waive flaky tests

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

Author: xinhe-nv

tests/integration/defs/accuracy/test_disaggregated_serving.py

@@ -358,8 +358,8 @@ class TestLlama3_1_8BInstruct(LlmapiAccuracyTestHarness):
     MODEL_NAME = "meta-llama/Llama-3.1-8B-Instruct"
     MODEL_PATH = f"{llm_models_root()}/llama-3.1-model/Llama-3.1-8B-Instruct"
 
+    @skip_pre_hopper
     @pytest.mark.skip_less_device(2)
-    @pytest.mark.skip_less_device_memory(32000)
     @pytest.mark.parametrize("disable_overlap_scheduler", [False, True])
     @pytest.mark.parametrize("ctx_enable_block_reuse", [True, False])
     @pytest.mark.parametrize("gen_enable_block_reuse", [True, False])

tests/integration/test_lists/waives.txt

@@ -388,3 +388,9 @@ examples/test_phi.py::test_phi_fp8_with_bf16_lora[Phi-3.5-MoE-instruct] SKIP (ht
 unittest/llmapi/test_llm_multi_gpu_pytorch.py::test_llm_rpc_tp2 SKIP (https://nvbugs/5594753)
 accuracy/test_llm_api_pytorch.py::TestQwen2_VL_7B::test_auto_dtype SKIP (https://nvbugs/5602338)
 unittest/llmapi/test_memory_profiling.py SKIP (https://nvbugs/5580781)
+accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=1-ctx_pp=2] SKIP (https://nvbugs/5582277)
+accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=1-ctx_pp=4] SKIP (https://nvbugs/5582277)
+accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=2-ctx_pp=2] SKIP (https://nvbugs/5582277)
+accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=2-ctx_pp=4] SKIP (https://nvbugs/5582277)
+accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[MMLU-gen_tp=2-ctx_pp=2] SKIP (https://nvbugs/5582277)
+accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[MMLU-gen_tp=2-ctx_pp=4] SKIP (https://nvbugs/5582277)