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

[Core] Parse vLLM engine required fields from hf_config to model_arch_config (#28454)

Browse Source 
Signed-off-by: Xingyu Liu <charlotteliu12x@gmail.com>
Signed-off-by: Xingyu Liu <38244988+charlotte12l@users.noreply.github.com>
This commit is contained in:
Xingyu Liu
2026-01-02 16:13:15 -07:00
committed by GitHub
parent a0e9ee83c7
commit 0eee877f67
11 changed files with 1121 additions and 287 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/models/utils.py
View File

@@ -471,12 +471,16 @@ def dummy_hf_overrides(
"num_kv_shared_layers": 1,
}
_hf_config = hf_config
class DummyConfig:
hf_config = _hf_config
hf_text_config = text_config
model_arch_config = ModelConfig.get_model_arch_config(DummyConfig)
# Only set MoE related config when the model has MoE layers.
# Otherwise all models detected as MoE by _get_transformers_backend_cls.
if ModelConfig.get_num_experts(DummyConfig) > 0:
if model_arch_config.num_experts > 0:
update_dict.update(
{
"num_experts": num_experts,