mirror of
https://github.com/wassname/jaxtyping.git
synced 2026-08-27 12:10:13 +08:00
Patrick Kidger
12d540794f
Phew, this ended up being a pretty complicated change! The basic summary is that we now support the syntax ``` @jaxtyped(typechecker=typechecker) def f(...): ... ``` and when using this, we now get pretty error messages about what went wrong. ( The old syntax, i.e. ``` @jaxtyped @typechecker def f(...): ... ``` is still supported, but doesn't give much information. ) The internals of this do quite a lot of magic! In particular we dynamically create quite a lot of functions and test the provided arguments against their signatures. The overhead should still be minimal under `jax.jit`, though. (TODO: what's the overhead like in non-jit situations, e.g. PyTorch? I've tried to minimise the overhead throughout just to be sure, but perhaps PyTorch users should stick to the old syntax?)
345 lines
8.9 KiB
Python
345 lines
8.9 KiB
Python
# Copyright (c) 2022 Google LLC
|
|
#
|
|
# Permission is hereby granted, free of charge, to any person obtaining a copy of
|
|
# this software and associated documentation files (the "Software"), to deal in
|
|
# the Software without restriction, including without limitation the rights to
|
|
# use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
|
|
# the Software, and to permit persons to whom the Software is furnished to do so,
|
|
# subject to the following conditions:
|
|
#
|
|
# The above copyright notice and this permission notice shall be included in all
|
|
# copies or substantial portions of the Software.
|
|
#
|
|
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
|
|
# FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
|
|
# COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
|
|
# IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
|
|
# CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
|
|
|
from typing import NamedTuple, Tuple, Union
|
|
|
|
import equinox as eqx
|
|
import jax
|
|
import jax.numpy as jnp
|
|
import jax.random as jr
|
|
import pytest
|
|
|
|
import jaxtyping
|
|
from jaxtyping import Array, Float, PyTree
|
|
|
|
from .helpers import make_mlp, ParamError
|
|
|
|
|
|
def test_direct(typecheck):
|
|
@typecheck
|
|
def g(x: PyTree):
|
|
pass
|
|
|
|
g(1)
|
|
g({"a": jnp.array(1), "b": [object()]})
|
|
g(object())
|
|
|
|
@typecheck
|
|
def h() -> PyTree:
|
|
return object()
|
|
|
|
h()
|
|
|
|
|
|
def test_subscript(getkey, typecheck):
|
|
@typecheck
|
|
def g(x: PyTree[int]):
|
|
pass
|
|
|
|
g(1)
|
|
g([1, 2, {"a": 3}])
|
|
g(jax.tree_map(lambda _: 1, make_mlp(getkey())))
|
|
|
|
with pytest.raises(ParamError):
|
|
g(object())
|
|
with pytest.raises(ParamError):
|
|
g("hi")
|
|
with pytest.raises(ParamError):
|
|
g([1, 2, {"a": 3}, "bye"])
|
|
|
|
|
|
def test_leaf_pytrees(getkey, typecheck):
|
|
@typecheck
|
|
def g(x: PyTree[eqx.nn.MLP]):
|
|
pass
|
|
|
|
g(make_mlp(getkey()))
|
|
g([make_mlp(getkey()), make_mlp(getkey()), {"a": make_mlp(getkey())}])
|
|
|
|
with pytest.raises(ParamError):
|
|
g([1, 2])
|
|
with pytest.raises(ParamError):
|
|
g([1, 2, make_mlp()])
|
|
|
|
|
|
def test_nested_pytrees(getkey, typecheck):
|
|
# PyTree[...] is logically equivalent to PyTree[PyTree[...]]
|
|
@typecheck
|
|
def g(x: PyTree[PyTree[eqx.nn.MLP]]):
|
|
pass
|
|
|
|
g(make_mlp(getkey()))
|
|
g([make_mlp(getkey()), make_mlp(getkey()), {"a": make_mlp(getkey())}])
|
|
|
|
with pytest.raises(ParamError):
|
|
g([1, 2])
|
|
with pytest.raises(ParamError):
|
|
g([1, 2, make_mlp()])
|
|
|
|
|
|
def test_pytree_array(jaxtyp, typecheck):
|
|
@jaxtyp(typecheck)
|
|
def g(x: PyTree[Float[jnp.ndarray, "..."]]):
|
|
pass
|
|
|
|
g(jnp.array(1.0))
|
|
g([jnp.array(1.0), jnp.array(1.0), {"a": jnp.array(1.0)}])
|
|
with pytest.raises(ParamError):
|
|
g(jnp.array(1))
|
|
with pytest.raises(ParamError):
|
|
g(1.0)
|
|
|
|
|
|
def test_pytree_shaped_array(jaxtyp, typecheck, getkey):
|
|
@jaxtyp(typecheck)
|
|
def g(x: PyTree[Float[jnp.ndarray, "b c"]]):
|
|
pass
|
|
|
|
g(jnp.array([[1.0]]))
|
|
g([jr.normal(getkey(), (1, 1)), jr.normal(getkey(), (1, 1))])
|
|
g([jr.normal(getkey(), (3, 4)), jr.normal(getkey(), (3, 4))])
|
|
with pytest.raises(ParamError):
|
|
g(jnp.array(1.0))
|
|
with pytest.raises(ParamError):
|
|
g(jnp.array([[1]]))
|
|
with pytest.raises(ParamError):
|
|
g([jr.normal(getkey(), (3, 4)), jr.normal(getkey(), (1, 1))])
|
|
with pytest.raises(ParamError):
|
|
g([jr.normal(getkey(), (1, 1)), jr.normal(getkey(), (3, 4))])
|
|
|
|
|
|
def test_pytree_union(typecheck):
|
|
@typecheck
|
|
def g(x: PyTree[Union[int, str]]):
|
|
pass
|
|
|
|
g([1])
|
|
g(["hi"])
|
|
g([1, "hi"])
|
|
with pytest.raises(ParamError):
|
|
g(object())
|
|
|
|
|
|
def test_pytree_tuple(typecheck):
|
|
@typecheck
|
|
def g(x: PyTree[Tuple[int, int]]):
|
|
pass
|
|
|
|
g((1, 1))
|
|
g([(1, 1)])
|
|
g([(1, 1), {"a": (1, 1)}])
|
|
with pytest.raises(ParamError):
|
|
g(object())
|
|
with pytest.raises(ParamError):
|
|
g(1)
|
|
with pytest.raises(ParamError):
|
|
g((1, 2, 3))
|
|
with pytest.raises(ParamError):
|
|
g([1, 1])
|
|
with pytest.raises(ParamError):
|
|
g([(1, 1), "hi"])
|
|
|
|
|
|
def test_pytree_namedtuple(typecheck):
|
|
class CustomNamedTuple(NamedTuple):
|
|
x: Float[jnp.ndarray, "a b"]
|
|
y: Float[jnp.ndarray, "b c"]
|
|
|
|
class OtherCustomNamedTuple(NamedTuple):
|
|
x: Float[jnp.ndarray, "a b"]
|
|
y: Float[jnp.ndarray, "b c"]
|
|
|
|
@typecheck
|
|
def g(x: PyTree[CustomNamedTuple]):
|
|
...
|
|
|
|
g(
|
|
CustomNamedTuple(
|
|
x=jax.random.normal(jax.random.PRNGKey(42), (3, 2)),
|
|
y=jax.random.normal(jax.random.PRNGKey(420), (2, 5)),
|
|
)
|
|
)
|
|
with pytest.raises(ParamError):
|
|
g(object())
|
|
with pytest.raises(ParamError):
|
|
g(
|
|
OtherCustomNamedTuple(
|
|
x=jax.random.normal(jax.random.PRNGKey(42), (3, 2)),
|
|
y=jax.random.normal(jax.random.PRNGKey(420), (2, 5)),
|
|
)
|
|
)
|
|
|
|
|
|
def test_subclass_pytree():
|
|
x = PyTree
|
|
y = PyTree[int]
|
|
assert issubclass(x, PyTree)
|
|
assert issubclass(y, PyTree)
|
|
assert not issubclass(int, PyTree)
|
|
|
|
|
|
def test_structure_match(jaxtyp, typecheck):
|
|
@jaxtyp(typecheck)
|
|
def f(x: PyTree[int, " T"], y: PyTree[str, " T"]):
|
|
pass
|
|
|
|
f(1, "hi")
|
|
f((3, 4, {"a": 5}), ("a", "b", {"a": "c"}))
|
|
|
|
with pytest.raises(ParamError):
|
|
f(1, ("hi",))
|
|
|
|
|
|
def test_structure_prefix(jaxtyp, typecheck):
|
|
@jaxtyp(typecheck)
|
|
def f(x: PyTree[int, " T"], y: PyTree[str, "T ..."]):
|
|
pass
|
|
|
|
f(1, "hi")
|
|
f((3, 4, {"a": 5}), ("a", "b", {"a": "c"}))
|
|
f(1, ("hi",))
|
|
f((1, 2), ({"a": "hi"}, {"a": "bye"}))
|
|
f((1, 2), ({"a": "hi"}, {"not-a": "bye"}))
|
|
|
|
with pytest.raises(ParamError):
|
|
f((1, 2), ({"a": "hi"}, {"a": "bye"}, {"a": "oh-no"}))
|
|
|
|
with pytest.raises(ParamError):
|
|
f((3, 4, 5), {"a": ("hi", "bye")})
|
|
|
|
|
|
def test_structure_suffix(jaxtyp, typecheck):
|
|
@jaxtyp(typecheck)
|
|
def f(x: PyTree[int, " T"], y: PyTree[str, "... T"]):
|
|
pass
|
|
|
|
f(1, "hi")
|
|
f((3, 4, {"a": 5}), ("a", "b", {"a": "c"}))
|
|
f(1, ("hi",))
|
|
|
|
with pytest.raises(ParamError):
|
|
f((3, 4), {"a": (1, 2)})
|
|
|
|
with pytest.raises(ParamError):
|
|
f((3, 4, 5), {"a": ("hi", "bye")})
|
|
|
|
|
|
def test_structure_compose(jaxtyp, typecheck):
|
|
@jaxtyp(typecheck)
|
|
def f(x: PyTree[int, " T"], y: PyTree[int, " S"], z: PyTree[str, "S T"]):
|
|
pass
|
|
|
|
f(1, 2, "hi")
|
|
f((1, 2), 2, ("a", "b"))
|
|
|
|
with pytest.raises(ParamError):
|
|
f((1, 2), 2, (1, 2))
|
|
|
|
f((1, 2), {"a": 3}, {"a": ("hi", "bye")})
|
|
|
|
with pytest.raises(ParamError):
|
|
f((1, 2), {"a": 3}, ({"a": "hi"}, {"a": "bye"}))
|
|
|
|
@jaxtyp(typecheck)
|
|
def g(x: PyTree[int, " T"], y: PyTree[int, " S"], z: PyTree[str, "T S"]):
|
|
pass
|
|
|
|
with pytest.raises(ParamError):
|
|
g((1, 2), {"a": 3}, {"a": ("hi", "bye")})
|
|
|
|
g((1, 2), {"a": 3}, ({"a": "hi"}, {"a": "bye"}))
|
|
|
|
|
|
@pytest.mark.parametrize("variadic", (False, True))
|
|
def test_treepath_dependence_function(variadic, jaxtyp, typecheck, getkey):
|
|
if variadic:
|
|
jtshape = "*?foo"
|
|
shape = (2, 3)
|
|
else:
|
|
jtshape = "?foo"
|
|
shape = (4,)
|
|
|
|
@jaxtyp(typecheck)
|
|
def f(
|
|
x: PyTree[Float[Array, jtshape], " T"], y: PyTree[Float[Array, jtshape], " T"]
|
|
):
|
|
pass
|
|
|
|
x1 = jr.normal(getkey(), shape)
|
|
y1 = jr.normal(getkey(), shape)
|
|
x2 = jr.normal(getkey(), (5,))
|
|
y2 = jr.normal(getkey(), (5,))
|
|
f(x1, y1)
|
|
f((x1, x2), (y1, y2))
|
|
|
|
with pytest.raises(ParamError):
|
|
f(x1, y2)
|
|
|
|
with pytest.raises(ParamError):
|
|
f((x1, x2), (y2, y1))
|
|
|
|
|
|
@pytest.mark.parametrize("variadic", (False, True))
|
|
def test_treepath_dependence_dataclass(variadic, typecheck, getkey):
|
|
if variadic:
|
|
jtshape = "*?foo"
|
|
shape = (2, 3)
|
|
else:
|
|
jtshape = "?foo"
|
|
shape = (4,)
|
|
|
|
@jaxtyping.jaxtyped(typechecker=typecheck)
|
|
class A(eqx.Module):
|
|
x: PyTree[Float[Array, jtshape], " T"]
|
|
y: PyTree[Float[Array, jtshape], " T"]
|
|
|
|
x1 = jr.normal(getkey(), shape)
|
|
y1 = jr.normal(getkey(), shape)
|
|
x2 = jr.normal(getkey(), (5,))
|
|
y2 = jr.normal(getkey(), (5,))
|
|
A(x1, y1)
|
|
A((x1, x2), (y1, y2))
|
|
|
|
with pytest.raises(ParamError):
|
|
A(x1, y2)
|
|
|
|
with pytest.raises(ParamError):
|
|
A((x1, x2), (y2, y1))
|
|
|
|
|
|
def test_treepath_dependence_missing_structure_annotation(jaxtyp, typecheck, getkey):
|
|
@jaxtyp(typecheck)
|
|
def f(x: PyTree[Float[Array, "?foo"], " T"], y: PyTree[Float[Array, "?foo"]]):
|
|
pass
|
|
|
|
x1 = jr.normal(getkey(), (2,))
|
|
y1 = jr.normal(getkey(), (2,))
|
|
with pytest.raises(ValueError, match="except when contained with structured"):
|
|
f(x1, y1)
|
|
|
|
|
|
def test_treepath_dependence_multiple_structure_annotation(jaxtyp, typecheck, getkey):
|
|
@jaxtyp(typecheck)
|
|
def f(x: PyTree[PyTree[Float[Array, "?foo"], " S"], " T"]):
|
|
pass
|
|
|
|
x1 = jr.normal(getkey(), (2,))
|
|
with pytest.raises(ValueError, match="ambiguous which PyTree"):
|
|
f(x1)
|