EngineX-Cambricon/enginex-mlu370-vllm
10
0
Fork 1
Code Issues Pull Requests Actions Projects Releases Wiki Activity

add deepseekv3 and llama4

Browse Source
This commit is contained in:
Chranos
2026-02-11 14:39:48 +08:00
parent d93c740e4d
commit d9d1f3a724
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm-v0.6.2/vllm/model_executor/model_loader/utils.py
View File

@@ -20,7 +20,7 @@ def set_default_torch_dtype(dtype: torch.dtype):
def get_model_architecture( def get_model_architecture(
model_config: ModelConfig) -> Tuple[Type[nn.Module], str]: model_config: ModelConfig) -> Tuple[Type[nn.Module], str]:
architectures = getattr(model_config.hf_config, "architectures", []) architectures = getattr(model_config.hf_config, "architectures", None) or []
# Special handling for quantized Mixtral. # Special handling for quantized Mixtral.
# FIXME(woosuk): This is a temporary hack. # FIXME(woosuk): This is a temporary hack.
mixtral_supported = [ mixtral_supported = [