[Test] - Waive or fix few known test failures (NVIDIA#5769)

Signed-off-by: Yanchao Lu <[email protected]>

Author: chzblych

jenkins/L0_Test.groovy

@@ -1825,9 +1825,10 @@ def launchTestJobs(pipeline, testFilter, dockerNode=null)
 
     multiNodesSBSAConfigs = [
         // Each stage test 1 testcase with 8 GPUs and 2 nodes.
-        "GB200-8_GPUs-2_Nodes-PyTorch-Post-Merge-1": ["gb200-multi-node", "l0_gb200_multi_nodes", 1, 3, 8, 2],
-        "GB200-8_GPUs-2_Nodes-PyTorch-Post-Merge-2": ["gb200-multi-node", "l0_gb200_multi_nodes", 2, 3, 8, 2],
-        "GB200-8_GPUs-2_Nodes-PyTorch-Post-Merge-3": ["gb200-multi-node", "l0_gb200_multi_nodes", 3, 3, 8, 2],
+        "GB200-8_GPUs-2_Nodes-PyTorch-Post-Merge-1": ["gb200-multi-node", "l0_gb200_multi_nodes", 1, 4, 8, 2],
+        "GB200-8_GPUs-2_Nodes-PyTorch-Post-Merge-2": ["gb200-multi-node", "l0_gb200_multi_nodes", 2, 4, 8, 2],
+        "GB200-8_GPUs-2_Nodes-PyTorch-Post-Merge-3": ["gb200-multi-node", "l0_gb200_multi_nodes", 3, 4, 8, 2],
+        "GB200-8_GPUs-2_Nodes-PyTorch-Post-Merge-4": ["gb200-multi-node", "l0_gb200_multi_nodes", 4, 4, 8, 2],
     ]
     fullSet += multiNodesSBSAConfigs.keySet()
 

tests/integration/test_lists/test-db/l0_dgx_b200.yml

@@ -49,4 +49,4 @@ l0_dgx_b200:
   - disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_attention_dp_overlap_cuda_graph[DeepSeek-V3-Lite-fp8]
   - disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_overlap_cuda_graph[DeepSeek-V3-Lite-fp8]
   - accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[4gpu_latency_moe_trtllm]
-  - accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[4gpu_latency_moe_cutlass] TIMEOUT (90)
+  - accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[4gpu_latency_moe_cutlass]

tests/integration/test_lists/waives.txt

@@ -453,3 +453,6 @@ accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_fp8_block_scales[latency] S
 full:GH200/disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_attention_dp_one[DeepSeek-V3-Lite-fp8] SKIP (https://nvbugs/5375966)
 accuracy/test_llm_api_pytorch.py::TestGemma3_1BInstruct::test_auto_dtype SKIP (https://nvbugs/5375620)
 disaggregated/test_workers.py::test_workers_kv_cache_aware_router_eviction[TinyLlama-1.1B-Chat-v1.0] SKIP (https://nvbugs/5375723)
+test_e2e.py::test_ptp_quickstart_advanced[Mixtral-8x7B-NVFP4-nvfp4-quantized/Mixtral-8x7B-Instruct-v0.1] SKIP (https://nvbugs/5377465)
+test_e2e.py::test_ptp_quickstart_advanced[Llama3.1-70B-FP8-llama-3.1-model/Llama-3.1-70B-Instruct-FP8] SKIP (https://nvbugs/5377465)
+accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[4gpu_latency_moe_cutlass] SKIP (https://nvbugs/5369010)

tests/unittest/_torch/multi_gpu_modeling/test_deepseek.py

@@ -17,7 +17,6 @@ def similar(a, b, threshold=0.9):
     return SequenceMatcher(None, a, b).ratio() >= threshold
 
 
-@pytest.mark.skip(reason="https://nvbugs/5370968")
 @pytest.mark.parametrize("model_name", ["DeepSeek-V3-Lite"],
                          ids=["deepseekv3_lite"])
 @pytest.mark.parametrize("backend", ["TRTLLM"], ids=["trtllm"])