update test case

crazydemo · crazydemo · commit 8b325def11f4 · 2026-01-02T10:46:22.000+08:00
Signed-off-by: Ivy Zhang &lt;25222398+crazydemo@users.noreply.github.com&gt;
diff --git a/tests/integration/defs/accuracy/test_llm_api_pytorch.py b/tests/integration/defs/accuracy/test_llm_api_pytorch.py
@@ -2265,6 +2265,7 @@ def test_nvfp4_multi_gpus(self, tp_size, pp_size, ep_size, mtp_nextn, fp8kv,
             #               extra_evaluator_kwargs=dict(apply_chat_template=True))
 
     @skip_pre_blackwell
+    @pytest.mark.skip_less_device_memory(95000)
     @pytest.mark.parametrize(
         "tp_size,pp_size,ep_size,mtp_nextn,fp8kv,attention_dp,cuda_graph,overlap_scheduler,max_batch_size,moe_backend",
         [

Original file line number	Diff line number	Diff line change
`@@ -2265,6 +2265,7 @@ def test_nvfp4_multi_gpus(self, tp_size, pp_size, ep_size, mtp_nextn, fp8kv,`
`2265`	`2265`	`# extra_evaluator_kwargs=dict(apply_chat_template=True))`
`2266`	`2266`
`2267`	`2267`	`@skip_pre_blackwell`
	`2268`	`+ @pytest.mark.skip_less_device_memory(95000)`
`2268`	`2269`	`@pytest.mark.parametrize(`
`2269`	`2270`	`"tp_size,pp_size,ep_size,mtp_nextn,fp8kv,attention_dp,cuda_graph,overlap_scheduler,max_batch_size,moe_backend",`
`2270`	`2271`	`[`