mirror of
https://github.com/ROCm/jax.git
synced 2025-04-16 11:56:07 +00:00
c41d271175
This PR adds basic support to remat to allow transferring intermediates (activations) to destination memory in the forward pass. Currently JAX only support host memory kind but the API allows to transfer to other memories too. Remat will automatically load the residuals back to the source memory in the backward pass. Introduce two singletons called `Recompute`, `Saveable` and a NamedTuple (`Offloadable`) that each policy can return. Currently policies return a bool which if True means saveable else recompute on backward pass. This is a backwards compatible change i.e. policies can still return a bool. A very basic offloadable policy can look like this: ``` def policy(prim, *avals, **params): return ad_checkpoint.Offloadable(src='tpu_hbm', dst='unpinned_host') ``` Co-authored-by: Matthew Johnson <mattjj@google.com> PiperOrigin-RevId: 564914301
2634 lines
116 KiB
Python
2634 lines
116 KiB
Python
# Copyright 2018 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
|
|
|
|
from collections import namedtuple
|
|
from collections.abc import Sequence, Hashable
|
|
from contextlib import contextmanager, AbstractContextManager
|
|
from functools import partial
|
|
import inspect
|
|
import itertools as it
|
|
import operator as op
|
|
from typing import (Any, Callable, NamedTuple, Optional,
|
|
Union)
|
|
from weakref import ref
|
|
|
|
import numpy as np
|
|
|
|
from jax._src import linear_util as lu
|
|
from jax._src.config import config
|
|
from jax._src import ad_util
|
|
from jax._src import api_util
|
|
from jax._src import core
|
|
from jax._src import effects
|
|
from jax._src import dtypes
|
|
from jax._src import profiler
|
|
from jax._src import source_info_util
|
|
from jax._src.api_util import (flattened_fun_in_tree, flatten_fun_nokwargs,
|
|
fun_sourceinfo)
|
|
from jax._src.core import (Trace, Tracer, Jaxpr, Literal, get_aval,
|
|
AbstractValue, ClosedJaxpr, new_jaxpr_eqn,
|
|
ConcreteArray, Var, DropVar, raise_to_shaped, Atom,
|
|
JaxprEqn, Primitive, ShapedArray, DShapedArray,
|
|
mapped_aval, unmapped_aval, DBIdx, InDBIdx, OutDBIdx,
|
|
InputType, OutputType, get_referent)
|
|
from jax._src.state.types import AbstractRef
|
|
from jax._src.tree_util import (PyTreeDef, treedef_tuple, tree_unflatten,
|
|
KeyPath, generate_key_paths, keystr)
|
|
from jax._src.util import (unzip2, safe_zip, safe_map, toposort, split_list,
|
|
merge_lists, partition_list, OrderedSet,
|
|
as_hashable_function, weakref_lru_cache)
|
|
|
|
|
|
map, unsafe_map = safe_map, map
|
|
zip, unsafe_zip = safe_zip, zip
|
|
def identity(x): return x
|
|
|
|
TracerId = int
|
|
AvalId = int
|
|
ConstId = int
|
|
|
|
def _update_annotation_known(
|
|
f: lu.WrappedFun,
|
|
orig_type: InputType | None,
|
|
in_knowns: list[bool]
|
|
) -> lu.WrappedFun:
|
|
if orig_type is None: return f
|
|
# orig_type might contain DBIdx, but we're tossing out some args so we have to
|
|
# re-index. moreover some of the implicit args may not be needed anymore.
|
|
# so we basically just re-infer the lambda input type
|
|
if (all(e for _, e in orig_type) and
|
|
not any(type(d) is DBIdx for a, _ in orig_type for d in a.shape
|
|
if type(a) is DShapedArray)):
|
|
new_type = [ty for ty, known in zip(orig_type, in_knowns) if known]
|
|
return lu.annotate(f, tuple(new_type))
|
|
|
|
# Replace DBIdx with names, prune down to explicit only.
|
|
class Name:
|
|
def __init__(self, a): self.a = a
|
|
names = [Name(a) for a, _ in orig_type]
|
|
avals = [a.update(shape=tuple(names[d.val] if type(d) is DBIdx else d # type: ignore
|
|
for d in a.shape))
|
|
if type(a) is DShapedArray else a for a, e in orig_type if e]
|
|
avals = [a for a, known in zip(avals, in_knowns) if known]
|
|
# Figure out the implicit part: names which aren't explicit and known.
|
|
expl_names = [o for o, (_, e) in zip(names, orig_type) if e]
|
|
expl_names = [o for o, k in zip(expl_names, in_knowns) if k]
|
|
expl_names_ = set(expl_names)
|
|
impl_names = {d for a in avals if type(a) is DShapedArray for d in a.shape
|
|
if type(d) is Name and d not in expl_names_}
|
|
impl_part = [(n.a, False) for n in impl_names] # type: ignore
|
|
# Figure out the explicit part: known explicit avals, replacing names w/ dbidx
|
|
name_map = {n: DBIdx(i) for i, n in enumerate((*impl_names, *expl_names))}
|
|
expl_part = [(a.update(shape=tuple(name_map.get(d, d) for d in a.shape))
|
|
if type(a) is DShapedArray else a, True) for a in avals]
|
|
return lu.annotate(f, (*impl_part, *expl_part))
|
|
|
|
class PartialVal(tuple):
|
|
"""Partial value: either a known value or an unknown (abstract) value.
|
|
|
|
Represented as a pair `(aval_opt, const)` of one of two kinds:
|
|
* `(None, <Constant>)` indicates a known value, where the constant is either a
|
|
Tracer or satisfies `core.valid_jaxtype(const)`;
|
|
* `(<AbstractValue>, None)` indicates an unknown value characterized by an
|
|
abstract value.
|
|
"""
|
|
def __new__(cls, xs: tuple[AbstractValue | None, core.Value]):
|
|
pv, const = xs
|
|
if config.jax_enable_checks:
|
|
# type checks
|
|
assert isinstance(pv, (AbstractValue, type(None))), xs
|
|
assert (const is None or isinstance(const, core.Tracer) or
|
|
core.valid_jaxtype(const)), const
|
|
# invariant checks
|
|
assert (pv is None) ^ (const is None)
|
|
return tuple.__new__(cls, xs)
|
|
|
|
@classmethod
|
|
def known(cls, const: core.Value) -> PartialVal:
|
|
return PartialVal((None, const))
|
|
|
|
@classmethod
|
|
def unknown(cls, aval: AbstractValue) -> PartialVal:
|
|
return PartialVal((aval, None))
|
|
|
|
def is_known(self) -> bool:
|
|
return self[0] is None
|
|
|
|
def get_known(self) -> core.Value | None:
|
|
"""Get the known value, if known, else None."""
|
|
return self[1] if self[0] is None else None
|
|
|
|
def get_aval(self) -> AbstractValue:
|
|
"""Get AbstractValue directly (if unknown) or from the constant (known)."""
|
|
known = self.get_known()
|
|
if known is not None:
|
|
return get_aval(known)
|
|
else:
|
|
return self[0]
|
|
|
|
|
|
class JaxprTrace(Trace['JaxprTracer']):
|
|
|
|
def __init__(self, *args, name_stack: source_info_util.NameStack):
|
|
super().__init__(*args)
|
|
self.name_stack = name_stack
|
|
|
|
def pure(self, val: Any) -> JaxprTracer:
|
|
return self.new_const(val)
|
|
|
|
def lift(self, val: Tracer) -> JaxprTracer:
|
|
return self.new_const(val)
|
|
|
|
def sublift(self, val: JaxprTracer) -> JaxprTracer:
|
|
return JaxprTracer(self, val.pval, FreeVar(val))
|
|
|
|
def new_const(self, val) -> JaxprTracer:
|
|
if isinstance(val, Tracer) and val._trace.level == self.level:
|
|
raise Exception
|
|
return JaxprTracer(self, PartialVal.known(val), None)
|
|
|
|
def new_instantiated_literal(self, val) -> JaxprTracer:
|
|
aval = get_aval(val)
|
|
return JaxprTracer(self, PartialVal.unknown(aval),
|
|
Literal(val, raise_to_shaped(aval)))
|
|
|
|
def new_instantiated_const(self, val) -> JaxprTracer:
|
|
aval = get_aval(val)
|
|
if isinstance(aval, DShapedArray):
|
|
shape = [self.new_instantiated_const(d)
|
|
if isinstance(d, Tracer) and d._trace.level < self.level else d
|
|
for d in aval.shape]
|
|
aval = aval.update(shape=tuple(shape))
|
|
return JaxprTracer(self, PartialVal.unknown(aval), ConstVar(val))
|
|
|
|
def new_arg(self, pval: PartialVal) -> JaxprTracer:
|
|
const = pval.get_known()
|
|
# XXX: Think twice before changing this constant argument pruning!
|
|
# This has really important consequences for partial_eval_jaxpr.
|
|
# Most importantly, this guarantees that the unknown jaxpr never uses
|
|
# known inputs (if it needs them, then they get passed through residuals).
|
|
if const is None:
|
|
aval = pval.get_aval()
|
|
if type(aval) is DShapedArray:
|
|
shape = [self.new_instantiated_const(d)
|
|
if isinstance(d, Tracer) and d._trace.level < self.level else d
|
|
for d in aval.shape]
|
|
aval = aval.update(shape=tuple(shape))
|
|
return JaxprTracer(self, PartialVal.unknown(aval), LambdaBinding())
|
|
else:
|
|
return self.new_const(const)
|
|
|
|
def instantiate_const(self, tracer: JaxprTracer) -> JaxprTracer:
|
|
const = tracer.pval.get_known()
|
|
if const is None:
|
|
return tracer
|
|
else:
|
|
if type(const) in core.literalable_types and np.shape(const) == ():
|
|
return self.new_instantiated_literal(const)
|
|
else:
|
|
return self.new_instantiated_const(const)
|
|
|
|
def instantiate_const_abstracted(self, tracer) -> JaxprTracer:
|
|
const = tracer.pval.get_known()
|
|
if const is None:
|
|
return tracer
|
|
else:
|
|
aval = raise_to_shaped(get_aval(const), np.isscalar(const))
|
|
return JaxprTracer(self, PartialVal.unknown(aval), ConstVar(const))
|
|
|
|
def process_primitive(self, primitive, tracers, params):
|
|
if primitive in custom_partial_eval_rules:
|
|
return custom_partial_eval_rules[primitive](self, *tracers, **params)
|
|
else:
|
|
return self.default_process_primitive(primitive, tracers, params)
|
|
|
|
def default_process_primitive(self, primitive, tracers, params):
|
|
# By default, if all the input tracers are known, then bind the primitive
|
|
# and consider all outputs known. Otherwise, stage the application into the
|
|
# jaxpr and consider all outputs unknown.
|
|
consts = [t.pval.get_known() for t in tracers]
|
|
if all(c is not None for c in consts):
|
|
return primitive.bind(*consts, **params)
|
|
tracers = map(self.instantiate_const, tracers)
|
|
avals = [t.aval for t in tracers]
|
|
out_aval, effects = primitive.abstract_eval(*avals, **params)
|
|
name_stack = self._current_truncated_name_stack()
|
|
source = source_info_util.current().replace(name_stack=name_stack)
|
|
if primitive.multiple_results:
|
|
out_tracers = [JaxprTracer(self, PartialVal.unknown(aval), None)
|
|
for aval in out_aval]
|
|
eqn = new_eqn_recipe(tracers, out_tracers, primitive, params, effects, source)
|
|
for t in out_tracers: t.recipe = eqn
|
|
return out_tracers
|
|
else:
|
|
out_tracer = JaxprTracer(self, PartialVal.unknown(out_aval), None)
|
|
out_tracer.recipe = new_eqn_recipe(tracers, [out_tracer], primitive,
|
|
params, effects, source)
|
|
return out_tracer
|
|
|
|
def process_call(self, primitive, f, tracers, params):
|
|
rule = call_partial_eval_rules.get(primitive)
|
|
if rule:
|
|
return rule(self, primitive, f, tracers, params)
|
|
|
|
update_params = call_param_updaters.get(primitive) or (lambda p, _, __: p)
|
|
in_knowns, in_avals, in_consts = partition_pvals([t.pval for t in tracers])
|
|
# TODO(mattjj): check in_avals are consistent with f.in_type
|
|
|
|
# We want to partially evaluate this call into two calls: one evaluated now
|
|
# taking known values (in_consts) as inputs and producing known values
|
|
# (out_consts) as outputs, and the other staged out as an eqn into the jaxpr
|
|
# being built. The latter takes as input residuals (res) produced as outputs
|
|
# of the first call, shared closed-over values (env), and explicit arguments
|
|
# which were unknown to the first call (corresponding to in_avals).
|
|
|
|
# Wrap f to perform the partial evaluation and plumb out aux data.
|
|
if not config.jax_dynamic_shapes:
|
|
f_ = trace_to_subjaxpr_nounits_fwd(f, self.main, False)
|
|
f_, aux = partial_eval_wrapper_nounits(f_, tuple(in_knowns),
|
|
tuple(in_avals))
|
|
else:
|
|
if f.in_type is None:
|
|
f = lu.annotate(f, tuple((a, True) for a in in_avals))
|
|
f_, aux = trace_to_subjaxpr_nounits_dyn(f, self.main, tuple(in_knowns),
|
|
f.in_type, False)
|
|
# Adjust parameters (e.g. donated_invars) for the call to be evaluated now.
|
|
const_params = update_params(params, in_knowns, 0)
|
|
|
|
# Run the call, getting known out vals and aux data used for staged-out call
|
|
out = primitive.bind(_update_annotation_known(f_, f.in_type, in_knowns),
|
|
*in_consts, **const_params)
|
|
fwds, out_knowns, out_type, jaxpr, env = aux()
|
|
# Split apart known outputs from the original call and non-fwded residuals.
|
|
out_consts, non_fwd_res_ = split_list(out, [sum(out_knowns)])
|
|
|
|
# Form the complete list of residuals by forwarding some inputs.
|
|
if config.jax_dynamic_shapes:
|
|
# With dynamic shapes, we may need to forward implicit arguments.
|
|
in_consts_, in_knowns_ = iter(in_consts), iter(in_knowns)
|
|
in_consts_full = [None] * len(f.in_type)
|
|
for idx, (aval, explicit) in enumerate(f.in_type):
|
|
if explicit and next(in_knowns_):
|
|
c = in_consts_full[idx] = next(in_consts_)
|
|
if aval.shape:
|
|
for d1, d2 in zip(aval.shape, c.shape):
|
|
if type(d1) is DBIdx:
|
|
in_consts_full[d1.val] = d2
|
|
else:
|
|
in_consts_full = in_consts
|
|
non_fwd_res = iter(non_fwd_res_)
|
|
res = [next(non_fwd_res) if i is None else in_consts_full[i] for i in fwds]
|
|
sentinel = object()
|
|
assert next(non_fwd_res, sentinel) is sentinel
|
|
|
|
# Create the input tracers for the staged-out (unknown-value) call.
|
|
res_tracers = map(self.instantiate_const, map(self.new_const, res))
|
|
env_tracers = map(self.full_raise, env)
|
|
unknown_arg_tracers = [t for t in tracers if not t.is_known()]
|
|
# Adjust parameters (e.g. donated_invars) for the staged-out call's args.
|
|
num_new_args = len(res_tracers) + len(env_tracers)
|
|
staged_params = dict(params, call_jaxpr=convert_constvars_jaxpr(jaxpr))
|
|
staged_params = update_params(staged_params, map(op.not_, in_knowns),
|
|
num_new_args)
|
|
# The outputs of the staged-out call are Tracers with the new eqn as recipe.
|
|
if config.jax_dynamic_shapes:
|
|
# With dynamic shapes, we may need to substitute Tracers into avals.
|
|
out_tracers = []
|
|
for aval, _ in out_type:
|
|
assert not isinstance(aval, ConcreteArray)
|
|
if type(aval) is DShapedArray:
|
|
shape = [[*res_tracers, *env_tracers, *unknown_arg_tracers][d.val]
|
|
if type(d) is InDBIdx else d for d in aval.shape]
|
|
aval = aval.update(shape=tuple(shape))
|
|
out_tracers.append(JaxprTracer(self, PartialVal.unknown(aval), None))
|
|
else:
|
|
out_tracers = [JaxprTracer(self, PartialVal.unknown(a), None)
|
|
for a in out_type]
|
|
name_stack = self._current_truncated_name_stack()
|
|
source = source_info_util.current().replace(name_stack=name_stack)
|
|
eqn = new_eqn_recipe((*res_tracers, *env_tracers, *unknown_arg_tracers),
|
|
out_tracers, primitive, staged_params, jaxpr.effects,
|
|
source)
|
|
for t in out_tracers: t.recipe = eqn
|
|
return merge_lists(out_knowns, out_tracers, out_consts)
|
|
|
|
def process_map(self, primitive, f: lu.WrappedFun, tracers, params):
|
|
update_params = call_param_updaters.get(primitive) or (lambda p, _, __: p)
|
|
in_knowns, in_avals, in_consts = partition_pvals([t.pval for t in tracers])
|
|
|
|
# This method is like process_call above, except:
|
|
# 1. we delete an axis from mapped-over input avals' shapes, and
|
|
# analogously add an axis to mapped-over output avals' shapes;
|
|
# 2. we update the in_axes and out_axes/out_axes_thunk parameters to
|
|
# reflect the inputs and outputs pruned from the unknown/known sides.
|
|
|
|
# Map (delete an axis from) unknown inputs' avals as dictated by in_axes.
|
|
unk_in_axes, const_in_axes = partition_list(in_knowns, params['in_axes'])
|
|
in_avals_mapped = [mapped_aval(params['axis_size'], ax, aval)
|
|
for ax, aval in zip(unk_in_axes, in_avals)]
|
|
|
|
# Wrap f to perform partial evaluation and plumb out aux data.
|
|
f = trace_to_subjaxpr_nounits(f, self.main, False)
|
|
f, aux = partial_eval_wrapper_nounits(f, tuple(in_knowns),
|
|
tuple(in_avals_mapped))
|
|
# Adjust params for knowns (e.g. donated_invars, in_axes, out_axes_thunk)
|
|
const_params = update_params(params, in_knowns, 0) # handles donated_invars
|
|
out_axes_thunk = params['out_axes_thunk']
|
|
@as_hashable_function(closure=out_axes_thunk)
|
|
def const_out_axes_thunk():
|
|
out_knowns, _, jaxpr, _ = aux()
|
|
_, out_axes = partition_list(out_knowns, out_axes_thunk())
|
|
return tuple(out_axes) + (0,) * len(jaxpr.constvars) # res mapped axis 0
|
|
const_params = dict(const_params, in_axes=tuple(const_in_axes),
|
|
out_axes_thunk=const_out_axes_thunk)
|
|
|
|
# Run the map, getting known out vals and aux data used for staged-out map.
|
|
out = primitive.bind(f, *in_consts, **const_params)
|
|
out_knowns, out_avals_mapped, jaxpr, env = aux()
|
|
# Split apart known outputs from the original call and residuals.
|
|
out_consts, res = split_list(out, [len(out) - len(jaxpr.constvars)])
|
|
|
|
# We can only check_jaxpr with the dynamic axis environment extended:
|
|
with core.extend_axis_env(params['axis_name'], params['axis_size'], None):
|
|
call_jaxpr = convert_constvars_jaxpr(jaxpr)
|
|
|
|
# Compute staged and const out_axes, taking into account residuals.
|
|
out_axes = params['out_axes_thunk']()
|
|
staged_out_axes, _ = partition_list(out_knowns, out_axes)
|
|
staged_in_axes = (0,) * len(res) + (None,) * len(env) + (*unk_in_axes,)
|
|
|
|
# Create the input tracers for the staged-out (unkonwn-value) call.
|
|
const_tracers = map(self.new_instantiated_const, res)
|
|
env_tracers = map(self.full_raise, env)
|
|
unknown_arg_tracers = [t for t in tracers if not t.is_known()]
|
|
# Adjust params for staged-out call on unknown values.
|
|
num_new_args = len(const_tracers) + len(env_tracers)
|
|
staged_params = update_params(params, map(op.not_, in_knowns), num_new_args)
|
|
staged_params = dict(staged_params, in_axes=staged_in_axes,
|
|
out_axes=tuple(staged_out_axes), call_jaxpr=call_jaxpr)
|
|
del staged_params['out_axes_thunk']
|
|
# The outputs of the staged-out call are Tracers with the new eqn as recipe.
|
|
out_avals = [unmapped_aval(params['axis_size'], params['axis_name'], ax, a)
|
|
for ax, a in zip(staged_out_axes, out_avals_mapped)]
|
|
out_tracers = [JaxprTracer(self, PartialVal.unknown(a), None)
|
|
for a in out_avals]
|
|
eqn = new_eqn_recipe((*const_tracers, *env_tracers, *unknown_arg_tracers), # type: ignore[arg-type]
|
|
out_tracers, primitive, staged_params,
|
|
jaxpr.effects,
|
|
source_info_util.current())
|
|
for t in out_tracers: t.recipe = eqn
|
|
|
|
return merge_lists(out_knowns, out_tracers, out_consts)
|
|
|
|
def post_process_call(self, primitive, out_tracers, params):
|
|
unknown_out_tracers = [t for t in out_tracers if not t.is_known()]
|
|
jaxpr, res, env = tracers_to_jaxpr([], unknown_out_tracers)
|
|
out_pvals = [t.pval for t in out_tracers]
|
|
out_knowns, out_avals, out_consts = partition_pvals(out_pvals)
|
|
out = [*out_consts, *res]
|
|
main = self.main
|
|
|
|
def todo(out):
|
|
trace = main.with_cur_sublevel()
|
|
out_consts, res = split_list(out, [len(out) - len(jaxpr.constvars)])
|
|
const_tracers = map(trace.new_instantiated_const, res)
|
|
in_tracers = (*const_tracers, *map(trace.full_raise, env))
|
|
out_tracers = [JaxprTracer(trace, PartialVal.unknown(a), None)
|
|
for a in out_avals]
|
|
update_params = call_param_updaters.get(primitive) or (lambda p, _, __: p)
|
|
new_params = update_params(params, [], len(in_tracers))
|
|
new_params = dict(new_params, call_jaxpr=convert_constvars_jaxpr(jaxpr))
|
|
name_stack = self._current_truncated_name_stack()
|
|
source = source_info_util.current().replace(name_stack=name_stack)
|
|
eqn = new_eqn_recipe(in_tracers, out_tracers, primitive, new_params,
|
|
jaxpr.effects, source)
|
|
for t in out_tracers: t.recipe = eqn
|
|
return merge_lists(out_knowns, out_tracers, out_consts)
|
|
|
|
return out, todo
|
|
|
|
def post_process_map(self, primitive, out_tracers, params):
|
|
unknown_out_tracers = [t for t in out_tracers if not t.is_known()]
|
|
jaxpr, res, env = tracers_to_jaxpr([], unknown_out_tracers)
|
|
out_pvals = [t.pval for t in out_tracers]
|
|
out_knowns, out_avals_mapped, out_consts = partition_pvals(out_pvals)
|
|
out = [*out_consts, *res]
|
|
main = self.main
|
|
|
|
with core.extend_axis_env(params['axis_name'], params['axis_size'], None):
|
|
call_jaxpr = convert_constvars_jaxpr(jaxpr)
|
|
|
|
def todo(out):
|
|
trace = main.with_cur_sublevel()
|
|
out_consts, res = split_list(out, [len(out) - len(jaxpr.constvars)])
|
|
const_tracers = map(trace.new_instantiated_const, res)
|
|
env_tracers = map(trace.full_raise, env)
|
|
|
|
staged_out_axes = tuple(out_axes_unknown) # set by out_axes_transform
|
|
staged_in_axes = (0,) * len(res) + (None,) * len(env)
|
|
|
|
update_params = call_param_updaters.get(primitive) or (lambda p, _, __: p)
|
|
staged_params = update_params(params, [], len(res) + len(env))
|
|
staged_params = dict(staged_params, in_axes=staged_in_axes,
|
|
out_axes=tuple(staged_out_axes),
|
|
call_jaxpr=call_jaxpr)
|
|
|
|
out_avals = [unmapped_aval(params['axis_size'], params['axis_name'], d, a)
|
|
for d, a in zip(staged_out_axes, out_avals_mapped)]
|
|
out_tracers = [JaxprTracer(trace, PartialVal.unknown(a), None)
|
|
for a in out_avals]
|
|
name_stack = self._current_truncated_name_stack()
|
|
source = source_info_util.current().replace(name_stack=name_stack)
|
|
eqn = new_eqn_recipe((*const_tracers, *env_tracers), out_tracers,
|
|
primitive, staged_params, jaxpr.effects, source)
|
|
for t in out_tracers: t.recipe = eqn
|
|
return merge_lists(out_knowns, out_tracers, out_consts)
|
|
|
|
def out_axes_transform(out_axes):
|
|
nonlocal out_axes_unknown
|
|
out_axes_unknown, out_axes_known = partition_list(out_knowns, out_axes)
|
|
return tuple(out_axes_known) + (0,) * len(jaxpr.constvars)
|
|
out_axes_unknown: list | None = None
|
|
|
|
return out, (todo, out_axes_transform)
|
|
|
|
def _current_truncated_name_stack(self):
|
|
return source_info_util.current_name_stack()[len(self.name_stack):]
|
|
|
|
def process_custom_jvp_call(self, prim, fun, jvp, tracers, *, symbolic_zeros):
|
|
# We assume partial evaluation is only performed to build linear functions,
|
|
# and hence we don't need to keep the custom JVP rule around anymore.
|
|
del jvp, symbolic_zeros
|
|
assert not all(t.is_known() for t in tracers)
|
|
return fun.call_wrapped(*tracers)
|
|
|
|
def post_process_custom_jvp_call(self, out_tracers, _):
|
|
# This path should only be reachable if we expose a partial eval API
|
|
# unrelated to autodiff, since we raise an error when differentiation with
|
|
# respect to values over which a custom_jvp function closes is detected.
|
|
raise NotImplementedError # TODO(mattjj)
|
|
|
|
def process_custom_transpose(self, prim, call, tracers, **params):
|
|
res_ts, lin_ts = split_list(tracers, [params['res_tree'].num_leaves])
|
|
assert all(t.is_known() for t in res_ts)
|
|
lin_all_known = all(t.is_known() for t in lin_ts)
|
|
if lin_all_known:
|
|
res_cvals = [t.pval[1] for t in res_ts]
|
|
lin_cvals = [t.pval[1] for t in lin_ts]
|
|
return prim.bind(call, *res_cvals, *lin_cvals, **params)
|
|
else:
|
|
out_tracers = [JaxprTracer(self, PartialVal.unknown(aval), None)
|
|
for aval in params['out_types']]
|
|
in_tracers = map(self.instantiate_const, tracers)
|
|
new_params = dict(params, call=call)
|
|
eqn = new_eqn_recipe(in_tracers, out_tracers, prim, new_params,
|
|
core.no_effects, source_info_util.current())
|
|
for t in out_tracers: t.recipe = eqn
|
|
return out_tracers
|
|
|
|
def process_custom_vjp_call(self, prim, f, fwd, bwd, tracers, out_trees,
|
|
symbolic_zeros):
|
|
# TODO(mattjj): after old remat is deleted, make this method trivial.
|
|
# Because we instantiate all tracers, in_knowns is all False.
|
|
tracers = map(self.instantiate_const_abstracted, tracers)
|
|
in_knowns, in_avals, () = partition_pvals([t.pval for t in tracers])
|
|
f = trace_to_subjaxpr_nounits(f, self.main, True)
|
|
f, aux = partial_eval_wrapper_nounits(f, tuple(in_knowns), tuple(in_avals))
|
|
out_flat = prim.bind(f, fwd, bwd, out_trees=out_trees,
|
|
symbolic_zeros=symbolic_zeros)
|
|
out_knowns, out_avals, jaxpr, env = aux()
|
|
out_consts, res = split_list(out_flat, [len(out_flat)-len(jaxpr.constvars)])
|
|
res_tracers = map(self.new_instantiated_const, res)
|
|
env_tracers = map(self.full_raise, env)
|
|
out_tracers = [JaxprTracer(self, PartialVal.unknown(a), None)
|
|
for a in out_avals]
|
|
closed_jaxpr = core.ClosedJaxpr(convert_constvars_jaxpr(jaxpr), ())
|
|
|
|
@_memoize
|
|
def fwd_jaxpr_thunk(*zeros):
|
|
fwd_ = _interleave_fun(fwd, zeros)
|
|
fwd_ = trace_to_subjaxpr_nounits(fwd_, self.main, True)
|
|
fwd_, aux = partial_eval_wrapper_nounits(
|
|
fwd_, tuple(in_knowns), tuple(in_avals))
|
|
with core.new_sublevel():
|
|
out_flat = fwd_.call_wrapped()
|
|
out_knowns, out_avals, jaxpr, env = aux()
|
|
_, res = split_list(out_flat, [len(out_flat)-len(jaxpr.constvars)])
|
|
converted_jaxpr = convert_envvars_to_constvars(jaxpr, len(env))
|
|
return converted_jaxpr, (*res, *env)
|
|
|
|
name_stack = self._current_truncated_name_stack()
|
|
source = source_info_util.current().replace(name_stack=name_stack)
|
|
eqn = new_eqn_recipe((*res_tracers, *env_tracers, *tracers),
|
|
out_tracers, prim.initial_style,
|
|
dict(fun_jaxpr=closed_jaxpr,
|
|
fwd_jaxpr_thunk=fwd_jaxpr_thunk,
|
|
num_consts=len(res) + len(env),
|
|
bwd=bwd, out_trees=out_trees,
|
|
symbolic_zeros=symbolic_zeros),
|
|
jaxpr.effects, source)
|
|
for t in out_tracers: t.recipe = eqn
|
|
return merge_lists(out_knowns, out_tracers, out_consts)
|
|
|
|
def post_process_custom_vjp_call(self, out_tracers, _):
|
|
# This path should only be reachable if we expose a partial eval API
|
|
# unrelated to autodiff, since we raise an error when differentiation with
|
|
# respect to values over which a custom_vjp function closes is detected.
|
|
raise NotImplementedError # TODO(mattjj)
|
|
|
|
def partition_pvals(
|
|
pvals: list[PartialVal]
|
|
) -> tuple[list[bool], list[AbstractValue], list[Any]]:
|
|
knowns = [pval.is_known() for pval in pvals ]
|
|
avals = [pval.get_aval() for pval in pvals if not pval.is_known()]
|
|
consts = [pval.get_known() for pval in pvals if pval.is_known()]
|
|
return knowns, avals, consts
|
|
|
|
@lu.transformation_with_aux
|
|
def partial_eval_wrapper_nounits(
|
|
in_knowns: Sequence[bool], in_avals: Sequence[AbstractValue],
|
|
*in_consts: Any):
|
|
in_avals_, in_consts_ = iter(in_avals), iter(in_consts)
|
|
in_pvals = [PartialVal.known(next(in_consts_)) if known else
|
|
PartialVal.unknown(next(in_avals_)) for known in in_knowns]
|
|
sentinel = object()
|
|
assert next(in_avals_, sentinel) is next(in_consts_, sentinel) is sentinel
|
|
jaxpr, (*maybe_fwds, out_pvals, res, env) = yield (in_pvals,), {}
|
|
out_knowns, out_avals, out_consts = partition_pvals(out_pvals)
|
|
yield (*out_consts, *res), (*maybe_fwds, out_knowns, out_avals, jaxpr, env)
|
|
|
|
@lu.transformation_with_aux
|
|
def trace_to_subjaxpr_nounits_dyn(
|
|
main: core.MainTrace, in_knowns: Sequence[bool], in_type: InputType,
|
|
instantiate: bool | Sequence[bool],
|
|
*in_consts: Any):
|
|
trace = main.with_cur_sublevel()
|
|
in_avals, which_explicit = unzip2(in_type)
|
|
|
|
# To form input tracers from in_type, we need to first build ConstVar tracers
|
|
# for all axis sizes, so that we can then use those tracers in the shapes of
|
|
# avals for unknown inputs' tracers. We use ConstVar recipes for on-the-fly
|
|
# type agreement checking via get_referent.
|
|
in_consts_full: list[JaxprTracer | None] = [None] * len(in_type)
|
|
in_consts_iter, in_knowns_iter = iter(in_consts), iter(in_knowns)
|
|
for idx, (aval, explicit) in enumerate(in_type):
|
|
if explicit and next(in_knowns_iter):
|
|
constval = next(in_consts_iter)
|
|
if isinstance(aval, DShapedArray):
|
|
for i, d in enumerate(aval.shape):
|
|
if isinstance(d, DBIdx):
|
|
if in_consts_full[d.val] is None:
|
|
in_consts_full[d.val] = \
|
|
JaxprTracer(trace, PartialVal.unknown(in_avals[d.val]),
|
|
ConstVar(constval.shape[i]))
|
|
assert core.same_referent(constval.shape[i], in_consts_full[d.val])
|
|
shape = [in_consts_full[d.val] if type(d) is DBIdx else d # type: ignore
|
|
for d in aval.shape]
|
|
aval = aval.update(shape=tuple(shape))
|
|
in_consts_full[idx] = JaxprTracer(trace, PartialVal.unknown(aval),
|
|
ConstVar(constval))
|
|
# Check that we covered all axis sizes with ConstVar tracers.
|
|
for idx, (aval, explicit) in enumerate(in_type):
|
|
if not explicit: assert in_consts_full[idx] is not None
|
|
if isinstance(aval, DShapedArray):
|
|
assert all(type(d) is not DBIdx or in_consts_full[d.val] is not None # type: ignore
|
|
for d in aval.shape)
|
|
|
|
# Next, build tracers for all unknown inputs, using the in_consts_full list
|
|
# for axis size tracers when necessary.
|
|
in_tracers = []
|
|
in_knowns_iter = iter(in_knowns)
|
|
for aval, explicit in in_type:
|
|
if explicit and not next(in_knowns_iter):
|
|
if isinstance(aval, DShapedArray):
|
|
shape = [in_consts_full[d.val] if type(d) is DBIdx else d # type: ignore
|
|
for d in aval.shape]
|
|
aval = aval.update(shape=tuple(shape))
|
|
tracer = JaxprTracer(trace, PartialVal.unknown(aval), LambdaBinding())
|
|
in_tracers.append(tracer)
|
|
|
|
# Merge in_consts and in_tracers and call wrapped fn with explicit arguments.
|
|
in_args = merge_lists(in_knowns, in_tracers, in_consts)
|
|
ans = yield in_args, {}
|
|
|
|
# Instantiate outputs and build jaxpr.
|
|
if isinstance(instantiate, bool):
|
|
instantiate = [instantiate] * len(ans)
|
|
out_tracers = map(trace.full_raise, map(core.full_lower, ans))
|
|
out_tracers = [trace.instantiate_const(trace.full_raise(t)) if inst else t
|
|
for inst, t in zip(instantiate, out_tracers)]
|
|
|
|
# Collect known outputs.
|
|
out_knowns: list[bool] = [t.is_known() for t in out_tracers]
|
|
out_consts: list[Any] = [t.pval.get_known() for t in out_tracers
|
|
if t.is_known()]
|
|
|
|
# Build the jaxpr.
|
|
out_tracers = [t for t in out_tracers if not t.is_known()]
|
|
jaxpr, res, env = tracers_to_jaxpr(in_tracers, out_tracers)
|
|
out_avals = [v.aval for v in jaxpr.outvars]
|
|
idx_map = {v: InDBIdx(i)
|
|
for i, v in enumerate(it.chain(jaxpr.constvars, jaxpr.invars))}
|
|
out_type = [(a.update(shape=tuple(idx_map.get(d, d) for d in a.shape)) # type: ignore
|
|
if type(a) is DShapedArray else a, True) for a in out_avals]
|
|
|
|
# Which residuals are just forwarded inputs? Check obj id, then prune.
|
|
id_map = {id(c.recipe.val): i for i, c in enumerate(in_consts_full) # type: ignore
|
|
if c is not None}
|
|
fwds: list[int | None] = [id_map.get(id(c)) for c in res]
|
|
res = tuple([c for c, fwd in zip(res, fwds) if fwd is None])
|
|
|
|
del main, in_consts, trace, in_consts_iter, in_knowns_iter, in_consts_full, \
|
|
in_tracers, in_args, ans, out_tracers, out_avals
|
|
yield (*out_consts, *res), (fwds, out_knowns, tuple(out_type), jaxpr, env)
|
|
|
|
|
|
custom_partial_eval_rules: dict[Primitive, Callable] = {}
|
|
call_partial_eval_rules: dict[Primitive, Callable] = {}
|
|
call_param_updaters: dict[Primitive, Callable] = {}
|
|
|
|
def _closed_call_param_updater(params, _, __):
|
|
jaxpr = params.get('call_jaxpr')
|
|
if jaxpr is None: return params
|
|
assert type(jaxpr) is core.Jaxpr
|
|
return dict(params, call_jaxpr=core.ClosedJaxpr(jaxpr, ()))
|
|
call_param_updaters[core.closed_call_p] = _closed_call_param_updater
|
|
|
|
def abstract_eval_fun(fun, *avals, debug_info=None, **params):
|
|
_, avals_out, _ = trace_to_jaxpr_dynamic(
|
|
lu.wrap_init(fun, params), avals, debug_info)
|
|
assert all(isinstance(aval, AbstractValue) for aval in avals_out)
|
|
return avals_out
|
|
|
|
|
|
JaxprTracerRecipe = Union['JaxprEqnRecipe', 'LambdaBinding', 'FreeVar',
|
|
'ConstVar', Literal]
|
|
|
|
class JaxprTracer(Tracer):
|
|
__slots__ = ['pval', 'recipe']
|
|
|
|
def __init__(self, trace: JaxprTrace, pval: PartialVal,
|
|
recipe: JaxprTracerRecipe | None):
|
|
assert isinstance(pval, PartialVal)
|
|
pv, const = pval
|
|
if isinstance(const, Tracer) and const._trace.level >= trace.level:
|
|
raise core.escaped_tracer_error(
|
|
const, f"Tracer from a higher level: {const} in trace {trace}")
|
|
if isinstance(pv, DShapedArray):
|
|
assert all(not isinstance(d, Tracer) or isinstance(d, JaxprTracer) and
|
|
d._trace.level == trace.level for d in pv.shape)
|
|
self._trace = trace
|
|
self.pval = pval
|
|
self.recipe = recipe
|
|
|
|
def __repr__(self):
|
|
return f'Traced<{self.aval}:{self._trace}>'
|
|
|
|
@property
|
|
def aval(self) -> AbstractValue:
|
|
return self.pval.get_aval()
|
|
|
|
@property
|
|
def parents(self) -> Sequence[JaxprTracer]:
|
|
if isinstance(self.recipe, JaxprEqnRecipe):
|
|
# TODO broadcast_in_dim can create a new tracer...
|
|
return self.recipe.in_tracers
|
|
elif isinstance(self.aval, DShapedArray):
|
|
return [d for d in self.aval.shape if isinstance(d, JaxprTracer)]
|
|
else:
|
|
return []
|
|
|
|
def full_lower(self):
|
|
known = self.pval.get_known()
|
|
if known is not None:
|
|
return core.full_lower(known)
|
|
else:
|
|
return self
|
|
|
|
def is_known(self):
|
|
return self.pval.is_known()
|
|
|
|
def get_referent(self):
|
|
if self.pval.is_known():
|
|
return get_referent(self.pval.get_known())
|
|
elif isinstance(self.recipe, (FreeVar, ConstVar, Literal)):
|
|
return get_referent(self.recipe.val) # pytype: disable=attribute-error
|
|
else:
|
|
return self
|
|
|
|
|
|
@profiler.annotate_function
|
|
def trace_to_jaxpr(
|
|
fun: lu.WrappedFun, pvals: Sequence[PartialVal],
|
|
instantiate: bool | Sequence[bool] = False,
|
|
) -> tuple[Jaxpr, list[PartialVal], list[core.Value]]:
|
|
"""
|
|
Partially evaluate a function, building a jaxpr for un-evaluated computation.
|
|
|
|
Args:
|
|
fun: lu.WrappedFun representing the function to be partially evaluated. The
|
|
function must be flattened, in the sense of accepting jaxpr type arguments
|
|
and returning a flat list of jaxpr type outputs.
|
|
pvals: sequence of PartialVals of length equal to the number of inputs to
|
|
`fun` indicating which inputs are known or unknown.
|
|
instantiate: optional bool or sequence of bools of length equal to the
|
|
number of outputs of `fun` indicating which outputs should be forced to be
|
|
treated as unknown and hence instantiated in the jaxpr. If a single bool,
|
|
the value is applied to all outputs. Default False.
|
|
|
|
Returns:
|
|
A triple where the first element is a jaxpr representing the computation
|
|
which depends on unknown inputs; the second element is a list of PartialVals
|
|
of length equal to the length of the output of `fun` representing which
|
|
outputs are known and unknown (along with their values and abstract values,
|
|
respectively); the third element is a list of known residual values. The
|
|
returned jaxpr takes as inputs the known residual values followed by values
|
|
of the originally unknown inputs.
|
|
"""
|
|
current_name_stack = source_info_util.current_name_stack()
|
|
with core.new_main(JaxprTrace, name_stack=current_name_stack) as main:
|
|
fun = trace_to_subjaxpr(fun, main, instantiate)
|
|
jaxpr, (out_pvals, consts, env) = fun.call_wrapped(pvals)
|
|
assert not env
|
|
del main, fun, env
|
|
|
|
return jaxpr, out_pvals, consts
|
|
|
|
@profiler.annotate_function
|
|
def trace_to_jaxpr_nounits(
|
|
fun: lu.WrappedFun, pvals: Sequence[PartialVal],
|
|
instantiate: bool | Sequence[bool] = False,
|
|
) -> tuple[Jaxpr, list[PartialVal], list[core.Value]]:
|
|
current_name_stack = source_info_util.current_name_stack()
|
|
with core.new_main(JaxprTrace, name_stack=current_name_stack) as main:
|
|
fun = trace_to_subjaxpr_nounits(fun, main, instantiate)
|
|
jaxpr, (out_pvals, consts, env) = fun.call_wrapped(pvals)
|
|
assert not env
|
|
del main, fun, env
|
|
return jaxpr, out_pvals, consts
|
|
|
|
|
|
@lu.transformation
|
|
def trace_to_subjaxpr_nounits(
|
|
main: core.MainTrace,
|
|
instantiate: bool | Sequence[bool],
|
|
in_pvals: Sequence[PartialVal]):
|
|
assert all([isinstance(pv, PartialVal) for pv in in_pvals]), in_pvals
|
|
out_tracers, jaxpr, out_consts, env = yield from _trace_to_subjaxpr_nounits(
|
|
main, instantiate, in_pvals)
|
|
out_pvals = [t.pval for t in out_tracers]
|
|
del out_tracers
|
|
yield jaxpr, (out_pvals, out_consts, env)
|
|
|
|
def _trace_to_subjaxpr_nounits(main, instantiate, in_pvals):
|
|
trace = main.with_cur_sublevel()
|
|
in_knowns = [pval.is_known() for pval in in_pvals]
|
|
in_consts = [pval.get_known() for pval in in_pvals if pval.is_known()]
|
|
in_tracers = [trace.new_arg(pval) for pval in in_pvals if not pval.is_known()]
|
|
in_args = merge_lists(in_knowns, in_tracers, in_consts)
|
|
ans = yield in_args, {}
|
|
assert isinstance(ans, (list, tuple)), (
|
|
f"Got unexpected return type when tracing function to jaxpr: {ans}")
|
|
assert all(isinstance(x, core.Tracer) or core.valid_jaxtype(x) for x in ans), (
|
|
f"Got unexpected return type when tracing function to jaxpr: {ans}")
|
|
if isinstance(instantiate, bool):
|
|
instantiate = [instantiate] * len(ans)
|
|
out_tracers = map(trace.full_raise, map(core.full_lower, ans))
|
|
out_tracers = [trace.instantiate_const(trace.full_raise(t)) if inst else t
|
|
for inst, t in zip(instantiate, out_tracers)]
|
|
out_tracers_ = [t for t in out_tracers if not t.is_known()]
|
|
jaxpr, out_consts, env = tracers_to_jaxpr(in_tracers, out_tracers_)
|
|
return out_tracers, jaxpr, out_consts, env
|
|
|
|
# The below variant implements an optimization where residuals which are also
|
|
# inputs are indicated in auxiliary data rather than passed as outputs.
|
|
# TODO(mattjj): update all callers to use this version, delete other version.
|
|
@lu.transformation
|
|
def trace_to_subjaxpr_nounits_fwd(
|
|
main: core.MainTrace,
|
|
instantiate: bool | Sequence[bool],
|
|
in_pvals: Sequence[PartialVal]):
|
|
assert all([isinstance(pv, PartialVal) for pv in in_pvals]), in_pvals
|
|
out_tracers, jaxpr, out_consts, env = yield from _trace_to_subjaxpr_nounits(
|
|
main, instantiate, in_pvals)
|
|
out_pvals = [t.pval for t in out_tracers]
|
|
|
|
# Which out_consts (aka residuals) are just forwarded inputs? Check obj id.
|
|
in_consts = [pval.get_known() for pval in in_pvals if pval.is_known()]
|
|
id_map = {id(c): i for i, c in enumerate(in_consts)}
|
|
fwds: list[int | None] = [id_map.get(id(c)) for c in out_consts]
|
|
pruned_consts = [c for c, fwd in zip(out_consts, fwds) if fwd is None]
|
|
|
|
del out_tracers
|
|
yield jaxpr, (fwds, out_pvals, pruned_consts, env)
|
|
|
|
|
|
FreeVar = namedtuple('FreeVar', ['val'])
|
|
ConstVar = namedtuple('ConstVar', ['val'])
|
|
LambdaBinding = namedtuple('LambdaBinding', [])
|
|
class JaxprEqnRecipe(NamedTuple):
|
|
eqn_id: Any
|
|
in_tracers: Sequence[JaxprTracer]
|
|
out_tracer_refs: Sequence[ref[JaxprTracer]]
|
|
out_avals: Sequence[core.AbstractValue]
|
|
primitive: Primitive
|
|
params: dict[str, Any]
|
|
effects: core.Effects
|
|
source_info: source_info_util.SourceInfo
|
|
|
|
def new_eqn_recipe(in_tracers: Sequence[JaxprTracer],
|
|
out_tracers: Sequence[JaxprTracer],
|
|
primitive: Primitive,
|
|
params: dict[str, Any],
|
|
effects: core.Effects,
|
|
source_info: source_info_util.SourceInfo
|
|
) -> JaxprEqnRecipe:
|
|
# TODO(necula): move these checks to core.check_jaxpr, and call in more places
|
|
if primitive.call_primitive or primitive.map_primitive:
|
|
assert "call_jaxpr" in params
|
|
assert ("donated_invars" not in params or
|
|
len(params["donated_invars"]) == len(params["call_jaxpr"].invars))
|
|
if primitive.map_primitive:
|
|
assert ("in_axes" in params and
|
|
len(params["in_axes"]) == len(params["call_jaxpr"].invars))
|
|
assert ("donated_invars" in params and
|
|
len(params["donated_invars"]) == len(params["call_jaxpr"].invars))
|
|
out_avals = [core.raise_to_shaped(t.aval) for t in out_tracers]
|
|
return JaxprEqnRecipe(object(), tuple(in_tracers), map(ref, out_tracers),
|
|
out_avals, primitive, params, effects, source_info)
|
|
|
|
|
|
def recipe_to_eqn(getvar: Callable[[JaxprTracer], Atom],
|
|
recipe: JaxprEqnRecipe) -> core.JaxprEqn:
|
|
(_, in_tracers, out_tracer_refs, out_avals, prim, params, eff, src) = recipe
|
|
invars = [getvar(t) for t in in_tracers]
|
|
out_tracers = [t_ref() for t_ref in out_tracer_refs]
|
|
outvars = [DropVar(a) if t is None else getvar(t) # type: ignore
|
|
for a, t in zip(out_avals, out_tracers)]
|
|
return new_jaxpr_eqn(invars, outvars, prim, params, eff, src)
|
|
|
|
def tracers_to_jaxpr(
|
|
in_tracers: Sequence[JaxprTracer],
|
|
out_tracers: Sequence[JaxprTracer]
|
|
) -> tuple[Jaxpr, tuple[Any, ...], tuple[Any, ...]]:
|
|
"""Constructs Jaxpr given tracers for inputs and outputs.
|
|
|
|
Params:
|
|
in_tracers: the tracers that were created for the function inputs
|
|
out_tracers: the tracers that were output by the function.
|
|
|
|
Returns: a triple of a `Jaxpr`, a list of constant values corresponding to
|
|
the `constvars` in the returned Jaxps, and a list of environment values.
|
|
The vars for the environment values have been prepended to the Jaxpr's
|
|
`invars`.
|
|
"""
|
|
gensym = core.gensym()
|
|
|
|
t_to_var: dict[TracerId, Var] = {}
|
|
consts: dict[Var, Any] = {}
|
|
env: dict[Var, JaxprTracer] = {}
|
|
constid_to_var: dict[ConstId, Var] = {} # for deduplication
|
|
|
|
def get_atom(t: JaxprTracer) -> Atom:
|
|
return t.recipe if type(t.recipe) is Literal else t_to_var[id(t)]
|
|
|
|
def newvar(t: JaxprTracer | None) -> Var:
|
|
assert t is not None
|
|
var = gensym(type_substitute(t.aval))
|
|
var_ = t_to_var.setdefault(id(t), var)
|
|
assert var is var_
|
|
return var
|
|
|
|
def type_substitute(aval: AbstractValue) -> AbstractValue:
|
|
if isinstance(aval, DShapedArray):
|
|
# Replace any Tracers in aval.shape with Vars or Literal values
|
|
shape = [get_atom(d) if type(d) is JaxprTracer else d for d in aval.shape]
|
|
shape = [d.val if type(d) is Literal else d for d in shape]
|
|
aval = aval.update(shape=tuple(shape))
|
|
return aval
|
|
|
|
processed_eqn_ids = set()
|
|
eqns: list[core.JaxprEqn] = []
|
|
for t in toposort([*in_tracers, *out_tracers]):
|
|
r = t.recipe
|
|
if isinstance(r, JaxprEqnRecipe):
|
|
# TODO broadcast_in_dim can create a new tracer, not present in parents
|
|
if r.eqn_id not in processed_eqn_ids:
|
|
in_atoms = map(get_atom, r.in_tracers)
|
|
outvars = [DropVar(type_substitute(a)) if rf() is None else newvar(rf())
|
|
for a, rf in zip(r.out_avals, r.out_tracer_refs)]
|
|
eqns.append(new_jaxpr_eqn(in_atoms, outvars, r.primitive, r.params,
|
|
r.effects, r.source_info))
|
|
processed_eqn_ids.add(r.eqn_id)
|
|
elif isinstance(r, LambdaBinding):
|
|
if not any(t is in_tracer for in_tracer in in_tracers):
|
|
raise core.escaped_tracer_error(t, f"Tracer not in input tracers: {t}")
|
|
newvar(t)
|
|
elif isinstance(r, ConstVar):
|
|
var = constid_to_var.get(id(r.val))
|
|
if var is None:
|
|
var = constid_to_var[id(r.val)] = newvar(t)
|
|
consts[var] = r.val
|
|
t_to_var[id(t)] = var
|
|
elif isinstance(r, FreeVar):
|
|
env[newvar(t)] = r.val # type: ignore
|
|
elif isinstance(r, Literal):
|
|
pass
|
|
elif r is None:
|
|
assert False
|
|
else:
|
|
raise TypeError(r)
|
|
|
|
env_vars, env_vals = unzip2(env.items())
|
|
invars = [*env_vars, *map(get_atom, in_tracers)]
|
|
const_vars, const_vals = unzip2(consts.items())
|
|
outvars = map(get_atom, out_tracers) # type: ignore[arg-type]
|
|
jaxpr_effects = make_jaxpr_effects(const_vars, invars, outvars, eqns)
|
|
jaxpr = Jaxpr(const_vars, invars, # type: ignore[list-item,arg-type]
|
|
outvars, eqns, jaxpr_effects)
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr)
|
|
# del getvar # needed to avoid cyclic-reference closure, apparently!
|
|
return jaxpr, const_vals, env_vals
|
|
|
|
@weakref_lru_cache
|
|
def convert_constvars_jaxpr(jaxpr: Jaxpr) -> Jaxpr:
|
|
"""Moves the constvars to the start of invars."""
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr)
|
|
dbg = jaxpr.debug_info and jaxpr.debug_info._replace(
|
|
arg_names=(None,) * len(jaxpr.constvars) + jaxpr.debug_info.arg_names)
|
|
lifted_jaxpr = Jaxpr(constvars=(),
|
|
invars=jaxpr.constvars + jaxpr.invars,
|
|
outvars=jaxpr.outvars, eqns=jaxpr.eqns,
|
|
effects=jaxpr.effects, debug_info=dbg)
|
|
config.jax_enable_checks and core.check_jaxpr(lifted_jaxpr)
|
|
return lifted_jaxpr
|
|
|
|
@weakref_lru_cache
|
|
def convert_invars_to_constvars(jaxpr: Jaxpr, n: int) -> Jaxpr:
|
|
"""Move n invars to constvars. Like an inverse of convert_constvars_Jaxpr."""
|
|
if any(isinstance(eff, effects.JaxprInputEffect) for eff in jaxpr.effects):
|
|
raise NotImplementedError
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr)
|
|
constvars, invars = split_list(jaxpr.invars, [n])
|
|
dbg = jaxpr.debug_info and jaxpr.debug_info._replace(
|
|
arg_names=jaxpr.debug_info.arg_names[n:])
|
|
lifted_jaxpr = jaxpr.replace(constvars=tuple(constvars), invars=invars,
|
|
debug_info=dbg)
|
|
config.jax_enable_checks and core.check_jaxpr(lifted_jaxpr)
|
|
return lifted_jaxpr
|
|
|
|
def convert_envvars_to_constvars(jaxpr: Jaxpr, num_env_vars: int) -> Jaxpr:
|
|
if any(isinstance(eff, effects.JaxprInputEffect) for eff in jaxpr.effects):
|
|
raise NotImplementedError
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr)
|
|
env_vars, invars = split_list(jaxpr.invars, [num_env_vars])
|
|
converted_jaxpr = Jaxpr(constvars=jaxpr.constvars + env_vars,
|
|
invars=invars, outvars=jaxpr.outvars, eqns=jaxpr.eqns,
|
|
effects=jaxpr.effects)
|
|
config.jax_enable_checks and core.check_jaxpr(converted_jaxpr)
|
|
return converted_jaxpr
|
|
|
|
|
|
def partial_eval_jaxpr_nounits(
|
|
jaxpr: ClosedJaxpr, unknowns: Sequence[bool],
|
|
instantiate: bool | Sequence[bool],
|
|
) -> tuple[ClosedJaxpr, ClosedJaxpr, list[bool], list[AbstractValue]]:
|
|
"""Unzip a jaxpr in two by data dependence into 'known' and 'unknown' parts.
|
|
|
|
That is, given a jaxpr and a sequence of booleans indicating which jaxpr
|
|
inputs (i.e. invars) are considered unknown, produce two jaxprs, a list of
|
|
booleans representing which of the original jaxpr's outputs are unknown (i.e.
|
|
have a data dependence on an unknown input), and a list of abstract values
|
|
representing residuals (part of the first jaxpr's output and the second
|
|
jaxpr's input). The two jaxprs result from partitioning the original jaxpr's
|
|
first-order primitive applications based on whether all the inputs to the
|
|
application are known (in which case the application is represented in the
|
|
'known' jaxpr and its result is considered known) or whether any inputs to the
|
|
application are unknown (in which case the application is represented in the
|
|
'unknown' jaxpr and its result is considered unknown). Higher-order primitives
|
|
are recursively unzipped in two.
|
|
|
|
The `instantiate` argument can be used to ensure some outputs are lifted into
|
|
the 'unknown' jaxpr.
|
|
|
|
For example, give an input jaxpr:
|
|
|
|
{ lambda ; a:f32[] b:f32[]. let
|
|
c:f32[] = cos a
|
|
d:f32[] = sin a
|
|
e:f32[] = neg d
|
|
f:f32[] = mul e b
|
|
in (c, f) }
|
|
|
|
then applying this function with `unknowns=[False, True]` and
|
|
`instantiate=False` produces as an output triple:
|
|
|
|
# jaxpr_known
|
|
{ lambda ; a:f32[]. let
|
|
b:f32[] = cos a
|
|
c:f32[] = sin a
|
|
d:f32[] = neg c
|
|
in (b, d) }
|
|
|
|
# jaxpr_unknown
|
|
{ lambda ; a:f32[] b:f32[]. let c:f32[] = mul b a in (c,) }
|
|
|
|
# out_unknowns
|
|
[False, True]
|
|
|
|
Notice in particular that the first output (jaxpr_known) contains all the
|
|
primitive applications which do not have a data dependence on an unknown
|
|
input. Also notice the input and output types: the input type of the first
|
|
jaxpr produced represents the type of the known inputs of the original jaxpr,
|
|
and the output type of the second jaxpr produced represents the type of the
|
|
unknown outputs of the original jaxpr.
|
|
|
|
In the above example, the output of jaxpr_known named `d` is a _residual_
|
|
output, and corresponds to the input named `a` in jaxpr_unknown. In general,
|
|
jaxpr_known will produce extra outputs (at the end of its output list)
|
|
corresponding to intermeidate values of the original jaxpr which must be
|
|
passed to jaxpr_unknown (as leading inputs).
|
|
"""
|
|
instantiate = tuple(instantiate) if isinstance(instantiate, list) else instantiate
|
|
return _partial_eval_jaxpr_nounits(jaxpr, tuple(unknowns), instantiate)
|
|
|
|
@weakref_lru_cache
|
|
def _partial_eval_jaxpr_nounits(jaxpr, in_unknowns, instantiate):
|
|
f = lu.wrap_init(core.jaxpr_as_fun(jaxpr))
|
|
|
|
cell = []
|
|
def fun(*known_vals_in):
|
|
known_vals_in = iter(known_vals_in)
|
|
unknown_avals = (a for a, uk in zip(jaxpr.in_avals, in_unknowns) if uk)
|
|
in_pvals = [PartialVal.unknown(next(unknown_avals)) if uk
|
|
else PartialVal.known(next(known_vals_in)) for uk in in_unknowns]
|
|
assert next(known_vals_in, None) is next(unknown_avals, None) is None
|
|
jaxpr_unknown_, out_pvals, residuals = trace_to_jaxpr_nounits(
|
|
f, in_pvals, instantiate=instantiate)
|
|
jaxpr_unknown = convert_constvars_jaxpr(jaxpr_unknown_)
|
|
out_unknowns = [not pval.is_known() for pval in out_pvals]
|
|
res_avals = [core.raise_to_shaped(core.get_aval(r)) for r in residuals]
|
|
cell.append((out_unknowns, jaxpr_unknown, res_avals))
|
|
known_vals_out = [pval.get_known() for pval in out_pvals if pval.is_known()]
|
|
return [*known_vals_out, *residuals]
|
|
|
|
known_avals = [a for a, uk in zip(jaxpr.in_avals, in_unknowns) if not uk]
|
|
jaxpr_known, _, consts_known = trace_to_jaxpr_dynamic(lu.wrap_init(fun), known_avals)
|
|
(out_unknowns, jaxpr_unknown, res_avals), = cell # pytype: disable=bad-unpacking
|
|
|
|
# check jaxpr_known and jaxpr_unknown in isolation
|
|
# TODO(mattjj): enable weak type checking here
|
|
if config.jax_enable_checks:
|
|
core.check_jaxpr(jaxpr_known)
|
|
core.check_jaxpr(jaxpr_unknown)
|
|
# check jaxpr_known has input type corresponding to known inputs of jaxpr
|
|
assert ([v.aval for v in jaxpr_known.invars] ==
|
|
[a for a, uk in zip(jaxpr.in_avals, in_unknowns) if not uk])
|
|
# check jaxpr_known has out type corresponding to known outs of jaxpr plus res
|
|
assert ([v.aval.strip_weak_type() for v in jaxpr_known.outvars] ==
|
|
[a.strip_weak_type() for a, uk in zip(jaxpr.out_avals, out_unknowns)
|
|
if not uk] + [a.strip_weak_type() for a in res_avals])
|
|
# check jaxpr_unknown has input type corresponding to res plus unknown inputs
|
|
assert ([v.aval.strip_weak_type() for v in jaxpr_unknown.invars] ==
|
|
[a.strip_weak_type() for a in res_avals] +
|
|
[a.strip_weak_type() for a, uk in zip(jaxpr.in_avals, in_unknowns)
|
|
if uk])
|
|
# check jaxpr_unknown has output type corresponding to unknown outputs
|
|
assert ([v.aval.strip_weak_type() for v in jaxpr_unknown.outvars] ==
|
|
[a.strip_weak_type() for a, uk in zip(jaxpr.out_avals, out_unknowns)
|
|
if uk])
|
|
|
|
closed_jaxpr_known = ClosedJaxpr(jaxpr_known, consts_known)
|
|
closed_jaxpr_unknown = ClosedJaxpr(jaxpr_unknown, ())
|
|
return closed_jaxpr_known, closed_jaxpr_unknown, out_unknowns, res_avals
|
|
|
|
|
|
def partial_eval_jaxpr_custom(
|
|
jaxpr: Jaxpr,
|
|
in_unknowns: Sequence[bool],
|
|
in_inst: bool | Sequence[bool],
|
|
ensure_out_unknowns: bool | Sequence[bool],
|
|
ensure_out_inst: bool | Sequence[bool],
|
|
saveable: Callable[..., RematCases_],
|
|
) -> tuple[Jaxpr, Jaxpr, list[bool], list[bool], int]:
|
|
if type(in_inst) is bool:
|
|
in_inst = (in_inst,) * len(jaxpr.invars)
|
|
if type(ensure_out_unknowns) is bool:
|
|
ensure_out_unknowns = (ensure_out_unknowns,) * len(jaxpr.outvars)
|
|
if type(ensure_out_inst) is bool:
|
|
ensure_out_inst = (ensure_out_inst,) * len(jaxpr.outvars)
|
|
jaxpr_known, jaxpr_staged, out_unknowns, out_inst, num_res, num_res_ref = \
|
|
_partial_eval_jaxpr_custom_cached(jaxpr, tuple(in_unknowns),
|
|
tuple(in_inst),
|
|
tuple(ensure_out_unknowns),
|
|
tuple(ensure_out_inst), saveable)
|
|
if num_res_ref > 0:
|
|
raise ValueError(
|
|
"Cannot use `partial_eval_jaxpr_custom` with stateful jaxprs.")
|
|
return jaxpr_known, jaxpr_staged, out_unknowns, out_inst, num_res
|
|
|
|
def partial_eval_jaxpr_stateful(
|
|
jaxpr: Jaxpr,
|
|
in_unknowns: Sequence[bool],
|
|
in_inst: bool | Sequence[bool],
|
|
ensure_out_unknowns: bool | Sequence[bool],
|
|
ensure_out_inst: bool | Sequence[bool],
|
|
saveable: Callable[..., RematCases_],
|
|
) -> tuple[Jaxpr, Jaxpr, list[bool], list[bool], int, int]:
|
|
if type(in_inst) is bool:
|
|
in_inst = (in_inst,) * len(jaxpr.invars)
|
|
if type(ensure_out_unknowns) is bool:
|
|
ensure_out_unknowns = (ensure_out_unknowns,) * len(jaxpr.outvars)
|
|
if type(ensure_out_inst) is bool:
|
|
ensure_out_inst = (ensure_out_inst,) * len(jaxpr.outvars)
|
|
jaxpr_known, jaxpr_staged, out_unknowns, out_inst, num_res, num_res_ref = \
|
|
_partial_eval_jaxpr_custom_cached(jaxpr, tuple(in_unknowns),
|
|
tuple(in_inst),
|
|
tuple(ensure_out_unknowns),
|
|
tuple(ensure_out_inst), saveable)
|
|
return jaxpr_known, jaxpr_staged, out_unknowns, out_inst, num_res, num_res_ref
|
|
|
|
@weakref_lru_cache
|
|
def _partial_eval_jaxpr_custom_cached(
|
|
jaxpr: Jaxpr,
|
|
in_unknowns: tuple[bool, ...],
|
|
in_inst: tuple[bool, ...],
|
|
ensure_out_unknowns: tuple[bool, ...],
|
|
ensure_out_inst: tuple[bool, ...],
|
|
saveable: Callable[..., RematCases_],
|
|
) -> tuple[Jaxpr, Jaxpr, list[bool], list[bool], int, int]:
|
|
env: dict[Var, tuple[bool, bool]] = {}
|
|
residuals: OrderedSet[Var] = OrderedSet()
|
|
residual_refs: OrderedSet[Var] = OrderedSet()
|
|
|
|
def read(x: Atom) -> tuple[bool, bool]:
|
|
if type(x) is Var:
|
|
return env[x]
|
|
return (False, True)
|
|
|
|
def write(unk: bool, inst: bool, v: Var) -> None:
|
|
assert (unk, inst) != (True, False)
|
|
env[v] = (unk, inst)
|
|
|
|
def ensure_instantiated(inst: bool, x: Atom) -> Atom:
|
|
if type(x) is Var and not inst:
|
|
residuals.add(x)
|
|
return x
|
|
|
|
newvar = core.gensym(suffix='_offload')
|
|
known_eqns, staged_eqns = [], []
|
|
map(write, in_unknowns, in_inst, jaxpr.invars)
|
|
map(partial(write, False, True), jaxpr.constvars)
|
|
for eqn in jaxpr.eqns:
|
|
unks_in, inst_in = unzip2(map(read, eqn.invars))
|
|
rule = partial_eval_jaxpr_custom_rules.get(eqn.primitive)
|
|
if rule:
|
|
eqn1, eqn2, unks_out, inst_out, res = rule(saveable, unks_in, inst_in, eqn)
|
|
eqn1 and known_eqns.append(eqn1); eqn2 and staged_eqns.append(eqn2) # type: ignore
|
|
for r in res:
|
|
if isinstance(r.aval, AbstractRef):
|
|
residual_refs.add(r)
|
|
else:
|
|
residuals.add(r)
|
|
map(write, unks_out, inst_out, eqn.outvars)
|
|
elif any(unks_in):
|
|
inputs = map(ensure_instantiated, inst_in, eqn.invars)
|
|
staged_eqns.append(eqn.replace(invars=inputs))
|
|
map(partial(write, True, True), eqn.outvars)
|
|
else:
|
|
known_eqns.append(eqn)
|
|
# If it's an effectful primitive, we always to run and avoid staging it.
|
|
policy = ensure_enum(saveable(
|
|
eqn.primitive, *[x.aval for x in eqn.invars], **eqn.params))
|
|
if eqn.effects or isinstance(policy, SaveableType):
|
|
map(partial(write, False, False), eqn.outvars)
|
|
elif isinstance(policy, Offloadable):
|
|
from jax._src.dispatch import device_put_p, TransferToMemoryKind # type: ignore
|
|
resvars = [newvar(v.aval) for v in eqn.outvars]
|
|
offload_eqn = core.JaxprEqn(
|
|
eqn.outvars, resvars, device_put_p,
|
|
dict(device=TransferToMemoryKind(policy.dst), src=None),
|
|
set(), source_info_util.new_source_info())
|
|
known_eqns.append(offload_eqn)
|
|
# resvars are known and available in the backward jaxpr.
|
|
map(partial(write, False, True), resvars)
|
|
residuals.update(resvars)
|
|
reload_eqn = core.JaxprEqn(
|
|
resvars, eqn.outvars, device_put_p, # type: ignore
|
|
dict(device=TransferToMemoryKind(policy.src), src=None),
|
|
set(), source_info_util.new_source_info())
|
|
staged_eqns.append(reload_eqn)
|
|
# outvars are known and available in the backward jaxpr.
|
|
map(partial(write, False, True), eqn.outvars)
|
|
else:
|
|
assert isinstance(policy, RecomputeType)
|
|
inputs = map(ensure_instantiated, inst_in, eqn.invars)
|
|
staged_eqns.append(eqn.replace(invars=inputs))
|
|
map(partial(write, False, True), eqn.outvars)
|
|
unzipped = unzip2(map(read, jaxpr.outvars))
|
|
out_unknowns, out_inst = list(unzipped[0]), list(unzipped[1])
|
|
assert all(type(v) is Var for v in residuals), residuals
|
|
|
|
for x, inst, ensure_inst in zip(jaxpr.outvars, out_inst, ensure_out_inst):
|
|
if ensure_inst: ensure_instantiated(inst, x)
|
|
out_unknowns = map(op.or_, out_unknowns, ensure_out_unknowns)
|
|
out_inst = map(op.or_, out_inst, ensure_out_inst)
|
|
|
|
ins_known, _ = partition_list(in_unknowns, jaxpr.invars)
|
|
outs_known, _ = partition_list(out_unknowns, jaxpr.outvars)
|
|
ref_res_is_input = [r in ins_known for r in residual_refs]
|
|
non_input_res_refs, _ = partition_list(ref_res_is_input, list(residual_refs))
|
|
ins_known_and_ref_res = [*ins_known, *non_input_res_refs]
|
|
known_outvars = [*outs_known, *residuals]
|
|
known_effects = make_jaxpr_effects(jaxpr.constvars, ins_known_and_ref_res,
|
|
known_outvars, known_eqns)
|
|
jaxpr_known = Jaxpr(jaxpr.constvars, ins_known_and_ref_res, known_outvars,
|
|
known_eqns, known_effects)
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr_known)
|
|
|
|
_, ins_staged = partition_list(in_inst, jaxpr.invars)
|
|
_, outs_staged = partition_list(out_inst, jaxpr.outvars)
|
|
staged_invars = [*residuals, *non_input_res_refs, *ins_staged]
|
|
staged_effects = make_jaxpr_effects(jaxpr.constvars, staged_invars,
|
|
outs_staged, staged_eqns)
|
|
jaxpr_staged = Jaxpr(jaxpr.constvars, staged_invars,
|
|
outs_staged, staged_eqns, staged_effects)
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr_staged)
|
|
|
|
return (jaxpr_known, jaxpr_staged, out_unknowns, out_inst, len(residuals),
|
|
len(non_input_res_refs))
|
|
|
|
|
|
MemoryKind = str
|
|
|
|
class RecomputeType: pass
|
|
Recompute = RecomputeType()
|
|
|
|
class SaveableType: pass
|
|
Saveable = SaveableType()
|
|
|
|
class Offloadable(NamedTuple):
|
|
src: MemoryKind
|
|
dst: MemoryKind
|
|
|
|
RematCases = Union[RecomputeType, SaveableType, Offloadable]
|
|
RematCases_ = Union[RematCases, bool]
|
|
|
|
def ensure_enum(case: bool | RematCases) -> RematCases:
|
|
if isinstance(case, bool):
|
|
return Saveable if case else Recompute
|
|
return case
|
|
|
|
# A primitive rule for policy-driven partial evaluation returns a 5-tuple
|
|
# with the components representing, respectively:
|
|
# * the JaxprEqn for the 'known' side (or None if there is no known component),
|
|
# * the JaxprEqn for the 'unknown' side (or None),
|
|
# * a list of booleans indicating which of the original outputs are unknown,
|
|
# * a list of booleans indicating which of the original outputs are
|
|
# instantiated (i.e. available) in the 'unknown' side,
|
|
# * a list of Var instances representing residuals to be added (i.e. to be
|
|
# plumbed as outputs of the 'known' side jaxpr and added as input binders to
|
|
# the 'unknown' jaxpr).
|
|
PartialEvalCustomResult = tuple[Optional[JaxprEqn], Optional[JaxprEqn],
|
|
Sequence[bool], Sequence[bool], list[Var]]
|
|
PartialEvalCustomRule = Callable[
|
|
[Callable[..., RematCases_], Sequence[bool], Sequence[bool], JaxprEqn],
|
|
PartialEvalCustomResult]
|
|
partial_eval_jaxpr_custom_rules: dict[Primitive, PartialEvalCustomRule] = {}
|
|
|
|
def partial_eval_jaxpr_custom_rule_not_implemented(
|
|
name: str, saveable: Callable[..., RematCases_], unks_in: Sequence[bool],
|
|
inst_in: Sequence[bool], eqn: JaxprEqn) -> PartialEvalCustomResult:
|
|
msg = (f'custom-policy remat rule not implemented for {name}, '
|
|
'open a feature request at https://github.com/google/jax/issues!')
|
|
raise NotImplementedError(msg)
|
|
|
|
|
|
ParamsUpdater = Callable[[Sequence[bool], Sequence[bool], Sequence[bool],
|
|
Sequence[bool], int, dict, dict],
|
|
tuple[dict, dict]]
|
|
ResAvalUpdater = Callable[[dict[str, Any], AbstractValue], AbstractValue]
|
|
def _default_res_aval_updater(
|
|
params: dict[str, Any], aval: AbstractValue) -> AbstractValue:
|
|
return aval
|
|
|
|
@contextmanager
|
|
def trivial_ctx(_): yield
|
|
|
|
def call_partial_eval_custom_rule(
|
|
jaxpr_param_name: str, params_updater: ParamsUpdater,
|
|
saveable: Callable[..., RematCases_], unks_in: list[bool], inst_in: list[bool],
|
|
eqn: JaxprEqn, *, res_aval: ResAvalUpdater = _default_res_aval_updater,
|
|
ctx: Callable[[core.ParamDict], AbstractContextManager[None]] = trivial_ctx,
|
|
) -> tuple[JaxprEqn, JaxprEqn, Sequence[bool], Sequence[bool], list[Var]]:
|
|
jaxpr = eqn.params[jaxpr_param_name]
|
|
with ctx(eqn.params):
|
|
jaxpr_known, jaxpr_staged, unks_out, inst_out, num_res = \
|
|
partial_eval_jaxpr_custom(jaxpr, unks_in, inst_in, False, False, saveable)
|
|
ins_known, _ = partition_list(unks_in, eqn.invars)
|
|
out_binders_known, _ = partition_list(unks_out, eqn.outvars)
|
|
_, ins_staged = partition_list(inst_in, eqn.invars)
|
|
_, out_binders_staged = partition_list(inst_out, eqn.outvars)
|
|
newvar = core.gensym([jaxpr_known, jaxpr_staged])
|
|
params_known = {**eqn.params, jaxpr_param_name: jaxpr_known}
|
|
params_staged = {**eqn.params, jaxpr_param_name: jaxpr_staged}
|
|
params_known, params_staged = params_updater(
|
|
unks_in, inst_in, map(op.not_, unks_out), inst_out, num_res, params_known,
|
|
params_staged)
|
|
residuals = [newvar(res_aval(params_known, var.aval))
|
|
for var in jaxpr_staged.invars[:num_res]]
|
|
eqn_known = new_jaxpr_eqn(ins_known, [*out_binders_known, *residuals],
|
|
eqn.primitive, params_known, jaxpr_known.effects, eqn.source_info)
|
|
eqn_staged = new_jaxpr_eqn([*residuals, *ins_staged], out_binders_staged,
|
|
eqn.primitive, params_staged,
|
|
jaxpr_staged.effects, eqn.source_info)
|
|
assert len(eqn_staged.invars) == len(jaxpr_staged.invars)
|
|
new_inst = [x for x, inst in zip(eqn.invars, inst_in)
|
|
if type(x) is Var and not inst]
|
|
return eqn_known, eqn_staged, unks_out, inst_out, new_inst + residuals
|
|
|
|
def closed_call_partial_eval_custom_rule(
|
|
jaxpr_param_name: str, params_updater: ParamsUpdater,
|
|
saveable: Callable[..., RematCases_], unks_in: list[bool], inst_in: list[bool],
|
|
eqn: JaxprEqn, *, res_aval: ResAvalUpdater = _default_res_aval_updater,
|
|
) -> tuple[JaxprEqn, JaxprEqn, Sequence[bool], Sequence[bool], list[Var]]:
|
|
# TODO(sharadmv,mattjj): dedup this rule with call_partial_eval_custom_rule.
|
|
closed_jaxpr = eqn.params[jaxpr_param_name]
|
|
jaxpr_known_, jaxpr_staged_, unks_out, inst_out, num_res_out, num_res_ref = \
|
|
partial_eval_jaxpr_stateful(closed_jaxpr.jaxpr, unks_in, inst_in,
|
|
False, False, saveable)
|
|
num_res = num_res_ref + num_res_out
|
|
# Forming these fresh ClosedJaxprs defeats caching, but caller handles caching
|
|
jaxpr_known = core.ClosedJaxpr(jaxpr_known_, closed_jaxpr.consts)
|
|
jaxpr_staged = core.ClosedJaxpr(jaxpr_staged_, closed_jaxpr.consts)
|
|
ins_known, _ = partition_list(unks_in, eqn.invars)
|
|
out_binders_known, _ = partition_list(unks_out, eqn.outvars)
|
|
_, ins_staged = partition_list(inst_in, eqn.invars)
|
|
_, out_binders_staged = partition_list(inst_out, eqn.outvars)
|
|
newvar = core.gensym([jaxpr_known.jaxpr, jaxpr_staged.jaxpr])
|
|
params_known = {**eqn.params, jaxpr_param_name: jaxpr_known}
|
|
params_staged = {**eqn.params, jaxpr_param_name: jaxpr_staged}
|
|
params_known, params_staged = params_updater(
|
|
unks_in, inst_in, map(op.not_, unks_out), inst_out, num_res, params_known,
|
|
params_staged)
|
|
residuals, ref_residuals = split_list(
|
|
[newvar(res_aval(params_known, v)) for v
|
|
in jaxpr_staged.in_avals[:num_res]], [num_res_out])
|
|
eqn_known = new_jaxpr_eqn([*ins_known, *ref_residuals],
|
|
[*out_binders_known, *residuals],
|
|
eqn.primitive, params_known, jaxpr_known.effects,
|
|
eqn.source_info)
|
|
eqn_staged = new_jaxpr_eqn([*residuals, *ref_residuals, *ins_staged],
|
|
out_binders_staged,
|
|
eqn.primitive, params_staged, jaxpr_staged.effects,
|
|
eqn.source_info)
|
|
assert len(eqn_staged.invars) == len(jaxpr_staged.in_avals)
|
|
assert len(ins_known) + len(ref_residuals) == len(jaxpr_known.jaxpr.invars)
|
|
assert len(ins_staged) + len(ref_residuals) + len(residuals) == len(jaxpr_staged.jaxpr.invars)
|
|
assert len(out_binders_known) + len(residuals) == len(jaxpr_known.jaxpr.outvars)
|
|
new_inst = [x for x, inst in zip(eqn.invars, inst_in)
|
|
if type(x) is Var and not inst]
|
|
new_vars = [*new_inst, *residuals, *ref_residuals]
|
|
return eqn_known, eqn_staged, unks_out, inst_out, new_vars
|
|
|
|
partial_eval_jaxpr_custom_rules[core.call_p] = \
|
|
partial(call_partial_eval_custom_rule, 'call_jaxpr',
|
|
lambda _, __, ___, ____, _____, x, y: (x, y))
|
|
partial_eval_jaxpr_custom_rules[core.closed_call_p] = \
|
|
partial(closed_call_partial_eval_custom_rule, 'call_jaxpr',
|
|
lambda _, __, ___, ____, _____, x, y: (x, y))
|
|
|
|
|
|
def _jaxpr_forwarding(jaxpr: Jaxpr) -> list[int | None]:
|
|
# Compute which inputs are just forwarded to outputs.
|
|
fwds: dict[Var, Var] = dict(zip(jaxpr.invars, jaxpr.invars))
|
|
for eqn in jaxpr.eqns:
|
|
if eqn.primitive in forwarding_rules:
|
|
eqn = eqn.replace(invars=[a if type(a) is Literal else fwds.get(a, a) # type: ignore
|
|
for a in eqn.invars])
|
|
fwd_vars, _ = forwarding_rules[eqn.primitive](eqn)
|
|
for v_orig, v_new in zip(eqn.outvars, fwd_vars):
|
|
if v_new is not None:
|
|
fwds[v_orig] = v_new
|
|
idxs: dict[Var, int] = {v: i for i, v in enumerate(jaxpr.invars)}
|
|
return [None if type(v) is Literal else idxs.get(fwds.get(v)) # type: ignore
|
|
for v in jaxpr.outvars]
|
|
|
|
|
|
def dce_jaxpr(jaxpr: Jaxpr, used_outputs: Sequence[bool],
|
|
instantiate: bool | Sequence[bool] = False,
|
|
) -> tuple[Jaxpr, list[bool]]:
|
|
if type(instantiate) is bool:
|
|
instantiate = (instantiate,) * len(jaxpr.invars)
|
|
return _dce_jaxpr(jaxpr, tuple(used_outputs), tuple(instantiate))
|
|
|
|
|
|
def dce_jaxpr_consts(jaxpr: Jaxpr, used_outputs: Sequence[bool],
|
|
instantiate: bool | Sequence[bool] = False,
|
|
) -> tuple[Jaxpr, list[bool], list[bool]]:
|
|
jaxpr_ = convert_constvars_jaxpr(jaxpr)
|
|
new_jaxpr_, used_inputs_ = dce_jaxpr(jaxpr_, used_outputs)
|
|
used_consts, used_inputs = split_list(used_inputs_, [len(jaxpr.constvars)])
|
|
new_jaxpr = convert_invars_to_constvars(new_jaxpr_, sum(used_consts))
|
|
return new_jaxpr, used_consts, used_inputs
|
|
|
|
|
|
@weakref_lru_cache
|
|
def _dce_jaxpr(jaxpr: Jaxpr, used_outputs: tuple[bool, ...],
|
|
instantiate: tuple[bool, ...]
|
|
) -> tuple[Jaxpr, list[bool]]:
|
|
env: dict[Var, bool] = {}
|
|
|
|
def read(v: Var) -> bool:
|
|
return env.get(v, False)
|
|
|
|
def write(x: Atom, b: bool) -> None:
|
|
if type(x) is Var:
|
|
env[x] = read(x) or b
|
|
|
|
def has_effects(e: JaxprEqn) -> bool:
|
|
return bool(e.effects) or core.primitive_uses_outfeed(e.primitive, e.params)
|
|
|
|
new_eqns = []
|
|
map(write, jaxpr.outvars, used_outputs)
|
|
for eqn in jaxpr.eqns[::-1]:
|
|
used_outs = map(read, eqn.outvars)
|
|
if not any(used_outs) and not has_effects(eqn):
|
|
used_ins = [False] * len(eqn.invars)
|
|
else:
|
|
rule = dce_rules.get(eqn.primitive, _default_dce_rule)
|
|
used_ins, new_eqn = rule(used_outs, eqn)
|
|
if new_eqn is not None:
|
|
new_eqns.append(new_eqn)
|
|
map(write, eqn.invars, used_ins)
|
|
used_inputs = map(read, jaxpr.invars)
|
|
used_inputs = map(op.or_, instantiate, used_inputs)
|
|
|
|
invars = [v for v, b in zip(jaxpr.invars, used_inputs) if b]
|
|
outvars = [v for v, b in zip(jaxpr.outvars, used_outputs) if b]
|
|
eqns = new_eqns[::-1]
|
|
jaxpr_effects = make_jaxpr_effects(jaxpr.constvars, invars, outvars, eqns)
|
|
|
|
dbg = jaxpr.debug_info and core.JaxprDebugInfo(
|
|
jaxpr.debug_info.traced_for, jaxpr.debug_info.func_src_info,
|
|
tuple(v for v, b in zip(jaxpr.debug_info.arg_names, used_inputs) if b),
|
|
tuple(v for v, b in zip(jaxpr.debug_info.result_paths, used_outputs) if b))
|
|
new_jaxpr = Jaxpr(jaxpr.constvars, invars, outvars, eqns, jaxpr_effects, dbg)
|
|
config.jax_enable_checks and core.check_jaxpr(new_jaxpr)
|
|
|
|
return new_jaxpr, used_inputs
|
|
|
|
DCERule = Callable[[list[bool], JaxprEqn], tuple[list[bool], Optional[JaxprEqn]]]
|
|
|
|
def _default_dce_rule(
|
|
used_outs: list[bool], eqn: JaxprEqn
|
|
) -> tuple[list[bool], JaxprEqn]:
|
|
return [True] * len(eqn.invars), eqn
|
|
|
|
dce_rules: dict[Primitive, DCERule] = {}
|
|
|
|
|
|
def dce_jaxpr_call_rule(used_outputs: list[bool], eqn: JaxprEqn
|
|
) -> tuple[list[bool], JaxprEqn | None]:
|
|
new_jaxpr, used_inputs = dce_jaxpr(eqn.params['call_jaxpr'], used_outputs)
|
|
new_params = dict(eqn.params, call_jaxpr=new_jaxpr)
|
|
update_params = call_param_updaters.get(eqn.primitive)
|
|
if update_params:
|
|
new_params = update_params(new_params, used_inputs, 0)
|
|
if not any(used_inputs) and not any(used_outputs) and not new_jaxpr.effects:
|
|
return used_inputs, None
|
|
else:
|
|
new_eqn = new_jaxpr_eqn(
|
|
[v for v, used in zip(eqn.invars, used_inputs) if used],
|
|
[v for v, used in zip(eqn.outvars, used_outputs) if used],
|
|
eqn.primitive, new_params, new_jaxpr.effects, eqn.source_info)
|
|
return used_inputs, new_eqn
|
|
dce_rules[core.call_p] = dce_jaxpr_call_rule
|
|
|
|
|
|
def dce_jaxpr_closed_call_rule(used_outputs: list[bool], eqn: JaxprEqn
|
|
) -> tuple[list[bool], JaxprEqn]:
|
|
# TODO(mattjj): de-duplicate with above rule?
|
|
jaxpr_ = eqn.params['call_jaxpr']
|
|
jaxpr, consts = jaxpr_.jaxpr, jaxpr_.consts
|
|
new_jaxpr, used_inputs = dce_jaxpr(jaxpr, used_outputs)
|
|
new_params = dict(eqn.params, call_jaxpr=core.ClosedJaxpr(new_jaxpr, consts))
|
|
new_eqn = new_jaxpr_eqn(
|
|
[v for v, used in zip(eqn.invars, used_inputs) if used],
|
|
[v for v, used in zip(eqn.outvars, used_outputs) if used],
|
|
eqn.primitive, new_params, new_jaxpr.effects, eqn.source_info)
|
|
return used_inputs, new_eqn
|
|
dce_rules[core.closed_call_p] = dce_jaxpr_closed_call_rule
|
|
|
|
@weakref_lru_cache
|
|
def close_jaxpr(jaxpr: Jaxpr) -> ClosedJaxpr:
|
|
return ClosedJaxpr(jaxpr, ())
|
|
|
|
def move_binders_to_front(closed_jaxpr: ClosedJaxpr, to_move: Sequence[bool]
|
|
) -> ClosedJaxpr:
|
|
"""Reorder `invars` by moving those indicated in `to_move` to the front."""
|
|
return _move_binders_to_front(closed_jaxpr, tuple(to_move))
|
|
|
|
@weakref_lru_cache
|
|
def _move_binders_to_front(closed_jaxpr: ClosedJaxpr, to_move: tuple[bool, ...]
|
|
) -> ClosedJaxpr:
|
|
assert len(closed_jaxpr.in_avals) == len(to_move)
|
|
new_invars = _move_to_front(closed_jaxpr.jaxpr.invars, to_move)
|
|
new_jaxpr = Jaxpr(closed_jaxpr.jaxpr.constvars, new_invars,
|
|
closed_jaxpr.jaxpr.outvars, closed_jaxpr.jaxpr.eqns,
|
|
closed_jaxpr.jaxpr.effects)
|
|
new_closed_jaxpr = core.ClosedJaxpr(new_jaxpr, closed_jaxpr.consts)
|
|
return new_closed_jaxpr
|
|
|
|
def _move_to_front(lst: Sequence, to_move: Sequence[bool]) -> Sequence:
|
|
return ([elt for elt, move in zip(lst, to_move) if move] +
|
|
[elt for elt, move in zip(lst, to_move) if not move])
|
|
|
|
def move_binders_to_back(closed_jaxpr: ClosedJaxpr, to_move: Sequence[bool]
|
|
) -> ClosedJaxpr:
|
|
"""Reorder `invars` by moving those indicated in `to_move` to the back."""
|
|
return move_binders_to_front(closed_jaxpr, map(op.not_, to_move))
|
|
|
|
class DynamicJaxprTracer(core.Tracer):
|
|
__slots__ = ['aval', '_debug_info']
|
|
|
|
def __init__(self, trace, aval, line_info=None):
|
|
self._trace = trace
|
|
self._line_info = line_info
|
|
# Needed for UnexpectedTracerError.
|
|
self._debug_info = self._trace.frame.debug_info
|
|
self.aval = aval
|
|
|
|
def full_lower(self):
|
|
return self
|
|
|
|
def _contents(self):
|
|
return ()
|
|
|
|
def _origin_msg(self):
|
|
if not self._trace.main.jaxpr_stack: # type: ignore
|
|
# If this Tracer has been leaked the jaxpr stack may no longer be
|
|
# available. So we can't print as much origin information.
|
|
return ("\nThis DynamicJaxprTracer was created on line "
|
|
f"{source_info_util.summarize(self._line_info)}")
|
|
else:
|
|
invar_pos, progenitor_eqns = self._trace.frame.find_progenitors(self)
|
|
dbg = self._debug_info
|
|
if dbg is None:
|
|
return ""
|
|
|
|
origin = ("The error occurred while tracing the function "
|
|
f"{dbg.func_src_info or '<unknown>'} for {dbg.traced_for}. ")
|
|
arg_info = arg_info_all(dbg)
|
|
if invar_pos and arg_info:
|
|
arg_info = [arg_info[i] for i in invar_pos]
|
|
arg_names = [f'{name}{keystr(path)}' for name, path in arg_info]
|
|
if len(arg_names) == 1:
|
|
arg_info_str = f"the argument {arg_names[0]}"
|
|
elif len(arg_names) == 2:
|
|
arg_info_str = f"the arguments {arg_names[0]} and {arg_names[1]}"
|
|
else:
|
|
*rest, last = arg_names
|
|
arg_info_str = f"the arguments {', '.join(rest)}, and {last}"
|
|
origin += ("This concrete value was not available in Python because it "
|
|
f"depends on the value{'s' if len(invar_pos) > 1 else ''} "
|
|
f"of {arg_info_str}.")
|
|
elif progenitor_eqns:
|
|
msts = [" operation "
|
|
f"{core.pp_eqn(eqn, core.JaxprPpContext(), core.JaxprPpSettings(print_shapes=True))}\n"
|
|
f" from line {source_info_util.summarize(eqn.source_info)}"
|
|
for eqn in progenitor_eqns[:5]] # show at most 5
|
|
origin += ("This value became a tracer due to JAX operations on these lines:"
|
|
"\n\n" + "\n\n".join(msts))
|
|
if len(progenitor_eqns) > 5:
|
|
origin += "\n\n(Additional originating lines are not shown.)"
|
|
return "\n" + origin
|
|
|
|
def _assert_live(self) -> None:
|
|
if not self._trace.main.jaxpr_stack: # type: ignore
|
|
raise core.escaped_tracer_error(self, None)
|
|
|
|
def get_referent(self):
|
|
frame = self._trace.frame
|
|
val = frame.constvar_to_val.get(frame.tracer_to_var.get(id(self)))
|
|
return self if val is None else get_referent(val)
|
|
api_util._shaped_abstractify_handlers[DynamicJaxprTracer] = op.attrgetter("aval")
|
|
|
|
def make_jaxpr_effects(constvars, invars, outvars, eqns) -> effects.Effects:
|
|
jaxpr_effects = set()
|
|
all_vars = [*constvars, *invars]
|
|
for eqn in eqns:
|
|
for eff in eqn.effects:
|
|
if isinstance(eff, effects.JaxprInputEffect):
|
|
if eff.input_index >= len(eqn.invars):
|
|
raise ValueError(
|
|
f"`JaxprInputEffect` {eff} is invalid."
|
|
f"\n Equation: {eqn}\n"
|
|
"\n Jaxpr: "
|
|
f"{core.Jaxpr(constvars, invars, outvars, eqns, set())}")
|
|
invar = eqn.invars[eff.input_index]
|
|
if invar not in all_vars:
|
|
raise ValueError(
|
|
f"`JaxprInputEffect` {eff} does not have "
|
|
f"corresponding input: {invar}."
|
|
f"\n Equation: {eqn}\n"
|
|
"\n Jaxpr: "
|
|
f"{core.Jaxpr(constvars, invars, outvars, eqns, set())}")
|
|
eff = eff.replace(input_index=all_vars.index(invar))
|
|
jaxpr_effects.add(eff)
|
|
return jaxpr_effects
|
|
|
|
|
|
class JaxprStackFrame:
|
|
gensym: Callable[[AbstractValue], Var]
|
|
tracer_to_var: dict[TracerId, Var]
|
|
constid_to_tracer: dict[ConstId, Tracer]
|
|
constvar_to_val: dict[Var, Any]
|
|
tracers: list[DynamicJaxprTracer] # hold onto strong refs for all tracers
|
|
eqns: list[JaxprEqn]
|
|
invars: list[Var]
|
|
effects: core.Effects
|
|
debug_info: DebugInfo | None
|
|
|
|
def __init__(self):
|
|
self.gensym = core.gensym()
|
|
self.tracer_to_var = {}
|
|
self.constid_to_tracer = {}
|
|
self.constvar_to_val = {}
|
|
self.tracers = [] # circ refs, frame->tracer->trace->main->frame,
|
|
self.eqns = [] # cleared when we pop frame from main
|
|
self.invars = []
|
|
self.effects = set()
|
|
self.debug_info = None
|
|
|
|
def add_eqn(self, eqn: core.JaxprEqn):
|
|
self.eqns.append(eqn)
|
|
|
|
def to_jaxpr(self, out_tracers: Sequence[Tracer]) -> tuple[Jaxpr, list[Any]]:
|
|
# It's not necessary, but we keep the tracer-to-var mapping injective:
|
|
assert len(self.tracer_to_var) == len(set(self.tracer_to_var.values()))
|
|
outvars = [self.tracer_to_var[id(t)] for t in out_tracers]
|
|
constvals: Sequence[Any]
|
|
constvars, constvals = unzip2(self.constvar_to_val.items())
|
|
jaxpr_effects = make_jaxpr_effects(constvars, self.invars, outvars,
|
|
self.eqns)
|
|
jaxpr = Jaxpr(constvars, self.invars, outvars, self.eqns, jaxpr_effects)
|
|
jaxpr, constvals = _const_folding_and_forwarding(jaxpr, constvals)
|
|
jaxpr, constvals = _inline_literals(jaxpr, constvals)
|
|
return jaxpr, list(constvals)
|
|
|
|
def to_jaxpr2(self, out_tracers):
|
|
# It's not necessary, but we keep the tracer-to-var mapping injective:
|
|
assert len(self.tracer_to_var) == len(set(self.tracer_to_var.values()))
|
|
constvars, constvals = unzip2(self.constvar_to_val.items())
|
|
expl_outvars = [self.tracer_to_var[id(t)] for t in out_tracers]
|
|
jaxpr_effects = make_jaxpr_effects(constvars, self.invars, expl_outvars,
|
|
self.eqns)
|
|
jaxpr = Jaxpr(constvars, self.invars, expl_outvars, self.eqns,
|
|
jaxpr_effects)
|
|
# We can't run check_jaxpr until after we normalize.
|
|
jaxpr, constvals = _const_folding_and_forwarding(jaxpr, constvals)
|
|
jaxpr, constvals = _inline_literals(jaxpr, constvals)
|
|
jaxpr, out_type = _add_implicit_outputs(jaxpr)
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr)
|
|
return jaxpr, out_type, constvals
|
|
|
|
def newvar(self, aval):
|
|
if isinstance(aval, DShapedArray):
|
|
# this aval may have tracers in it, so we replace those with variables
|
|
new_shape = [self.tracer_to_var[id(d)] if isinstance(d, Tracer) else d
|
|
for d in aval.shape]
|
|
aval = aval.update(shape=tuple(new_shape))
|
|
return self.gensym(aval)
|
|
|
|
def find_progenitors(self, tracer):
|
|
var = self.tracer_to_var.get(id(tracer))
|
|
if not var:
|
|
return None, None
|
|
active_vars = {var}
|
|
for eqn in self.eqns[::-1]:
|
|
produced = set(eqn.outvars) & active_vars
|
|
if produced:
|
|
active_vars.difference_update(produced)
|
|
active_vars.update(eqn.invars)
|
|
invar_positions = [i for i, v in enumerate(self.invars) if v in active_vars]
|
|
constvars = active_vars & set(self.constvar_to_val)
|
|
const_eqns = [eqn for eqn in self.eqns if set(eqn.invars) & constvars]
|
|
return invar_positions, const_eqns
|
|
|
|
def _const_folding_and_forwarding(
|
|
jaxpr: Jaxpr, constvals: Sequence[Any]) -> tuple[Jaxpr, tuple[Any, ...]]:
|
|
consts: dict[Var, Any] = dict(zip(jaxpr.constvars, constvals))
|
|
var_subs: dict[Var, Var] = {} # not Dict[Var, Atom] b/c literals not inlined
|
|
new_eqns = []
|
|
for eqn in jaxpr.eqns:
|
|
# always apply invar substitutions
|
|
eqn = eqn.replace(invars=[var_subs.get(v, v) for v in eqn.invars])
|
|
# if any inputs are constants and we have a constant-folding rule, apply it
|
|
has_input_effect = any(isinstance(eff, effects.JaxprInputEffect)
|
|
for eff in eqn.effects)
|
|
if (eqn.primitive in const_fold_rules and any(v in consts for v in eqn.invars)
|
|
and not has_input_effect):
|
|
consts_in = [consts.get(v) for v in eqn.invars]
|
|
consts_out, new_eqn = const_fold_rules[eqn.primitive](consts_in, eqn)
|
|
assert (new_eqn is None) == all(c is not None for c in consts_out)
|
|
for v, c in zip(eqn.outvars, consts_out):
|
|
if c is not None: consts[v] = c
|
|
if new_eqn is None: continue
|
|
else: eqn = new_eqn
|
|
# if the application trivially maps some inputs to outputs, simplify
|
|
if eqn.primitive in forwarding_rules and not has_input_effect:
|
|
fwd_vars, new_eqn = forwarding_rules[eqn.primitive](eqn)
|
|
assert (new_eqn is None) == all(v is not None for v in fwd_vars)
|
|
for v_orig, v_new in zip(eqn.outvars, fwd_vars):
|
|
if v_new is not None: var_subs[v_orig] = v_new
|
|
if new_eqn is None: continue
|
|
else: eqn = new_eqn
|
|
new_eqns.append(eqn)
|
|
new_constvars, new_constvals = unzip2(consts.items())
|
|
new_outvars = [var_subs.get(v, v) for v in jaxpr.outvars]
|
|
jaxpr_effects = make_jaxpr_effects(new_constvars, jaxpr.invars, new_outvars,
|
|
new_eqns)
|
|
new_jaxpr = Jaxpr(new_constvars, jaxpr.invars, new_outvars, new_eqns,
|
|
jaxpr_effects, jaxpr.debug_info)
|
|
return new_jaxpr, new_constvals
|
|
|
|
ConstFoldRule = Callable[[list[Optional[Any]], JaxprEqn],
|
|
tuple[list[Optional[Any]], Optional[JaxprEqn]]]
|
|
const_fold_rules: dict[Primitive, ConstFoldRule] = {}
|
|
|
|
ForwardingRule = Callable[[JaxprEqn],
|
|
tuple[list[Optional[Var]], Optional[JaxprEqn]]]
|
|
forwarding_rules: dict[Primitive, ForwardingRule] = {}
|
|
|
|
|
|
def _inline_literals(
|
|
jaxpr: Jaxpr, constvals: Sequence[Any]
|
|
) -> tuple[Jaxpr, list[Any]]:
|
|
# This function also prunes unused constants and inserts `dropvar` symbols.
|
|
input_effects = {eff for eff in jaxpr.effects
|
|
if isinstance(eff, effects.JaxprInputEffect)}
|
|
# Don't inline any literal with an input effect
|
|
has_input_effect = [any(eff.input_index == i for eff in input_effects)
|
|
for i in range(len(constvals))]
|
|
lits = {v: Literal(c, v.aval) for v, c, e in zip(jaxpr.constvars, constvals,
|
|
has_input_effect)
|
|
if type(c) in core.literalable_types and not np.shape(c) and not e}
|
|
lit: Callable[[Var], Literal | None] = lits.get
|
|
newname: Callable[[AbstractValue], Var] = core.gensym()
|
|
newvars: dict[Var, Var] = {}
|
|
newvar = lambda aval: newname(_substitute_vars_in_type(lits, newvars, aval))
|
|
var = lambda v: newvars.get(v) or newvars.setdefault(v, newvar(v.aval))
|
|
dropvar = lambda aval: DropVar(_substitute_vars_in_type(lits, newvars, aval))
|
|
|
|
def vars_in_shape(aval: AbstractValue) -> Sequence[Var]:
|
|
if isinstance(aval, DShapedArray):
|
|
return [d for d in aval.shape if isinstance(d, Var)]
|
|
return []
|
|
|
|
used = {v for eqn in jaxpr.eqns for invar in eqn.invars
|
|
for v in it.chain([invar], vars_in_shape(invar.aval))}
|
|
used |= {v for outvar in jaxpr.outvars
|
|
for v in it.chain([outvar], vars_in_shape(outvar.aval))}
|
|
new_constvars = [var(v) for v in jaxpr.constvars if v in used and not lit(v)]
|
|
new_constvals = [c for v, c in zip(jaxpr.constvars, constvals)
|
|
if v in used and not lit(v)]
|
|
new_invars = [var(v) for v in jaxpr.invars]
|
|
new_eqns = []
|
|
for eqn in jaxpr.eqns:
|
|
invars = [lit(v) or var(v) for v in eqn.invars]
|
|
outvars = [var(v) if v in used else dropvar(v.aval) for v in eqn.outvars]
|
|
new_eqns.append(eqn.replace(invars=invars, outvars=outvars))
|
|
new_outvars = [lit(v) or var(v) for v in jaxpr.outvars]
|
|
jaxpr_effects = make_jaxpr_effects(new_constvars, new_invars, new_outvars,
|
|
new_eqns)
|
|
new_jaxpr = Jaxpr(new_constvars, new_invars, new_outvars, new_eqns,
|
|
jaxpr_effects, jaxpr.debug_info)
|
|
return new_jaxpr, new_constvals
|
|
|
|
|
|
class DynamicJaxprTrace(core.Trace):
|
|
__slots__ = [] # type: ignore
|
|
|
|
@property
|
|
def frame(self):
|
|
return self.main.jaxpr_stack[-1] # pytype: disable=attribute-error
|
|
|
|
def new_arg(self, aval):
|
|
tracer = DynamicJaxprTracer(self, aval, source_info_util.current())
|
|
self.frame.tracers.append(tracer)
|
|
self.frame.tracer_to_var[id(tracer)] = var = self.frame.newvar(aval)
|
|
self.frame.invars.append(var)
|
|
return tracer
|
|
|
|
def new_const(self, c):
|
|
# TODO(mattjj): for ints, or hashable consts, don't rely on id
|
|
tracer = self.frame.constid_to_tracer.get(id(c))
|
|
if tracer is None:
|
|
aval = raise_to_shaped(get_aval(c), weak_type=dtypes.is_weakly_typed(c))
|
|
aval = self._lift_tracers_in_aval(aval)
|
|
tracer = self._new_const(aval, c)
|
|
return tracer
|
|
|
|
pure = lift = new_const
|
|
|
|
def _new_const(self, aval, c):
|
|
tracer = DynamicJaxprTracer(self, aval, source_info_util.current())
|
|
self.frame.tracers.append(tracer)
|
|
self.frame.tracer_to_var[id(tracer)] = var = self.frame.newvar(aval)
|
|
self.frame.constid_to_tracer[id(c)] = tracer
|
|
self.frame.constvar_to_val[var] = c
|
|
return tracer
|
|
|
|
def sublift(self, t):
|
|
# When lifting closed-over tracers corresponding to this same trace, the
|
|
# variable to lift could have tracers (representing axis size variables) in
|
|
# its shape. We must lift those too!
|
|
tracer = self.frame.constid_to_tracer.get(id(t))
|
|
if tracer is None:
|
|
aval = raise_to_shaped(get_aval(t), weak_type=dtypes.is_weakly_typed(t))
|
|
aval = self._lift_tracers_in_aval(aval)
|
|
tracer = self._new_const(aval, t)
|
|
return tracer
|
|
|
|
def _lift_tracers_in_aval(self, aval):
|
|
if (not isinstance(aval, DShapedArray) or
|
|
not any(isinstance(d, Tracer) for d in aval.shape)):
|
|
return aval
|
|
shape = [self.full_raise(d) if isinstance(d, Tracer) else d
|
|
for d in aval.shape]
|
|
return aval.update(shape=tuple(shape))
|
|
|
|
def getvar(self, tracer):
|
|
var = self.frame.tracer_to_var.get(id(tracer))
|
|
if var is None:
|
|
raise core.escaped_tracer_error(tracer)
|
|
return var
|
|
|
|
def makevar(self, tracer):
|
|
var = self.frame.tracer_to_var.get(id(tracer))
|
|
assert var is None, "a jaxpr variable must be created only once per tracer"
|
|
self.frame.tracers.append(tracer)
|
|
var = self.frame.tracer_to_var[id(tracer)] = self.frame.newvar(tracer.aval)
|
|
return var
|
|
|
|
def instantiate_const(self, val):
|
|
if (isinstance(val, Tracer) and val._trace.main is self.main
|
|
and val._trace.sublevel == self.sublevel):
|
|
return val
|
|
else:
|
|
return self.new_const(val)
|
|
|
|
def process_primitive(self, primitive, tracers, params):
|
|
if primitive in custom_staging_rules:
|
|
return custom_staging_rules[primitive](self, *tracers, **params)
|
|
return self.default_process_primitive(primitive, tracers, params)
|
|
|
|
def default_process_primitive(self, primitive, tracers, params):
|
|
avals = [t.aval for t in tracers]
|
|
out_avals, effects = primitive.abstract_eval(*avals, **params)
|
|
# == serve as a "not xor" here.
|
|
if not (isinstance(out_avals, (tuple,list)) == primitive.multiple_results):
|
|
raise ValueError(f"{primitive}.abstract_eval() method should return"
|
|
f" a tuple or a list if {primitive}.multiple_results"
|
|
" is true. Otherwise it shouldn't.")
|
|
out_avals = [out_avals] if not primitive.multiple_results else out_avals
|
|
source_info = source_info_util.current()
|
|
out_tracers = [DynamicJaxprTracer(self, a, source_info) for a in out_avals]
|
|
invars = map(self.getvar, tracers)
|
|
outvars = map(self.makevar, out_tracers)
|
|
eqn = new_jaxpr_eqn(invars, outvars, primitive, params, effects, source_info)
|
|
self.frame.add_eqn(eqn)
|
|
return out_tracers if primitive.multiple_results else out_tracers.pop()
|
|
|
|
def process_call(self, call_primitive, f, explicit_tracers, params):
|
|
if f.in_type is None:
|
|
f = lu.annotate(f, tuple((raise_to_shaped(t.aval), True)
|
|
for t in explicit_tracers))
|
|
implicit_tracers = _extract_implicit_args(self, f.in_type, explicit_tracers)
|
|
in_tracers = [*implicit_tracers, *explicit_tracers]
|
|
# TODO(mattjj): check in_tracers are consistent with f.in_type annotation
|
|
with core.new_sublevel():
|
|
# TODO(lenamartens): Make call_primitive name -> API function name mapping.
|
|
# (currently this will display eg. 'xla_call' instead of `jit`)
|
|
dbg = debug_info_final(f, call_primitive.name)
|
|
jaxpr, out_type, consts = trace_to_subjaxpr_dynamic2(f, self.main, debug_info=dbg)
|
|
if params.get('inline', False):
|
|
return core.eval_jaxpr(jaxpr, consts, *in_tracers,
|
|
propagate_source_info=False)
|
|
source_info = source_info_util.current()
|
|
out_tracers = []
|
|
for aval, _ in out_type:
|
|
if type(aval) is DShapedArray:
|
|
shape = [[*consts, *in_tracers][d.val] if type(d) is InDBIdx else
|
|
out_tracers[d.val] if type(d) is OutDBIdx else
|
|
d for d in aval.shape]
|
|
aval = aval.update(shape=tuple(get_referent(d) for d in shape))
|
|
out_tracers.append(DynamicJaxprTracer(self, aval, source_info))
|
|
invars = map(self.getvar, in_tracers)
|
|
constvars = map(self.getvar, map(self.instantiate_const, consts))
|
|
outvars = map(self.makevar, out_tracers)
|
|
new_params = dict(params, call_jaxpr=convert_constvars_jaxpr(jaxpr))
|
|
update_params = call_param_updaters.get(call_primitive)
|
|
if update_params:
|
|
new_params = update_params(new_params, [True] * len(explicit_tracers),
|
|
len(consts) + len(implicit_tracers))
|
|
eqn = new_jaxpr_eqn([*constvars, *invars], outvars, call_primitive,
|
|
new_params, new_params['call_jaxpr'].effects,
|
|
source_info)
|
|
self.frame.add_eqn(eqn)
|
|
return [t for t, (_, keep) in zip(out_tracers, out_type) if keep]
|
|
|
|
def post_process_call(self, call_primitive, out_tracers, params):
|
|
assert False # unreachable
|
|
|
|
def process_map(self, map_primitive, f, tracers, params):
|
|
in_avals = [t.aval for t in tracers]
|
|
axis_name, axis_size = params['axis_name'], params['axis_size']
|
|
reduced_in_avals = [core.mapped_aval(axis_size, in_axis, a)
|
|
if in_axis is not None else a
|
|
for a, in_axis in zip(in_avals, params['in_axes'])]
|
|
with core.extend_axis_env(axis_name, params["global_axis_size"], None): # type: ignore
|
|
with core.new_sublevel():
|
|
jaxpr, reduced_out_avals, consts = trace_to_subjaxpr_dynamic(
|
|
f, self.main, reduced_in_avals,
|
|
debug_info=debug_info_final(f, map_primitive.name))
|
|
ordered_effects = effects.ordered_effects.filter_in(jaxpr.effects)
|
|
if ordered_effects:
|
|
raise ValueError("Ordered effects not supported for "
|
|
f"map primitives: {ordered_effects}")
|
|
out_axes = params['out_axes_thunk']()
|
|
out_avals = [core.unmapped_aval(axis_size, axis_name, out_axis, a)
|
|
if out_axis is not None else a
|
|
for a, out_axis in zip(reduced_out_avals, out_axes)]
|
|
source_info = source_info_util.current()
|
|
out_tracers = [DynamicJaxprTracer(self, a, source_info) for a in out_avals]
|
|
invars = map(self.getvar, tracers)
|
|
constvars = map(self.getvar, map(self.instantiate_const, consts))
|
|
outvars = map(self.makevar, out_tracers)
|
|
new_in_axes = (None,) * len(consts) + params['in_axes']
|
|
new_params = dict(params, in_axes=new_in_axes, out_axes=out_axes,
|
|
call_jaxpr=convert_constvars_jaxpr(jaxpr))
|
|
del new_params['out_axes_thunk']
|
|
update_params = call_param_updaters.get(map_primitive)
|
|
if update_params:
|
|
new_params = update_params(new_params, [True] * len(tracers), len(consts))
|
|
eqn = new_jaxpr_eqn([*constvars, *invars], outvars, map_primitive,
|
|
new_params, jaxpr.effects, source_info)
|
|
self.frame.add_eqn(eqn)
|
|
return out_tracers
|
|
|
|
def post_process_map(self, map_primitive, out_tracers, params):
|
|
assert False # unreachable
|
|
|
|
def process_custom_jvp_call(self, prim, fun, jvp, tracers, *, symbolic_zeros):
|
|
in_avals = [t.aval for t in tracers]
|
|
with core.new_sublevel():
|
|
fun_jaxpr, out_avals, consts = trace_to_subjaxpr_dynamic(fun, self.main, in_avals)
|
|
closed_fun_jaxpr = core.ClosedJaxpr(convert_constvars_jaxpr(fun_jaxpr), ())
|
|
main_ = ref(self.main)
|
|
|
|
@_memoize
|
|
def jvp_jaxpr_thunk(*in_zeros):
|
|
for store in jvp.stores: store and store.reset()
|
|
nz_tangent_avals, zero_avals = partition_list(in_zeros, in_avals)
|
|
jvp_, out_zeros = _jvp_jaxpr_zeros(jvp, in_zeros, tuple(zero_avals))
|
|
in_avals_ = (*in_avals, *nz_tangent_avals)
|
|
jaxpr, _, out_consts = trace_to_subjaxpr_dynamic(jvp_, main_(), in_avals_)
|
|
return jaxpr, out_consts, out_zeros()
|
|
|
|
out_tracers = [DynamicJaxprTracer(self, a) for a in out_avals]
|
|
invars = map(self.getvar, tracers)
|
|
constvars = map(self.getvar, map(self.instantiate_const, consts))
|
|
outvars = map(self.makevar, out_tracers)
|
|
eqn = new_jaxpr_eqn([*constvars, *invars], outvars, prim,
|
|
dict(call_jaxpr=closed_fun_jaxpr,
|
|
jvp_jaxpr_thunk=jvp_jaxpr_thunk,
|
|
num_consts=len(consts),
|
|
symbolic_zeros=symbolic_zeros),
|
|
fun_jaxpr.effects,
|
|
source_info_util.current())
|
|
self.frame.add_eqn(eqn)
|
|
return out_tracers
|
|
|
|
def post_process_custom_jvp_call(self, out_tracers, _):
|
|
assert False # unreachable
|
|
|
|
def process_custom_vjp_call(self, prim, fun, fwd, bwd, tracers, out_trees,
|
|
symbolic_zeros):
|
|
in_avals = [t.aval for t in tracers]
|
|
with core.new_sublevel():
|
|
fun_jaxpr, out_avals, consts = trace_to_subjaxpr_dynamic(fun, self.main, in_avals)
|
|
closed_fun_jaxpr = core.ClosedJaxpr(convert_constvars_jaxpr(fun_jaxpr), ())
|
|
|
|
main_ = ref(self.main)
|
|
|
|
@_memoize
|
|
def fwd_jaxpr_from_zeros(*zeros):
|
|
for store in fwd.stores: store and store.reset()
|
|
fwd_ = _interleave_fun(fwd, zeros)
|
|
return trace_to_subjaxpr_dynamic(fwd_, main_(), in_avals)[::2]
|
|
|
|
out_tracers = [DynamicJaxprTracer(self, a) for a in out_avals]
|
|
invars = map(self.getvar, tracers)
|
|
constvars = map(self.getvar, map(self.instantiate_const, consts))
|
|
outvars = map(self.makevar, out_tracers)
|
|
eqn = new_jaxpr_eqn([*constvars, *invars], outvars, prim.initial_style,
|
|
dict(fun_jaxpr=closed_fun_jaxpr,
|
|
fwd_jaxpr_thunk=fwd_jaxpr_from_zeros,
|
|
num_consts=len(consts),
|
|
bwd=bwd, out_trees=out_trees,
|
|
symbolic_zeros=symbolic_zeros),
|
|
fun_jaxpr.effects,
|
|
source_info_util.current())
|
|
self.frame.add_eqn(eqn)
|
|
return out_tracers
|
|
|
|
def post_process_custom_vjp_call(self, out_tracers, _):
|
|
assert False # unreachable
|
|
|
|
def process_custom_transpose(self, prim, call, tracers, *,
|
|
transpose, out_types,
|
|
lin_tree, res_tree, out_tree):
|
|
tracers_res, tracers_lin = split_list(tracers, [res_tree.num_leaves])
|
|
|
|
in_avals_p = [t.aval for t in tracers]
|
|
in_avals_t = [*[t.aval for t in tracers_res], *out_types]
|
|
|
|
with core.new_sublevel():
|
|
call_jaxpr, out_avals, call_consts = trace_to_subjaxpr_dynamic(
|
|
call, self.main, in_avals_p)
|
|
closed_call_jaxpr = core.ClosedJaxpr(
|
|
convert_constvars_jaxpr(call_jaxpr), ())
|
|
|
|
transpose_flat, in_tree2 = flatten_fun_nokwargs(
|
|
lu.wrap_init(transpose), treedef_tuple((res_tree, out_tree)))
|
|
|
|
main_ = ref(self.main)
|
|
# the following thunk evaluates to a pair: transpose_jaxpr, transpose_consts
|
|
@_memoize
|
|
def transpose_jaxpr_thunk():
|
|
for store in transpose_flat.stores: store.reset()
|
|
jaxpr, _, consts = trace_to_subjaxpr_dynamic(transpose_flat, main_(),
|
|
in_avals_t)
|
|
return jaxpr, consts
|
|
|
|
out_tracers = [DynamicJaxprTracer(self, a) for a in out_avals]
|
|
invars = map(self.getvar, tracers)
|
|
constvars = map(self.getvar, map(self.instantiate_const, call_consts))
|
|
outvars = map(self.makevar, out_tracers)
|
|
eqn = new_jaxpr_eqn([*constvars, *invars], outvars, prim,
|
|
dict(call_jaxpr=closed_call_jaxpr,
|
|
transpose_jaxpr_thunk=transpose_jaxpr_thunk,
|
|
out_types=out_types, res_tree=res_tree,
|
|
lin_tree=lin_tree, out_tree=out_tree),
|
|
closed_call_jaxpr.effects,
|
|
source_info_util.current())
|
|
self.frame.add_eqn(eqn)
|
|
return out_tracers
|
|
|
|
|
|
custom_staging_rules: dict[Primitive, Callable] = {}
|
|
|
|
@lu.transformation
|
|
def _interleave_fun(every_others, *args, **kwargs):
|
|
args_ = [x for pair in zip(args, every_others) for x in pair]
|
|
yield (yield (args_, kwargs))
|
|
|
|
def _memoize(fn):
|
|
cells = {}
|
|
saved_state = core.thread_local_state.trace_state.copy()
|
|
sentinel = object()
|
|
def memoized(*args):
|
|
out = cells.get(args, sentinel)
|
|
if out is sentinel:
|
|
prev_state = core.thread_local_state.trace_state
|
|
core.thread_local_state.trace_state = saved_state
|
|
try:
|
|
out = cells[args] = fn(*args)
|
|
finally:
|
|
core.thread_local_state.trace_state = prev_state
|
|
return out
|
|
return memoized
|
|
|
|
@lu.transformation_with_aux
|
|
def _jvp_jaxpr_zeros(in_zeros, zero_avals, *primal_tangent_avals):
|
|
in_primals, nz_in_tangents = split_list(primal_tangent_avals, [len(in_zeros)])
|
|
symbolic_zeros = map(ad_util.SymbolicZero, zero_avals)
|
|
tangents = merge_lists(in_zeros, nz_in_tangents, symbolic_zeros)
|
|
out = yield (*in_primals, *tangents), {}
|
|
n, ragged = divmod(len(out), 2)
|
|
assert not ragged
|
|
out_primals, out_tangents = out[:n], out[n:]
|
|
out_zeros = [type(t) is ad_util.SymbolicZero for t in out_tangents]
|
|
out_nz_tangents, _ = partition_list(out_zeros, out_tangents)
|
|
yield [*out_primals, *out_nz_tangents], out_zeros
|
|
|
|
# TODO(mattjj): remove this DebugInfo and helper functions, replace with
|
|
# api_util.py versions
|
|
|
|
class DebugInfo(NamedTuple):
|
|
func_src_info: str | None # f'{fun.__name__} at {filename}:{lineno}'
|
|
signature: inspect.Signature | None # inspect.signature(fun)
|
|
in_tree: PyTreeDef | None # caller/constructor might not have this info
|
|
out_tree: Callable[[], PyTreeDef] | None # lazy, not avail at trace time
|
|
has_kwargs: bool # whether in_tree corresponds to (args, kwargs) or args
|
|
traced_for: str # "jit", "scan", "make_jaxpr", etc
|
|
|
|
def debug_info(fn: Callable, in_tree: PyTreeDef | None,
|
|
out_tree_thunk: Callable[[], PyTreeDef] | None,
|
|
has_kwargs: bool, traced_for: str) -> DebugInfo:
|
|
try: sig = inspect.signature(fn)
|
|
except (ValueError, TypeError): sig = None
|
|
src_info = fun_sourceinfo(fn)
|
|
return DebugInfo(src_info, sig, in_tree, out_tree_thunk, has_kwargs,
|
|
traced_for)
|
|
|
|
def debug_info_final(fn: lu.WrappedFun, traced_for: str) -> DebugInfo:
|
|
"Make a DebugInfo from data available to final-style primitives like pmap."
|
|
in_tree, out_tree, has_kws = flattened_fun_in_tree(fn) or (None, None, False)
|
|
return debug_info(fn.f, in_tree, out_tree, has_kws, traced_for)
|
|
|
|
def arg_info_all(dbg: DebugInfo) -> list[tuple[str, KeyPath]] | None:
|
|
ba = None if dbg.in_tree is None else sig_info(dbg)
|
|
if ba is None: return None
|
|
return [(name, key_path) for name, dummy_arg in ba.arguments.items()
|
|
for key_path, _ in generate_key_paths(dummy_arg)]
|
|
|
|
def sig_info(dbg: DebugInfo) -> inspect.BoundArguments | None:
|
|
if dbg.in_tree is None or dbg.signature is None: return None
|
|
try:
|
|
dummy_args = tree_unflatten(dbg.in_tree, [False] * dbg.in_tree.num_leaves)
|
|
except:
|
|
return None
|
|
args, kwargs = dummy_args if dbg.has_kwargs else (dummy_args, {})
|
|
try:
|
|
return dbg.signature.bind(*args, **kwargs)
|
|
except (TypeError, ValueError):
|
|
return None
|
|
|
|
def result_info(dbg: DebugInfo) -> list[KeyPath] | None:
|
|
if dbg.out_tree is None: return None
|
|
try:
|
|
num_leaves = dbg.out_tree().num_leaves
|
|
dummy_result = tree_unflatten(dbg.out_tree(), [False] * num_leaves)
|
|
except:
|
|
return None
|
|
else:
|
|
return [path for path, _ in generate_key_paths(dummy_result)]
|
|
|
|
|
|
@profiler.annotate_function
|
|
def trace_to_jaxpr_dynamic(
|
|
fun: lu.WrappedFun,
|
|
in_avals: Sequence[AbstractValue],
|
|
debug_info: DebugInfo | None = None,
|
|
*,
|
|
keep_inputs: list[bool] | None = None,
|
|
) -> tuple[Jaxpr, list[AbstractValue], list[Any]]:
|
|
with core.new_main(DynamicJaxprTrace, dynamic=True) as main: # type: ignore
|
|
main.jaxpr_stack = () # type: ignore
|
|
jaxpr, out_avals, consts = trace_to_subjaxpr_dynamic(
|
|
fun, main, in_avals, keep_inputs=keep_inputs, debug_info=debug_info)
|
|
del main, fun
|
|
return jaxpr, out_avals, consts
|
|
|
|
|
|
def trace_to_subjaxpr_dynamic(
|
|
fun: lu.WrappedFun,
|
|
main: core.MainTrace,
|
|
in_avals: Sequence[AbstractValue],
|
|
*,
|
|
keep_inputs: Sequence[bool] | None = None,
|
|
debug_info: DebugInfo | None = None,
|
|
) -> tuple[Jaxpr, list[AbstractValue], list[Any]]:
|
|
keep_inputs = [True] * len(in_avals) if keep_inputs is None else keep_inputs
|
|
|
|
frame = JaxprStackFrame()
|
|
frame.debug_info = debug_info
|
|
with extend_jaxpr_stack(main, frame), source_info_util.reset_name_stack():
|
|
trace = DynamicJaxprTrace(main, core.cur_sublevel())
|
|
in_tracers = _input_type_to_tracers(trace.new_arg, in_avals)
|
|
in_tracers_ = [t for t, keep in zip(in_tracers, keep_inputs) if keep]
|
|
ans = fun.call_wrapped(*in_tracers_)
|
|
out_tracers = map(trace.full_raise, ans)
|
|
jaxpr, consts = frame.to_jaxpr(out_tracers)
|
|
del fun, main, trace, frame, in_tracers, out_tracers, ans
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr)
|
|
return jaxpr, [v.aval for v in jaxpr.outvars], consts
|
|
|
|
|
|
@profiler.annotate_function
|
|
def trace_to_jaxpr_dynamic2(
|
|
fun: lu.WrappedFun, debug_info: DebugInfo | None = None
|
|
) -> tuple[Jaxpr, OutputType, list[Any]]:
|
|
with core.new_main(DynamicJaxprTrace, dynamic=True) as main: # type: ignore
|
|
main.jaxpr_stack = () # type: ignore
|
|
jaxpr, out_type, consts = trace_to_subjaxpr_dynamic2(fun, main, debug_info)
|
|
del main, fun
|
|
return jaxpr, out_type, consts
|
|
|
|
def trace_to_subjaxpr_dynamic2(
|
|
fun: lu.WrappedFun, main: core.MainTrace,
|
|
debug_info: DebugInfo | None = None
|
|
) -> tuple[Jaxpr, OutputType, list[Any]]:
|
|
in_avals, keep_inputs = unzip2(fun.in_type)
|
|
frame = JaxprStackFrame()
|
|
frame.debug_info = debug_info
|
|
with extend_jaxpr_stack(main, frame), source_info_util.reset_name_stack():
|
|
trace = DynamicJaxprTrace(main, core.cur_sublevel())
|
|
in_tracers = _input_type_to_tracers(trace.new_arg, in_avals)
|
|
in_tracers_ = [t for t, keep in zip(in_tracers, keep_inputs) if keep]
|
|
ans = fun.call_wrapped(*in_tracers_)
|
|
out_tracers = map(trace.full_raise, ans)
|
|
jaxpr, out_type, consts = frame.to_jaxpr2(out_tracers)
|
|
del fun, main, trace, frame, in_tracers, out_tracers, ans
|
|
return jaxpr, out_type, consts
|
|
|
|
|
|
@contextmanager
|
|
def extend_jaxpr_stack(main, frame):
|
|
main.jaxpr_stack = main.jaxpr_stack + (frame,)
|
|
try:
|
|
yield
|
|
finally:
|
|
assert frame is main.jaxpr_stack[-1]
|
|
main.jaxpr_stack = main.jaxpr_stack[:-1]
|
|
|
|
|
|
@profiler.annotate_function
|
|
def trace_to_jaxpr_final(
|
|
fun: lu.WrappedFun,
|
|
in_avals: Sequence[AbstractValue],
|
|
debug_info: DebugInfo | None = None,
|
|
keep_inputs: Sequence[bool] | None = None,
|
|
) -> tuple[Jaxpr, list[AbstractValue], list[Any]]:
|
|
with core.new_base_main(DynamicJaxprTrace) as main: # type: ignore
|
|
main.jaxpr_stack = () # type: ignore
|
|
with core.new_sublevel():
|
|
jaxpr, out_avals, consts = trace_to_subjaxpr_dynamic(
|
|
fun, main, in_avals, keep_inputs=keep_inputs, debug_info=debug_info)
|
|
del fun, main
|
|
return jaxpr, out_avals, consts
|
|
|
|
|
|
@profiler.annotate_function
|
|
def trace_to_jaxpr_final2(
|
|
fun: lu.WrappedFun, debug_info: DebugInfo | None = None
|
|
) -> tuple[Jaxpr, OutputType, list[Any]]:
|
|
with core.new_base_main(DynamicJaxprTrace) as main: # type: ignore
|
|
main.jaxpr_stack = () # type: ignore
|
|
with core.new_sublevel():
|
|
jaxpr, out_type, consts = trace_to_subjaxpr_dynamic2(fun, main, debug_info)
|
|
del fun, main
|
|
return jaxpr, out_type, consts
|
|
|
|
|
|
AbstractedAxisName = Hashable
|
|
AbstractedAxesSpec = Union[dict[int, AbstractedAxisName],
|
|
tuple[AbstractedAxisName, ...]]
|
|
def infer_lambda_input_type(
|
|
axes_specs: Sequence[AbstractedAxesSpec] | None,
|
|
args: Sequence[Any]
|
|
) -> InputType:
|
|
ndims = [getattr(get_aval(x), 'ndim', 0) for x in args]
|
|
partial_specs = _canonicalize_specs(ndims, axes_specs)
|
|
specs = _complete_specs(args, partial_specs)
|
|
idxs, implicit_types = _collect_implicit(args, specs)
|
|
implicit_sig = [(ty, False) for ty in implicit_types]
|
|
explicit_sig = [(_arg_type(idxs, x, s), True) for x, s in zip(args, specs)]
|
|
input_type = (*implicit_sig, *explicit_sig)
|
|
lu._check_input_type(input_type)
|
|
return input_type
|
|
|
|
def _spec_to_dict(spec: AbstractedAxesSpec) -> dict[int, AbstractedAxisName]:
|
|
if isinstance(spec, tuple):
|
|
return {i: d for i, d in enumerate(spec) if d is not None}
|
|
else:
|
|
return spec
|
|
|
|
def _canonicalize_specs(
|
|
ndims: Sequence[int], specs: Sequence[AbstractedAxesSpec] | None
|
|
) -> list[dict[int, AbstractedAxisName]]:
|
|
if specs is None:
|
|
return [{}] * len(ndims)
|
|
else:
|
|
return [_spec_to_dict(s) for n, s in zip(ndims, specs)]
|
|
|
|
def _complete_specs(
|
|
args: Sequence[Any], partial_specs: list[dict[int, AbstractedAxisName]]
|
|
) -> list[dict[int, AbstractedAxisName]]:
|
|
# The abstracted axes specification in `partial_specs` is partial in the sense
|
|
# that there could be additional axis abstraction represented in `args` due to
|
|
# Tracers existing in the shapes of elements of `args`. The purpose of this
|
|
# function is to produce a full specification, for each argument mapping any
|
|
# abstracted axis positions to a name, introducing new names as needed for
|
|
# Tracers in axis sizes which don't already correspond to abstracted axis
|
|
# names (with one new name per unique Tracer object id).
|
|
|
|
# Identify each user-supplied name in partial_specs with a size.
|
|
sizes: dict[AbstractedAxisName, int | DynamicJaxprTracer] = {}
|
|
for x, spec in zip(args, partial_specs):
|
|
for i, name in spec.items():
|
|
d = sizes.setdefault(name, x.shape[i])
|
|
if d is not x.shape[i] and d != x.shape[i]: raise TypeError
|
|
|
|
# Introduce new names as needed for Tracers in shapes.
|
|
named_tracers: dict[TracerId, AbstractedAxisName] = {
|
|
id(d): name for name, d in sizes.items() if isinstance(d, Tracer)}
|
|
specs: list[dict[int, AbstractedAxisName]] = []
|
|
for x, spec in zip(args, partial_specs):
|
|
if isinstance(get_aval(x), DShapedArray):
|
|
spec = dict(spec)
|
|
for i, d in enumerate(x.shape):
|
|
if isinstance(d, Tracer):
|
|
spec[i] = named_tracers.get(id(d), TracerAsName(d))
|
|
specs.append(spec)
|
|
|
|
# Assert that `specs` is now complete in the sense that there are no Tracers
|
|
# which don't correspond to an AbstractedAxisName.
|
|
assert all(not spec or not any(isinstance(d, Tracer) and i not in spec
|
|
for i, d in enumerate(x.shape))
|
|
for x, spec in zip(args, specs))
|
|
return specs
|
|
|
|
|
|
def _collect_implicit(
|
|
args: Sequence[Any], specs: list[dict[int, AbstractedAxisName]]
|
|
) -> tuple[dict[AbstractedAxisName, DBIdx], list[AbstractValue]]:
|
|
# Given an explicit argument list and a specification of abstracted axes, we
|
|
# want to produce an InputType by identifying AbstractedAxisNames with DBIdxs
|
|
# and figuring out which AbstractedAxisNames correspond to implicit arguments.
|
|
|
|
idxs: dict[AbstractedAxisName, DBIdx] = {}
|
|
implicit_types: list[AbstractValue] = []
|
|
explicit_tracers: dict[TracerId, int] = {}
|
|
counter = it.count()
|
|
|
|
# Add implicit arguments to idxs.
|
|
for explicit_idx, (x, spec) in enumerate(zip(args, specs)):
|
|
for i, name in spec.items():
|
|
if name not in idxs and id(x.shape[i]) not in explicit_tracers:
|
|
idxs[name] = DBIdx(next(counter))
|
|
implicit_types.append(raise_to_shaped(get_aval(x.shape[i])))
|
|
if isinstance(x, Tracer):
|
|
explicit_tracers.setdefault(id(x), explicit_idx) # use the first
|
|
|
|
# Now that we know the implicit args, add explicit args to idxs.
|
|
offset = len(implicit_types)
|
|
for x, spec in zip(args, specs):
|
|
for i, name in spec.items():
|
|
if id(x.shape[i]) in explicit_tracers:
|
|
idxs.setdefault(name, DBIdx(offset + explicit_tracers[id(x.shape[i])]))
|
|
|
|
return idxs, implicit_types
|
|
|
|
def _arg_type(
|
|
idxs: dict[AbstractedAxisName, DBIdx], x: Any,
|
|
spec: dict[int, AbstractedAxisName]
|
|
) -> AbstractValue:
|
|
# Produce an AbstractValue by substituting DBIdxs for AbstractedAxisNames.
|
|
aval = get_aval(x) # aval.shape could contain Tracers
|
|
if not spec: return core.raise_to_shaped(aval)
|
|
shape: list[int | DBIdx] = [idxs[spec[i]] if i in spec else d
|
|
for i, d in enumerate(aval.shape)]
|
|
assert not any(isinstance(d, Tracer) for d in shape)
|
|
return DShapedArray(tuple(shape), aval.dtype, False)
|
|
|
|
def _add_implicit_outputs(jaxpr: Jaxpr) -> tuple[Jaxpr, OutputType]:
|
|
invars = [*jaxpr.constvars, *jaxpr.invars]
|
|
expl_outvars = jaxpr.outvars
|
|
|
|
# First do a pass to collect implicit outputs, meaning variables which occurr
|
|
# in explicit_outvars types but not in invars or to the left in outvars.
|
|
seen: set[Var] = set(invars)
|
|
impl_outvars = [seen.add(d) or d for x in expl_outvars if type(x) is Var and # type: ignore
|
|
(seen.add(x) or type(x.aval) is DShapedArray) # type: ignore
|
|
for d in x.aval.shape if type(d) is Var and d not in seen]
|
|
outvars = [*impl_outvars, *expl_outvars]
|
|
|
|
# Now assemble an OutputType by mapping vars in shapes to InDBIdx/OutDBIdx.
|
|
in_map : dict[Var, InDBIdx] = {v: InDBIdx(i) for i, v in enumerate( invars)}
|
|
out_map: dict[Var, OutDBIdx] = {x: OutDBIdx(i) for i, x in enumerate(outvars)
|
|
if type(x) is Var}
|
|
out_avals_ = (x.aval for x in outvars)
|
|
out_avals = [a.update(shape=tuple(in_map.get(d, out_map.get(d))
|
|
if type(d) is Var else d for d in a.shape))
|
|
if type(a) is DShapedArray else a for a in out_avals_]
|
|
kept_outs = [False] * len(impl_outvars) + [True] * len(expl_outvars)
|
|
out_type = tuple(zip(out_avals, kept_outs))
|
|
|
|
new_jaxpr = Jaxpr(jaxpr.constvars, jaxpr.invars, outvars, jaxpr.eqns,
|
|
jaxpr.effects, jaxpr.debug_info)
|
|
config.jax_enable_checks and core.check_jaxpr(jaxpr)
|
|
return new_jaxpr, out_type
|
|
|
|
|
|
class TracerAsName:
|
|
ref: Any
|
|
def __init__(self, tracer):
|
|
self.ref = core.get_referent(tracer)
|
|
def __eq__(self, other):
|
|
return isinstance(other, TracerAsName) and self.ref is other.ref
|
|
def __hash__(self):
|
|
return id(self.ref)
|
|
|
|
def _extract_implicit_args(
|
|
trace: DynamicJaxprTrace, in_type: Sequence[tuple[AbstractValue, bool]],
|
|
explicit_tracers: Sequence[DynamicJaxprTracer]
|
|
) -> Sequence[DynamicJaxprTracer]:
|
|
# First, construct a list to represent the full argument list, leaving the
|
|
# implicit arguments as Nones for now.
|
|
explicit_tracers_ = iter(explicit_tracers)
|
|
tracers = [next(explicit_tracers_) if expl else None for _, expl in in_type]
|
|
assert next(explicit_tracers_, None) is None
|
|
del explicit_tracers_
|
|
|
|
# Next, populate the implicit arguments using DBIdxs in in_type.
|
|
for i, (aval, explicit) in enumerate(in_type):
|
|
if not explicit or not isinstance(aval, DShapedArray):
|
|
continue # can't populate an implicit argument
|
|
tracer = tracers[i]
|
|
assert tracer is not None
|
|
for d1, d2 in zip(aval.shape, tracer.aval.shape):
|
|
if isinstance(d1, DBIdx):
|
|
if tracers[d1.val] is None:
|
|
tracers[d1.val] = trace.instantiate_const(d2)
|
|
assert tracers[d1.val] is trace.instantiate_const(d2)
|
|
assert all(t is not None for t in tracers)
|
|
return [t for t, (_, e) in zip(tracers, in_type) if not e] # type: ignore
|
|
|
|
def _input_type_to_tracers(
|
|
new_arg: Callable[[AbstractValue], Tracer],
|
|
in_avals: Sequence[AbstractValue]
|
|
) -> Sequence[Tracer]:
|
|
# Create input Tracers given input AbstractValues, each of which can contain
|
|
# DeBruijn indices which refer to positions in the input argument list. That
|
|
# is, each element `a` of `in_avals` can have DBIdx instances in its shape,
|
|
# which must refer to positions left of `a`'s.
|
|
in_tracers: list[Tracer] = []
|
|
|
|
def _substitute_tracers_in_aval(a: AbstractValue) -> AbstractValue:
|
|
if isinstance(a, DShapedArray) and any(type(d) is DBIdx for d in a.shape):
|
|
shape = [in_tracers[d.val] if type(d) is DBIdx else d for d in a.shape] # type: ignore
|
|
return a.update(shape=tuple(shape))
|
|
return a
|
|
|
|
for a in in_avals:
|
|
in_tracers.append(new_arg(_substitute_tracers_in_aval(a)))
|
|
return in_tracers
|
|
|
|
def _substitute_vars_in_type(
|
|
consts: dict[Var, Literal], env: dict[Var, Var], a: AbstractValue
|
|
) -> AbstractValue:
|
|
if isinstance(a, DShapedArray) and any(isinstance(d, Var) for d in a.shape):
|
|
shape = [consts[d].val if d in consts else env[d] # type: ignore
|
|
if isinstance(d, Var) else d for d in a.shape]
|
|
return a.update(shape=tuple(shape))
|
|
else:
|
|
return a
|
|
|
|
Const = Any
|
|
Val = Any
|
|
|
|
def pad_jaxpr(jaxpr: Jaxpr, consts: Sequence[Const]
|
|
) -> tuple[Jaxpr, list[Const]]:
|
|
bounds = {v: v.aval.dtype.bound for v in jaxpr.invars
|
|
if isinstance(v.aval, core.UnshapedArray) and
|
|
type(v.aval.dtype) is core.bint and not v.aval.shape}
|
|
idxs = {v: DBIdx(i) for i, v in enumerate(jaxpr.invars)}
|
|
|
|
def substitute(aval: AbstractValue) -> AbstractValue:
|
|
if (isinstance(aval, core.UnshapedArray) and type(aval.dtype) is core.bint
|
|
and not aval.shape):
|
|
return ShapedArray((), dtypes._scalar_type_to_dtype(int))
|
|
elif isinstance(aval, DShapedArray):
|
|
shape = [bounds.get(d, idxs.get(d, d)) for d in aval.shape] # type: ignore
|
|
typ = ShapedArray if all(type(d) is int for d in shape) else DShapedArray
|
|
return typ(tuple(shape), aval.dtype, aval.weak_type)
|
|
else:
|
|
return aval
|
|
|
|
in_avals = [substitute(v.aval) for v in jaxpr.invars]
|
|
eval_padded = lu.wrap_init(partial(_eval_jaxpr_padded, jaxpr, consts))
|
|
padded_jaxpr, _, padded_consts = trace_to_jaxpr_dynamic(eval_padded, in_avals)
|
|
return padded_jaxpr, padded_consts
|
|
|
|
class BoundedAxisSize(NamedTuple):
|
|
val: int | DynamicJaxprTracer
|
|
bound: int
|
|
|
|
def _eval_jaxpr_padded(
|
|
jaxpr: Jaxpr, consts: list[Const], *args: DynamicJaxprTracer
|
|
) -> list[Const | DynamicJaxprTracer]:
|
|
env: dict[Var, Val] = {}
|
|
|
|
def read(x):
|
|
return x.val if type(x) is Literal else env[x]
|
|
|
|
def write(v, val) -> None:
|
|
env[v] = val
|
|
|
|
map(write, jaxpr.constvars, consts)
|
|
map(write, jaxpr.invars, args)
|
|
last_used = core.last_used(jaxpr)
|
|
for eqn in jaxpr.eqns:
|
|
in_avals = [_substitute_axis_sizes(env, v.aval) for v in eqn.invars]
|
|
out_avals = [_substitute_axis_sizes(env, v.aval) for v in eqn.outvars]
|
|
rule = padding_rules[eqn.primitive]
|
|
outs = rule(in_avals, out_avals, *map(read, eqn.invars), **eqn.params)
|
|
map(write, eqn.outvars, outs)
|
|
core.clean_up_dead_vars(eqn, env, last_used)
|
|
return map(read, jaxpr.outvars)
|
|
|
|
def _substitute_axis_sizes(env: dict, aval: AbstractValue) -> AbstractValue:
|
|
if isinstance(aval, DShapedArray):
|
|
shp = []
|
|
for d in aval.shape:
|
|
if isinstance(d, core.DArray):
|
|
assert not d.shape and type(d.dtype) is core.bint
|
|
shp.append(BoundedAxisSize(int(d._data), int(d.dtype.bound)))
|
|
elif (type(d) is core.Var and isinstance(d.aval, core.DShapedArray) and
|
|
type(d.aval.dtype) is core.bint):
|
|
assert not d.aval.shape
|
|
shp.append(BoundedAxisSize(env[d], d.aval.dtype.bound))
|
|
else:
|
|
shp.append(env.get(d, d))
|
|
return DShapedArray(tuple(shp), aval.dtype, aval.weak_type)
|
|
else:
|
|
return aval
|
|
|
|
def _is_bint_axis_size(d: int | core.DArray | core.Var) -> bool:
|
|
if isinstance(d, core.DArray):
|
|
assert not d.shape # pytype: disable=attribute-error
|
|
return type(d.dtype) is core.bint # pytype: disable=attribute-error
|
|
elif isinstance(d, core.Var):
|
|
return (isinstance(d.aval, core.DShapedArray) and # pytype: disable=attribute-error
|
|
type(d.aval.dtype) is core.bint) # pytype: disable=attribute-error
|
|
return False
|
|
|
|
|
|
padding_rules: dict[Primitive, Callable] = {}
|
|
|
|
def def_trivial_padding(prim: Primitive) -> None:
|
|
if prim.multiple_results:
|
|
padding_rules[prim] = partial(_trivial_padding_rule_multi, prim)
|
|
else:
|
|
padding_rules[prim] = partial(_trivial_padding_rule, prim)
|
|
|
|
def _trivial_padding_rule(prim, _, __, *args, **params):
|
|
return [prim.bind(*args, **params)]
|
|
|
|
def _trivial_padding_rule_multi(prim, _, __, *args, **params):
|
|
return prim.bind(*args, **params)
|
|
|
|
def call_padding_rule(prim, in_avals, out_avals, *args, call_jaxpr, **params):
|
|
if call_jaxpr.constvars: raise NotImplementedError
|
|
padded_jaxpr, padded_consts = pad_jaxpr(call_jaxpr, ())
|
|
if padded_consts: raise NotImplementedError
|
|
new_params = dict(params, call_jaxpr=padded_jaxpr)
|
|
subfuns, bind_params = prim.get_bind_params(new_params)
|
|
return prim.bind(*subfuns, *args, **bind_params)
|
|
|
|
|
|
# TODO(mattjj): the following are deprecated; update callers to _nounits version
|
|
# See https://github.com/google/jax/pull/9498
|
|
@lu.transformation
|
|
def trace_to_subjaxpr(main: core.MainTrace, instantiate: bool | Sequence[bool],
|
|
pvals: Sequence[PartialVal]):
|
|
assert all([isinstance(pv, PartialVal) for pv in pvals]), pvals
|
|
trace = main.with_cur_sublevel()
|
|
in_tracers = map(trace.new_arg, pvals)
|
|
ans = yield in_tracers, {}
|
|
assert isinstance(ans, (list, tuple)), (
|
|
f"Got unexpected return type when tracing function to jaxpr: {ans}")
|
|
assert all(isinstance(x, core.Tracer) or core.valid_jaxtype(x) for x in ans), (
|
|
f"Got unexpected return type when tracing function to jaxpr: {ans}")
|
|
instantiate = [instantiate] * len(ans) if isinstance(instantiate, bool) else instantiate
|
|
out_tracers = map(trace.full_raise, map(core.full_lower, ans))
|
|
out_tracers = map(partial(instantiate_const_at, trace), instantiate, out_tracers)
|
|
jaxpr, consts, env = tracers_to_jaxpr(in_tracers, out_tracers)
|
|
out_pvals = [t.pval for t in out_tracers]
|
|
del trace, in_tracers, out_tracers
|
|
yield jaxpr, (out_pvals, consts, env)
|
|
|
|
partial_eval_jaxpr: Callable
|
|
|
|
def instantiate_const_at(trace: JaxprTrace, instantiate: bool, tracer):
|
|
if instantiate:
|
|
return trace.instantiate_const(trace.full_raise(tracer))
|
|
else:
|
|
return tracer
|