From 07d9390e9b050bf16ea7cc99980211a084ed1ed5 Mon Sep 17 00:00:00 2001 From: xinhe-nv <200704525+xinhe-nv@users.noreply.github.com> Date: Wed, 14 Jan 2026 11:43:00 +0800 Subject: [PATCH] [None][test] add test into qa test list (#10627) Signed-off-by: Xin He (SW-GPU) <200704525+xinhe-nv@users.noreply.github.com> --- .../accuracy/test_disaggregated_serving.py | 1 + .../test_lists/qa/llm_function_core.txt | 5 + .../qa/llm_function_core_sanity.txt | 321 +++++++++--------- 3 files changed, 170 insertions(+), 157 deletions(-) diff --git a/tests/integration/defs/accuracy/test_disaggregated_serving.py b/tests/integration/defs/accuracy/test_disaggregated_serving.py index 14bb0cb811..26763fcf56 100644 --- a/tests/integration/defs/accuracy/test_disaggregated_serving.py +++ b/tests/integration/defs/accuracy/test_disaggregated_serving.py @@ -1131,6 +1131,7 @@ class TestGPTOSS(LlmapiAccuracyTestHarness): @pytest.mark.timeout(DEFAULT_TEST_TIMEOUT) +@skip_pre_blackwell class TestDeepSeekV32Exp(LlmapiAccuracyTestHarness): MODEL_NAME = "deepseek-ai/DeepSeek-V3.2-Exp" MODEL_PATH = f"{llm_models_root()}/DeepSeek-V3.2-Exp-FP4-v2" diff --git a/tests/integration/test_lists/qa/llm_function_core.txt b/tests/integration/test_lists/qa/llm_function_core.txt index 8e7873f61c..d56861e986 100644 --- a/tests/integration/test_lists/qa/llm_function_core.txt +++ b/tests/integration/test_lists/qa/llm_function_core.txt @@ -104,6 +104,8 @@ accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_ accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_adp_lmtp] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_trtllmgen_adp_lmtp] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_bs8_mtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_mtp] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_tp8] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_tp4] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_pp4_mtp] @@ -111,6 +113,8 @@ accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_chunked_ accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_chunked_prefill[throughput_tp4] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_corner_case accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale[throughput] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale[throughput_mtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale[throughput_mtp_trtllm] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale_chunked_prefill[latency] accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale_chunked_prefill[throughput] accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[baseline] @@ -264,6 +268,7 @@ accuracy/test_llm_api_pytorch_multimodal.py::TestGemma3_27BInstruct::test_fp8_pr accuracy/test_llm_api_pytorch_multimodal.py::TestQwen3VL_MOE::test_auto_dtype # disaggregated serving accuracy test +accuracy/test_disaggregated_serving.py::TestDeepSeekV32Exp::test_auto_dtype[False] accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=0-overlap_scheduler=True] accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=0-overlap_scheduler=False] accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=2-overlap_scheduler=True] diff --git a/tests/integration/test_lists/qa/llm_function_core_sanity.txt b/tests/integration/test_lists/qa/llm_function_core_sanity.txt index 75282580e0..8a276d4615 100644 --- a/tests/integration/test_lists/qa/llm_function_core_sanity.txt +++ b/tests/integration/test_lists/qa/llm_function_core_sanity.txt @@ -1,119 +1,73 @@ # text generation accuracy test -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8B::test_nvfp4 -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_chunked_prefill[attn_backend=FLASHINFER] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_eagle3[sampler_async_worker=False-eagle3_one_model=False-overlap_scheduler=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_eagle3[sampler_async_worker=False-eagle3_one_model=True-overlap_scheduler=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_eagle3[sampler_async_worker=True-eagle3_one_model=True-overlap_scheduler=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_llm_sampler -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_4gpus[xgrammar] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding[xgrammar] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_ngram -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=True-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=True-sampler_async_worker=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_2_1B::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestLlama3_2_1B::test_fp8_prequantized -accuracy/test_llm_api_pytorch.py::TestLlama3_2_3B::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestLlama3_2_3B::test_fp8_prequantized -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp4_tp2pp2[torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp4_tp2pp2[torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_tp4[torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_tp4[torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_nvfp4_tp4[torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_nvfp4_tp4[torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=True-torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=True-torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=False-torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=False-torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_chunked_prefill[attn_backend=FLASHINFER] -accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_chunked_prefill[attn_backend=TRTLLM] -accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8_chunked_prefill[tp8ep8-cuda_graph=True] -accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp4_chunked_prefill[tp4ep4-cuda_graph=True] -accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp8_chunked_prefill[tp4ep4-cuda_graph=True] +accuracy/test_llm_api_pytorch.py::TestBielik11BInstruct::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestBielik11BInstruct::test_fp8 +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale[throughput] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale[throughput_mtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale[throughput_mtp_trtllm] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale_chunked_prefill[latency] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale_chunked_prefill[throughput] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_adp_lmtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_trtllmgen] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_trtllmgen_adp_lmtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_bs8_mtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_mtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_pp4_mtp] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_tp4] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_tp8] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_chunked_prefill[latency] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_chunked_prefill[throughput_tp4] +accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_corner_case +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[baseline] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[baseline_fp8kv] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[baseline_mtp1] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[disable_skip_indexer] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[latency] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline_fp8kv] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline_mtp1] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline_pp4_mtp1] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[disable_skip_indexer] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[latency] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus_chunked_prefill[baseline_fp8kv] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus_chunked_prefill[latency] accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_online_eplb[mtp_nextn=0-moe_backend=WIDEEP] accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_online_eplb[mtp_nextn=2-moe_backend=WIDEEP] accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_fp8_block_scales_4gpus_static_eplb -accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4_4gpus_online_eplb[fp8kv=False-moe_backend=WIDEEP] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4_4gpus_online_eplb[fp8kv=True-moe_backend=WIDEEP] accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4[moe_backend=CUTLASS-mtp_nextn=2-fp8kv=False-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=False] accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4[moe_backend=CUTLASS-mtp_nextn=2-fp8kv=False-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_trtllmgen] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_adp_lmtp] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency_trtllmgen_adp_lmtp] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_tp4] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_tp8] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_pp4_mtp] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_chunked_prefill[latency] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_chunked_prefill[throughput_tp4] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale[throughput] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale_chunked_prefill[latency] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_fp8_blockscale_chunked_prefill[throughput] -accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus_corner_case -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[baseline] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[baseline_mtp1] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[baseline_fp8kv] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[latency] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_fp8_blockscale[disable_skip_indexer] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline_mtp1] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline_fp8kv] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[latency] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[disable_skip_indexer] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus[baseline_pp4_mtp1] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus_chunked_prefill[baseline_fp8kv] -accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus_chunked_prefill[latency] -accuracy/test_llm_api_pytorch.py::TestQwen2_7BInstruct::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestQwen3_4B::test_eagle3 -accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_fp8_block_scales[latency] -accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_w4a8_mxfp4[fp8-latency] -accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_w4a8_mxfp4[mxfp8-latency] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_fp8_block_scales[latency-torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_fp8_block_scales[latency-torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_cutlass-torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_cutlass-torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_trtllm-torch_compile=False] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_trtllm-torch_compile=True] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a16_mxfp4[latency-TRTLLM] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[fp8-latency-CUTLASS] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[fp8-latency-TRITON] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[fp8-latency-TRTLLM] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[mxfp8-latency-CUTLASS] -accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[mxfp8-latency-TRTLLM] -accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_fp8[latency] -accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_fp8[throughput_latency] -accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_nvfp4_4gpus[latency_moe_trtllm_eagle3] -accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_nvfp4[latency_moe_cutlass] -accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_nvfp4[latency_moe_trtllm] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4_4gpus_online_eplb[fp8kv=False-moe_backend=WIDEEP] +accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4_4gpus_online_eplb[fp8kv=True-moe_backend=WIDEEP] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-one_model-no_overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-one_model-overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-two_model-no_overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-two_model-overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-one_model-no_overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-one_model-overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-two_model-no_overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-two_model-overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-one_model-no_overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-one_model-overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-two_model-no_overlap_scheduler] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-two_model-overlap_scheduler] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[True-True-cutlass-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[True-True-cutlass-fp8] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[True-True-triton-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[True-True-trtllm-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[True-True-trtllm-fp8] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-cutlass-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-trtllm-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-trtllm-fp8] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-triton-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-cutlass-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-trtllm-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-trtllm-fp8] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-triton-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[dp2-cutlass-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[dp2-triton-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[dp2-trtllm-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[dp2-trtllm-fp8] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[dp2-triton-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-cutlass-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-triton-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-trtllm-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[ep2-trtllm-fp8] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-cutlass-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-triton-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-trtllm-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_2gpus[tp2-trtllm-fp8] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[dp4-cutlass-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[dp4-triton-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[dp4-trtllm-auto] @@ -126,53 +80,109 @@ accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[tp4-cutlass-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[tp4-triton-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[tp4-trtllm-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[tp4-trtllm-fp8] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[cutlass-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[triton-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[trtllm-auto] +accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[trtllm-fp8] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4a16[dp4-auto] accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4a16[dp4-fp8] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[cutlass-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[trtllm-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[triton-auto] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_chunked_prefill[trtllm-fp8] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-one_model-overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-one_model-no_overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-two_model-overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[cutlass-two_model-no_overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-one_model-overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-one_model-no_overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-two_model-overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[trtllm-two_model-no_overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-one_model-overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-one_model-no_overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-two_model-overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[triton-two_model-no_overlap_scheduler] -accuracy/test_llm_api_pytorch.py::TestMistral7B::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestMistralSmall24B::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestMixtral8x7B::test_fp8_tp2 -accuracy/test_llm_api_pytorch.py::TestMixtral8x7B::test_nvfp4_tp2 -accuracy/test_llm_api_pytorch.py::TestMinistral8BInstruct::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestMinistral8BInstruct::test_fp8 accuracy/test_llm_api_pytorch.py::TestGemma3_1BInstruct::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestGemma3_27BInstruct::test_fp8_prequantized accuracy/test_llm_api_pytorch.py::TestGemma3_27BInstruct::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestMinitron4BBaseInstruct::test_fp8_prequantized -accuracy/test_llm_api_pytorch.py::TestNemotronNas::test_auto_dtype_tp8 -accuracy/test_llm_api_pytorch.py::TestSeedOss_36B::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestGemma3_27BInstruct::test_fp8_prequantized accuracy/test_llm_api_pytorch.py::TestKanana_Instruct::test_auto_dtype accuracy/test_llm_api_pytorch.py::TestKimiK2::test_fp8_blockscale[latency] accuracy/test_llm_api_pytorch.py::TestKimiK2::test_nvfp4[4gpus] accuracy/test_llm_api_pytorch.py::TestKimiK2::test_nvfp4[8gpus] -accuracy/test_llm_api_pytorch.py::TestBielik11BInstruct::test_auto_dtype -accuracy/test_llm_api_pytorch.py::TestBielik11BInstruct::test_fp8 +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8B::test_nvfp4 +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_auto_dtype_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=True-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_chunked_prefill[attn_backend=FLASHINFER] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_eagle3[sampler_async_worker=False-eagle3_one_model=False-overlap_scheduler=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_eagle3[sampler_async_worker=False-eagle3_one_model=True-overlap_scheduler=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_eagle3[sampler_async_worker=True-eagle3_one_model=True-overlap_scheduler=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=False-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=False-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=False-disable_overlap_scheduler=True-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=False-sampler_async_worker=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_beam_search[enable_cuda_graph=True-enable_padding=True-disable_overlap_scheduler=True-sampler_async_worker=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_llm_sampler +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding[xgrammar] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_4gpus[xgrammar] +accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_ngram +accuracy/test_llm_api_pytorch.py::TestLlama3_2_1B::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestLlama3_2_1B::test_fp8_prequantized +accuracy/test_llm_api_pytorch.py::TestLlama3_2_3B::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestLlama3_2_3B::test_fp8_prequantized +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp4_tp2pp2[torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp4_tp2pp2[torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=False-torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=False-torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=True-torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_eagle3_tp8[eagle3_one_model=True-torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_tp4[torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp8_tp4[torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_nvfp4_tp4[torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_nvfp4_tp4[torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestMinistral8BInstruct::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestMinistral8BInstruct::test_fp8 +accuracy/test_llm_api_pytorch.py::TestMinitron4BBaseInstruct::test_fp8_prequantized +accuracy/test_llm_api_pytorch.py::TestMistral7B::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestMistralSmall24B::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestMixtral8x7B::test_fp8_tp2 +accuracy/test_llm_api_pytorch.py::TestMixtral8x7B::test_nvfp4_tp2 +accuracy/test_llm_api_pytorch.py::TestNemotronNas::test_auto_dtype_tp8 +accuracy/test_llm_api_pytorch.py::TestNemotronV3Nano::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestNemotronV3Nano::test_fp8 +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-4-False-True-True] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-1-True-True-True] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-4-False-True-False] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-1-False-False-False] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-4-True-False-False] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-1-True-True-False] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-1-False-False-True] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_auto_dtype_4gpus[4-4-True-False-True] +accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_8gpus accuracy/test_llm_api_pytorch.py::TestPhi4::test_auto_dtype accuracy/test_llm_api_pytorch.py::TestPhi4::test_fp8 accuracy/test_llm_api_pytorch.py::TestPhi4MiniInstruct::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestQwen2_7BInstruct::test_auto_dtype +accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_fp8[latency] +accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_fp8[throughput_latency] +accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_nvfp4[latency_moe_cutlass] +accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_nvfp4[latency_moe_trtllm] +accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_nvfp4_4gpus[latency_moe_trtllm_eagle3] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_fp8_block_scales[latency-torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_fp8_block_scales[latency-torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_cutlass-torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_cutlass-torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_trtllm-torch_compile=False] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_nvfp4[latency_moe_trtllm-torch_compile=True] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a16_mxfp4[latency-TRTLLM] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[fp8-latency-CUTLASS] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[fp8-latency-TRITON] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[fp8-latency-TRTLLM] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[mxfp8-latency-CUTLASS] +accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a8_mxfp4[mxfp8-latency-TRTLLM] +accuracy/test_llm_api_pytorch.py::TestQwen3_4B::test_eagle3 +accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_fp8_block_scales[latency] +accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_w4a8_mxfp4[fp8-latency] +accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_w4a8_mxfp4[mxfp8-latency] +accuracy/test_llm_api_pytorch.py::TestSeedOss_36B::test_auto_dtype # multimodal accuracy tests -accuracy/test_llm_api_pytorch_multimodal.py::TestQwen2_VL_7B::test_auto_dtype -accuracy/test_llm_api_pytorch_multimodal.py::TestQwen2_5_VL_7B::test_auto_dtype +accuracy/test_llm_api_pytorch_multimodal.py::TestGemma3_27BInstruct::test_fp8_prequantized accuracy/test_llm_api_pytorch_multimodal.py::TestLlava_V1_6_Mistral_7B::test_auto_dtype accuracy/test_llm_api_pytorch_multimodal.py::TestNVILA_8B::test_auto_dtype +accuracy/test_llm_api_pytorch_multimodal.py::TestQwen2_5_VL_7B::test_auto_dtype +accuracy/test_llm_api_pytorch_multimodal.py::TestQwen2_VL_7B::test_auto_dtype accuracy/test_llm_api_pytorch_multimodal.py::TestVILA1_5_3B::test_auto_dtype -accuracy/test_llm_api_pytorch_multimodal.py::TestGemma3_27BInstruct::test_fp8_prequantized # disaggregated serving accuracy test accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=0-overlap_scheduler=False] @@ -180,6 +190,9 @@ accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_ accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=2-overlap_scheduler=False] accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=2-overlap_scheduler=True] accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_nixl_backend +accuracy/test_disaggregated_serving.py::TestDeepSeekV32Exp::test_auto_dtype[False] +accuracy/test_disaggregated_serving.py::TestGPTOSS::test_auto_dtype[False] +accuracy/test_disaggregated_serving.py::TestGPTOSS::test_auto_dtype[True] accuracy/test_disaggregated_serving.py::TestGemma3_1BInstruct::test_auto_dtype[False] accuracy/test_disaggregated_serving.py::TestGemma3_1BInstruct::test_auto_dtype[True] accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_auto_dtype[False-False-False] @@ -201,64 +214,58 @@ accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symm accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[MMLU-tp1pp2] accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[MMLU-tp2pp1] accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[MMLU-tp2pp2] -accuracy/test_disaggregated_serving.py::TestLlama4ScoutInstruct::test_auto_dtype[False] -accuracy/test_disaggregated_serving.py::TestLlama4ScoutInstruct::test_auto_dtype[True] accuracy/test_disaggregated_serving.py::TestQwen3_30B_A3B::test_mixed_ctx_gen_model[ctxpp2gentp2] accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_nixl_backend -accuracy/test_disaggregated_serving.py::TestGPTOSS::test_auto_dtype[True] -accuracy/test_disaggregated_serving.py::TestGPTOSS::test_auto_dtype[False] # e2e test +test_e2e.py::test_eagle3_output_consistency_4gpus[Qwen3/saved_models_Qwen3-235B-A22B_fp8_hf-Qwen3/qwen3-235B-eagle3] +test_e2e.py::test_eagle3_output_consistency_4gpus[Qwen3/saved_models_Qwen3-235B-A22B_nvfp4_hf-Qwen3/qwen3-235B-eagle3] test_e2e.py::test_openai_chat_harmony test_e2e.py::test_openai_consistent_chat test_e2e.py::test_openai_multi_chat_example test_e2e.py::test_ptp_quickstart -test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Llama3.1-405B-FP8-llama-3.1-model/Llama-3.1-405B-Instruct-FP8-8] -test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Llama3.1-70B-BF16-llama-3.1-model/Meta-Llama-3.1-70B-8] -test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Llama3.1-70B-FP8-llama-3.1-model/Llama-3.1-70B-Instruct-FP8-2] -test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Mixtral-8x7B-BF16-Mixtral-8x7B-v0.1-8] -test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Mixtral-8x7B-NVFP4-nvfp4-quantized/Mixtral-8x7B-Instruct-v0.1-8] -test_e2e.py::test_ptp_quickstart_advanced_deepseek_r1_8gpus[DeepSeek-R1-DeepSeek-R1/DeepSeek-R1] -test_e2e.py::test_ptp_quickstart_advanced_ngram[Llama-3.1-8B-Instruct-llama-3.1-model/Llama-3.1-8B-Instruct] test_e2e.py::test_ptp_quickstart_advanced[Llama3.1-8B-BF16-llama-3.1-model/Meta-Llama-3.1-8B] test_e2e.py::test_ptp_quickstart_advanced[Llama3.1-8B-FP8-llama-3.1-model/Llama-3.1-8B-Instruct-FP8] test_e2e.py::test_ptp_quickstart_advanced[Llama3.1-8B-NVFP4-nvfp4-quantized/Meta-Llama-3.1-8B] test_e2e.py::test_ptp_quickstart_advanced[Llama3.2-11B-BF16-llama-3.2-models/Llama-3.2-11B-Vision] test_e2e.py::test_ptp_quickstart_advanced[Qwen3-30B-A3B-Qwen3/Qwen3-30B-A3B] +test_e2e.py::test_ptp_quickstart_advanced_deepseek_r1_8gpus[DeepSeek-R1-DeepSeek-R1/DeepSeek-R1] +test_e2e.py::test_ptp_quickstart_advanced_deepseek_r1_w4afp8_8gpus[DeepSeek-R1-W4AFP8-DeepSeek-R1/DeepSeek-R1-W4AFP8] +test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Llama3.1-405B-FP8-llama-3.1-model/Llama-3.1-405B-Instruct-FP8-8] +test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Llama3.1-70B-BF16-llama-3.1-model/Meta-Llama-3.1-70B-8] +test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Llama3.1-70B-FP8-llama-3.1-model/Llama-3.1-70B-Instruct-FP8-2] +test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Mixtral-8x7B-BF16-Mixtral-8x7B-v0.1-8] +test_e2e.py::test_ptp_quickstart_advanced_multi_gpus[Mixtral-8x7B-NVFP4-nvfp4-quantized/Mixtral-8x7B-Instruct-v0.1-8] +test_e2e.py::test_ptp_quickstart_advanced_ngram[Llama-3.1-8B-Instruct-llama-3.1-model/Llama-3.1-8B-Instruct] test_e2e.py::test_ptp_quickstart_bert[TRTLLM-BertForSequenceClassification-bert/bert-base-uncased-yelp-polarity] test_e2e.py::test_ptp_quickstart_bert[VANILLA-BertForSequenceClassification-bert/bert-base-uncased-yelp-polarity] test_e2e.py::test_ptp_scaffolding[DeepSeek-R1-Distill-Qwen-7B-DeepSeek-R1/DeepSeek-R1-Distill-Qwen-7B] -test_e2e.py::test_ptp_quickstart_advanced_deepseek_r1_w4afp8_8gpus[DeepSeek-R1-W4AFP8-DeepSeek-R1/DeepSeek-R1-W4AFP8] test_e2e.py::test_ptp_star_attention_example[Llama3.1-8B-BF16-llama-3.1-model/Meta-Llama-3.1-8B] test_e2e.py::test_qwen_e2e_cpprunner_large_new_tokens[DeepSeek-R1-Distill-Qwen-1.5B-DeepSeek-R1-Distill-Qwen-1.5B] test_e2e.py::test_relaxed_acceptance_quickstart_advanced_deepseek_r1_8gpus[DeepSeek-R1-DeepSeek-R1/DeepSeek-R1] test_e2e.py::test_trtllm_bench_pytorch_backend_sanity[meta-llama/Llama-3.1-8B-llama-3.1-8b-hf-nvfp4-False-False] test_e2e.py::test_trtllm_benchmark_serving[gpt_oss/gpt-oss-20b] test_e2e.py::test_trtllm_multimodal_benchmark_serving -test_e2e.py::test_eagle3_output_consistency_4gpus[Qwen3/saved_models_Qwen3-235B-A22B_fp8_hf-Qwen3/qwen3-235B-eagle3] -test_e2e.py::test_eagle3_output_consistency_4gpus[llama4-models/nvidia/Llama-4-Maverick-17B-128E-Instruct-FP8-Llama-4-Maverick-17B-128E-Eagle3] -test_e2e.py::test_eagle3_output_consistency_4gpus[Qwen3/saved_models_Qwen3-235B-A22B_nvfp4_hf-Qwen3/qwen3-235B-eagle3] # e2e disaggregated serving test -disaggregated/test_disaggregated.py::test_disaggregated_cache_aware_balance[TinyLlama-1.1B-Chat-v1.0] +disaggregated/test_auto_scaling.py::test_disagg_server_restart[etcd-round_robin] +disaggregated/test_auto_scaling.py::test_minimal_instances[etcd-round_robin] +disaggregated/test_auto_scaling.py::test_service_discovery[etcd-round_robin] +disaggregated/test_auto_scaling.py::test_worker_restart[etcd-round_robin] disaggregated/test_disaggregated.py::test_disaggregated_cache_aware_balance[TinyLlama-1.1B-Chat-v1.0] disaggregated/test_disaggregated.py::test_disaggregated_cuda_graph[TinyLlama-1.1B-Chat-v1.0] -disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_attention_dp_one_mtp[DeepSeek-V3-Lite-fp8] -disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_attention_dp_one[DeepSeek-V3-Lite-fp8] disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_attention_dp[DeepSeek-V3-Lite-fp8] +disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_attention_dp_one[DeepSeek-V3-Lite-fp8] +disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_attention_dp_one_mtp[DeepSeek-V3-Lite-fp8] disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_mpi[DeepSeek-V3-Lite-fp8] disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_nixl[DeepSeek-V3-Lite-fp8] disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_ucx[DeepSeek-V3-Lite-fp8] disaggregated/test_disaggregated.py::test_disaggregated_load_balance[TinyLlama-1.1B-Chat-v1.0] disaggregated/test_disaggregated.py::test_disaggregated_multi_gpu_with_mpirun[TinyLlama-1.1B-Chat-v1.0] -disaggregated/test_disaggregated.py::test_disaggregated_single_gpu_with_mpirun_trt_backend[TinyLlama-1.1B-Chat-v1.0] disaggregated/test_disaggregated.py::test_disaggregated_single_gpu_with_mpirun[TinyLlama-1.1B-Chat-v1.0] +disaggregated/test_disaggregated.py::test_disaggregated_single_gpu_with_mpirun_trt_backend[TinyLlama-1.1B-Chat-v1.0] disaggregated/test_disaggregated.py::test_disaggregated_trtllm_sampler[TinyLlama-1.1B-Chat-v1.0] disaggregated/test_workers.py::test_workers_conditional_disaggregation[TinyLlama-1.1B-Chat-v1.0] -disaggregated/test_workers.py::test_workers_kv_cache_aware_router_eviction[TinyLlama-1.1B-Chat-v1.0] disaggregated/test_workers.py::test_workers_kv_cache_aware_router[TinyLlama-1.1B-Chat-v1.0] +disaggregated/test_workers.py::test_workers_kv_cache_aware_router_eviction[TinyLlama-1.1B-Chat-v1.0] disaggregated/test_workers.py::test_workers_kv_cache_events[TinyLlama-1.1B-Chat-v1.0] -disaggregated/test_auto_scaling.py::test_service_discovery[etcd-round_robin] -disaggregated/test_auto_scaling.py::test_minimal_instances[etcd-round_robin] -disaggregated/test_auto_scaling.py::test_worker_restart[etcd-round_robin] -disaggregated/test_auto_scaling.py::test_disagg_server_restart[etcd-round_robin]