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

[V1] Support LLM.apply_model (#18465)

Browse Source 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2025-09-20 15:14:35 +08:00
committed by GitHub
parent be874c0201
commit 3d9a1d2de5
17 changed files with 194 additions and 169 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
tests/quantization/test_fp8.py
View File

@@ -60,8 +60,8 @@ def test_kv_cache_model_load_and_run(vllm_runner, model_id: str,
if use_rocm_aiter:
monkeypatch.setenv("VLLM_ROCM_USE_AITER", "1")
# vllm_runner.apply_model() relies on V0 internals.
monkeypatch.setenv("VLLM_USE_V1", "0")
# `LLM.apply_model` requires pickling a function.
monkeypatch.setenv("VLLM_ALLOW_INSECURE_SERIALIZATION", "1")
with vllm_runner(model_id, kv_cache_dtype="fp8") as llm:
def check_model(model):
@@ -104,8 +104,8 @@ def test_load_fp16_model(vllm_runner, kv_cache_dtype: str, force_marlin: bool,
if use_rocm_aiter:
monkeypatch.setenv("VLLM_ROCM_USE_AITER", "1")
# vllm_runner.apply_model() relies on V0 internals.
monkeypatch.setenv("VLLM_USE_V1", "0")
# `LLM.apply_model` requires pickling a function.
monkeypatch.setenv("VLLM_ALLOW_INSECURE_SERIALIZATION", "1")
if force_marlin:
monkeypatch.setenv("VLLM_TEST_FORCE_FP8_MARLIN", "1")