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

Fix AOPerModuleConfig name changes (#18869)

Browse Source 
Signed-off-by: Jerry Zhang <jerryzh168@gmail.com>
This commit is contained in:
Jerry Zhang
2025-06-05 21:51:32 -04:00
committed by GitHub
parent cb6d572e85
commit c8134bea15
3 changed files with 25 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/quantization/test_torchao.py
View File

@@ -13,7 +13,7 @@ TORCHAO_AVAILABLE = importlib.util.find_spec("torchao") is not None
@pytest.mark.skipif(not TORCHAO_AVAILABLE, reason="torchao is not available")
def test_pre_quantized_model(vllm_runner):
with vllm_runner("drisspg/float8_dynamic_act_float8_weight-opt-125m",
with vllm_runner("drisspg/fp8-opt-125m",
quantization="torchao",
dtype="bfloat16",
enforce_eager=True) as llm:
@@ -30,10 +30,10 @@ def test_pre_quantized_model(vllm_runner):
"cuda:0",
# {"": "cuda"},
])
def test_opt_125m_int4wo_model_loading_with_params(vllm_runner,
def test_opt_125m_int8wo_model_loading_with_params(vllm_runner,
pt_load_map_location):
torch._dynamo.reset()
model_name = "jerryzh168/opt-125m-int4wo"
model_name = "jerryzh168/opt-125m-int8wo-partial-quant"
with vllm_runner(model_name=model_name,
quantization="torchao",
dtype="bfloat16",