A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[CI/Build] Remove redundant register in model init tests (#23715)
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Cyrus Leung committed
Aug 27, 2025 at 16:11 UTC
91e382c935c2905c29f3ca22c658e03e8f02deaa
1 file changed
-5
tests/models/test_initialization.py
-5
@@ -38,11 +38,6 @@ def can_initialize(model_arch: str, monkeypatch: pytest.MonkeyPatch,
38
model_arch=model_arch,
39
exist_overrides=model_info.hf_overrides)
40
41
- if model_arch in ("Llama4ForCausalLM", "EagleLlama4ForCausalLM"):
42
- from vllm.model_executor.models.llama4 import Llama4ForCausalLM
43
- from vllm.model_executor.models.registry import ModelRegistry
44
- ModelRegistry.register_model("Llama4ForCausalLM", Llama4ForCausalLM)
45
-
41
# Avoid calling model.forward()
42
def _initialize_kv_caches_v0(self) -> None:
43
self.cache_config.num_gpu_blocks = 0