax.grad(
lambda *args:
jnp.sum(lax.scatter( # type: ignore
*args,
indices_are_sorted=False,
unique_indices=False,
))
)(*args),
arg_descriptors=[RandArg((7, 4), _f32), # : [b, 4]
np.array([[1], [2]], np.int32), # indices: [2, 1]
RandArg((7, 2), _f32), # updates: [b, 2]
StaticArg(lax.ScatterDimensionNumbers((0,), (1,), (1,))),
],
polymorphic_shapes=["b, ...", None, "b, ..."]),
PolyHarness("scatter_grad", "poly_indices",
lambda *args: jax.grad(
lambda *args:
jnp.sum(lax.scatter( # type: ignore
*args,
indices_are_sorted=False,
unique_indices=False))
)(*args),
arg_descriptors=[RandArg((7, 4), _f32), # op: [b, 4]
# indices: [b, 2]
np.array(
[[1, 2], [-2, 0], [6, 4], [7, -1], [1, 0],
[3, 0], [0, 5]], np.int32),
RandArg((7, 1), _f32), # updates: [b, 1]
StaticArg(lax.ScatterDimensionNumbers((1,), (0,), (0, 1))),
],
polymorphic_shapes=["b, ...", "b, ...", "b, ..."]),
[
PolyHarness("segment", f"{name}_bucket_size_{bucket_size}",
lambda x, segment_ids: op(
x, # f32[b, s, buckets]
segment_ids, # i32[b]
num_segments=x.shape[1],
bucket_size=(x.shape[2]
if bucket_size == "poly" else bucket_size)),
arg_descriptors=[RandArg((8, 5, 2), _f32),
np.array([0, 0, 0, 1, 1, 3, 3, 3],
dtype=np.int32)
],
polymorphic_shapes=["b, s, buckets", "b"])
for name, op in [
("max", ops.segment_max),
("min", ops.segment_min),
("sum", ops.segment_sum),
("prod", ops.segment_prod),
]
for bucket_size in [None, 2, "poly"]
],
[
PolyHarness("schur",
f"shape={jtu.format_shape_dtype_string(shape, dtype)}_{poly=}_{compute_schur_vectors=}",
lambda a, compute_schur_vectors: lax.linalg.schur(
a, compute_schur_vectors=compute_schur_vectors),
arg_descriptors=[RandArg(shape, dtype),
StaticArg(compute_schur_vectors)],
polymorphic_shapes=[poly])
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for compute_schur_vectors in [True, False]
for (shape, poly) in [
((3, 3), "w, w"),
((3, 4, 4), "b, w, w"),
]
],
PolyHarness("select", "0",
# x.shape = (b, 3)
lambda x: lax.select(x > 5., x, x),
arg_descriptors=[RandArg((7, 3), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("select", "1",
# x.shape = (b, 3); y.shape = (3,)
jax.vmap(lambda x, y: lax.select(x > 5., x, y), in_axes=[0, None]),
arg_descriptors=[RandArg((7, 3), _f32), RandArg((3,), _f32)],
polymorphic_shapes=["b, ...", None]),
PolyHarness("slice", "entire_axis",
lambda x: lax.slice(x, start_indices=(0, 1), limit_indices=(x.shape[0], 3)),
arg_descriptors=[RandArg((7, 3), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("slice_in_dim", "entire_axis",
lambda x: lax.slice_in_dim(x, 0, x.shape[0], stride=1, axis=0),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("slice_in_dim", "start=neg",
lambda x: lax.slice_in_dim(x, -1, x.shape[0], stride=1, axis=0),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("slice_in_dim", "limit=neg",
lambda x: lax.slice_in_dim(x, 0, -1, stride=1, axis=0),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("slice_in_dim", "stride=2_even",
lambda x: lax.slice_in_dim(x, 0, x.shape[0], stride=2, axis=0),
arg_descriptors=[RandArg((12, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("slice_in_dim", "stride=2_odd",
lambda x: lax.slice_in_dim(x, 0, x.shape[0], stride=2, axis=0),
arg_descriptors=[RandArg((13, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("slice_in_dim", "stride_sym",
lambda x: lax.slice_in_dim(x, 0, x.shape[0], stride=1 + x.shape[0] // 4, axis=0),
arg_descriptors=[RandArg((13, 4), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("sort", "",
lambda a: lax.sort(a),
arg_descriptors=[RandArg((16,), _f32)],
polymorphic_shapes=["b"]),
PolyHarness("jvp_sort", "",
lambda a: jax.jvp(lax.sort, (a,), (a,)),
arg_descriptors=[RandArg((16,), _f32)],
polymorphic_shapes=["b"]),
PolyHarness("jnp_split", "idx_tuple_ct",
# The indices are a tuple with constants
lambda a: jnp.split(a, (2,)),
arg_descriptors=[RandArg((16,), _f32)],
polymorphic_shapes=["b + 4"]),
PolyHarness("jnp_split", "idx_tuple_poly",
# The indices are a tuple with poly expressions
lambda a: jnp.split(a, (a.shape[0] - 2,), axis=0),
arg_descriptors=[RandArg((16,), _f32)],
polymorphic_shapes=["b + 4"]),
PolyHarness("jnp_split", "idx_ct",
# A constant num_sections
lambda a: jnp.split(a, 2, axis=0),
arg_descriptors=[RandArg((16,), _f32)],
polymorphic_shapes=["2*b + 4"]),
PolyHarness("jnp_split", "idx_poly",
# A poly expression as num_sections
lambda a: jnp.split(a, a.shape[0] // 2, axis=0),
arg_descriptors=[RandArg((16,), _f32)],
polymorphic_shapes=["b + 4"],
expect_error=(ValueError,
"jax.numpy.split with a symbolic number of sections is not supported")),
PolyHarness("squeeze", "axis=empty",
jnp.squeeze,
arg_descriptors=[RandArg((5,), _f32), StaticArg(())],
polymorphic_shapes=["b, ..."]),
PolyHarness("squeeze", "axis=None",
jnp.squeeze,
arg_descriptors=[RandArg((5,), _f32), StaticArg(None)],
polymorphic_shapes=["b, ..."],
expect_error=(ValueError, "jnp.squeeze with axis=None is not supported with shape polymorphism")),
PolyHarness("squeeze", "axis=1",
jnp.squeeze,
arg_descriptors=[RandArg((4, 1), _f32), StaticArg((1,))],
polymorphic_shapes=["b, ..."]),
PolyHarness("squeeze", "axis=1_2",
jnp.squeeze,
arg_descriptors=[RandArg((4, 1, 1), _f32), StaticArg((1, 2))],
polymorphic_shapes=["b, ..."]),
PolyHarness("squeeze", "error",
jnp.squeeze,
arg_descriptors=[RandArg((3, 33), _f32), StaticArg(-1)],
polymorphic_shapes=["b0, b1"],
expect_error=(ValueError,
re.escape(
"cannot select an axis to squeeze out which has size not equal to one, got shape=(b0, b1) and dimensions=(1,)"))
),
PolyHarness("take", "",
lambda a, i: jnp.take(a, i, axis=1),
arg_descriptors=[RandArg((3, 4, 5), _f32), np.array([1, 2], np.int32)],
polymorphic_shapes=["b, ...", None]),
PolyHarness("take_along_axis", "0",
lambda x, y: jnp.take_along_axis(x, y, axis=0),
arg_descriptors=[RandArg((5, 2), _f32), RandArg((5, 1), np.int32)],
polymorphic_shapes=["b, ...", "b, ..."]),
PolyHarness("take_along_axis", "1",
lambda x, y: jnp.take_along_axis(x, y, axis=1),
arg_descriptors=[RandArg((5, 2), _f32), RandArg((5, 1), np.int32)],
polymorphic_shapes=["b, ...", "b, ..."]),
PolyHarness("tile", "0",
lambda x: jnp.tile(x, (1, 2)),
arg_descriptors=[RandArg((4, 3), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("tile", "1",
# The repetitions are polys
lambda x: jnp.tile(x, (1, x.shape[0])),
arg_descriptors=[RandArg((4, 2), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("lax_top_k", "",
lambda x: jax.lax.top_k(x, x.shape[-1] - 1),
arg_descriptors=[RandArg((16,), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("tri", "N=poly_M=None",
lambda x: jnp.tri(x.shape[0]) + x,
arg_descriptors=[RandArg((3, 1), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("tri", "N=poly_M=poly",
lambda x: jnp.tri(x.shape[0], M=x.shape[0] + 2) + x,
arg_descriptors=[RandArg((3, 1), _f32)],
polymorphic_shapes=["b, ..."]),
PolyHarness("tril", "",
lambda x: jnp.tril(jnp.ones((x.shape[0], x.shape[0] + x.shape[1]),
dtype=_f32),
k=x.shape[1]),
arg_descriptors=[RandArg((3, 4), _f32)],
polymorphic_shapes=["m, n"]),
[
PolyHarness("triangular_solve",
f"shape={jtu.format_shape_dtype_string(a_shape, dtype)}_{left_side=}_{a_poly=}_{b_poly=}",
lambda a, b, left_side: lax.linalg.triangular_solve(
jnp.tril(a) + 5 * jnp.eye(a.shape[-1], dtype=a.dtype),
b, left_side=left_side,
lower=True, transpose_a=False, conjugate_a=False,
unit_diagonal=False),
arg_descriptors=[RandArg(a_shape, dtype),
RandArg(b_shape, dtype),
StaticArg(left_side)],
polymorphic_shapes=[a_poly, b_poly],
# Some of the cases have batch dimensions for `a`, so the
# "+" needs rank promotion.
override_jax_config_flags={"jax_numpy_rank_promotion": "allow"})
for dtype in {np.float32, np.float64, np.complex64, np.complex128} & jtu.supported_dtypes()
for (left_side, a_shape, b_shape, a_poly, b_poly) in [
(True, (3, 4, 4), (3, 4, 5), "b, ...", "b, ..."),
(True, (3, 4, 4), (3, 4, 5), "b, k, k", "b, k, m"),
(False, (3, 4, 4), (3, 5, 4), "b, ...", "b, ..."),
(False, (3, 4, 4), (3, 5, 4), "b, k, k", "b, m, k"),
# We use custom calls on CPU if not batched
(True, (4, 4), (4, 5), "k, k", "k, m"),
(False, (4, 4), (5, 4), "k, k", "m, k"),
]
],
[
PolyHarness("var",
f"{axis=}_{keepdims=}_where=None",
lambda x, axis, keepdims: jnp.var(x, axis=axis, keepdims=keepdims, where=None),
arg_descriptors=[RandArg((7, 8, 4), _f32),
StaticArg(axis),
StaticArg(keepdims)],
polymorphic_shapes=["b, ..."])
for keepdims in [False, True]
for axis in [None, (0,), (0, 1), (1,)]
],
[
PolyHarness("var",
f"{axis=}_{keepdims=}_where=Some",
lambda x, where, axis, keepdims: jnp.var(x, axis=axis, keepdims=keepdims, where=where),
arg_descriptors=[RandArg((7, 8, 4), _f32),
RandArg((7, 8, 4), np.bool_),
StaticArg(axis),
StaticArg(keepdims)],
polymorphic_shapes=["b, ...", "b, ..."])
for keepdims in [False, True]
for axis in [None, (0,), (0, 1), (1,)]
],
PolyHarness("where", "",
jnp.where,
arg_descriptors=[RandArg((2,), np.bool_), RandArg((), _f32), RandArg((2,), _f32)],
polymorphic_shapes=["b, ...", None, "b, ..."]),
]
### We add to the test harnesses some that are obtained from the
### primitive harnesses by applying vmap to the function and then asserting
### that we can convert shape polymorphically the result.
def _make_vmap_primitive_harnesses() -> Sequence[PolyHarness]:
"""For each harness group, pick a single dtype.
See PolyHarness for documentation.
"""
all_h = test_harnesses.all_harnesses
res = []
# Index by group
harness_groups: dict[
str, Sequence[test_harnesses.Harness]] = collections.defaultdict(list)
device = jtu.device_under_test()
for h in all_h:
# Drop the JAX limitations
if not h.filter(device_under_test=device, include_jax_unimpl=False):
continue
harness_groups[h.group_name].append(h)
selected_harnesses = []
for _, hlist in harness_groups.items():
# Pick the dtype with the most harnesses in this group. Some harness
# groups only test different use cases at a few dtypes.
c = collections.Counter([h.dtype for h in hlist])
(_, max_count), = c.most_common(1)
# Pick the first alphabetically among those with max_count, to ensure
# that we generate deterministic tests.
dtypes_with_max_count = (dtype for dtype, count in c.items()
if count == max_count)
dtype, *_ = sorted(dtypes_with_max_count, key=str)
selected_harnesses.extend([h for h in hlist if h.dtype == dtype])
batch_size = 3
for h in selected_harnesses:
if h.group_name in [
"tridiagonal_solve", # batching not implemented in JAX
]:
continue
def make_batched_arg_descriptor(
ad: test_harnesses.ArgDescriptor) -> test_harnesses.ArgDescriptor | None:
if isinstance(ad, RandArg):
return RandArg((batch_size,) + ad.shape, ad.dtype)
elif isinstance(ad, CustomArg):
def wrap_custom(rng):
arg = ad.make(rng)
return np.stack([arg] * batch_size)
return CustomArg(wrap_custom)
else:
assert isinstance(ad, np.ndarray), ad
return np.stack([ad] * batch_size)
new_args = [make_batched_arg_descriptor(ad)
for ad in h.arg_descriptors
if not isinstance(ad, StaticArg)]
# This test does not make sense for nullary functions
if not new_args:
continue
vmap_harness = PolyHarness("vmap_" + h.group_name, h.name,
jax.vmap(h.dyn_fun, in_axes=0, out_axes=0),
arg_descriptors=new_args,
polymorphic_shapes=["b, ..."] * len(new_args))
vmap_harness.original_harness = h
res.append(vmap_harness)
return res
_POLY_SHAPE_TEST_HARNESSES.append(_make_vmap_primitive_harnesses())
def _flatten_harnesses(harnesses):
res = []
for h in harnesses:
if isinstance(h, Sequence):
res.extend(_flatten_harnesses(h))
else:
res.append(h)
return res
class ShapePolyHarnessesTest(jtu.JaxTestCase):
"""This test runs for all _POLY_SHAPE_PRIMITIVE_HARNESSES."""
def setUp(self):
_start_profile(self)
super().setUp()
def tearDown(self):
super().tearDown()
_stop_profile(self)
# For each primitive "xxx" the test will be called "test_harness_xxx_...".
# If you want to run this test for only one harness that includes "foo"
# in the name (after test_harness), add parameter `one_containing="foo"`
# to parameterized below.
@test_harnesses.parameterized(
_flatten_harnesses(_POLY_SHAPE_TEST_HARNESSES),
#one_containing="",
)
def test_harness(self, harness: PolyHarness):
# We do not expect the associative scan error on TPUs
if harness.expect_error == expect_error_associative_scan and jtu.test_device_matches(["tpu"]):
harness.expect_error = None
if harness.group_name == "schur" and not jtu.test_device_matches(["cpu"]):
raise unittest.SkipTest("schur decomposition is only implemented on CPU.")
if "fft_fft_type" in harness.fullname:
if "nr_fft_lengths_2" in harness.fullname:
raise unittest.SkipTest("native serialization with shape polymorphism not implemented for fft with non-constant fft_lengths on GPU and TPU")
if harness.group_name == "vmap_eigh":
raise unittest.SkipTest(
"Should not compare eigendecompositions for equality directly"
"because eigenvalues are sorted.")
if harness.group_name == "vmap_tan":
# Tan (b/274462307) require support for custom call stablehlo.tan.
raise unittest.SkipTest(
"native lowering with shape polymorphism requires additional StableHLO feature support")
if (jtu.test_device_matches(["cpu", "gpu"]) and
harness.fullname in [
"cumsum_reduce_axis_poly", "cumprod_reduce_axis_poly",
"cummin_reduce_axis_poly", "cummax_reduce_axis_poly",
"cumlogsumexp_reduce_axis_poly",
"jnp_insert_insert_constant", "jnp_insert_insert_poly",
"jnp_nonzero_size_constant", "jnp_nonzero_size_poly"]):
# Need associative scan reductions on CPU and GPU. On TPU we use the
# reduce_window HLO, but on CPU and GPU (with axis size >= 32) we use
# a recursive associative scan that we cannot express with shape
# polymorphism.
raise unittest.SkipTest(
"native serialization with shape polymorphism not implemented for window_reductions on CPU and GPU")
if harness.group_name == "vmap_conv_general_dilated":
# https://github.com/openxla/stablehlo/issues/1268
raise unittest.SkipTest("Need more dynamism for DynamicConvOp")
if harness.group_name == "eig" and not jtu.test_device_matches(["cpu"]):
raise unittest.SkipTest("JAX implements eig only on CPU.")
if (harness.group_name in ("eigh", "svd") and
not harness.polymorphic_shapes[0].endswith("...") and
jtu.test_device_matches(["tpu"])):
raise unittest.SkipTest(
"Shape polymorphism for Eigh and Svd is only supported for batch dimensions on TPU.")
config_flags = harness.override_jax_config_flags
# Update this here rather than in harness object because vmap_random_gamma is derived
# from test_harnesses.all_harnesses, which strips override_jax_config_flags.
if "random_gamma" in harness.group_name:
config_flags = {**config_flags, "jax_debug_key_reuse": False}
# TPU precision is a little lower since we swap the order of matmul operands.
if "cholesky" in harness.group_name and jtu.test_device_matches(["tpu"]):
harness.tol = 5e-5
with jtu.thread_local_config_context(**config_flags):
harness.run_test(self)
if __name__ == "__main__":
absltest.main(testLoader=jtu.JaxTestLoader())
Reference in New Issue View Git Blame Copy Permalink