[Core] Support model loader plugins (#21067)

Signed-off-by: 22quinn <33176974+22quinn@users.noreply.github.com>
This commit is contained in:
22quinn
2025-07-24 01:49:44 -07:00
committed by GitHub
parent f0f4de8f26
commit 610852a423
9 changed files with 159 additions and 86 deletions
@@ -2,7 +2,6 @@
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
from vllm import SamplingParams
from vllm.config import LoadFormat
test_model = "openai-community/gpt2"
@@ -17,7 +16,6 @@ sampling_params = SamplingParams(temperature=0.8, top_p=0.95, seed=0)
def test_model_loader_download_files(vllm_runner):
with vllm_runner(test_model,
load_format=LoadFormat.FASTSAFETENSORS) as llm:
with vllm_runner(test_model, load_format="fastsafetensors") as llm:
deserialized_outputs = llm.generate(prompts, sampling_params)
assert deserialized_outputs