parallelize_llama.py

# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

# This file applies the PT-D parallelisms (except pipeline parallelism) and various
# training techniques (e.g. activation checkpointing and compile) to the Llama model.

from collections import defaultdict

import torch
import torch.nn as nn
from torch.distributed import DeviceMesh
from torch.distributed._composable.fsdp import fully_shard, MixedPrecisionPolicy
from torch.distributed._composable.replicate import replicate
from torch.distributed._tensor import Replicate, Shard
from torch.distributed.algorithms._checkpoint.checkpoint_wrapper import (
    checkpoint_wrapper as ptd_checkpoint_wrapper,
)
from torch.distributed.tensor.parallel import (
    ColwiseParallel,
    parallelize_module,
    PrepareModuleInput,
    RowwiseParallel,
    SequenceParallel,
)

from torchtitan.config_manager import JobConfig, TORCH_DTYPE_MAP
from torchtitan.logging import logger
from torchtitan.parallelisms.parallel_dims import ParallelDims


# NOTE(lty): experimental for the PT-D 24 research internship project
def torch_spmd_parallelize(
    model: nn.Module,
    world_mesh: DeviceMesh,
    parallel_dims: ParallelDims,
    job_config: JobConfig,
):
    # ensure full graph compile on non-llama models
    torch._dynamo.config.capture_scalar_outputs = True
    torch._dynamo.config.capture_dynamic_output_shape_ops = True

    # simple fsdp configs
    torch._inductor.config.simplefsdp.bucket_mode = "greedy"
    torch._inductor.config.simplefsdp.enable_reorder = True
    torch._inductor.config.simplefsdp.degree = parallel_dims.dp
    torch._inductor.config.simplefsdp.tp_enabled = parallel_dims.tp_enabled
    torch._inductor.config.simplefsdp.pp_enabled = parallel_dims.pp_enabled
    torch._inductor.config.simplefsdp.device_mesh = world_mesh.mesh.tolist()

    print("enable reorder", torch._inductor.config.simplefsdp.enable_reorder)
    if torch._inductor.config.simplefsdp.bucket_mode == "transformer_block":
        print("enable block-level bucket")
    elif torch._inductor.config.simplefsdp.bucket_mode == "greedy":
        print("enable greedy auto bucket")
        print(
            "ag_comm_time_multiplier",
            torch._inductor.config.simplefsdp.ag_comm_time_multiplier,
        )
        print(
            "rs_comm_time_multiplier",
            torch._inductor.config.simplefsdp.rs_comm_time_multiplier,
        )

    if parallel_dims.tp_enabled:
        apply_tp(
            model,
            world_mesh["tp"],
            loss_parallel=parallel_dims.loss_parallel_enabled,
            enable_float8=job_config.float8.enable_float8_linear,
            enable_async_tp=job_config.experimental.enable_async_tensor_parallel,
        )

    ac_config = job_config.activation_checkpoint
    if ac_config.mode != "none":
        apply_ac(model, ac_config)

    if parallel_dims.dp_enabled:
        from torch_spmd.data_parallel import data_parallel, MixedPrecisionPolicy

        mp_policy = MixedPrecisionPolicy(
            param_dtype=TORCH_DTYPE_MAP[job_config.training.mixed_precision_param],
            reduce_dtype=TORCH_DTYPE_MAP[job_config.training.mixed_precision_reduce],
        )
        dp_mesh = world_mesh["dp"] if world_mesh.ndim > 1 else world_mesh

        model = data_parallel(
            model,
            dp_mesh,
            mode="fully_shard",
            ac_mode=ac_config.mode,
            mp_policy=mp_policy,
        )
        logger.info("Applied Simple FSDP to the model")

    if job_config.training.compile:
        model = torch.compile(model, fullgraph=True)
        logger.info("Compiling with torch.compile")

    return model


def parallelize_llama(
    model: nn.Module,
    world_mesh: DeviceMesh,
    parallel_dims: ParallelDims,
    job_config: JobConfig,
):
    """
    Apply tensor parallelism, activation checkpointing, torch.compile, and data
    parallelism to the model.

    NOTE: The passed-in model preferably should be on meta device. Otherwise,
    the model must fit on GPU or CPU memory.
    """
    # NOTE(lty): experimental for the PT-D 24 research internship project
    if job_config.experimental.torch_spmd:
        return torch_spmd_parallelize(model, world_mesh, parallel_dims, job_config)

    if parallel_dims.tp_enabled:
        if (
            job_config.experimental.enable_async_tensor_parallel
            and not job_config.training.compile
        ):
            raise RuntimeError("Async TP requires --training.compile")
        apply_tp(
            model,
            world_mesh["tp"],
            loss_parallel=parallel_dims.loss_parallel_enabled,
            enable_float8=job_config.float8.enable_float8_linear,
            enable_async_tp=job_config.experimental.enable_async_tensor_parallel,
        )

    if job_config.activation_checkpoint.mode != "none":
        apply_ac(model, job_config.activation_checkpoint)

    # turn on per-TransformerBlock compile after AC wrapping and before FSDP
    if job_config.training.compile:
        if job_config.model.norm_type == "fused_rmsnorm":
            raise NotImplementedError(
                "fused_rmsnorm is not compatible with torch.compile yet. "
                "Please use rmsnorm or layernorm."
            )
        apply_compile(model)

    if parallel_dims.dp_enabled:
        if parallel_dims.dp_type == "fsdp":
            dp_mesh = world_mesh["dp"] if world_mesh.ndim > 1 else world_mesh
            assert dp_mesh.mesh_dim_names == ("dp",), dp_mesh.mesh_dim_names

            apply_fsdp(
                model,
                dp_mesh,
                param_dtype=TORCH_DTYPE_MAP[job_config.training.mixed_precision_param],
                reduce_dtype=TORCH_DTYPE_MAP[
                    job_config.training.mixed_precision_reduce
                ],
                tp_enabled=parallel_dims.tp_enabled,
                pp_enabled=parallel_dims.pp_enabled,
            )
        else:
            if world_mesh.ndim > 1:
                raise RuntimeError("DDP has not supported > 1D parallelism")
            apply_ddp(
                model,
                world_mesh,
                enable_compile=job_config.training.compile,
                enable_compiled_autograd=job_config.experimental.enable_compiled_autograd,
            )


def apply_tp(
    model: nn.Module,
    tp_mesh: DeviceMesh,
    loss_parallel: bool,
    enable_float8: bool,
    enable_async_tp: bool,
):
    """Apply tensor parallelism."""
    # 1. Parallelize the embedding and shard its outputs (which are the first
    # transformer block's inputs)
    # 2. Parallelize the root norm layer over the sequence dim
    # 3. Parallelize the final linear output layer
    parallelize_module(
        model,
        tp_mesh,
        {
            "tok_embeddings": RowwiseParallel(
                input_layouts=Replicate(),
                output_layouts=Shard(1),
            ),
            "norm": SequenceParallel(),
            "output": ColwiseParallel(
                input_layouts=Shard(1),
                output_layouts=Shard(-1) if loss_parallel else Replicate(),
                use_local_output=not loss_parallel,
            ),
        },
    )

    # Parallel styles used for transformer block linear weights and their
    # inputs may be different for float8 linears
    if enable_float8:
        # TODO(vkuzo): once float8 configuration supports delayed scaling,
        # add a check here to enforce supported float8 all-gather configurations
        # TODO(vkuzo): add the items below to __init__.py of torchao.float8 and import from there
        from torchao.float8.float8_tensor_parallel import (
            Float8ColwiseParallel,
            Float8RowwiseParallel,
            PrepareFloat8ModuleInput,
        )

        rowwise_parallel, colwise_parallel, prepare_module_input = (
            Float8RowwiseParallel,
            Float8ColwiseParallel,
            PrepareFloat8ModuleInput,
        )
    else:
        rowwise_parallel, colwise_parallel, prepare_module_input = (
            RowwiseParallel,
            ColwiseParallel,
            PrepareModuleInput,
        )

    # Apply tensor + sequence parallelism to every transformer block
    # NOTE: At the cost of model code change, we can accelerate Sequence Parallel
    #       by folding (and unfolding) the batch dimension and the sequence dimension.
    #       Examples can be found at https://github.com/pytorch/torchtitan/pull/437
    for layer_id, transformer_block in model.layers.items():
        layer_plan = {
            "attention_norm": SequenceParallel(),
            "attention": prepare_module_input(
                input_layouts=(Shard(1), None),
                desired_input_layouts=(Replicate(), None),
            ),
            "attention.wq": colwise_parallel(),
            "attention.wk": colwise_parallel(),
            "attention.wv": colwise_parallel(),
            "attention.wo": rowwise_parallel(output_layouts=Shard(1)),
            "ffn_norm": SequenceParallel(),
            "feed_forward": prepare_module_input(
                input_layouts=(Shard(1),),
                desired_input_layouts=(Replicate(),),
            ),
            "feed_forward.w1": colwise_parallel(),
            "feed_forward.w2": rowwise_parallel(output_layouts=Shard(1)),
            "feed_forward.w3": colwise_parallel(),
        }

        parallelize_module(
            module=transformer_block,
            device_mesh=tp_mesh,
            parallelize_plan=layer_plan,
        )

    if enable_async_tp:
        from torch.distributed._symmetric_memory import enable_symm_mem_for_group

        torch._inductor.config._micro_pipeline_tp = True
        enable_symm_mem_for_group(tp_mesh.get_group().group_name)

    logger.info(
        f"Applied {'Float8 ' if enable_float8 else ''}{'Async ' if enable_async_tp else ''}"
        "Tensor Parallelism to the model"
    )


# for selective op activation checkpointing
_save_list = {
    torch.ops.aten.mm.default,
    torch.ops.aten._scaled_dot_product_efficient_attention.default,
    torch.ops.aten._scaled_dot_product_flash_attention.default,
    torch.ops._c10d_functional.reduce_scatter_tensor.default,
}


def _apply_ac_to_transformer_block(module: nn.Module, ac_config):
    valid_ac_modes = ("full", "selective")
    if ac_config.mode not in valid_ac_modes:
        raise ValueError(
            f"Invalid AC mode: {ac_config.mode}. Valid modes: {valid_ac_modes}"
        )

    if ac_config.mode == "full":
        return ptd_checkpoint_wrapper(module, preserve_rng_state=False)

    assert ac_config.mode == "selective", f"{ac_config.mode}"
    use_op_sac = ac_config.selective_ac_option == "op"
    use_layer_sac = ac_config.selective_ac_option.isdigit()
    if not use_op_sac and not use_layer_sac:
        raise ValueError(
            f"Invalid selective AC option: {ac_config.selective_ac_option}. "
            f"Valid options: 'op' or a positive int representing layer frequency"
        )
    if use_op_sac:
        from torch.utils.checkpoint import (
            CheckpointPolicy,
            create_selective_checkpoint_contexts,
        )

        def _get_custom_policy(meta):
            def _custom_policy(ctx, func, *args, **kwargs):
                mode = "recompute" if ctx.is_recompute else "forward"
                mm_count_key = f"{mode}_mm_count"
                if func == torch.ops.aten.mm.default:
                    meta[mm_count_key] += 1
                # Saves output of all compute ops, except every second mm
                to_save = func in _save_list and not (
                    func == torch.ops.aten.mm.default and meta[mm_count_key] % 2 == 0
                )
                return (
                    CheckpointPolicy.MUST_SAVE
                    if to_save
                    else CheckpointPolicy.PREFER_RECOMPUTE
                )

            return _custom_policy

        def selective_checkpointing_context_fn():
            meta = defaultdict(int)
            return create_selective_checkpoint_contexts(_get_custom_policy(meta))

        return ptd_checkpoint_wrapper(
            module,
            context_fn=selective_checkpointing_context_fn,
            preserve_rng_state=False,
        )
    elif use_layer_sac:
        # Checkpoint every `ac_freq` of the modules passed to this function
        ac_freq = int(ac_config.selective_ac_option)
        ptd_checkpoint_wrapper.__dict__.setdefault("_count", 0)
        ptd_checkpoint_wrapper._count += 1
        if not ac_freq or ptd_checkpoint_wrapper._count % ac_freq == 0:
            return ptd_checkpoint_wrapper(module, preserve_rng_state=False)
        else:
            return module


def apply_ac(model: nn.Module, ac_config):
    """Apply activation checkpointing to the model."""
    for layer_id, transformer_block in model.layers.named_children():
        transformer_block = _apply_ac_to_transformer_block(transformer_block, ac_config)
        model.layers.register_module(layer_id, transformer_block)

    logger.info(f"Applied {ac_config.mode} activation checkpointing to the model")


def apply_compile(model: nn.Module):
    """
    Apply torch.compile to each TransformerBlock, which makes compilation efficient due to
    repeated structure. Alternatively one can compile the whole model (after applying DP).
    """
    # TODO(ruisizhang123): avoid recompiling error on 70B model
    torch._dynamo.config.cache_size_limit = 128
    for layer_id, transformer_block in model.layers.named_children():
        transformer_block = torch.compile(transformer_block, fullgraph=True)
        model.layers.register_module(layer_id, transformer_block)

    logger.info("Compiling each TransformerBlock with torch.compile")


def apply_fsdp(
    model: nn.Module,
    dp_mesh: DeviceMesh,
    param_dtype: torch.dtype,
    reduce_dtype: torch.dtype,
    tp_enabled: bool,
    pp_enabled: bool,
):
    """
    Apply data parallelism to the model. FSDP2 is used here.
    """
    mp_policy = MixedPrecisionPolicy(param_dtype=param_dtype, reduce_dtype=reduce_dtype)
    fsdp_config = {"mesh": dp_mesh, "mp_policy": mp_policy}

    # TODO(lty): the check below requires the latest PyTorch nightly; remove for now
    # TODO: remove this check once PyTorch 2.5 is released. We can safely assume
    # that users won't use a nightly build which is older than 20240809 by then.
    # if tp_enabled:
    #     # check if strided sharding is enabled, which is necessary for 2D/3D DCP
    #     check_strided_sharding_enabled()

    for layer_id, transformer_block in model.layers.items():
        if pp_enabled:
            # For PP, do not reshard after forward to avoid per-microbatch
            # all-gathers, which can be expensive and non-overlapped
            # NOTE: set reshard_after_forward to True for fair comparison with simple FSDP
            reshard_after_forward = True
        else:
            # As an optimization, do not reshard after forward for the last
            # transformer block since FSDP would prefetch it immediately
            reshard_after_forward = int(layer_id) < len(model.layers) - 1
        fully_shard(
            transformer_block,
            **fsdp_config,
            reshard_after_forward=reshard_after_forward,
        )
    fully_shard(model, **fsdp_config, reshard_after_forward=not pp_enabled)

    logger.info("Applied FSDP to the model")


def apply_ddp(
    model: nn.Module,
    dp_mesh: DeviceMesh,
    enable_compile: bool,
    enable_compiled_autograd: bool,
):
    if enable_compile:
        if enable_compiled_autograd:
            torch._dynamo.config.optimize_ddp = (
                "python_reducer_without_compiled_forward"
            )
        else:
            torch._dynamo.config.optimize_ddp = "ddp_optimizer"

    replicate(model, device_mesh=dp_mesh, bucket_cap_mb=100)

    logger.info("Applied DDP to the model")