biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Misc] Qwen MoE model supports LoRA (#20932)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li
2025-07-18 02:32:52 +08:00
committed by GitHub
parent 90bd2ab6e3
commit a3a6c695f4
4 changed files with 20 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
vllm/model_executor/models/qwen2_moe.py
View File

@@ -53,7 +53,7 @@ from vllm.model_executor.model_loader.weight_utils import default_weight_loader
from vllm.model_executor.sampling_metadata import SamplingMetadata
from vllm.sequence import IntermediateTensors
from .interfaces import SupportsPP
from .interfaces import SupportsLoRA, SupportsPP
from .utils import (AutoWeightsLoader, extract_layer_index,
is_pp_missing_parameter,
make_empty_intermediate_tensors_factory, make_layers,
@@ -448,8 +448,7 @@ class Qwen2MoeModel(nn.Module):
if weight_name not in name:
continue
name = name.replace(weight_name, param_name)
if "layers.13.mlp.experts.w2_weight" in name:
pass
# Skip layers on other devices.
if is_pp_missing_parameter(name, self):
continue
@@ -494,7 +493,7 @@ class Qwen2MoeModel(nn.Module):
return loaded_params
class Qwen2MoeForCausalLM(nn.Module, SupportsPP):
class Qwen2MoeForCausalLM(nn.Module, SupportsPP, SupportsLoRA):
fall_back_to_pt_during_load = False
packed_modules_mapping = {