Support expert parallel load balancing in Transformers backend (#26287)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-06 11:20:16 +00:00
committed by GitHub
parent 19a00eb210
commit 0340f45553
2 changed files with 76 additions and 27 deletions
+63 -16
View File
@@ -30,6 +30,7 @@ from vllm.model_executor.layers.fused_moe import FusedMoE
from vllm.platforms import current_platform
from vllm.utils import direct_register_custom_op
from .interfaces import MixtureOfExperts
from .transformers import (
TransformersBase,
TransformersForCausalLM,
@@ -116,17 +117,41 @@ direct_register_custom_op(
)
class TransformersMoEBase(TransformersBase):
class TransformersMoEBase(TransformersBase, MixtureOfExperts):
def __init__(self, *, vllm_config, prefix=""):
self.check_version("4.57.0.dev0", "MoE models support")
self.ep_group = get_ep_group()
super().__init__(vllm_config=vllm_config, prefix=prefix)
if self.parallel_config.enable_eplb:
raise NotImplementedError(
"Transformers backend does not support expert parallel load "
"balancing yet."
def set_eplb_state(
self,
expert_load_view: torch.Tensor,
logical_to_physical_map: torch.Tensor,
logical_replica_count: torch.Tensor,
):
for moe_layer_idx, mlp_layer in enumerate(self.mlp_layers):
mlp_layer.experts.set_eplb_state(
moe_layer_idx=moe_layer_idx,
expert_load_view=expert_load_view,
logical_to_physical_map=logical_to_physical_map,
logical_replica_count=logical_replica_count,
)
def update_physical_experts_metadata(
self,
num_physical_experts: int,
num_local_physical_experts: int,
):
assert self.num_local_physical_experts == num_local_physical_experts
self.num_physical_experts = num_physical_experts
self.num_local_physical_experts = num_local_physical_experts
self.num_redundant_experts = num_physical_experts - self.num_logical_experts
for mlp in self.mlp_layers:
mlp.n_local_physical_experts = num_local_physical_experts
mlp.n_physical_experts = num_physical_experts
mlp.n_redundant_experts = self.num_redundant_experts
mlp.experts.update_expert_map()
def get_expert_mapping(self) -> list[tuple[str, str, int, str]]:
"""
Params for weights, fp8 weight scales, fp8 activation scales
@@ -138,6 +163,8 @@ class TransformersMoEBase(TransformersBase):
("w1", "w2", "w3"), # Granite, Mixtral, Phi MoE style
("linear", "linear_1", "linear_v"), # Grok1 style
]
num_experts = self.model_config.get_num_experts()
num_redundant_experts = self.parallel_config.eplb_config.num_redundant_experts
expert_mapping = []
for gate_proj, down_proj, up_proj in ckpt_names:
expert_mapping.extend(
@@ -145,8 +172,8 @@ class TransformersMoEBase(TransformersBase):
ckpt_gate_proj_name=gate_proj,
ckpt_down_proj_name=down_proj,
ckpt_up_proj_name=up_proj,
num_experts=self.model_config.get_num_experts(),
num_redundant_experts=0, # TODO: enable EPLB
num_experts=num_experts,
num_redundant_experts=num_redundant_experts,
)
)
return expert_mapping
@@ -167,12 +194,15 @@ class TransformersMoEBase(TransformersBase):
# If there are shared experts, the results are
# reduced after mlp.forward() not inside FusedMoE
num_experts_shared = getattr_iter(
num_shared_experts = getattr_iter(
text_config,
["num_experts_shared", "n_shared_experts", "moe_num_shared_experts"],
[
"n_shared_experts", # DeepSeek, Docs, GLM
"moe_num_shared_experts", # Aria, Ernie
],
0,
)
reduce_results = num_experts_shared == 0
reduce_results = num_shared_experts == 0
def add_all_reduce(mlp: nn.Module):
"""Adds an all-reduce to the output of `mlp.forward()`."""
@@ -207,13 +237,23 @@ class TransformersMoEBase(TransformersBase):
# Expert mapping for `AutoWeightsLoader`
expert_mapping = self.get_expert_mapping()
# Configs
parallel_config = self.parallel_config
eplb_config = parallel_config.eplb_config
# Expert parallel load balancing kwargs
enable_eplb = parallel_config.enable_eplb
num_redundant_experts = eplb_config.num_redundant_experts
enable_eplb = self.parallel_config.enable_eplb
num_redundant_experts = self.parallel_config.eplb_config.num_redundant_experts
# MixtureOfExperts mixin settings
ep_size = self.ep_group.world_size
self.mlp_layers = [] # Used for MixtureOfExperts methods
self.expert_weights = []
self.num_moe_layers = 0
self.num_expert_groups = 1 if num_expert_group is None else num_expert_group
self.num_logical_experts = num_experts
self.num_physical_experts = num_experts + num_redundant_experts
self.num_local_physical_experts = self.num_physical_experts // ep_size
self.num_routed_experts = num_experts
self.num_shared_experts = num_shared_experts
self.num_redundant_experts = num_redundant_experts
# Recursively fuse MoE layers
def _recursive_replace(module: nn.Module, prefix: str):
@@ -235,6 +275,9 @@ class TransformersMoEBase(TransformersBase):
for mlp_param_name, _ in mlp.named_parameters():
if "shared_expert" in mlp_param_name:
reduce_results = False
# If the config does not specify num_shared_experts, but
# the model has shared experts, we assume there is one.
self.num_shared_experts = 1
break
# Replace experts module with FusedMoE
fused_experts = TransformersFusedMoE(
@@ -258,6 +301,10 @@ class TransformersMoEBase(TransformersBase):
)
mlp.experts = fused_experts
log_replacement(qual_name, experts, fused_experts)
# Update MixtureOfExperts mixin state
self.mlp_layers.append(mlp)
self.expert_weights.append(fused_experts.get_expert_weights())
self.num_moe_layers += 1
# If results are not all-reduced in FusedMoE, ensure they
# are all-reduced at the end of mlp.forward() if tensor
# parallel or expert parallel is enabled