# Copyright 2024 The Flax 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
#
# http://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 dataclasses
import inspect
import os
import threading
import typing as tp
from abc import ABCMeta
from copy import deepcopy
from flax.nnx import variablelib
import jax
import numpy as np
import treescope # type: ignore[import-untyped]
from treescope import rendering_parts
from flax import errors, nnx
from flax.nnx import (
graph,
reprlib,
tracers,
visualization,
)
from flax import config
from flax.nnx.variablelib import Variable
from flax.typing import SizeBytes
BUILDING_DOCS = 'FLAX_DOC_BUILD' in os.environ
A = tp.TypeVar('A')
P = tp.TypeVar('P', bound='Pytree')
DataAnnotation = '__data__'
Data = tp.Annotated[A, DataAnnotation]
Data.__doc__ = """Data marks attributes of a class as pytree data using type annotations.
Data annotations must be used at the class level and will apply to all instances.
The usage of Data is recommended when type annotations are used already present
or required e.g. for dataclasses.
Example::
from flax import nnx
import jax
import dataclasses
@dataclasses.dataclass
class Foo(nnx.Pytree):
a: nnx.Data[int] # Annotates `a` as pytree data
b: str # `b` is not pytree data
foo = Foo(a=42, b='hello')
assert jax.tree.leaves(foo) == [42]
"""
DATA_REGISTRY: set[type] = set()
@dataclasses.dataclass(frozen=True, slots=True)
class DataAttr:
value: tp.Any
[docs]def data(value: A, /) -> A:
"""Annotates a an attribute as pytree data.
The return value from `data` must be directly assigned to an Object attribute
which will be registered as a pytree data attribute.
Example::
from flax import nnx
import jax
class Foo(nnx.Pytree):
def __init__(self):
self.data_attr = nnx.data(42) # pytree data
self.static_attr = "hello" # static attribute
foo = Foo()
assert jax.tree.leaves(foo) == [42]
Args:
value: The value to annotate as data.
Returns:
A value which will register the attribute as data on assignment.
"""
return DataAttr(value) # type: ignore[return-value]
[docs]def register_data_type(type_: type, /) -> None:
"""Registers a type as pytree data type recognized by Object.
Custom types registered as data will be automatically recognized
as data attributes when assigned to an Object attribute. This means
that values of this type do not need to be wrapped in `nnx.data(...)`
for Object to mark the attribute its being assigned to as data.
Example::
from flax import nnx
from dataclasses import dataclass
@dataclass(frozen=True)
class MyType:
value: int
nnx.register_data_type(MyType)
class Foo(nnx.Pytree):
def __init__(self, a):
self.a = MyType(a) # Automatically registered as data
self.b = "hello" # str not registered as data
foo = Foo(42)
assert nnx.is_data(foo.a) # True
assert jax.tree.leaves(foo) == [MyType(value=42)]
"""
DATA_REGISTRY.add(type_)
[docs]def is_data(value: tp.Any, /) -> bool:
"""Checks if a value is a registered data type.
This function checks a the value is registered data type, which means it is
automatically recognized as data when assigned a nnx.Pytree attribute.
Data types are:
- jax.Arrays
- np.ndarrays
- ArrayRefs
- Variables (Param, BatchStat, RngState, etc.)
- All graph nodes (Object, Module, Rngs, etc.)
- Any type registered with `nnx.register_data_type`
- Any pytree that contains at least one node or leaf element of the above
Example::
>>> from flax import nnx
>>> import jax.numpy as jnp
... # ------ DATA ------------
>>> assert nnx.is_data( jnp.array(0) ) # Arrays
>>> assert nnx.is_data( nnx.Param(1) ) # Variables
>>> assert nnx.is_data( nnx.Rngs(2) ) # nnx.Pytrees
>>> assert nnx.is_data( nnx.Linear(1, 1,rngs=nnx.Rngs(0)) ) # Modules
... # ------ STATIC ------------
>>> assert not nnx.is_data( 'hello' ) # strings, arbitrary objects
>>> assert not nnx.is_data( 42 ) # int, float, bool, complex, etc.
>>> assert not nnx.is_data( [1, 2.0, 3j, jnp.array(1)] ) # list, dict, tuple, pytrees
Args:
value: The value to check.
Returns:
A string representing the attribute status.
"""
return (
graph.is_node_leaf(value)
or graph.is_graph_node(value)
or type(value) in DATA_REGISTRY
)
StaticAnnotation = '__static__'
Static = tp.Annotated[A, StaticAnnotation]
Static.__doc__ = """Static marks attributes of a class as static using type annotations.
Static annotations must be used at the class level and will apply to all instances.
The usage of Static is recommended when type annotations are used already present
or required e.g. for dataclasses.
"""
@dataclasses.dataclass(frozen=True, slots=True)
class StaticAttr:
value: tp.Any
[docs]def static(value: A, /) -> A:
"""Annotates a an attribute as static.
The return value from `static` must be directly assigned to an Object
attribute
which will be registered as static attribute.
Example::
from flax import nnx
class Foo(nnx.Pytree):
def __init__(self, a, b):
self.a = nnx.static(a) # pytree metadata
self.b = nnx.data(b) # pytree data
foo = Foo("one", "two")
assert jax.tree.leaves(foo) == ["two"]
By default ``nnx.Pytree`` will ...
"""
return StaticAttr(value) # type: ignore[return-value]
def _collect_stats(
node: tp.Any, node_stats: dict[int, dict[type[Variable], SizeBytes]]
):
if not graph.is_node(node) and not isinstance(node, Variable):
raise ValueError(f'Expected a graph node or Variable, got {type(node)!r}.')
if id(node) in node_stats:
return
stats: dict[type[Variable], SizeBytes] = {}
node_stats[id(node)] = stats
if isinstance(node, Variable):
var_type = type(node)
if issubclass(var_type, nnx.RngState):
var_type = nnx.RngState
size_bytes = SizeBytes.from_any(node.raw_value)
if size_bytes:
stats[var_type] = size_bytes
else:
node_impl = graph.get_node_impl(node)
assert node_impl is not None
node_dict = node_impl.node_dict(node)
for key, value in node_dict.items():
if id(value) in node_stats:
continue
if graph.is_node(value) or isinstance(value, Variable):
_collect_stats(value, node_stats)
child_stats = node_stats[id(value)]
for var_type, size_bytes in child_stats.items():
if var_type in stats:
stats[var_type] += size_bytes
else:
stats[var_type] = size_bytes
@dataclasses.dataclass
class ObjectContext(threading.local):
seen_modules_repr: set[int] | None = None
node_stats: dict[int, dict[type[Variable], SizeBytes]] | None = None
OBJECT_CONTEXT = ObjectContext()
class PytreeState(reprlib.Representable):
__slots__ = ('_trace_state', '_initializing', '_is_setup')
def __init__(self, initializing: bool = False, is_setup: bool = False):
self._trace_state = tracers.TraceState()
self._initializing = initializing
self._is_setup = is_setup
@property
def trace_state(self) -> tracers.TraceState:
return self._trace_state
@property
def initializing(self) -> bool:
return self._initializing
@property
def is_setup(self) -> bool:
return self._is_setup
def __nnx_repr__(self):
yield reprlib.Object(type(self))
yield reprlib.Attr('trace_state', self._trace_state)
def __treescope_repr__(self, path, subtree_renderer):
return visualization.render_object_constructor(
object_type=type(self),
attributes={'trace_state': self._trace_state},
path=path,
subtree_renderer=subtree_renderer,
)
def _flatten_pytree_state(state: PytreeState):
return (), (state.initializing, state.is_setup)
def _unflatten_pytree_state(static: tuple[bool, bool], _):
initializing, setup = static
return PytreeState(initializing, setup)
jax.tree_util.register_pytree_node(
PytreeState,
_flatten_pytree_state,
_unflatten_pytree_state,
)
[docs]def check_pytree(pytree):
"""Checks if a pytree is valid."""
if not isinstance(pytree, Pytree):
raise TypeError(f'Expected a Pytree, got {type(pytree)}.')
for name, value in vars(pytree).items():
pytree._check_value(name, value, new_status=None)
class PytreeMeta(ABCMeta):
if not tp.TYPE_CHECKING:
def __call__(cls, *args: Any, **kwargs: Any) -> Any:
return _graph_node_meta_call(cls, *args, **kwargs)
def _pytree_meta_construct(cls, self, *args, **kwargs):
self.__init__(*args, **kwargs)
ObjectMeta = PytreeMeta
def _graph_node_meta_call(cls: tp.Type[P], *args, **kwargs) -> P:
node = cls.__new__(cls, *args, **kwargs)
vars_obj = vars(node)
vars_obj['_pytree__state'] = PytreeState()
vars_obj['_pytree__nodes'] = cls._pytree__nodes
cls._pytree_meta_construct(node, *args, **kwargs)
if cls._pytree__is_pytree:
missing: dict[str, bool] = {}
for name, value in vars(node).items():
if name not in vars_obj['_pytree__nodes']:
missing[name] = is_data(value)
if missing:
vars_obj['_pytree__nodes'] = vars_obj['_pytree__nodes'].update(missing)
check_pytree(node)
return node
@dataclasses.dataclass(frozen=True, repr=False)
class ArrayRepr(reprlib.Representable):
shape: tp.Tuple[int, ...]
dtype: tp.Any
@staticmethod
def from_array(array: jax.Array | np.ndarray) -> ArrayRepr:
return ArrayRepr(array.shape, array.dtype)
def __nnx_repr__(self):
yield reprlib.Object(type='Array', same_line=True)
yield reprlib.Attr('shape', self.shape)
yield reprlib.Attr('dtype', self.dtype)
@dataclasses.dataclass(frozen=True, repr=False)
class MutableArrayRepr(reprlib.Representable):
shape: tp.Tuple[int, ...]
dtype: tp.Any
@staticmethod
def from_array(array: jax.Array | np.ndarray) -> MutableArrayRepr:
return MutableArrayRepr(array.shape, array.dtype)
def __nnx_repr__(self):
yield reprlib.Object(type='ArrayRef', same_line=True)
yield reprlib.Attr('shape', self.shape)
yield reprlib.Attr('dtype', self.dtype)
class AttributeStatus(tp.NamedTuple):
is_data: bool
explicit: bool
[docs]class Pytree(reprlib.Representable, metaclass=PytreeMeta):
"""Base class for all NNX objects."""
if tp.TYPE_CHECKING:
_pytree__nodes: graph.HashableMapping[tp.Any, bool]
_pytree__state: PytreeState
_pytree__is_pytree: bool
def __init_subclass__(
cls,
*,
pytree: bool = config.flax_pytree_module,
**kwargs,
) -> None:
super().__init_subclass__(**kwargs)
cls._pytree__is_pytree = pytree
graph.register_graph_node_type(
type=cls,
flatten=cls._graph_node_flatten,
set_key=cls._graph_node_set_key, # type: ignore
pop_key=cls._graph_node_pop_key, # type: ignore
create_empty=cls._graph_node_create_empty,
clear=cls._graph_node_clear,
init=cls._graph_node_init, # type: ignore
)
nodes: dict[str, bool] = dict(getattr(cls, '_pytree__nodes', ()))
nodes['_pytree__state'] = True
try:
type_hints = tp.get_type_hints(
cls, globals(), {cls.__name__: cls}, include_extras=True
)
except NameError:
type_hints = cls.__annotations__
# add annotation attributes
for name, type_ in type_hints.items():
if isinstance(type_, str):
if type_.startswith('nnx.Data'):
nodes[name] = True
elif type_.startswith('nnx.Static'):
nodes[name] = False
else:
type_metadata = getattr(type_, '__metadata__', ())
if DataAnnotation in type_metadata:
nodes[name] = True
elif StaticAnnotation in type_metadata:
nodes[name] = False
cls._pytree__nodes = graph.HashableMapping(nodes, copy=False)
if pytree:
jax.tree_util.register_pytree_with_keys(
cls,
flatten_with_keys=cls._pytree__flatten_with_paths,
unflatten_func=cls._pytree__unflatten,
flatten_func=cls._pytree__flatten,
)
if BUILDING_DOCS:
# set correct signature for sphinx
cls.__signature__ = inspect.signature(cls.__init__)
# Backward compatibility with PR #4863
@property
def _object__nodes(self):
return self._pytree__nodes
@property
def _object__state(self):
return self._pytree__state
if not tp.TYPE_CHECKING:
def __setattr__(self, name: str, value: Any) -> None:
self._setattr(name, value)
def _setattr(self, name, value: tp.Any) -> None:
self._check_valid_context(
lambda: f"Cannot mutate '{type(self).__name__}' from different trace level"
)
data: bool = False
explicit: bool = False
if type(value) is DataAttr:
value = value.value
if self._pytree__is_pytree:
data = True
explicit = True
elif type(value) is StaticAttr:
value = value.value
if self._pytree__is_pytree:
data = False
explicit = True
elif self._pytree__is_pytree:
data = is_data(value)
if self._pytree__is_pytree:
self._check_value(name, value, AttributeStatus(data, explicit))
if name not in self._pytree__nodes or (
explicit and self._pytree__nodes[name] != data
):
vars(self)['_pytree__nodes'] = self._pytree__nodes.update({name: data})
if isinstance(name, str):
object.__setattr__(self, name, value)
else:
vars(self)[name] = value
def _check_value(self, key, value, new_status: AttributeStatus | None):
def _has_arrays(leaves):
return any(
isinstance(leaf, (np.ndarray, jax.Array))
or variablelib.is_array_ref(leaf)
for leaf in leaves
)
def _get_annotations(leaves):
return {
'data' if type(leaf) is DataAttr else 'static'
for leaf in leaves
if type(leaf) is DataAttr or type(leaf) is StaticAttr
}
def _has_visited(x):
if id(x) in visited:
return True
visited.add(id(x))
return False
visited: set[int] = set()
leaves = jax.tree.leaves(value, is_leaf=_has_visited)
current_is_data = (
self._pytree__nodes[key] if key in self._pytree__nodes else False
)
existing_attr = key in vars(self)
if (
new_status is not None
and not new_status.explicit
and new_status.is_data
and existing_attr
and not current_is_data
):
raise ValueError(
f"Cannot assign data value of type '{type(value)}' to static"
f" attribute '{key}' of Pytree type '{type(self)}'. To override the"
' status explicitly wrap the value with nnx.data on assignment:\n\n '
f' _.{key} = nnx.data(...)\n\n'
)
if _has_arrays(leaves):
# check no data in nnx.static assignments
if new_status is not None:
if not new_status.is_data and new_status.explicit:
raise ValueError(
f"Found Arrays in value of type '{type(value)}' annotated with "
f"nnx.static(...) when setting attribute '{key}' of Pytree type "
f"'{type(self)}'."
)
if not new_status.explicit and not current_is_data and existing_attr:
base_pytree_type = Pytree
for t in type(self).mro()[1:]:
if issubclass(t, nnx.Pytree):
base_pytree_type = t
break
raise ValueError(
f"Found Arrays on value of type '{type(value)}' assigned to"
f" static attribute '{key}' of Pytree type '{type(self)}'. Static"
' attributes should not contain Array values. Consider one of'
' the following options:\n\n1. If the attribute is meant to be'
' static, either remove the Array value or wrap it in a static'
' container (e.g., using `nnx.static(...)`).\n2. If the'
' attribute is meant to be data, wrap the value with nnx.data on'
f' assignment:\n\n _.{key} = nnx.data(...)\n\n3. Alternatively,'
' annotate the class attribute with nnx.Data:\n\n class'
f' {type(self).__name__}({base_pytree_type.__name__}):\n '
f' {key}: nnx.Data[{type(value).__name__}]\n\n4. Disable pytree'
' for this class:\n\n class'
f' {type(self).__name__}({base_pytree_type.__name__},'
' pytree=False):\n\n'
)
# check no data in static attributes after __init__
elif not current_is_data:
base_pytree_type = Pytree
for t in type(self).mro()[1:]:
if issubclass(t, nnx.Pytree):
base_pytree_type = t
break
raise ValueError(
f'Found unexpected Arrays on value of type {type(value)} in static'
f" attribute '{key}' of Pytree type '{type(self)}'. This is an"
' error starting from Flax version 0.12.0.\nConsider one of the'
' following options:\n\n1. If the attribute is meant to be static,'
' either remove the Array value or wrap it in a static'
' container.\n2. Wrap the value with nnx.data on'
f' assignment:\n\n _.{key} = nnx.data(...)\n\n3. Annotate the'
' class attribute with nnx.Data:\n\n class'
f' {type(self).__name__}({base_pytree_type.__name__}):\n {key}:'
f' nnx.Data[{type(value).__name__}]\n\n4. If the container is a'
' list or dict, try using nnx.List(...) or nnx.Dict(...)'
' instead.\n5. Disable pytree for this class:\n\n class'
f' {type(self).__name__}({base_pytree_type.__name__},'
f' pytree=False):\n\n'
)
if tags := _get_annotations(leaves):
raise ValueError(
f'Found unexpected tags {tags} on attribute'
f" '{type(self).__name__}.{key}'. Values from nnx.data(...)"
' and\nnnx.static(...) should be assigned to nnx.Pytree attributes'
' directly, they should not be inside other structures. Got value of'
f" type '{type(value)}' on Pytree of type '{type(self)}'."
)
def _check_valid_context(self, error_msg: tp.Callable[[], str]) -> None:
if not self._pytree__state.trace_state.is_valid():
raise errors.TraceContextError(error_msg())
def __deepcopy__(self: P, memo=None) -> P:
graphdef, state = graph.split(self)
graphdef = deepcopy(graphdef)
state = deepcopy(state)
return graph.merge(graphdef, state)
def __nnx_repr__(self):
if OBJECT_CONTEXT.node_stats is None or id(self) not in OBJECT_CONTEXT.node_stats:
node_stats: dict[int, dict[type[Variable], SizeBytes]] = {}
_collect_stats(self, node_stats)
OBJECT_CONTEXT.node_stats = node_stats
stats = node_stats[id(self)]
clear_node_stats = True
else:
stats = OBJECT_CONTEXT.node_stats[id(self)]
clear_node_stats = False
if OBJECT_CONTEXT.seen_modules_repr is None:
OBJECT_CONTEXT.seen_modules_repr = set()
clear_seen = True
else:
clear_seen = False
if id(self) in OBJECT_CONTEXT.seen_modules_repr:
yield reprlib.Object(type=type(self), empty_repr='...')
return
try:
if stats:
stats_repr = ' # ' + ', '.join(
f'{var_type.__name__}: {size_bytes}'
for var_type, size_bytes in stats.items()
)
if len(stats) > 1:
total_bytes = sum(stats.values(), SizeBytes(0, 0))
stats_repr += f', Total: {total_bytes}'
else:
stats_repr = ''
yield reprlib.Object(type=type(self), comment=stats_repr)
OBJECT_CONTEXT.seen_modules_repr.add(id(self))
for name, value in vars(self).items():
if name.startswith('_'):
continue
def to_shape_dtype(value):
if isinstance(value, Variable):
return value.replace(
raw_value=jax.tree.map(to_shape_dtype, value.raw_value)
)
elif variablelib.is_array_ref(value) and np.prod(value.shape) > 1:
return MutableArrayRepr(value.shape, value.dtype)
elif (
isinstance(value, (np.ndarray, jax.Array))
and np.prod(value.shape) > 1
):
return ArrayRepr(value.shape, value.dtype)
return value
value = jax.tree.map(to_shape_dtype, value, is_leaf=graph.is_graph_node)
yield reprlib.Attr(name, value)
finally:
if clear_seen:
OBJECT_CONTEXT.seen_modules_repr = None
if clear_node_stats:
OBJECT_CONTEXT.node_stats = None
def __treescope_repr__(self, path, subtree_renderer):
from flax import nnx
if OBJECT_CONTEXT.node_stats is None:
node_stats: dict[int, dict[type[Variable], SizeBytes]] = {}
_collect_stats(self, node_stats)
OBJECT_CONTEXT.node_stats = node_stats
stats = node_stats[id(self)]
clear_node_stats = True
else:
stats = OBJECT_CONTEXT.node_stats[id(self)]
clear_node_stats = False
try:
if stats:
stats_repr = ' # ' + ', '.join(
f'{var_type.__name__}: {size_bytes}'
for var_type, size_bytes in stats.items()
)
if len(stats) > 1:
total_bytes = sum(stats.values(), SizeBytes(0, 0))
stats_repr += f', Total: {total_bytes}'
first_line_annotation = rendering_parts.comment_color(
rendering_parts.text(f'{stats_repr}')
)
else:
first_line_annotation = None
children = {}
for name, value in vars(self).items():
if name.startswith('_'):
continue
children[name] = value
if isinstance(self, nnx.Module):
color = treescope.formatting_util.color_from_string(
type(self).__qualname__
)
else:
color = None
return visualization.render_object_constructor(
object_type=type(self),
attributes=children,
path=path,
subtree_renderer=subtree_renderer,
first_line_annotation=first_line_annotation,
color=color,
)
finally:
if clear_node_stats:
OBJECT_CONTEXT.node_stats = None
# pickle support
def __getstate__(self):
return vars(self).copy()
def __setstate__(self, state):
vars(self).update(state)
# -------------------------
# Pytree Definition
# -------------------------
_pytree__key_sort_fn: tp.Callable | None = None
def _pytree__flatten_with_paths(self):
obj_vars = vars(self)
node_attributes = self._pytree__nodes
node_names: list[str] = []
node_attrs: list[tuple[tp.Any, tp.Any]] = []
static_attrs: list[tuple[str, tp.Any]] = []
for name, value in sorted(obj_vars.items(), key=self._pytree__key_sort_fn):
if name in node_attributes and node_attributes[name]:
node_names.append(name)
node_attrs.append((
jax.tree_util.GetAttrKey(name)
if isinstance(name, str)
else jax.tree_util.SequenceKey(name),
value,
))
else:
static_attrs.append((name, value))
return node_attrs, (tuple(node_names), tuple(static_attrs))
def _pytree__flatten(self):
obj_vars = vars(self)
node_attributes = self._pytree__nodes
node_names: list[str] = []
node_attrs: list[tp.Any] = []
static_attrs: list[tuple[str, tp.Any]] = []
for name, value in sorted(obj_vars.items(), key=self._pytree__key_sort_fn):
if name in node_attributes and node_attributes[name]:
node_names.append(name)
node_attrs.append(value)
else:
static_attrs.append((name, value))
return node_attrs, (tuple(node_names), tuple(static_attrs))
@classmethod
def _pytree__unflatten(
cls,
static: tuple[tuple[str, ...], tuple[tuple[str, tp.Any], ...]],
node_attrs: tp.Iterable[tp.Any],
):
node_names, static_attrs = static
obj = object.__new__(cls)
vars_obj = vars(obj)
vars_obj.update(zip(node_names, node_attrs, strict=True))
vars_obj.update(static_attrs)
return obj
# -------------------------
# Graph Definition
# -------------------------
def _graph_node_flatten(self):
nodes = vars(self)
nodes = sorted(nodes.items(), key=self._pytree__key_sort_fn)
return nodes, type(self)
def _graph_node_set_key(self, key: str, value: tp.Any):
if not isinstance(key, str):
raise KeyError(f'Invalid key: {key!r}')
elif (
hasattr(self, key)
and isinstance(variable := getattr(self, key), Variable)
and isinstance(value, Variable)
):
variable.update_from_state(value)
else:
setattr(self, key, value)
def _graph_node_pop_key(self, key: str):
if not isinstance(key, str):
raise KeyError(f'Invalid key: {key!r}')
return vars(self).pop(key)
@staticmethod
def _graph_node_create_empty(node_type: tp.Type[P]) -> P:
node = object.__new__(node_type)
return node
def _graph_node_clear(self):
vars(self).clear()
def _graph_node_init(self, attributes: tp.Iterable[tuple[str, tp.Any]]):
vars(self).update(attributes)
[docs]class Object(Pytree, pytree=False):
"""Base class for NNX objects that are not pytrees."""
def __init_subclass__(cls, **kwargs):
pytree = kwargs.pop('pytree', False)
if pytree is not False:
raise ValueError(
"Object is not a pytree, but 'pytree' was explicitly set to "
f'{pytree!r} for type {cls}.'
)
super().__init_subclass__(pytree=pytree, **kwargs)
