# 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[use_temperature=False-attn_backend=FLASHINFER]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_chunked_prefill[use_temperature=True-attn_backend=FLASHINFER]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_chunked_prefill[use_temperature=False-attn_backend=TRTLLM]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_chunked_prefill[use_temperature=True-attn_backend=TRTLLM]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_fp8_llm_sampler
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=False-eagle3_one_model=False-overlap_scheduler=False]
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_eagle3_sa
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_eagle3_sa_global_pool
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_ngram
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[llguidance]
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_4gpus[llguidance]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[xgrammar-eagle3_one_model=True]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[xgrammar-eagle3_one_model=False]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[llguidance-eagle3_one_model=True]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[llguidance-eagle3_one_model=False]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_ngram[xgrammar]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_ngram[llguidance]
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_1_8BInstruct::test_pard[overlap_scheduler=True]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_pard[overlap_scheduler=False]
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_pard_sa
accuracy/test_llm_api_pytorch.py::TestLlama3_1_8BInstruct::test_pard_sa_global_pool
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_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_fp4_tp2pp2[torch_compile=False-enable_gemm_allreduce_fusion=False]
accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp4_tp2pp2[torch_compile=False-enable_gemm_allreduce_fusion=True]
accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp4_tp2pp2[torch_compile=True-enable_gemm_allreduce_fusion=False]
accuracy/test_llm_api_pytorch.py::TestLlama3_3_70BInstruct::test_fp4_tp2pp2[torch_compile=True-enable_gemm_allreduce_fusion=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_auto_dtype[tp8-cuda_graph=False]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_auto_dtype[tp8ep4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_auto_dtype[tp8ep8-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_auto_dtype[tp4-cuda_graph=False]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_auto_dtype[tp4ep2-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_auto_dtype[tp4ep4-cuda_graph=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[tp8ep8-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8[tp8ep4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8[tp8-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8[tp4ep4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8[tp4ep2-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8[tp4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8_chunked_prefill[tp8ep8-cuda_graph=False]
accuracy/test_llm_api_pytorch.py::TestLlama4MaverickInstruct::test_fp8_chunked_prefill[tp8ep8-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_auto_dtype[tp8-cuda_graph=False]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_auto_dtype[tp8ep4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_auto_dtype[tp8ep8-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_auto_dtype[tp4-cuda_graph=False]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_auto_dtype[tp4ep2-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_auto_dtype[tp4ep4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp8[tp8ep8-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp8[tp4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp4[tp8ep8-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp4[tp4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp8_chunked_prefill[tp4ep4-cuda_graph=True]
accuracy/test_llm_api_pytorch.py::TestLlama4ScoutInstruct::test_fp4_chunked_prefill[tp4ep4-cuda_graph=True]
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-v2_kv_cache=False]
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-v2_kv_cache=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16[mtp_nextn=0-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=True-enable_chunked_prefill=False-v2_kv_cache=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16[mtp_nextn=0-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=True-enable_chunked_prefill=False-v2_kv_cache=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16[mtp_nextn=2-attention_dp=True-cuda_graph=True-overlap_scheduler=True-torch_compile=False-enable_chunked_prefill=True-v2_kv_cache=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16[mtp_nextn=2-attention_dp=True-cuda_graph=True-overlap_scheduler=True-torch_compile=False-enable_chunked_prefill=True-v2_kv_cache=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16[mtp_nextn=2-attention_dp=True-cuda_graph=True-overlap_scheduler=True-torch_compile=True-enable_chunked_prefill=True-v2_kv_cache=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16[mtp_nextn=2-attention_dp=True-cuda_graph=True-overlap_scheduler=True-torch_compile=True-enable_chunked_prefill=True-v2_kv_cache=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=0-attention_dp=False-cuda_graph=False-overlap_scheduler=False-enable_chunked_prefill=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=0-attention_dp=True-cuda_graph=True-overlap_scheduler=True-enable_chunked_prefill=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=0-attention_dp=False-cuda_graph=False-overlap_scheduler=False-enable_chunked_prefill=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=0-attention_dp=True-cuda_graph=True-overlap_scheduler=True-enable_chunked_prefill=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=2-attention_dp=False-cuda_graph=False-overlap_scheduler=False-enable_chunked_prefill=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=2-attention_dp=True-cuda_graph=True-overlap_scheduler=True-enable_chunked_prefill=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=2-attention_dp=False-cuda_graph=False-overlap_scheduler=False-enable_chunked_prefill=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_python_scheduler[mtp_nextn=2-attention_dp=True-cuda_graph=True-overlap_scheduler=True-enable_chunked_prefill=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_mtp_sa_global_pool
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_2_model_mtp
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_fp8_block_scales[mtp=disable-fp8kv=False-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_fp8_block_scales[mtp=disable-fp8kv=False-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4[moe_backend=CUTLASS-mtp_nextn=0-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=0-fp8kv=False-attention_dp=False-cuda_graph=False-overlap_scheduler=False-torch_compile=True]
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::TestDeepSeekV3Lite::test_nvfp4_batch_waiting[batch_wait_timeout_iters=10-batch_wait_max_tokens_ratio=0.75-mtp_nextn=0-fp8kv=False-cuda_graph=False-overlap_scheduler=False-torch_compile=False]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_nvfp4_batch_waiting[batch_wait_timeout_iters=10-batch_wait_max_tokens_ratio=0.75-mtp_nextn=0-fp8kv=False-cuda_graph=False-overlap_scheduler=False-torch_compile=True]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_python_scheduler[tp4-mtp_nextn=0]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_python_scheduler[tp4-mtp_nextn=2]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_python_scheduler[ep4-mtp_nextn=0]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_python_scheduler[ep4-mtp_nextn=2]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_kv_cache_aware_routing[mtp_nextn=0]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_bfloat16_4gpus_kv_cache_aware_routing[mtp_nextn=2]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_fp8_block_scales_4gpus_static_eplb
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_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_guided_decoding[xgrammar-mtp_nextn=0]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_guided_decoding[xgrammar-mtp_nextn=2]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_guided_decoding[llguidance-mtp_nextn=0]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_guided_decoding[llguidance-mtp_nextn=2]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_guided_decoding_4gpus[xgrammar-mtp_nextn=0]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_guided_decoding_4gpus[xgrammar-mtp_nextn=2]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_guided_decoding_4gpus[llguidance-mtp_nextn=0]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV3Lite::test_guided_decoding_4gpus[llguidance-mtp_nextn=2]
accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[latency]
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_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]
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::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]
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::TestDeepSeekV32::test_nvfp4_multi_gpus_chunked_prefill[latency_qsplit]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus_piecewise_cuda_graph[baseline]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_nvfp4_multi_gpus_piecewise_cuda_graph[mtp3_fp8kv_chunked]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_dsa_host_cache_offload[host_cache_offload]
accuracy/test_llm_api_pytorch.py::TestDeepSeekV32::test_dsa_host_cache_offload[host_cache_offload_mtp1]
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_eagle3[eagle3_one_model=True-enable_chunked_prefill=False-enable_max_concurrency=False-enable_draft_len_schedule=True]
accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_eagle3[eagle3_one_model=True-enable_chunked_prefill=False-enable_max_concurrency=True-enable_draft_len_schedule=False]
accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_fp8_block_scales[latency]
accuracy/test_llm_api_pytorch.py::TestQwen3_8B::test_bf16[multi_gpus_no_cache]
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_fp8[latency-torch_compile=False]
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_w4a8_mxfp4[fp8-latency-CUTLASS]
accuracy/test_llm_api_pytorch.py::TestQwen3_30B_A3B::test_w4a16_mxfp4[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-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_w4a16_mxfp4[latency-TRTLLM]
accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_fp8[throughput_latency]
accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_fp8[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[latency_moe_trtllm_attention_dp]
accuracy/test_llm_api_pytorch.py::TestQwen3_235B_A22B::test_nvfp4_4gpus[latency_moe_trtllm_eagle3]
accuracy/test_llm_api_pytorch.py::TestQwen3_5_35B_A3B::test_bf16
accuracy/test_llm_api_pytorch.py::TestQwen3_5_35B_A3B::test_fp8
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[v1_kv_cache-True-True-cutlass-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[v1_kv_cache-True-True-cutlass-fp8]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[v1_kv_cache-True-True-triton-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[v1_kv_cache-True-True-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[v1_kv_cache-True-True-trtllm-fp8]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[v2_kv_cache-True-True-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_1gpu[v2_kv_cache-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-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_4gpus[v1_kv_cache-dp4-cutlass-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-dp4-triton-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-dp4-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-dp4-trtllm-fp8]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-ep4-cutlass-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-ep4-triton-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-ep4-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-ep4-trtllm-fp8]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-tp4-cutlass-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-tp4-triton-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-tp4-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v1_kv_cache-tp4-trtllm-fp8]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache-dp4-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache-dp4-trtllm-fp8]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache-ep4-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache-ep4-trtllm-fp8]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache_no_reuse-tp4-trtllm-auto]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_w4_4gpus[v2_kv_cache_no_reuse-tp4-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[v1_kv_cache-cutlass-one_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-cutlass-one_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-cutlass-one_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-cutlass-one_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-cutlass-two_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-cutlass-two_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-cutlass-two_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-cutlass-two_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-trtllm-one_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-trtllm-one_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-trtllm-one_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-trtllm-one_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-trtllm-two_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-trtllm-two_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-trtllm-two_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-trtllm-two_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-triton-one_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-triton-one_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-triton-one_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-triton-one_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-triton-two_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-triton-two_model-overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v1_kv_cache-triton-two_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_4gpus[v2_kv_cache-triton-two_model-no_overlap_scheduler]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_vswa_reuse_4gpus[one_model]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_vswa_reuse_4gpus[two_model]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_guided_decoding_4gpus[one_model]
accuracy/test_llm_api_pytorch.py::TestGPTOSS::test_eagle3_guided_decoding_4gpus[two_model]
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::TestMistralSmall24B::test_fp8
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_1BInstruct::test_fp8_prequantized[torch_compile=False]
accuracy/test_llm_api_pytorch.py::TestGemma3_1BInstruct::test_fp8_prequantized[torch_compile=True]
accuracy/test_llm_api_pytorch.py::TestGemma3_1BInstruct::test_fp8_vswa_reuse
accuracy/test_llm_api_pytorch.py::TestGemma3_1BInstruct::test_fp8_guided_decoding_vswa_reuse[xgrammar]
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::TestGLM4_6::test_nvfp4_multi_gpus[throughput]
accuracy/test_llm_api_pytorch.py::TestGLM4_6::test_nvfp4_multi_gpus[throughput_trtllm]
accuracy/test_llm_api_pytorch.py::TestGLM4_6::test_nvfp4_2_model_mtp[2model]
accuracy/test_llm_api_pytorch.py::TestGLM4_6::test_nvfp4_2_model_mtp[2model_trtllm]
accuracy/test_llm_api_pytorch.py::TestGLM4_5Air::test_nvfp4_multi_gpus[throughput]
accuracy/test_llm_api_pytorch.py::TestGLM4_5Air::test_nvfp4_multi_gpus[throughput_trtllm]
accuracy/test_llm_api_pytorch.py::TestGLM4_5Air::test_nvfp4_2_model_mtp[2model]
accuracy/test_llm_api_pytorch.py::TestGLM4_5Air::test_nvfp4_2_model_mtp[2model_trtllm]
accuracy/test_llm_api_pytorch.py::TestSeedOss_36B::test_auto_dtype
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::TestKimiK25::test_nvfp4[tp8]
accuracy/test_llm_api_pytorch.py::TestKimiK25::test_nvfp4[tp8_attn_dp]
accuracy/test_llm_api_pytorch.py::TestKimiK25::test_nvfp4[ep8]
accuracy/test_llm_api_pytorch.py::TestKimiK25::test_nvfp4[dep8]
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::TestPhi4MM::test_auto_dtype
accuracy/test_llm_api_pytorch.py::TestPhi4MM::test_auto_dtype_long_rope
accuracy/test_llm_api_pytorch.py::TestPhi4MM::test_fp4
accuracy/test_llm_api_pytorch.py::TestPhi4MM::test_fp8
accuracy/test_llm_api_pytorch.py::TestPhi4MiniInstruct::test_auto_dtype
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::TestEXAONE4::test_auto_dtype
accuracy/test_llm_api_pytorch.py::TestMistralNemo12B::test_auto_dtype
accuracy/test_llm_api_pytorch.py::TestMistralNemo12B::test_auto_dtype_tp2
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_auto_dtype_4gpus[4-4-True-True-True]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_fp8_4gpus[attention_dp_off-python_mamba_cache]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_fp8_4gpus[attention_dp_off-cpp_mamba_cache]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_fp8_4gpus[attention_dp_on-python_mamba_cache]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_fp8_4gpus[attention_dp_on-cpp_mamba_cache]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_8gpus[attention_dp_on-trtllm]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_4gpu_mtp_ar
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_bf16_4gpu_mtp_ar
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_8gpus_mtp
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_parallelism[TP4_PP2]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_parallelism[TEP4_PP2]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_parallelism[TP8_PP1]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_parallelism[TEP8_PP1]
accuracy/test_llm_api_pytorch.py::TestNemotronV3Super::test_nvfp4_parallelism[TP8_PP1_ADP]
accuracy/test_llm_api_pytorch.py::TestGLM5FP8::test_8gpus[tp_size=8-ep_size=8]
accuracy/test_llm_api_pytorch.py::TestQwen3_5_397B_A17B::test_nvfp4[adp4_cutedsl]
accuracy/test_llm_api_pytorch.py::TestQwen3_5_397B_A17B::test_nvfp4[adp4_trtllm]
accuracy/test_llm_api_pytorch.py::TestQwen3_5_397B_A17B::test_nvfp4[tep4_cutedsl]
accuracy/test_llm_api_pytorch.py::TestQwen3_5_397B_A17B::test_nvfp4[tep4_trtllm]

# 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::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::TestVILA1_5_3B::test_auto_dtype
accuracy/test_llm_api_pytorch_multimodal.py::TestNemotron_Nano_12B_V2_VL::test_auto_dtype
accuracy/test_llm_api_pytorch_multimodal.py::TestPhi4MMFusedVisionLora::test_auto_dtype
accuracy/test_llm_api_pytorch_multimodal.py::TestGemma3_27BInstruct::test_fp8_prequantized
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]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype[mtp_nextn=2-overlap_scheduler=False]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_guided_decoding[xgrammar-mtp_nextn=0]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_guided_decoding[xgrammar-mtp_nextn=2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_guided_decoding[llguidance-mtp_nextn=0]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_guided_decoding[llguidance-mtp_nextn=2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp1tp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp1tp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp1tp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp1tp1cp4]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp1tp1cp4]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp1tp1cp4]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp2tp1cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp2tp1cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp2tp1cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp1dp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp1dp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp1dp2cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp1tp2cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp1tp2cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp1tp2cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp1tp1cp4]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp1tp1cp4]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp1tp1cp4]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp2tp1cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp2tp1cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp2tp1cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[nccl-cudagraph:with_padding-pp1dp2cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v1-cudagraph:with_padding-pp1dp2cp2]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype_with_helix[fifo_v2-cudagraph:with_padding-pp1dp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV32Exp::test_auto_dtype_with_helix[fifo-cudagraph:with_padding-pp1tp1cp4]
accuracy/test_disaggregated_serving.py::TestDeepSeekV32Exp::test_auto_dtype_with_helix[fifo-cudagraph:with_padding-pp1tp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV32Exp::test_auto_dtype_with_helix[fifo-cudagraph:with_padding-pp1dp2cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV32Exp::test_auto_dtype_with_helix[fifo-cudagraph:with_padding-pp2tp1cp2]
accuracy/test_disaggregated_serving.py::TestDeepSeekV3Lite::test_nixl_backend
accuracy/test_dwdp_disaggregated_serving.py::TestDwdpDeepSeekV3Lite::test_dwdp_accuracy
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::TestGPTOSS::test_auto_dtype[True]
accuracy/test_disaggregated_serving.py::TestGPTOSS::test_auto_dtype[False]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_auto_dtype[False-False-False-True]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_auto_dtype[True-True-True-True]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ngram
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_eagle3[eagle3_one_model=True-overlap_scheduler=True]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_eagle3[eagle3_one_model=False-overlap_scheduler=False]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_guided_decoding[xgrammar]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_guided_decoding[llguidance]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[xgrammar-eagle3_one_model=True]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[xgrammar-eagle3_one_model=False]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[llguidance-eagle3_one_model=True]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_guided_decoding_with_eagle3[llguidance-eagle3_one_model=False]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[GSM8K-tp1pp2]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[GSM8K-tp2pp1]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[GSM8K-tp2pp2]
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-tp2pp2]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_tp_pp_symmetric[MMLU-tp2pp1]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=1-ctx_pp=2]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=1-ctx_pp=4]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=2-ctx_pp=2]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[GSM8K-gen_tp=2-ctx_pp=4]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[MMLU-gen_tp=1-ctx_pp=2]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[MMLU-gen_tp=1-ctx_pp=4]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[MMLU-gen_tp=2-ctx_pp=2]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_ctx_pp_gen_tp_asymmetric[MMLU-gen_tp=2-ctx_pp=4]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_multi_instance[GSM8K]
accuracy/test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_multi_instance[MMLU]
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_8B::test_auto_dtype[False-True]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype[True-True]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_auto_dtype[False-False]
accuracy/test_disaggregated_serving.py::TestQwen3_8B::test_nixl_backend
accuracy/test_disaggregated_serving.py::TestKimiK2::test_nvfp4
accuracy/test_disaggregated_serving.py::TestNemotron3Super120B::test_auto_dtype
accuracy/test_disaggregated_serving.py::TestNemotron3Super120B::test_nixl_backend

# e2e test
test_e2e.py::test_llama_e2e[use_py_session-remove_input_padding-]
test_e2e.py::test_llama_e2e[use_py_session--]
test_e2e.py::test_mistral_e2e[use_py_session-remove_input_padding--]
test_e2e.py::test_mistral_e2e[use_py_session---]
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_openai_multi_chat_example
test_e2e.py::test_openai_consistent_chat
test_e2e.py::test_openai_chat_guided_decoding[meta-llama/Llama-3.1-8B-Instruct]
test_e2e.py::test_openai_chat_guided_decoding[openai/gpt-oss-120b]
test_e2e.py::test_openai_chat_harmony
test_e2e.py::test_trtllm_benchmark_serving[llama-3.1-model/Meta-Llama-3.1-8B]
test_e2e.py::test_trtllm_benchmark_serving[gpt_oss/gpt-oss-20b]
test_e2e.py::test_trtllm_multimodal_benchmark_serving
test_e2e.py::test_ptp_quickstart
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_ngram[Llama-3.1-8B-Instruct-llama-3.1-model/Llama-3.1-8B-Instruct]
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[Llama3.1-405B-FP8-llama-3.1-model/Llama-3.1-405B-Instruct-FP8-8]
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_8gpus_chunked_prefill_sq_22k[Llama-4-Maverick-17B-128E-Instruct-FP8-llama4-models/nvidia/Llama-4-Maverick-17B-128E-Instruct-FP8-False]
test_e2e.py::test_ptp_quickstart_advanced_8gpus_chunked_prefill_sq_22k[Llama-4-Maverick-17B-128E-Instruct-FP8-llama4-models/nvidia/Llama-4-Maverick-17B-128E-Instruct-FP8-True]
test_e2e.py::test_ptp_quickstart_advanced_8gpus_chunked_prefill_sq_22k[Llama-4-Scout-17B-16E-Instruct-FP8-llama4-models/Llama-4-Scout-17B-16E-Instruct-FP8-True]
test_e2e.py::test_ptp_quickstart_advanced_8gpus_chunked_prefill_sq_22k[Llama-4-Scout-17B-16E-Instruct-FP4-llama4-models/Llama-4-Scout-17B-16E-Instruct-FP4-True]
test_e2e.py::test_relaxed_acceptance_quickstart_advanced_deepseek_r1_8gpus[DeepSeek-R1-DeepSeek-R1/DeepSeek-R1]
test_e2e.py::test_ptp_quickstart_multimodal_kv_cache_reuse[mistral-small-3.1-24b-instruct-Mistral-Small-3.1-24B-Instruct-2503-0.6-image]
test_e2e.py::test_ptp_quickstart_multimodal_chunked_prefill[mistral-small-3.1-24b-instruct-Mistral-Small-3.1-24B-Instruct-2503-0.6-image]
test_e2e.py::test_ptp_quickstart_multimodal_phi4mm[phi4-multimodal-instruct-multimodals/Phi-4-multimodal-instruct-audio]
test_e2e.py::test_ptp_quickstart_multimodal_phi4mm[phi4-multimodal-instruct-multimodals/Phi-4-multimodal-instruct-image]
test_e2e.py::test_ptp_quickstart_multimodal_phi4mm[phi4-multimodal-instruct-multimodals/Phi-4-multimodal-instruct-image_audio]
test_e2e.py::test_ptp_quickstart_multimodal_phi4mm[phi4-multimodal-instruct-fp4-multimodals/Phi-4-multimodal-instruct-FP4-image_audio]
test_e2e.py::test_ptp_quickstart_multimodal_phi4mm[phi4-multimodal-instruct-fp8-multimodals/Phi-4-multimodal-instruct-FP8-image_audio]
test_e2e.py::test_ptp_quickstart_multimodal_multiturn[mistral-small-3.1-24b-instruct-Mistral-Small-3.1-24B-Instruct-2503]
test_e2e.py::test_ptp_quickstart_bert[VANILLA-BertForSequenceClassification-bert/bert-base-uncased-yelp-polarity]
test_e2e.py::test_ptp_quickstart_bert[TRTLLM-BertForSequenceClassification-bert/bert-base-uncased-yelp-polarity]
test_e2e.py::test_ptp_quickstart_advanced_pp_enabled[Llama3.3-70B-FP8-llama-3.3-models/Llama-3.3-70B-Instruct-FP8-2-2-False]
test_e2e.py::test_ptp_quickstart_advanced_pp_enabled[Llama3.3-70B-FP8-llama-3.3-models/Llama-3.3-70B-Instruct-FP8-2-2-True]
test_e2e.py::test_ptp_quickstart_advanced_pp_enabled[Llama3.3-70B-FP8-llama-3.3-models/Llama-3.3-70B-Instruct-FP8-2-4-False]
test_e2e.py::test_ptp_quickstart_advanced_pp_enabled[Llama3.3-70B-FP8-llama-3.3-models/Llama-3.3-70B-Instruct-FP8-2-4-True]
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_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_eagle3_output_repetition_4gpus[modelopt-hf-model-hub/Llama-3.3-70B-Instruct-fp8-EAGLE3-LLaMA3.3-Instruct-70B]
test_e2e.py::test_eagle3_output_repetition_4gpus[Qwen3/Qwen3-30B-A3B-Qwen3/Qwen3-30B-eagle3]
test_e2e.py::test_eagle3_output_repetition_4gpus[Qwen3/saved_models_Qwen3-235B-A22B_fp8_hf-Qwen3/qwen3-235B-eagle3]
test_e2e.py::test_eagle3_output_repetition_4gpus[llama4-models/nvidia/Llama-4-Maverick-17B-128E-Instruct-FP8-Llama-4-Maverick-17B-128E-Eagle3]
test_e2e.py::test_eagle3_output_repetition_4gpus[Qwen3/saved_models_Qwen3-235B-A22B_nvfp4_hf-Qwen3/qwen3-235B-eagle3]
unittest/llmapi/test_llm_pytorch.py::test_gemma3_1b_instruct_multi_lora
unittest/llmapi/test_llm_pytorch.py::test_lora_many_adapters_no_memory_leak
llmapi/test_llm_examples.py::test_llmapi_server_example

# e2e serve test
examples/serve/test_serve.py::test_config_file_loading[--extra_llm_api_options]
examples/serve/test_serve.py::test_config_file_loading[--config]
examples/serve/test_serve.py::test_env_overrides_pdl
examples/serve/test_serve_negative.py::test_invalid_max_tokens
examples/serve/test_serve_negative.py::test_invalid_temperature
examples/serve/test_serve_negative.py::test_invalid_top_p[-0.1]
examples/serve/test_serve_negative.py::test_invalid_top_p[1.1]
examples/serve/test_serve_negative.py::test_empty_messages_array
examples/serve/test_serve_negative.py::test_missing_message_role
examples/serve/test_serve_negative.py::test_invalid_token_ids
examples/serve/test_serve_negative.py::test_extremely_large_token_id
examples/serve/test_serve_negative.py::test_server_stability_under_invalid_requests
examples/serve/test_serve_negative.py::test_concurrent_invalid_requests
examples/serve/test_serve_negative.py::test_mixed_valid_invalid_requests
examples/serve/test_serve_negative.py::test_health_check_during_errors
examples/serve/test_serve_negative.py::test_request_exceeds_context_length
examples/serve/test_serve_negative.py::test_malformed_json_request
examples/serve/test_serve_negative.py::test_missing_content_type_header
examples/serve/test_serve_negative.py::test_extremely_large_batch

# e2e disaggregated serving test
disaggregated/test_disaggregated.py::test_disaggregated_single_gpu[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_multi_gpu[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_single_gpu_trt_backend[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_mpi[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_deepseek_v3_lite_fp8_nixl[DeepSeek-V3-Lite-fp8]
disaggregated/test_disaggregated.py::test_disaggregated_logprobs_serving[llama-3.1-8b-instruct]
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_attention_dp_gen_only[DeepSeek-V3-Lite-fp8]
disaggregated/test_disaggregated.py::test_disaggregated_load_balance[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_cache_aware_balance[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_trtllm_sampler[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_llama4_long_context_kv_cache_overflow[llama4-models/nvidia/Llama-4-Maverick-17B-128E-Instruct-FP8]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_simple_qwen3[False-False-Qwen3-8B-FP8]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_simple_qwen3[False-True-Qwen3-8B-FP8]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_simple_qwen3[True-False-Qwen3-8B-FP8]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_simple_qwen3[True-True-Qwen3-8B-FP8]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_logprobs[False-TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_logprobs[True-TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_cancel_gen_requests[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_logits[False-TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated_single_gpu.py::test_disaggregated_logits[True-TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_ctxtp2pp2_gentp2pp2[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_ctxpp4_genpp4[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_kv_cache_time_output[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_events[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_auto_scaling.py::test_service_discovery[etcd-round_robin]
disaggregated/test_auto_scaling.py::test_service_discovery[etcd-load_balancing]
disaggregated/test_auto_scaling.py::test_service_discovery[etcd-kv_cache_aware]
disaggregated/test_auto_scaling.py::test_service_discovery[http-round_robin]
disaggregated/test_auto_scaling.py::test_service_discovery[http-load_balancing]
disaggregated/test_auto_scaling.py::test_service_discovery[http-kv_cache_aware]
disaggregated/test_auto_scaling.py::test_minimal_instances[etcd-round_robin]
disaggregated/test_auto_scaling.py::test_minimal_instances[http-round_robin]
disaggregated/test_auto_scaling.py::test_worker_restart[etcd-round_robin]
disaggregated/test_auto_scaling.py::test_worker_restart[etcd-load_balancing]
disaggregated/test_auto_scaling.py::test_worker_restart[etcd-kv_cache_aware]
disaggregated/test_auto_scaling.py::test_worker_restart[http-round_robin]
disaggregated/test_auto_scaling.py::test_worker_restart[http-load_balancing]
disaggregated/test_auto_scaling.py::test_worker_restart[http-kv_cache_aware]
disaggregated/test_auto_scaling.py::test_disagg_server_restart[etcd-round_robin]
disaggregated/test_auto_scaling.py::test_disagg_server_restart[http-round_robin]
disaggregated/test_disaggregated.py::test_disaggregated_overlap_transceiver_runtime_python[TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_transceiver_runtime_python[DeepSeek-V3-Lite-fp8]
disaggregated/test_disaggregated.py::test_disaggregated_overlap_gen_first[ctx_pp1-TinyLlama-1.1B-Chat-v1.0]
disaggregated/test_disaggregated.py::test_disaggregated_overlap_gen_first[ctx_pp4-TinyLlama-1.1B-Chat-v1.0]

# llm-api promote pytorch to default
llmapi/test_llm_api_qa.py::TestLlmDefaultBackend::test_llm_args_logging
llmapi/test_llm_api_qa.py::TestLlmDefaultBackend::test_llm_args_type_tensorrt
llmapi/test_llm_api_qa.py::TestLlmDefaultBackend::test_llm_args_type_default
llmapi/test_llm_api_qa.py::TestLlmDefaultBackend::test_llm_args_logging

# AutoDeploy text generation accuracy tests
accuracy/test_llm_api_autodeploy.py::TestLlama3_1_8B_Instruct_Eagle3::test_eagle3_one_model
accuracy/test_llm_api_autodeploy.py::TestNemotronSuperV3::test_mtp[4]
