[https://nvbugs/5519462][fix] Add failed cases into waives.txt (NVIDIA#7817)

Signed-off-by: xinhe-nv <[email protected]>
Signed-off-by: Xin He (SW-GPU) <[email protected]>

Author: xinhe-nv

tests/integration/defs/accuracy/test_disaggregated_serving.py

@@ -697,9 +697,9 @@ def test_nixl_backend(self):
             task.evaluate(llm)
 
     @pytest.mark.skip_less_device(8)
+    @skip_pre_hopper
     @parametrize_with_ids("overlap_scheduler", [True, False])
-    @parametrize_with_ids("mtp_nextn",
-                          [0, pytest.param(2, marks=skip_pre_hopper)])
+    @parametrize_with_ids("mtp_nextn", [0, 2])
     @pytest.mark.skip_less_device(8)
     def test_auto_dtype(self, overlap_scheduler, mtp_nextn):
         ctx_server_config = {"disable_overlap_scheduler": True}

tests/integration/defs/accuracy/test_llm_api_pytorch.py

@@ -1130,23 +1130,21 @@ def test_nvfp4_tp2(self):
 # This class has extensively parameterized test methods, which yield totally 200 test cases.
 # This is because this model requires high test coverage over the feature combinations.
 # Normally we should not parameterize test methods so extensively -- just test on the typical/important feature combinations.
+@skip_pre_hopper
 class TestDeepSeekV3Lite(LlmapiAccuracyTestHarness):
     MODEL_NAME = "deepseek-ai/DeepSeek-V3-Lite"
     MODEL_PATH = f"{llm_models_root()}/DeepSeek-V3-Lite/bf16"
 
     @pytest.mark.skip_less_device_memory(60000)
     # Chunked Prefill for MLA can only be enabled on SM100
-    @parametrize_with_ids(
-        "enable_chunked_prefill",
-        [False, pytest.param(True, marks=skip_pre_hopper)])
+    @parametrize_with_ids("enable_chunked_prefill", [False, True])
     @parametrize_with_ids("torch_compile", [False, True])
     @parametrize_with_ids("attention_dp,cuda_graph,overlap_scheduler",
                           [(False, False, False), (True, False, False),
                            (False, True, False), (False, False, True),
                            (False, True, True), (True, True, True)])
     # Only Hopper and Blackwell MLA kernel supports MTP
-    @parametrize_with_ids("mtp_nextn",
-                          [0, pytest.param(2, marks=skip_pre_hopper)])
+    @parametrize_with_ids("mtp_nextn", [0, 2])
     def test_bfloat16(self, mtp_nextn, attention_dp, cuda_graph,
                       overlap_scheduler, torch_compile, enable_chunked_prefill):
         kv_cache_config = KvCacheConfig(free_gpu_memory_fraction=0.75)

tests/integration/test_lists/waives.txt

@@ -283,9 +283,7 @@ disaggregated/test_disaggregated.py::test_disaggregated_diff_max_tokens[TinyLlam
 disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_tp1_single_gpu_mtp[DeepSeek-V3-Lite-fp8] SKIP (https://nvbugs/5465642)
 examples/test_multimodal.py::test_llm_multimodal_general[Mistral-Small-3.1-24B-Instruct-2503-pp:1-tp:1-bfloat16-bs:1-cpp_e2e:False-nb:1] SKIP (https://nvbugs/5431146)
 accuracy/test_disaggregated_serving.py::TestQwen3_30B_A3B::test_mixed_ctx_gen_model[ctxpp2gentp2] SKIP (https://nvbugs/5470769)
-full:L40S/accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=0-overlap_scheduler=False] SKIP (https://nvbugs/5347051)
 full:L40S/accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[MMLU-tp2pp2] SKIP (https://nvbugs/5471108)
-full:L20/accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=0-overlap_scheduler=False] SKIP (https://nvbugs/5347051)
 full:L20/accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[MMLU-tp2pp2] SKIP (https://nvbugs/5471108)
 test_e2e.py::test_multi_nodes_eval[llama4-models/nvidia/Llama-4-Maverick-17B-128E-Instruct-FP8-tp8pp2-mmlu] SKIP (https://nvbugs/5473781)
 disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_llama_context_capacity[False-False-DeepSeek-V3-Lite-fp8/fp8] SKIP (https://nvbugs/5477404)
@@ -337,9 +335,6 @@ full:H100/accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8
 full:H100/accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8[tp8ep4-cuda_graph=True] SKIP (https://nvbugs/5512734)
 full:H100/accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8[tp8-cuda_graph=True] SKIP (https://nvbugs/5512734)
 full:H100/accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8_eagle3[tp8-torch_compile=True] SKIP (https://nvbugs/5483534)
-full:A100/accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=0-overlap_scheduler=False] SKIP (https://nvbugs/5519462)
-full:A100/accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=0-overlap_scheduler=True] SKIP (https://nvbugs/5519462)
-full:A100/accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16[mtp_nextn=0-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=False-enable_chunked_prefill=False] SKIP (https://nvbugs/5347051)
 full:A100/test_e2e.py::test_ptp_quickstart_multimodal[NVILA-8B-FP16-vila/NVILA-8B-video-False] SKIP (https://nvbugs/5453725)
 test_e2e.py::test_ptp_scaffolding[DeepSeek-R1-Distill-Qwen-7B-DeepSeek-R1/DeepSeek-R1-Distill-Qwen-7B] SKIP (https://nvbugs/5517260)
 accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp8_chunked_prefill[tp4ep4-cuda_graph=True] SKIP (https://nvbugs/5522462)