EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Bugfix] fix mtp profile run error where main model and mtp model use different quantization (#4102)

Browse Source 
### What this PR does / why we need it?
In PR https://github.com/vllm-project/vllm-ascend/pull/3420, we
initially placed the quantization type (quant_type) in the MoECommMethod
class. However, since MoECommMethod follows a singleton pattern, it
couldn't accommodate scenarios where different layers in the model might
use different quantization approaches (e.g., MTP modules using
floating-point computation while the main model employs quantized
computation).
In this PR, we've moved the quantization type to the AscendFusedMoe
class and pass it as a parameter to MoECommMethod.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
```bash
export HCCL_BUFFSIZE=1024
export VLLM_VERSION=0.11.0

vllm serve /home/data/DeepSeek-R1_w8a8/ \
 --data-parallel-size 2 \
 --tensor-parallel-size 8 \
 --enable-expert-parallel \
 --served-model-name dsv3 \
 --max-model-len 32768 \
 --max-num-batched-tokens 4096 \
 --max-num-seqs 16 \
 --quantization ascend \
 --trust-remote-code \
 --gpu-memory-utilization 0.9 \
 --speculative-config '{"num_speculative_tokens": 2, "method":"deepseek_mtp"}'
```


- vLLM version: v0.11.0
- vLLM main:
83f478bb19

---------

Signed-off-by: realliujiaxu <realliujiaxu@163.com>
This commit is contained in:
realliujiaxu
2025-11-13 11:02:31 +08:00
committed by GitHub
parent 17259cb265
commit 5093192769
6 changed files with 82 additions and 76 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
tests/ut/ops/test_moe_comm_method.py
View File

@@ -7,6 +7,7 @@ from tests.ut.base import TestBase
from vllm_ascend.ops.fused_moe.moe_comm_method import (AllGatherCommImpl, from vllm_ascend.ops.fused_moe.moe_comm_method import (AllGatherCommImpl,
AlltoAllCommImpl, AlltoAllCommImpl,
MC2CommImpl) MC2CommImpl)
from vllm_ascend.ops.fused_moe.prepare_finalize import QuantType
class TestMoECommMethod(TestBase): class TestMoECommMethod(TestBase):
@@ -67,7 +68,7 @@ class TestMoECommMethod(TestBase):
# Verify prepare was called with correct arguments # Verify prepare was called with correct arguments
mock_pf_instance.prepare.assert_called_once_with( mock_pf_instance.prepare.assert_called_once_with(
hidden_states, router_logits, False, False) hidden_states, router_logits, False, False, QuantType.NONE)
# Test finalize method # Test finalize method
comm_impl.finalize(h_out, comm_impl.finalize(h_out,
@@ -115,7 +116,7 @@ class TestMoECommMethod(TestBase):
# Verify prepare was called with correct arguments # Verify prepare was called with correct arguments
mock_pf_instance.prepare.assert_called_once_with( mock_pf_instance.prepare.assert_called_once_with(
hidden_states, router_logits, False, False) hidden_states, router_logits, False, False, QuantType.NONE)
# Test finalize method # Test finalize method
comm_impl.finalize(h_out, comm_impl.finalize(h_out,
@@ -165,7 +166,7 @@ class TestMoECommMethod(TestBase):
# Verify prepare was called with correct arguments # Verify prepare was called with correct arguments
mock_pf_instance.prepare.assert_called_once_with( mock_pf_instance.prepare.assert_called_once_with(
hidden_states, router_logits, False, False) hidden_states, router_logits, False, False, QuantType.NONE)
@patch("vllm_ascend.ops.fused_moe.moe_comm_method.get_current_vllm_config") @patch("vllm_ascend.ops.fused_moe.moe_comm_method.get_current_vllm_config")
@patch("vllm_ascend.ops.fused_moe.moe_comm_method.get_forward_context") @patch("vllm_ascend.ops.fused_moe.moe_comm_method.get_forward_context")

26
vllm_ascend/ops/fused_moe/fused_moe.py
View File

@@ -37,6 +37,11 @@ from vllm_ascend.eplb.core.eplb_utils import (determine_default_expert_map,
from vllm_ascend.ops.expert_load_balancer import ExpertLoadBalancer from vllm_ascend.ops.expert_load_balancer import ExpertLoadBalancer
from vllm_ascend.ops.fused_moe.experts_selector import select_experts from vllm_ascend.ops.fused_moe.experts_selector import select_experts
from vllm_ascend.ops.fused_moe.moe_comm_method import setup_moe_comm_method from vllm_ascend.ops.fused_moe.moe_comm_method import setup_moe_comm_method
from vllm_ascend.ops.fused_moe.prepare_finalize import QuantType
from vllm_ascend.quantization.w4a8_dynamic import \
AscendW4A8DynamicFusedMoEMethod
from vllm_ascend.quantization.w8a8_dynamic import \
AscendW8A8DynamicFusedMoEMethod
from vllm_ascend.utils import (ACL_FORMAT_FRACTAL_NZ, enable_sp, is_310p, from vllm_ascend.utils import (ACL_FORMAT_FRACTAL_NZ, enable_sp, is_310p,
is_enable_nz, npu_stream_switch, is_enable_nz, npu_stream_switch,
shared_expert_dp_enabled, shared_expert_dp_enabled,
@@ -289,7 +294,23 @@ class AscendFusedMoE(FusedMoE):
self.enable_shared_expert_dp = ascend_config.enable_shared_expert_dp self.enable_shared_expert_dp = ascend_config.enable_shared_expert_dp
setup_moe_comm_method(self.moe_config, self.quant_method) setup_moe_comm_method(self.moe_config)
self.quant_type = self._get_quant_type()
def _get_quant_type(self) -> QuantType:
quant_method = self.quant_method
if not hasattr(quant_method,
"quant_method") or quant_method.quant_method is None:
return QuantType.NONE
method = quant_method.quant_method
if isinstance(method, AscendW8A8DynamicFusedMoEMethod):
return QuantType.W8A8
elif isinstance(method, AscendW4A8DynamicFusedMoEMethod):
return QuantType.W4A8
else:
return QuantType.NONE
def update_expert_map(self, new_expert_map): def update_expert_map(self, new_expert_map):
self.expert_map = new_expert_map self.expert_map = new_expert_map
@@ -334,7 +355,8 @@ class AscendFusedMoE(FusedMoE):
hidden_states=hidden_states, hidden_states=hidden_states,
router_logits=router_logits, router_logits=router_logits,
replace_allreduce=forward_context.sp_enabled, replace_allreduce=forward_context.sp_enabled,
enable_shared_expert_dp=self.enable_shared_expert_dp) enable_shared_expert_dp=self.enable_shared_expert_dp,
quant_type=self.quant_type)
if isinstance(hidden_states, tuple): if isinstance(hidden_states, tuple):
hidden_states, pertoken_scale = hidden_states hidden_states, pertoken_scale = hidden_states

61
vllm_ascend/ops/fused_moe/moe_comm_method.py
View File

@@ -31,10 +31,6 @@ from vllm_ascend.ops.fused_moe.prepare_finalize import (
from vllm_ascend.ops.fused_moe.token_dispatcher import ( from vllm_ascend.ops.fused_moe.token_dispatcher import (
TokenDispatcherWithAll2AllV, TokenDispatcherWithAllGather, TokenDispatcherWithAll2AllV, TokenDispatcherWithAllGather,
TokenDispatcherWithMC2, TokenDispatcherWithMoge) TokenDispatcherWithMC2, TokenDispatcherWithMoge)
from vllm_ascend.quantization.w4a8_dynamic import \
AscendW4A8DynamicFusedMoEMethod
from vllm_ascend.quantization.w8a8_dynamic import \
AscendW8A8DynamicFusedMoEMethod
_MoECommMethods: Dict[Optional[MoECommType], MoECommMethod] = {} _MoECommMethods: Dict[Optional[MoECommType], MoECommMethod] = {}
@@ -44,54 +40,37 @@ def get_moe_comm_method(
return _MoECommMethods.get(moe_comm_type, None) return _MoECommMethods.get(moe_comm_type, None)
def setup_moe_comm_method(moe_config, quant_method): def setup_moe_comm_method(moe_config):
_MoECommMethods[MoECommType.ALLTOALL] = AlltoAllCommImpl( _MoECommMethods[MoECommType.ALLTOALL] = AlltoAllCommImpl(moe_config)
moe_config, quant_method) _MoECommMethods[MoECommType.ALLGATHER] = AllGatherCommImpl(moe_config)
_MoECommMethods[MoECommType.ALLGATHER] = AllGatherCommImpl( _MoECommMethods[MoECommType.MC2] = MC2CommImpl(moe_config)
moe_config, quant_method)
_MoECommMethods[MoECommType.MC2] = MC2CommImpl(moe_config, quant_method)
_MoECommMethods[MoECommType.NAIVE_MULTICAST] = NaiveMulticastCommImpl( _MoECommMethods[MoECommType.NAIVE_MULTICAST] = NaiveMulticastCommImpl(
moe_config, quant_method) moe_config)
class MoECommMethod(ABC): class MoECommMethod(ABC):
"""Base class for MoE communication methods.""" """Base class for MoE communication methods."""
def __init__(self, moe_config: FusedMoEConfig, quant_method=None): def __init__(self, moe_config: FusedMoEConfig):
self.model_type = get_current_vllm_config( self.model_type = get_current_vllm_config(
).model_config.hf_config.model_type ).model_config.hf_config.model_type
self.moe_config = moe_config self.moe_config = moe_config
self.token_dispatcher = self._get_token_dispatcher() self.token_dispatcher = self._get_token_dispatcher()
self.quant_type = self._get_quant_type(quant_method)
self.with_quant = self.quant_type != QuantType.NONE
self.prepare_finalize = self._get_prepare_finalize() self.prepare_finalize = self._get_prepare_finalize()
def _get_quant_type(self, quant_method) -> QuantType:
if not hasattr(quant_method,
"quant_method") or quant_method.quant_method is None:
return QuantType.NONE
method = quant_method.quant_method
if isinstance(method, AscendW8A8DynamicFusedMoEMethod):
return QuantType.W8A8
elif isinstance(method, AscendW4A8DynamicFusedMoEMethod):
return QuantType.W4A8
else:
return QuantType.NONE
def prepare( def prepare(
self, self,
hidden_states: torch.Tensor, hidden_states: torch.Tensor,
router_logits: torch.Tensor, router_logits: torch.Tensor,
enable_shared_expert_dp: bool = False, enable_shared_expert_dp: bool = False,
replace_allreduce: bool = False replace_allreduce: bool = False,
quant_type: QuantType = QuantType.NONE,
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
hidden_states, router_logits, mc2_mask, context_metadata = self.prepare_finalize.prepare( hidden_states, router_logits, mc2_mask, context_metadata = self.prepare_finalize.prepare(
hidden_states, router_logits, enable_shared_expert_dp, hidden_states, router_logits, enable_shared_expert_dp,
replace_allreduce) replace_allreduce, quant_type)
return hidden_states, router_logits, mc2_mask, context_metadata return hidden_states, router_logits, mc2_mask, context_metadata
def finalize(self, def finalize(self,
@@ -112,6 +91,8 @@ class MoECommMethod(ABC):
topk_ids: torch.Tensor, topk_ids: torch.Tensor,
activation: str = "silu", activation: str = "silu",
apply_router_weight_on_input: bool = False, apply_router_weight_on_input: bool = False,
use_int8_w8a8: bool = False,
use_int4_w4a8: bool = False,
global_num_experts: Optional[int] = None, global_num_experts: Optional[int] = None,
expert_map: Optional[torch.Tensor] = None, expert_map: Optional[torch.Tensor] = None,
w1_scale: Optional[torch.Tensor] = None, w1_scale: Optional[torch.Tensor] = None,
@@ -151,15 +132,14 @@ class MoECommMethod(ABC):
dynamic_scale_for_share=dynamic_scale_for_share, dynamic_scale_for_share=dynamic_scale_for_share,
mc2_mask=mc2_mask, mc2_mask=mc2_mask,
apply_router_weight_on_input=apply_router_weight_on_input, apply_router_weight_on_input=apply_router_weight_on_input,
with_quant=self.with_quant, with_quant=use_int8_w8a8 or use_int4_w4a8,
dynamic_eplb=dynamic_eplb, dynamic_eplb=dynamic_eplb,
pertoken_scale=pertoken_scale) pertoken_scale=pertoken_scale)
permuted_hidden_states, expert_tokens, dynamic_scale, group_list_type, topk_scales, context_metadata = \ permuted_hidden_states, expert_tokens, dynamic_scale, group_list_type, topk_scales, context_metadata = \
results["hidden_states"], results["group_list"], results.get("dynamic_scale"), results["group_list_type"], results.get("topk_scales"), results.get("context_metadata") results["hidden_states"], results["group_list"], results.get("dynamic_scale"), results["group_list_type"], results.get("topk_scales"), results.get("context_metadata")
mlp_output = unified_apply_mlp( mlp_output = unified_apply_mlp(hidden_states=permuted_hidden_states,
hidden_states=permuted_hidden_states,
w1=w1, w1=w1,
w1_scale=w1_scale, w1_scale=w1_scale,
w2=w2, w2=w2,
@@ -170,8 +150,9 @@ class MoECommMethod(ABC):
w1_scale_bias=w1_scale_bias, w1_scale_bias=w1_scale_bias,
w2_scale_bias=w2_scale_bias, w2_scale_bias=w2_scale_bias,
topk_scales=topk_scales, topk_scales=topk_scales,
with_quant=self.with_quant, with_quant=use_int8_w8a8
fusion=self.quant_type == QuantType.W8A8, or use_int4_w4a8,
fusion=use_int8_w8a8,
need_trans=need_trans, need_trans=need_trans,
dynamic_eplb=dynamic_eplb) dynamic_eplb=dynamic_eplb)
@@ -226,8 +207,7 @@ class AllGatherCommImpl(MoECommMethod):
num_local_experts=self.moe_config.num_local_experts) num_local_experts=self.moe_config.num_local_experts)
def _get_prepare_finalize(self): def _get_prepare_finalize(self):
return PrepareAndFinalizeWithAllGather(self.moe_config, return PrepareAndFinalizeWithAllGather(self.moe_config)
self.quant_type)
class MC2CommImpl(MoECommMethod): class MC2CommImpl(MoECommMethod):
@@ -244,7 +224,7 @@ class MC2CommImpl(MoECommMethod):
return TokenDispatcherWithMC2() return TokenDispatcherWithMC2()
def _get_prepare_finalize(self): def _get_prepare_finalize(self):
return PrepareAndFinalizeWithMC2(self.moe_config, self.quant_type) return PrepareAndFinalizeWithMC2(self.moe_config)
class AlltoAllCommImpl(MoECommMethod): class AlltoAllCommImpl(MoECommMethod):
@@ -264,7 +244,7 @@ class AlltoAllCommImpl(MoECommMethod):
num_local_experts=self.moe_config.num_local_experts) num_local_experts=self.moe_config.num_local_experts)
def _get_prepare_finalize(self): def _get_prepare_finalize(self):
return PrepareAndFinalizeWithAll2All(self.moe_config, self.quant_type) return PrepareAndFinalizeWithAll2All(self.moe_config)
class NaiveMulticastCommImpl(MoECommMethod): class NaiveMulticastCommImpl(MoECommMethod):
@@ -293,5 +273,4 @@ class NaiveMulticastCommImpl(MoECommMethod):
num_local_experts=self.moe_config.num_local_experts) num_local_experts=self.moe_config.num_local_experts)
def _get_prepare_finalize(self): def _get_prepare_finalize(self):
return PrepareAndFinalizeWithNaiveMulticast(self.moe_config, return PrepareAndFinalizeWithNaiveMulticast(self.moe_config)
self.quant_type)

38
vllm_ascend/ops/fused_moe/prepare_finalize.py
View File

@@ -53,11 +53,8 @@ class PrepareAndFinalize(ABC):
sizes, ranks, and communication settings. sizes, ranks, and communication settings.
""" """
def __init__(self, def __init__(self, moe_config: FusedMoEConfig):
moe_config: FusedMoEConfig,
quant_type: QuantType = QuantType.NONE):
self.moe_config = moe_config self.moe_config = moe_config
self.quant_type = quant_type
@abstractmethod @abstractmethod
def prepare( def prepare(
@@ -65,7 +62,8 @@ class PrepareAndFinalize(ABC):
hidden_states: torch.Tensor, hidden_states: torch.Tensor,
router_logits: torch.Tensor, router_logits: torch.Tensor,
enable_shared_expert_dp: bool = False, enable_shared_expert_dp: bool = False,
replace_allreduce: bool = False replace_allreduce: bool = False,
quant_type: QuantType = QuantType.NONE
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
""" """
@@ -79,6 +77,7 @@ class PrepareAndFinalize(ABC):
router_logits (torch.Tensor): Router outputs, shape [num_tokens, num_experts] router_logits (torch.Tensor): Router outputs, shape [num_tokens, num_experts]
enable_shared_expert_dp (bool): Skip DP communication for shared experts enable_shared_expert_dp (bool): Skip DP communication for shared experts
replace_allreduce (bool): Bypass default all-reduce behavior replace_allreduce (bool): Bypass default all-reduce behavior
quant_type: none, w8a8 or w4a8
Returns: Returns:
Tuple of: Tuple of:
@@ -117,10 +116,8 @@ class PrepareAndFinalizeWithAll2All(PrepareAndFinalize):
Will be used when num_tokens exceed mc2's limitation (512 tokens/rank). Will be used when num_tokens exceed mc2's limitation (512 tokens/rank).
""" """
def __init__(self, def __init__(self, moe_config: FusedMoEConfig):
moe_config: FusedMoEConfig, super().__init__(moe_config)
quant_type: QuantType = QuantType.NONE):
super().__init__(moe_config, quant_type)
self._restore_tp_across_dp() self._restore_tp_across_dp()
def _restore_tp_across_dp(self): def _restore_tp_across_dp(self):
@@ -133,7 +130,8 @@ class PrepareAndFinalizeWithAll2All(PrepareAndFinalize):
hidden_states: torch.Tensor, hidden_states: torch.Tensor,
router_logits: torch.Tensor, router_logits: torch.Tensor,
enable_shared_expert_dp: bool = False, enable_shared_expert_dp: bool = False,
replace_allreduce: bool = False replace_allreduce: bool = False,
quant_type=QuantType.NONE
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
""" """
@@ -211,10 +209,8 @@ class PrepareAndFinalizeWithMC2(PrepareAndFinalizeWithAll2All):
Relies on `mc2_mask` and `padded_num_tokens` from forward_context for alignment. Relies on `mc2_mask` and `padded_num_tokens` from forward_context for alignment.
""" """
def __init__(self, def __init__(self, moe_config: FusedMoEConfig):
moe_config: FusedMoEConfig, super().__init__(moe_config)
quant_type: QuantType = QuantType.NONE):
super().__init__(moe_config, quant_type)
self._restore_tp_across_dp() self._restore_tp_across_dp()
def _restore_tp_across_dp(self): def _restore_tp_across_dp(self):
@@ -231,7 +227,8 @@ class PrepareAndFinalizeWithMC2(PrepareAndFinalizeWithAll2All):
hidden_states: torch.Tensor, hidden_states: torch.Tensor,
router_logits: torch.Tensor, router_logits: torch.Tensor,
enable_shared_expert_dp: bool = False, enable_shared_expert_dp: bool = False,
replace_allreduce: bool = False replace_allreduce: bool = False,
quant_type=QuantType.NONE
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
""" """
@@ -312,6 +309,7 @@ class PrepareAndFinalizeWithAllGather(PrepareAndFinalize):
router_logits: torch.Tensor, router_logits: torch.Tensor,
enable_shared_expert_dp: bool = False, enable_shared_expert_dp: bool = False,
replace_allreduce: bool = False, replace_allreduce: bool = False,
quant_type=QuantType.NONE
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
""" """
@@ -322,7 +320,8 @@ class PrepareAndFinalizeWithAllGather(PrepareAndFinalize):
Tuple of (global_hidden_states, global_router_logits, None) Tuple of (global_hidden_states, global_router_logits, None)
""" """
if enable_sp(): if enable_sp():
return self._prepare_with_ep_group(hidden_states, router_logits) return self._prepare_with_ep_group(hidden_states, router_logits,
quant_type)
return self._prepare_with_dp_group(hidden_states, router_logits, return self._prepare_with_dp_group(hidden_states, router_logits,
enable_shared_expert_dp, enable_shared_expert_dp,
@@ -332,10 +331,11 @@ class PrepareAndFinalizeWithAllGather(PrepareAndFinalize):
self, self,
hidden_states: torch.Tensor, hidden_states: torch.Tensor,
router_logits: torch.Tensor, router_logits: torch.Tensor,
quant_type=QuantType.NONE
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
pertoken_scale = None pertoken_scale = None
if self.quant_type == QuantType.W8A8: if quant_type == QuantType.W8A8:
hidden_states, pertoken_scale = torch_npu.npu_dynamic_quant( hidden_states, pertoken_scale = torch_npu.npu_dynamic_quant(
hidden_states) hidden_states)
pertoken_scale = torch.ops.vllm.maybe_all_gather_and_maybe_unpad( pertoken_scale = torch.ops.vllm.maybe_all_gather_and_maybe_unpad(
@@ -356,6 +356,7 @@ class PrepareAndFinalizeWithAllGather(PrepareAndFinalize):
router_logits: torch.Tensor, router_logits: torch.Tensor,
enable_shared_expert_dp: bool = False, enable_shared_expert_dp: bool = False,
replace_allreduce: bool = False, replace_allreduce: bool = False,
quant_type=QuantType.NONE
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
""" """
@@ -500,7 +501,8 @@ class PrepareAndFinalizeWithNaiveMulticast(PrepareAndFinalize):
hidden_states: torch.Tensor, hidden_states: torch.Tensor,
router_logits: torch.Tensor, router_logits: torch.Tensor,
enable_shared_expert_dp: bool = False, enable_shared_expert_dp: bool = False,
replace_allreduce: bool = False replace_allreduce: bool = False,
quant_type=QuantType.NONE
) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor], ) -> tuple[torch.Tensor, torch.Tensor, Optional[torch.Tensor],
Optional[torch.Tensor]]: Optional[torch.Tensor]]:
""" """

1
vllm_ascend/quantization/w4a8_dynamic.py
View File

@@ -386,6 +386,7 @@ class AscendW4A8DynamicFusedMoEMethod:
w2_scale_bias=layer.w2_scale_bias, w2_scale_bias=layer.w2_scale_bias,
topk_weights=topk_weights, topk_weights=topk_weights,
topk_ids=topk_ids, topk_ids=topk_ids,
use_int4_w4a8=True,
expert_map=expert_map, expert_map=expert_map,
log2phy=log2phy, log2phy=log2phy,
global_redundant_expert_num=global_redundant_expert_num, global_redundant_expert_num=global_redundant_expert_num,

1
vllm_ascend/quantization/w8a8_dynamic.py
View File

@@ -256,6 +256,7 @@ class AscendW8A8DynamicFusedMoEMethod:
w2_scale=layer.w2_weight_scale, w2_scale=layer.w2_weight_scale,
topk_weights=topk_weights, topk_weights=topk_weights,
topk_ids=topk_ids, topk_ids=topk_ids,
use_int8_w8a8=True,
expert_map=expert_map, expert_map=expert_map,
log2phy=log2phy, log2phy=log2phy,
global_redundant_expert_num=global_redundant_expert_num, global_redundant_expert_num=global_redundant_expert_num,