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

[Model] Add Qwen3 and Qwen3MoE (#15289)

Browse Source 
Signed-off-by: YamPengLi <yampayne.lyp@alibaba-inc.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
This commit is contained in:
YamPengLi
2025-04-07 19:06:41 +08:00
committed by GitHub
parent e9ba99f296
commit 7699258ef0
6 changed files with 893 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
vllm/model_executor/models/qwen2.py
View File

@@ -263,7 +263,11 @@ class Qwen2DecoderLayer(nn.Module):
})
class Qwen2Model(nn.Module):
def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
def __init__(self,
*,
vllm_config: VllmConfig,
prefix: str = "",
decoder_layer_type: type[nn.Module] = Qwen2DecoderLayer):
super().__init__()
config = vllm_config.model_config.hf_config
@@ -297,12 +301,14 @@ class Qwen2Model(nn.Module):
else:
self.embed_tokens = PPMissingLayer()
# Use the provided decoder layer type or default to Qwen2DecoderLayer
decoder_layer_type = decoder_layer_type or Qwen2DecoderLayer
self.start_layer, self.end_layer, self.layers = make_layers(
config.num_hidden_layers,
lambda prefix: Qwen2DecoderLayer(config=config,
cache_config=cache_config,
quant_config=quant_config,
prefix=prefix),
lambda prefix: decoder_layer_type(config=config,
cache_config=cache_config,
quant_config=quant_config,
prefix=prefix),
prefix=f"{prefix}.layers",
)