# Copyright 2021 The JAX Authors. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # https://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. from __future__ import annotations from collections.abc import Sequence import functools from functools import partial import logging from typing import Any, Callable import types import numpy as np from jax._src import ad_util from jax._src import api from jax._src import config from jax._src import core from jax._src import dispatch from jax._src import linear_util as lu from jax._src import effects from jax._src import source_info_util from jax._src import traceback_util from jax._src import util from jax._src.api_util import flatten_fun, shaped_abstractify from jax._src.interpreters import ad from jax._src.interpreters import batching from jax._src.interpreters import mlir from jax._src.interpreters import partial_eval as pe from jax._src.lax import lax as lax_internal from jax._src.lax import convolution as lax_convolution from jax._src.lib.mlir.dialects import hlo from jax._src.traceback_util import api_boundary from jax._src.tree_util import tree_flatten, tree_unflatten, tree_structure, keystr from jax._src.util import (unzip2, wraps, split_list, partition_list, safe_map, safe_zip, merge_lists, weakref_lru_cache) source_info_util.register_exclusion(__file__) traceback_util.register_exclusion(__file__) map = safe_map zip = safe_zip logger = logging.getLogger(__name__) ### Policies def everything_saveable(*_, **__) -> bool: # This is the effective policy without any use of jax.remat. return True def nothing_saveable(*_, **__) -> bool: # This is the effective policy when using jax.remat without explicit policy. return False def dots_saveable(prim, *_, **__) -> bool: # Matrix multiplies are expensive, so let's save them (and nothing else). return prim in {lax_internal.dot_general_p, lax_convolution.conv_general_dilated_p} checkpoint_dots = dots_saveable def dot_with_no_batch_dims_saveable(prim, *_, **params) -> bool: # This is a useful heuristic for transformers. if prim is lax_internal.dot_general_p: (_, _), (lhs_b, rhs_b) = params['dimension_numbers'] if not lhs_b and not rhs_b: return True return False def offload_dot_with_no_batch_dims(offload_src, offload_dst): def policy(prim, *_, **params): # This is a useful heuristic for transformers. if prim is lax_internal.dot_general_p: (_, _), (lhs_b, rhs_b) = params['dimension_numbers'] if not lhs_b and not rhs_b: return pe.Offloadable(src=offload_src, dst=offload_dst) return pe.Recompute return policy name_p = core.Primitive('name') def save_anything_except_these_names(*names_not_to_save): """Save any values (not just named ones) excluding the names given.""" names_not_to_save = frozenset(names_not_to_save) def policy(prim, *_, **params): if prim is name_p: return params['name'] not in names_not_to_save return True # allow saving anything which is not named return policy def save_any_names_but_these(*names_not_to_save): """Save only named values, excluding the names given.""" names_not_to_save = frozenset(names_not_to_save) def policy(prim, *_, **params): if prim is name_p: return params['name'] not in names_not_to_save return False # only allow saving named values return policy def save_only_these_names(*names_which_can_be_saved): """Save only named values, and only among the names given.""" names_which_can_be_saved = set(names_which_can_be_saved) def policy(prim, *_, **params): if prim is name_p: return params['name'] in names_which_can_be_saved return False # not saveable unless it's in the allow-list return policy def save_and_offload_only_these_names( *, names_which_can_be_saved, names_which_can_be_offloaded, offload_src, offload_dst): names_which_can_be_saved = set(names_which_can_be_saved) names_which_can_be_offloaded = set(names_which_can_be_offloaded) intersection = names_which_can_be_saved.intersection(names_which_can_be_offloaded) if intersection: raise ValueError( "The names should be exclusive and should not intersect in" " `names_which_can_be_saved` and `names_which_can_be_offloaded`. Got" f" names_which_can_be_saved={names_which_can_be_saved}," f" names_which_can_be_offloaded={names_which_can_be_offloaded} and the" f" intersection={intersection}") def policy(prim, *_, **params): if prim is name_p and params['name'] in names_which_can_be_saved: return pe.Saveable if prim is name_p and params['name'] in names_which_can_be_offloaded: return pe.Offloadable(src=offload_src, dst=offload_dst) return pe.Recompute # not saveable unless it's in the allow-list return policy def save_from_both_policies(policy_1, policy_2): def policy(prim, *args, **params): return policy_1(prim, *args, **params) or policy_2(prim, *args, **params) return policy checkpoint_policies = types.SimpleNamespace( everything_saveable=everything_saveable, nothing_saveable=nothing_saveable, dots_saveable=dots_saveable, checkpoint_dots=dots_saveable, dots_with_no_batch_dims_saveable=dot_with_no_batch_dims_saveable, checkpoint_dots_with_no_batch_dims=dot_with_no_batch_dims_saveable, offload_dot_with_no_batch_dims=offload_dot_with_no_batch_dims, save_anything_except_these_names=save_anything_except_these_names, save_any_names_but_these=save_any_names_but_these, save_only_these_names=save_only_these_names, save_from_both_policies=save_from_both_policies, save_and_offload_only_these_names=save_and_offload_only_these_names) ### Main API @api_boundary def checkpoint(fun: Callable, *, prevent_cse: bool = True, policy: Callable[..., bool] | None = None, static_argnums: int | tuple[int, ...] = (), ) -> Callable: """Make ``fun`` recompute internal linearization points when differentiated. The :func:`jax.checkpoint` decorator, aliased to :func:`jax.remat`, provides a way to trade off computation time and memory cost in the context of automatic differentiation, especially with reverse-mode autodiff like :func:`jax.grad` and :func:`jax.vjp` but also with :func:`jax.linearize`. When differentiating a function in reverse-mode, by default all the linearization points (e.g. inputs to elementwise nonlinear primitive operations) are stored when evaluating the forward pass so that they can be reused on the backward pass. This evaluation strategy can lead to a high memory cost, or even to poor performance on hardware accelerators where memory access is much more expensive than FLOPs. An alternative evaluation strategy is for some of the linearization points to be recomputed (i.e. rematerialized) rather than stored. This approach can reduce memory usage at the cost of increased computation. This function decorator produces a new version of ``fun`` which follows the rematerialization strategy rather than the default store-everything strategy. That is, it returns a new version of ``fun`` which, when differentiated, doesn't store any of its intermediate linearization points. Instead, these linearization points are recomputed from the function's saved inputs. See the examples below. Args: fun: Function for which the autodiff evaluation strategy is to be changed from the default of storing all intermediate linearization points to recomputing them. Its arguments and return value should be arrays, scalars, or (nested) standard Python containers (tuple/list/dict) thereof. prevent_cse: Optional, boolean keyword-only argument indicating whether to prevent common subexpression elimination (CSE) optimizations in the HLO generated from differentiation. This CSE prevention has costs because it can foil other optimizations, and because it can incur high overheads on some backends, especially GPU. The default is True because otherwise, under a :func:`~jax.jit` or :func:`~jax.pmap`, CSE can defeat the purpose of this decorator. But in some settings, like when used inside a :func:`~jax.lax.scan`, this CSE prevention mechanism is unnecessary, in which case ``prevent_cse`` can be set to False. static_argnums: Optional, int or sequence of ints, a keyword-only argument indicating which argument values on which to specialize for tracing and caching purposes. Specifying arguments as static can avoid ConcretizationTypeErrors when tracing, but at the cost of more retracing overheads. See the example below. policy: Optional, callable keyword-only argument. It should be one of the attributes of ``jax.checkpoint_policies``. The callable takes as input a type-level specification of a first-order primitive application and returns a boolean indicating whether the corresponding output value(s) can be saved as residuals (or instead must be recomputed in the (co)tangent computation if needed). Returns: A function (callable) with the same input/output behavior as ``fun`` but which, when differentiated using e.g. :func:`jax.grad`, :func:`jax.vjp`, or :func:`jax.linearize`, recomputes rather than stores intermediate linearization points, thus potentially saving memory at the cost of extra computation. Here is a simple example: >>> import jax >>> import jax.numpy as jnp >>> @jax.checkpoint ... def g(x): ... y = jnp.sin(x) ... z = jnp.sin(y) ... return z ... >>> jax.value_and_grad(g)(2.0) (Array(0.78907233, dtype=float32, weak_type=True), Array(-0.2556391, dtype=float32, weak_type=True)) Here, the same value is produced whether or not the :func:`jax.checkpoint` decorator is present. When the decorator is not present, the values ``jnp.cos(2.0)`` and ``jnp.cos(jnp.sin(2.0))`` are computed on the forward pass and are stored for use in the backward pass, because they are needed on the backward pass and depend only on the primal inputs. When using :func:`jax.checkpoint`, the forward pass will compute only the primal outputs and only the primal inputs (``2.0``) will be stored for the backward pass. At that time, the value ``jnp.sin(2.0)`` is recomputed, along with the values ``jnp.cos(2.0)`` and ``jnp.cos(jnp.sin(2.0))``. While :func:`jax.checkpoint` controls what values are stored from the forward-pass to be used on the backward pass, the total amount of memory required to evaluate a function or its VJP depends on many additional internal details of that function. Those details include which numerical primitives are used, how they're composed, where jit and control flow primitives like scan are used, and other factors. The :func:`jax.checkpoint` decorator can be applied recursively to express sophisticated autodiff rematerialization strategies. For example: >>> def recursive_checkpoint(funs): ... if len(funs) == 1: ... return funs[0] ... elif len(funs) == 2: ... f1, f2 = funs ... return lambda x: f1(f2(x)) ... else: ... f1 = recursive_checkpoint(funs[:len(funs)//2]) ... f2 = recursive_checkpoint(funs[len(funs)//2:]) ... return lambda x: f1(jax.checkpoint(f2)(x)) ... If ``fun`` involves Python control flow that depends on argument values, it may be necessary to use the ``static_argnums`` parameter. For example, consider a boolean flag argument:: from functools import partial @partial(jax.checkpoint, static_argnums=(1,)) def foo(x, is_training): if is_training: ... else: ... Here, the use of ``static_argnums`` allows the ``if`` statement's condition to depends on the value of ``is_training``. The cost to using ``static_argnums`` is that it introduces re-tracing overheads across calls: in the example, ``foo`` is re-traced every time it is called with a new value of ``is_training``. In some situations, ``jax.ensure_compile_time_eval`` is needed as well:: @partial(jax.checkpoint, static_argnums=(1,)) def foo(x, y): with jax.ensure_compile_time_eval(): y_pos = y > 0 if y_pos: ... else: ... As an alternative to using ``static_argnums`` (and ``jax.ensure_compile_time_eval``), it may be easier to compute some values outside the :func:`jax.checkpoint`-decorated function and then close over them. """ @wraps(fun) @api_boundary def fun_remat(*args, **kwargs): fun_, args = _remat_static_argnums(fun, static_argnums, args) args_flat, in_tree = tree_flatten((args, kwargs)) in_avals = [shaped_abstractify(x) for x in args_flat] jaxpr, consts, out_tree = _trace_to_jaxpr(fun_, in_tree, tuple(in_avals)) out_flat = remat_p.bind( *consts, *args_flat, jaxpr=jaxpr, prevent_cse=prevent_cse, differentiated=False, policy=policy) return tree_unflatten(out_tree, out_flat) return fun_remat remat = checkpoint # alias # This function is similar to api_util.argnums_partial, except the error # messages are specific to jax.remat (and thus more actionable), the # hashing/caching behavior is slightly different, and this function accepts a # boolean for static_argnums. Perhaps the two could be de-duplicated. def _remat_static_argnums(fun, static_argnums, args): if type(static_argnums) is int: static_argnums = (static_argnums,) elif not (type(static_argnums) is tuple and all(type(d) is int for d in static_argnums)): raise TypeError("the `static_argnums` argument to `jax.checkpoint` / " "`jax.remat` must be an int, tuple of ints or, bool, but " f"got value {static_argnums}") if not all(-len(args) <= d < len(args) for d in static_argnums): raise ValueError("the `static_argnums` argument to `jax.checkpoint` / " "`jax.remat` can only take integer values greater than or " "equal to `-len(args)` and less than `len(args)`, but got " f"{static_argnums}") if not static_argnums: return fun, args nargs = len(args) static_argnums_ = frozenset(d % len(args) for d in static_argnums) dyn_args, static_args = [], [] for i, x in enumerate(args): if i in static_argnums_: static_args.append(WrapHashably(x)) else: dyn_args.append(x) new_fun = _dyn_args_fun(fun, static_argnums_, tuple(static_args), nargs) return new_fun, dyn_args class WrapHashably: val: Any hash: int hashable: bool def __init__(self, val): self.val = val try: self.hash = hash(val) self.hashable = True except: self.hash = id(val) self.hashable = False def __hash__(self): return self.hash def __eq__(self, other): if isinstance(other, WrapHashably): if self.hashable and other.hashable: return self.val == other.val else: return self.val is other.val return False # This caching is useful to avoid retracing even when static_argnums is used. # See api_benchmark.py:bench_remat_eager_retracing_overheads_static_argnums. # On that benchmark, including this caching makes a ~10x difference (which can # be made arbitrary large by involving larger functions to be traced). def _dyn_args_fun(fun: Callable, static_argnums: frozenset[int], static_args: tuple[WrapHashably, ...], nargs: int): if any(isinstance(x.val, core.Tracer) for x in static_args): return _dyn_args_fun_uncached(fun, static_argnums, static_args, nargs) return _dyn_args_fun_cached(fun, static_argnums, static_args, nargs) def _dyn_args_fun_uncached(fun: Callable, static_argnums: frozenset[int], static_args: tuple[WrapHashably, ...], nargs: int): def new_fun(*dyn_args, **kwargs): static_args_, dyn_args_ = iter(static_args), iter(dyn_args) full_args = [next(static_args_).val if i in static_argnums else next(dyn_args_) for i in range(nargs)] return fun(*full_args, **kwargs) return new_fun _dyn_args_fun_cached = weakref_lru_cache(_dyn_args_fun_uncached) # This helper is similar to those in control_flow/common.py, but with # remat-specific errors. @weakref_lru_cache def _trace_to_jaxpr(fun, in_tree, in_avals): flat_fun, out_tree = flatten_fun(lu.wrap_init(fun), in_tree) debug = pe.debug_info(fun, in_tree, out_tree, True, "checkpoint") try: jaxpr, _, consts, () = pe.trace_to_jaxpr_dynamic(flat_fun, in_avals, debug) except core.ConcretizationTypeError as e: msg, = e.args if 'for checkpoint' not in msg: raise new_msg = msg + "\n\n" + ( "Consider using the `static_argnums` parameter for `jax.remat` or " "`jax.checkpoint`. See the `jax.checkpoint` docstring and its example " "involving `static_argnums`:\n" "https://jax.readthedocs.io/en/latest/_autosummary/jax.checkpoint.html" "\n") new_e = core.ConcretizationTypeError.__new__(core.ConcretizationTypeError) new_e.args = (new_msg,) raise new_e from None return pe.convert_constvars_jaxpr(jaxpr), consts, out_tree() ### Utilities def saved_residuals(f, *args, **kwargs) -> list[tuple[core.AbstractValue, str]]: in_leaves, in_tree = tree_flatten((args, kwargs)) def f_(*args): args, kwargs = tree_unflatten(in_tree, args) return f(*args, **kwargs) out = api.make_jaxpr(lambda *args: api.linearize(f_, *args)[1], return_shape=True)(*in_leaves) assert isinstance(out, tuple) jaxpr_, out_shape = out jaxpr = jaxpr_.jaxpr out_tree = lambda: tree_structure(out_shape) assert len(jaxpr.invars) == len(in_leaves) dbg = pe.debug_info(f, in_tree, out_tree, True, "saved_residuals") arg_info = pe.arg_info_all(dbg) return _saved_residuals(jaxpr, arg_info) def _saved_residuals(jaxpr, arg_info) -> list[tuple[core.AbstractValue, str]]: res_lits = [x for x in jaxpr.outvars if isinstance(x, core.Literal)] res_vars = {x for x in jaxpr.outvars if not isinstance(x, core.Literal)} results = [] for x in res_lits: results.append((x.aval, 'from a literal')) for v in jaxpr.constvars: if v in res_vars: results.append((v.aval, 'from a constant')) for i, v in enumerate(jaxpr.invars): if v in res_vars: if arg_info is not None: arg_name, arg_path = arg_info[i] src = f'from the argument {arg_name}{keystr(arg_path)}' else: src = 'from the argument at flattened index {i}' results.append((v.aval, src)) named_vars = {v: e for e in jaxpr.eqns if e.primitive is name_p for v in e.invars} for eqn in jaxpr.eqns: src = source_info_util.summarize(eqn.source_info) for v in eqn.outvars: if v in res_vars: if eqn.primitive is name_p or v in named_vars and (eqn := named_vars[v]): results.append((v.aval, f"named '{eqn.params['name']}' from {src}")) elif str(eqn.primitive) == 'xla_call': results.append((v.aval, f"output of jitted function '{eqn.params['name']}' " f"from {src}")) else: results.append((v.aval, f'output of {eqn.primitive.name} from {src}')) assert len(results) == len(jaxpr.outvars) return results def print_saved_residuals(f, *args, **kwargs): for aval, src in saved_residuals(f, *args, **kwargs): print(f'{aval.str_short(short_dtypes=True)} {src}') ### Implementation remat_p = core.Primitive('remat2') remat_p.multiple_results = True @remat_p.def_impl def remat_impl(*args, jaxpr, prevent_cse, differentiated, policy): del prevent_cse, differentiated, policy # Unused. return core.eval_jaxpr(jaxpr, (), *args) @remat_p.def_effectful_abstract_eval def remat_abstract_eval(*args, jaxpr, prevent_cse, differentiated, policy): del args, prevent_cse, differentiated, policy # Unused. return [v.aval for v in jaxpr.outvars], jaxpr.effects def remat_jvp(primals, tangents, jaxpr, prevent_cse, differentiated, policy): assert not jaxpr.constvars in_nonzeros = [type(t) is not ad_util.Zero for t in tangents] jaxpr_jvp_, out_nz = ad.jvp_jaxpr(pe.close_jaxpr(jaxpr), in_nonzeros, False) nonzero_tangents = [t for t in tangents if type(t) is not ad_util.Zero] jaxpr_jvp = pe.convert_constvars_jaxpr(jaxpr_jvp_.jaxpr) outs = remat_p.bind( *jaxpr_jvp_.consts, *primals, *nonzero_tangents, jaxpr=jaxpr_jvp, prevent_cse=prevent_cse, differentiated=differentiated, policy=policy) out_primals, out_tangents_ = split_list(outs, [len(jaxpr.outvars)]) out_tangents_ = iter(out_tangents_) out_tangents = [next(out_tangents_) if nz else ad_util.Zero.from_value(p) for p, nz in zip(out_primals, out_nz)] return out_primals, out_tangents ad.primitive_jvps[remat_p] = remat_jvp effects.remat_allowed_effects.add_type(lax_internal.InOutFeedEffect) def remat_partial_eval(trace, *tracers, jaxpr, **params): assert not jaxpr.constvars disallowed_effects = effects.remat_allowed_effects.filter_not_in(jaxpr.effects) if disallowed_effects: raise NotImplementedError( 'Effects not supported in partial-eval of `checkpoint`/`remat`: ' f'{disallowed_effects}') policy = params['policy'] or nothing_saveable in_unknowns = [not t.is_known() for t in tracers] jaxpr_known, jaxpr_staged, out_unknowns, out_inst, num_res = \ pe.partial_eval_jaxpr_custom( jaxpr, in_unknowns, [True] * len(in_unknowns), False, False, policy) # DCE jaxpr_staged, keeping only instantiated outputs which are unknown _, out_inst_unknown = partition_list(out_inst, out_unknowns) jaxpr_unknown, in_used_staged = pe.dce_jaxpr(jaxpr_staged, out_inst_unknown) used_res, in_used_staged = split_list(in_used_staged, [num_res]) # DCE jaxpr_known, keeping all known outputs but discarding dce'd res out_used_known = [True] * (len(out_unknowns) - sum(out_unknowns)) + used_res jaxpr_known, in_used_known = pe.dce_jaxpr(jaxpr_known, out_used_known) num_res = sum(used_res) # compute known outputs and residuals (hoisted out of remat primitive) _, in_consts_ = unzip2(t.pval for t in tracers if t.pval.is_known()) _, in_consts = partition_list(in_used_known, in_consts_) out_consts = core.eval_jaxpr(jaxpr_known, (), *in_consts) out_knowns, residuals = split_list(out_consts, [len(out_consts)-num_res]) # set up unknown outputs with a recipe to call remat res_tracers = map(trace.new_instantiated_const, residuals) _, tracers_staged = partition_list(in_used_staged, tracers) in_jaxpr_tracers = res_tracers + map(trace.instantiate_const, tracers_staged) out_jaxpr_tracers = [pe.JaxprTracer(trace, pe.PartialVal.unknown(x.aval), None) for x in jaxpr_unknown.outvars] new_params = dict(params, jaxpr=jaxpr_unknown, differentiated=True) recipe = pe.new_eqn_recipe(in_jaxpr_tracers, out_jaxpr_tracers, remat_p, new_params, jaxpr_unknown.effects, source_info_util.current()) # log info about saved residuals log_level = logging.WARNING if config.log_checkpoint_residuals.value else logging.DEBUG if logger.isEnabledFor(log_level): try: _, staged_unk = partition_list(in_used_staged, in_unknowns) res_invars, _ = partition_list(staged_unk, jaxpr_unknown.invars[num_res:]) res_outvars = jaxpr_known.outvars[len(jaxpr_known.outvars) - num_res:] body_res = _saved_residuals(jaxpr_known.replace(outvars=res_outvars), None) logger.log(log_level, 'remat-decorated function ' + 'saving inputs with shapes:\n' * bool(res_invars) + ' %s\n' * len(res_invars) + 'and ' * bool(res_invars) * bool(body_res) + 'saving these intermediates:\n' * bool(body_res) + ' %s from %s\n' * len(body_res), *[v.aval.str_short() for v in res_invars], *[elt for (a, s) in body_res for elt in [a.str_short(), s]]) except: pass # just don't log anything on failure for t in out_jaxpr_tracers: t.recipe = recipe # zip together known and unknown outputs return merge_lists(out_unknowns, out_knowns, out_jaxpr_tracers) pe.custom_partial_eval_rules[remat_p] = remat_partial_eval def remat_partial_eval_custom_params_updater(*args): *_, params_known, params_staged = args return params_known, dict(params_staged, differentiated=True) pe.partial_eval_jaxpr_custom_rules[remat_p] = \ partial(pe.call_partial_eval_custom_rule, 'jaxpr', remat_partial_eval_custom_params_updater) def remat_transpose(out_cts, *in_primals, jaxpr, **params): assert not jaxpr.constvars in_linear = [ad.is_undefined_primal(x) for x in in_primals] out_zeros = [type(ct) is ad_util.Zero for ct in out_cts] transposed_jaxpr_, in_zeros = transpose_jaxpr( pe.close_jaxpr(jaxpr), in_linear, out_zeros) transposed_jaxpr, consts = transposed_jaxpr_.jaxpr, transposed_jaxpr_.consts transposed_jaxpr = pe.convert_constvars_jaxpr(transposed_jaxpr) args, _ = tree_flatten((in_primals, out_cts)) in_cts_nz = remat_p.bind(*consts, *args, jaxpr=transposed_jaxpr, **params) in_cts_nz_, in_zeros_ = iter(in_cts_nz), iter(in_zeros) in_cts = [None if not ad.is_undefined_primal(x) else ad_util.Zero(x.aval) if next(in_zeros_) else next(in_cts_nz_) for x in in_primals] assert next(in_cts_nz_, None) is next(in_zeros_, None) is None return in_cts ad.reducing_transposes[remat_p] = remat_transpose # TODO(mattjj): move this to ad.py def transpose_jaxpr(jaxpr: core.ClosedJaxpr, in_linear: bool | Sequence[bool], out_zeros: bool | Sequence[bool], ) -> tuple[core.ClosedJaxpr, list[bool]]: if type(in_linear) is bool: in_linear = (in_linear,) * len(jaxpr.in_avals) if type(out_zeros) is bool: out_zeros = (out_zeros,) * len(jaxpr.out_avals) return _transpose_jaxpr(jaxpr, tuple(in_linear), tuple(out_zeros)) @weakref_lru_cache def _transpose_jaxpr(jaxpr, in_lin, out_zeros): in_avals = ([a for a, lin in zip(jaxpr.in_avals, in_lin ) if not lin] + [a for a, zero in zip(jaxpr.out_avals, out_zeros) if not zero]) cell = lambda: None @lu.wrap_init def transposed(*args_flat): ins_flat, out_cts_flat = split_list(args_flat, [len(in_lin) - sum(in_lin)]) # Evaluate nonlinear parts using partial evaluation to get a linear jaxpr. ins_iter = iter(ins_flat) in_pvals = [pe.PartialVal.unknown(aval) if lin else pe.PartialVal.known(next(ins_iter)) for aval, lin in zip(jaxpr.in_avals, in_lin)] assert next(ins_iter, None) is None with source_info_util.extend_name_stack('rematted_computation'): lin_jaxpr, _, consts = pe.trace_to_jaxpr_nounits( lu.wrap_init(core.jaxpr_as_fun(jaxpr)), in_pvals, False) # Transpose the linear jaxpr (which only has linear inputs). out_cts_iter = iter(out_cts_flat) out_cts = [ad_util.Zero(aval) if zero else next(out_cts_iter) for aval, zero in zip(jaxpr.out_avals, out_zeros)] assert next(out_cts_iter, None) is None dummy_args = [ad.UndefinedPrimal(v.aval) for v in lin_jaxpr.invars] in_cts = ad.backward_pass(lin_jaxpr, False, consts, dummy_args, out_cts) # Identify symbolic zeros in the resulting cotangents, and return nonzeros. in_zeros = cell.in_cts_zero = [type(ct) is ad_util.Zero for ct in in_cts] in_cts_nz, _ = partition_list(in_zeros, in_cts) return in_cts_nz transposed_jaxpr_, _, consts, () = pe.trace_to_jaxpr_dynamic(transposed, in_avals) transposed_jaxpr = core.ClosedJaxpr(transposed_jaxpr_, consts) return transposed_jaxpr, cell.in_cts_zero # pytype: disable=attribute-error def remat_vmap(spmd_axis_name, axis_size, axis_name, main_type, args, dims, *, jaxpr, **params): assert not jaxpr.constvars jaxpr_batched_, out_batched = batching.batch_jaxpr_axes( pe.close_jaxpr(jaxpr), axis_size, dims, [batching.zero_if_mapped] * len(jaxpr.outvars), axis_name=axis_name, spmd_axis_name=spmd_axis_name, main_type=main_type) jaxpr_batched, consts = jaxpr_batched_.jaxpr, jaxpr_batched_.consts if consts: jaxpr_batched = pe.convert_constvars_jaxpr(jaxpr_batched) out_dims = [0 if b else None for b in out_batched] return remat_p.bind(*consts, *args, jaxpr=jaxpr_batched, **params), out_dims batching.axis_primitive_batchers[remat_p] = partial(remat_vmap, None) batching.spmd_axis_primitive_batchers[remat_p] = remat_vmap # TODO(mattjj,sharadmv): de-duplicate with pe.dce_jaxpr_call_rule def remat_dce(used_outputs: list[bool], eqn: core.JaxprEqn ) -> tuple[list[bool], core.JaxprEqn | None]: new_jaxpr, used_inputs = pe.dce_jaxpr(eqn.params['jaxpr'], used_outputs) new_params = dict(eqn.params, jaxpr=new_jaxpr) if (not any(used_inputs) and not any(used_outputs) and _has_effects(new_jaxpr.effects)): return used_inputs, None else: new_eqn = pe.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, eqn.ctx) return used_inputs, new_eqn pe.dce_rules[remat_p] = remat_dce def _has_effects(effects) -> bool: return bool({e for e in effects if not isinstance(e, core.NamedAxisEffect)}) def remat_lowering(*args, jaxpr: core.Jaxpr, prevent_cse: bool, differentiated: bool, is_gpu_platform: bool = False, **_): assert not jaxpr.constvars if differentiated and prevent_cse: if config.remat_opt_barrier.value: translation_rule = _remat_translation_using_opt_barrier elif is_gpu_platform: translation_rule = _remat_translation_using_while else: translation_rule = _remat_translation_using_cond else: translation_rule = lambda *args, jaxpr: core.eval_jaxpr(jaxpr, (), *args) return api.named_call(translation_rule, name="checkpoint")(*args, jaxpr=jaxpr) def _remat_translation_using_opt_barrier(*args, jaxpr: core.Jaxpr): args = _optimization_barrier(args) return core.eval_jaxpr(jaxpr, (), *args) # TODO(mattjj): add core utility for 'create dummy value for this type'? def _dummy_like(aval: core.AbstractValue) -> Any: if aval is core.abstract_token: return lax_internal.create_token() elif isinstance(aval, (core.ShapedArray, core.DShapedArray)): return lax_internal.broadcast(lax_internal.empty(aval.dtype), aval.shape) # type: ignore else: raise ValueError(aval) def _remat_translation_using_while(*args, jaxpr: core.Jaxpr): # Implements: # for(counter=0, result=0; counter < rng(1, 2); counter ++) { # result = eval_jaxpr(*args) # } # The loop carry is a tuple: (counter, result, args) from jax._src.lax import control_flow as lax_control_flow avals_out = tuple(v.aval for v in jaxpr.outvars) carry_init = (np.int32(0), tuple(map(_dummy_like, avals_out)), args) def cond(carry): counter, _, _ = carry unif = lax_internal.rng_uniform(np.int32(1), np.int32(2), shape=()) return counter < unif def body(carry): counter, _, args = carry results = core.eval_jaxpr(jaxpr, (), *args) return (counter + 1, tuple(results), args) carry_res = lax_control_flow.while_loop(cond, body, carry_init) return carry_res[1] def _remat_translation_using_cond(*args, jaxpr: core.Jaxpr): # Implements: # if(rng(0, 1) < 2) # return eval_jaxpr(*args) # else: # return 0 from jax._src.lax import control_flow as lax_control_flow avals_out = tuple(v.aval for v in jaxpr.outvars) def remat_comp(*args): return tuple(core.eval_jaxpr(jaxpr, (), *args)) def dummy_comp(*args): return tuple(map(_dummy_like, avals_out)) unif = lax_internal.rng_uniform(np.float32(0), np.float32(1), shape=()) return lax_control_flow.cond(unif < np.float32(2), remat_comp, dummy_comp, *args) mlir.register_lowering( remat_p, mlir.lower_fun(remat_lowering, multiple_results=True)) mlir.register_lowering( remat_p, mlir.lower_fun(partial(remat_lowering, is_gpu_platform=True), multiple_results=True), platform="gpu") def _optimization_barrier_abstract_eval(*args): return args def _optimization_barrier_lowering_rule(ctx, *args): barrier_types = map(mlir.aval_to_ir_types, ctx.avals_in) flat_args = mlir.flatten_lowering_ir_args(args) barrier_op = hlo.OptimizationBarrierOp(flat_args) return util.unflatten(barrier_op.results, map(len, barrier_types)) def _optimization_barrier(arg): flat_args, treedef = tree_flatten(arg) return tree_unflatten(treedef, optimization_barrier_p.bind(*flat_args)) optimization_barrier_p = core.Primitive('optimization_barrier') optimization_barrier_p.multiple_results = True optimization_barrier_p.def_impl( partial(dispatch.apply_primitive, optimization_barrier_p)) optimization_barrier_p.def_abstract_eval(_optimization_barrier_abstract_eval) mlir.register_lowering(optimization_barrier_p, _optimization_barrier_lowering_rule) def checkpoint_name(x, name): return name_p.bind(x, name=name) name_p.def_impl(lambda x, *, name: x) name_p.def_abstract_eval(lambda x, *, name: x) def name_jvp(primals, tangents, *, name): (x,), (xdot,) = primals, tangents return name_p.bind(x, name=name), xdot # don't name the tangent value ad.primitive_jvps[name_p] = name_jvp mlir.register_lowering(name_p, lambda ctx, x, *, name: [x]) def name_batcher(args, dims, *, name): (x,), (d,) = args, dims return name_p.bind(x, name=name), d batching.primitive_batchers[name_p] = name_batcher @functools.wraps(checkpoint) def checkpoint_wrapper( fun: Callable, *, concrete: bool = False, prevent_cse: bool = True, static_argnums: int | tuple[int, ...] = (), policy: Callable[..., bool] | None = None, ) -> Callable: if concrete: msg = ("The 'concrete' option to jax.checkpoint / jax.remat is deprecated; " "in its place, you can use its `static_argnums` option, and if " "necessary the `jax.ensure_compile_time_eval()` context manager.\n" "\n" "For example, if using `concrete=True` for an `is_training` flag:\n" "\n" " from functools import partial\n" "\n" " @partial(jax.checkpoint, concrete=True)\n" " def foo(x, is_training):\n" " if is_training:\n" " return f(x)\n" " else:\n" " return g(x)\n" "\n" "replace it with a use of `static_argnums`:\n" "\n" " @partial(jax.checkpoint, static_argnums=(1,))\n" " def foo(x, is_training):\n" " ...\n" "\n" "If jax.numpy operations need to be performed on static arguments, " "we can use the `jax.ensure_compile_time_eval()` context manager. " "For example, we can replace this use of `concrete=True`\n:" "\n" " @partial(jax.checkpoint, concrete=True)\n" " def foo(x, y):\n" " if y > 0:\n" " return f(x)\n" " else:\n" " return g(x)\n" "\n" "with this combination of `static_argnums` and " "`jax.ensure_compile_time_eval()`:\n" "\n" " @partial(jax.checkpoint, static_argnums=(1,))\n" " def foo(x, y):\n" " with jax.ensure_compile_time_eval():\n" " y_pos = y > 0\n" " if y_pos:\n" " return f(x)\n" " else:\n" " return g(x)\n" "\n" "See https://jax.readthedocs.io/en/latest/jep/11830-new-remat-checkpoint.html\n") raise NotImplementedError(msg) return checkpoint(fun, prevent_cse=prevent_cse, policy=policy, static_argnums=static_argnums)