# mypy: ignore-errors """ This module implements distributed training optimizations for TorchDynamo backends. It provides functionality to optimize models wrapped in DistributedDataParallel (DDP) by intelligently splitting compiled graphs to align with DDP's gradient synchronization boundaries. Key features include: - Graph partitioning based on parameter bucket sizes - Optimization of allreduce operations for distributed training - Support for parameter ignoring and buffer handling - Submodule compilation and management - Debugging utilities for distributed training The main component is the DDPOptimizer class, which handles graph splitting and recompilation to enable efficient distributed training while maintaining the benefits of compilation. """ import logging import traceback from dataclasses import dataclass, field from typing import Any, Optional from unittest import mock import torch from torch import fx from torch._dynamo.output_graph import GraphCompileReason from torch._dynamo.utils import deepcopy_to_fake_tensor, detect_fake_mode from torch._logging import trace_structured from torch.fx.node import Node # Regular log messages should go through 'log'. # ddp_graph_log is a separate artifact logger reserved for dumping graphs. # See docs/source/logging.rst for more info. log = logging.getLogger(__name__) ddp_graph_log = torch._logging.getArtifactLogger(__name__, "ddp_graphs") def args_str(args): # a debug helper if torch.is_tensor(args): return f"T[{args.shape}]" elif isinstance(args, tuple): return f"tuple({', '.join([args_str(x) for x in args])})" elif isinstance(args, list): return f"list({', '.join([args_str(x) for x in args])})" else: return str(args) @dataclass class Bucket: size: int = 0 params: list[str] = field(default_factory=list) nodes: list[fx.Node] = field(default_factory=list) # param_ids is just used for unit testing param_ids: list = field(default_factory=list) # keep track of any buckets that were extended for logging purposes opcount_increased_to_capture_external_output: int = 0 paramsize_before_opcount_increase: int = 0 def bucket_has_external_output(bucket: Bucket) -> bool: nodes_in_bucket = set() # we want to iterate in reverse order, but clumsi-luckily the bucket.nodes list was already created backwards # so we don't reverse it here for node in bucket.nodes: # assume node.op != output, since those are filtered in the original iteration nodes_in_bucket.add(node) for user in node.users: if user not in nodes_in_bucket: return True return False def pretty_print_buckets(buckets: list[Bucket], bucket_bytes_cap: int): headers = ("Index", "Size (b)", "Param Names") rows = [] extended_buckets = [] for idx, bucket in enumerate(reversed(buckets)): if len(bucket.params) > 0: rows.append((idx, bucket.size, bucket.params[0])) rows.extend((None, None, param) for param in bucket.params[1:]) if bucket.opcount_increased_to_capture_external_output > 0: extended_buckets.append( ( idx, bucket.opcount_increased_to_capture_external_output, bucket.size - bucket.paramsize_before_opcount_increase, ) ) if len(rows): log.info( "\nDDPOptimizer used bucket cap %s and created %d buckets. Enable debug logs for detailed bucket info.", bucket_bytes_cap, len(buckets), ) if len(extended_buckets): log.warning( "Some buckets were extended beyond their requested parameter capacities" " in order to ensure each subgraph has an output node, required for fx graph partitioning." " This can be the case when a subgraph would have only contained nodes performing inplace mutation," " and returning no logical outputs. This should not be a problem, unless it results in too few graph" " partitions for optimal DDP performance." ) try: from tabulate import tabulate log.debug( "\nDDPOptimizer produced the following bucket assignments:\n%s", tabulate(rows, headers=headers, tablefmt="simple_grid"), ) if len(extended_buckets): log.warning( "DDPOptimizer extended these buckets to ensure per-subgraph output nodes:\n%s", tabulate( extended_buckets, headers=("Index", "Extra Ops", "Extra Param Size (b)"), tablefmt="simple_grid", ), ) except ImportError: log.debug( "Please `pip install tabulate` in order to display ddp bucket sizes and diagnostic information." ) else: log.debug("DDPOptimizer captured no parameters and did not split this graph.") def has_higher_order_op(gm): # Check if there is a higher order op in the graph for node in gm.graph.nodes: if node.op == "get_attr": maybe_param = getattr(gm, node.target) if isinstance(maybe_param, torch.fx.GraphModule): return True return False # compile each of the partitioned submodules using the user-provided compiler class SubmodCompiler(torch.fx.interpreter.Interpreter): def __init__(self, module, compiler, fake_mode) -> None: super().__init__(module) self.compiler = compiler self.fake_mode = fake_mode def compile_submod(self, input_mod, args, kwargs): """ Compile the submodule, using a wrapper to make sure its output is always a tuple, which is required by AotAutograd based compilers """ assert len(kwargs) == 0, "We assume only args for these modules" class WrapperModule(torch.nn.Module): def __init__(self, submod, unwrap_singleton_tuple) -> None: super().__init__() self.submod = submod self.unwrap_singleton_tuple = unwrap_singleton_tuple def forward(self, *args): x = self.submod(*args) # TODO(whc) # for some reason the isinstance check is necessary if I split one node per submod # - even though I supposedly wrapped the output in a tuple in those cases, the real # compiled module was still returning a tensor if self.unwrap_singleton_tuple and isinstance(x, (tuple, list)): return x[0] return x unwrap_singleton_tuple = False for sn in input_mod.graph.nodes: if sn.op == "output": if not isinstance(sn.args[0], tuple): unwrap_singleton_tuple = True sn.args = (sn.args,) input_mod.recompile() input_mod.compile_subgraph_reason = GraphCompileReason( "DDPOptimizer intentional graph-break (See Note [DDPOptimizer])." " Set `torch._dynamo.config.optimize_ddp = False` to disable.", [ # it's close to useless to get a real stacktrace here, and quite verbose. traceback.FrameSummary(__file__, 0, DDPOptimizer), ], ) wrapper = WrapperModule( self.compiler(input_mod, args), unwrap_singleton_tuple, ) return wrapper # Note: # # The way distributed works today around fake tensors can be somewhat confusing. # Some of these codepaths are shared in both runtime, and compile time. The presence # of a fake_mode, read off of fake tensor inputs, dictates how we will operate. # # A few things to keep in mind: # # 1) We invoke `compile_submod` with a real module. The output of that gets stored # on the graph via `self.module.add_submodule(n.target, compiled_submod_real)`. # # 2) When running a call_module targeted node, if we have a fake_mode, we fakify the # module we got from self.fetch_attr(n.target). Regardless of fake_mode, we then execute it. # # 3) Fake tensors should always be around during compile time. # # 4) Fake tensors should never be around at runtime. # # 5) We end up with a compilation mode that takes a real submodule and fake tensors, # to match what aot_autograd expects. See Note: [Fake Modules and AOTAutograd] def run_node(self, n: Node) -> Any: args, kwargs = self.fetch_args_kwargs_from_env(n) new_args = [] assert self.fake_mode for arg in args: if isinstance(arg, torch.Tensor) and not isinstance( arg, torch._subclasses.FakeTensor ): new_args.append(torch._dynamo.utils.to_fake_tensor(arg, self.fake_mode)) else: new_args.append(arg) log.debug("run_node %s, %s got args %s", n.op, n.target, args_str(args)) assert isinstance(args, tuple) assert isinstance(kwargs, dict) if n.op == "call_module": real_mod = self.fetch_attr(n.target) if self.fake_mode: curr_submod = deepcopy_to_fake_tensor(real_mod, self.fake_mode) else: curr_submod = real_mod ddp_graph_log.debug("\n---%s graph---\n%s", n.target, curr_submod.graph) # When calling the compiler on the submod, inputs (new_args) are expected to # be FakeTensors already since Dynamo would have made them FakeTensors in the # non-DDP flow. However, the parameters are _not_ expected to be FakeTensors, # since this wrapping happens during compilation # Note: Returning Fake Tensors on First AOT Autograd Call # # Inductor will optimize strides of outputs when it deems it profitable. # For instance, converting to channels last. When we split the graph here # into multiple inductor compilations, we need to make sure that the # output strides of one compilation is appropriately passed to the subsequent # compilations. However, the mapping from inductor output to dynamo output # is non-trivial due to aot_autograd's deduping, de-aliasing, mutation, re-writing, # subclass handling, etc. In order to replay all this logic we set a flag such that # the first invocation of inductor in aot_autograd will return Fake Tensors with # appropriate strides. Then, all of aot autograd's runtime logic is replayed. # This gives us the appropriately strided outputs here which will reflect runtime strides. class FakeifyFirstAOTInvocationGuard: def __init__(self) -> None: self.tc = torch._guards.TracingContext.try_get() assert self.tc torch._guards.TracingContext.try_get().fakify_first_call = True def __del__(self) -> None: self.tc.fakify_first_call = False # For aot_eager and other backends, tracing context is not set has_tracing_context = torch._guards.TracingContext.try_get() is not None if has_tracing_context: g = FakeifyFirstAOTInvocationGuard() # noqa: F841 from torch._dynamo.utils import counters init = counters["aot_autograd"]["total"] compiled_submod_real = self.compile_submod(real_mod, new_args, kwargs) # TODO - better way of doing this? # Only aot autograd handles fakifying first call invoked_aot_autograd = init != counters["aot_autograd"]["total"] # We update the original (outer) graph with a call into the compiled module # instead of the uncompiled one. self.module.delete_submodule(n.target) n.target = "compiled_" + n.target self.module.add_submodule(n.target, compiled_submod_real) # Finally, we have to produce inputs for use compiling the next submodule, # and these need to be FakeTensors, so we execute the module under fake_mode # Because parameters are not fake we patch fake tensor mode to allow non fake inputs with ( self.fake_mode, mock.patch.object(self.fake_mode, "allow_non_fake_inputs", True), ): if has_tracing_context and invoked_aot_autograd: out = compiled_submod_real(*new_args, **kwargs) # output should be fake or subclass assert all( (not isinstance(t, torch.Tensor) or type(t) is not torch.Tensor) for t in (out if isinstance(out, (list, tuple)) else [out]) ) return out else: return curr_submod(*new_args, **kwargs) else: # placeholder or output nodes don't need to get compiled, just executed return getattr(self, n.op)(n.target, new_args, kwargs) class DDPOptimizer: """Note [DDPOptimizer] DDPOptimizer applies when dynamo compiles models wrapped in DistributedDataParallel (DDP), breaking the dynamo graph into chunks to compile separately, with the breaks aligning to the boundaries of gradient-allreduce buckets chosen by DDP. Background/Motivation - DDP uses allreduce collectives to synchronize partial gradients computed on different workers - DDP groups gradient allreduces into 'buckets' to optimize communication efficiency of all-reduce - Parameters grouped into buckets are assumed to be adjacent in time, so they become ready at around the same time during backward and thus can share the same allreduce efficiently - Allreduces must overlap with backward compute for optimal training performance - DDP schedules allreduces using 'hooks' fired from the c++ autograd engine in pytorch, which operates when individual grads become 'ready' - Dynamo+AOTAutograd produces a single fused graph that runs 'atomically' from the perspective of the autograd engine, such that all gradients become 'ready' at the same time. Hooks fire after the whole fused backward function executes, preventing any overlap of compute and communication Algorithm - DDPOptimizer starts off with an FX graph traced by dynamo which represents forward. It can traverse this graph in reverse order to determine the true order that gradients will become ready during backward. - Parameter sizes are counted in reverse order, up to a bucket size limit, at which point a new bucket is started and a graph break introduced - Each of the subgraphs is compiled by the compiler provided to dynamo by the user, and then fused back together into an outer module that is returned to the user Notes - It would be better to enforce (by adding an API to DDP) that the bucket splits chosen here are used by DDP, and that DDP does not need to detect or optimize bucket order by observing execution at runtime, as it does in eager. - If Dynamo can't capture a whole graph for the portion of the model wrapped by DDP, this algorithm will currently produce splits that do not necessarily align with the buckets used by DDP. This should result in performance degradation approaching the baseline case where graph-splits are not used, but not worse. - If the backend compiler fails to compile a single subgraph, it will execute eagerly despite the rest of the subgraphs being compiled - DDP has a 'parameters_and_buffers_to_ignore' field, which DDPOptimizer attempts to honor by reading markers left by DDP on individual parameters. In cases where other transformations, such as reparameterization, are also used, the ignore markers could be lost. If DDPOptimizer fails to ignore a parameter ignored by DDP, it is not catastrophic but could impact performance by choosing sub-optimal bucket splits. - DDPOptimizer always ignores all buffers, regardless of their ignore flag, since buffers do not require gradients, and therefore aren't allreduced by DDP. (They are broadcast during forward, but this is not covered by DDPOptimizer) Debugging - Generally, it is easiest to debug DDPOptimizer in a single process program, using pdb. - In many cases, the log messages are helpful (they show bucket size assignments)- just set TORCH_LOGS env to include any of 'dynamo', 'distributed', or 'dist_ddp'. - See `benchmarks/dynamo/distributed.py` for a simple harness that will run a toy model or a torchbench model in a single process (or with torchrun, in multiple processes) Args: bucket_bytes_cap (int): Controls the size of buckets, in bytes, used to determine graphbreaks. Should be set to match the equivalent parameter on the original DDP module. backend_compile_fn (callable): A dynamo compiler function, to be invoked to compile each subgraph. first_bucket_cap (int): Controls the size of the first bucket. Should match DDP's first bucket cap. DDP special-cases the first bucket size since it is sometimes optimal to start a small allreduce early. """ def __init__( self, bucket_bytes_cap: int, backend_compile_fn, first_bucket_cap: Optional[int] = None, ) -> None: if first_bucket_cap is not None: self.first_bucket_cap = first_bucket_cap elif torch.distributed.is_available(): # this constant comes from C10D lib which is not always built self.first_bucket_cap = torch.distributed._DEFAULT_FIRST_BUCKET_BYTES else: self.first_bucket_cap = bucket_bytes_cap self.bucket_bytes_cap = bucket_bytes_cap assert self.first_bucket_cap <= self.bucket_bytes_cap, ( "First bucket should be smaller/equal to other buckets to get comms warmed up ASAP" ) self.backend_compile_fn = backend_compile_fn def _ignore_parameter(self, parameter): return hasattr(parameter, "_ddp_ignored") and parameter._ddp_ignored def add_param(self, bucket, param, name): bucket.size += param.untyped_storage().nbytes() bucket.params.append(name) bucket.param_ids.append(id(param)) def add_module_params_to_bucket(self, mod, bucket, processed_modules, prefix): processed_modules.add(mod) for name, param in mod.named_parameters(): if param.requires_grad and not self._ignore_parameter(param): self.add_param(bucket, param, f"{prefix}_{name}") def add_param_args(self, bucket, node): for arg in node.args: if not isinstance(arg, torch.fx.node.Node): continue if arg.op != "placeholder": continue param = arg.meta["example_value"] if ( isinstance(param, torch.nn.Parameter) and param.requires_grad and not self._ignore_parameter(param) ): self.add_param(bucket, param, arg.target) def compile_fn(self, gm: fx.GraphModule, example_inputs: list[torch.Tensor]): """ Implements graph splitting, first determining a set of of buckets by counting parameter sizes in reverse graph order, then invoking the user/backend compiler to compile each subgraph. Finally, stiches compiled graphs into one graphmodule and returns its callable. """ # 1: compute the partition map according to DDP bucket logic buckets = [Bucket()] # (size, param_names) processed_modules = set() for node in reversed(gm.graph.nodes): if node.op in ("output", "placeholder"): continue if ( buckets[0].size >= self.bucket_bytes_cap or len(buckets) == 1 and buckets[0].size >= self.first_bucket_cap ): if bucket_has_external_output(buckets[0]): buckets.insert(0, Bucket()) else: # continue building this bucket past the point of filling its parameter capacity, # to increase chances it contains at least one node that is either a global output or # passed as input to a subsequent graph if buckets[0].opcount_increased_to_capture_external_output == 0: buckets[0].paramsize_before_opcount_increase = buckets[0].size buckets[0].opcount_increased_to_capture_external_output += 1 if node.op == "call_function": self.add_param_args(buckets[0], node) elif node.op == "call_module": target_mod = gm.get_submodule(node.target) if target_mod not in processed_modules: self.add_module_params_to_bucket( target_mod, buckets[0], processed_modules, node.target ) elif node.op == "call_method": if isinstance(node.args[0].target, str): target_mod = None try: target_mod = gm.get_submodule(node.args[0].target) except AttributeError: pass if target_mod is not None and target_mod not in processed_modules: self.add_module_params_to_bucket( target_mod, buckets[0], processed_modules, node.target ) # This handles situations like tmp = torch.mm(x, self.weight.t()) # t: "f32[512, 512]" = l_self_seq_2_weight.t(); l_self_seq_2_weight = None # tmp: "f32[512, 512]" = torch.mm(input_2, t); input_2 = t = None self.add_param_args(buckets[0], node) elif node.op == "get_attr": maybe_param = getattr(gm, node.target) if ( isinstance(maybe_param, torch.nn.Parameter) and maybe_param.requires_grad and not self._ignore_parameter(maybe_param) ): self.add_param(buckets[0], maybe_param, node.target) # All nodes have to be mapped to a bucket, even if they don't have their own params # Ignored params still end up in buckets, we just don't count them towards the capacity buckets[0].nodes.append(node) if len(buckets) > 1 and buckets[0].size == 0: # we collected a small preamble graph with ops that don't include parameters, fuse it back buckets[1].nodes.extend(buckets[0].nodes) assert len(buckets[0].params) == 0, "Params should be empty if size is 0" del buckets[0] # stash buckets for testing/debugging purposes self.buckets = buckets pretty_print_buckets(buckets, self.bucket_bytes_cap) if len(buckets) == 1: # bypass split/fuse logic if there is only one bucket return self.backend_compile_fn(gm, example_inputs) # 2: partition the graphmodule according to bucket capacity partition_map = {} for idx, b in enumerate(buckets): for node in b.nodes: partition_map[node] = idx split_gm = fx.passes.split_module.split_module( gm, None, lambda node: partition_map[node] ) debug_str = ( f"\n---orig graph---\n{gm.graph}\n" + f"\n---split graph---\n{split_gm.graph}\n" ) for name, module in split_gm.named_modules(): if "." not in name and len(name): # only print the submod graphs, not their children debug_str += f"\n---{name} graph---\n{module.graph}\n" debug_str += "\n---------------\n" ddp_graph_log.debug(debug_str) trace_structured( "optimize_ddp_split_graph", payload_fn=lambda: split_gm.print_readable(print_output=False), ) for name, module in split_gm.named_modules(): if "." not in name and len(name): trace_structured( "optimize_ddp_split_child", lambda: {"name": name}, payload_fn=lambda: module.print_readable(print_output=False), ) fake_mode = detect_fake_mode(example_inputs) if fake_mode is None: fake_mode = torch._subclasses.fake_tensor.FakeTensorMode() submod_compiler = SubmodCompiler(split_gm, self.backend_compile_fn, fake_mode) with torch._dynamo.utils._disable_saved_tensors_hooks_during_tracing(): submod_compiler.run(*example_inputs) split_gm.recompile() ddp_graph_log.debug( "\n---final graph---\n%s\n---------------\n", split_gm.graph ) return split_gm