From fd4e6132e5cb1d7d7397475c218a4d6cf623047a Mon Sep 17 00:00:00 2001 From: yufeiwu-nv <230315618+yufeiwu-nv@users.noreply.github.com> Date: Tue, 27 Jan 2026 13:05:42 +0800 Subject: [PATCH] [None][test] Fix missing test cases (#10881) Signed-off-by: yufeiwu-nv <230315618+yufeiwu-nv@users.noreply.github.com> Co-authored-by: Larry Xu <197874197+LarryXFly@users.noreply.github.com> Signed-off-by: Wangshanshan <30051912+dominicshanshan@users.noreply.github.com> --- .../test_lists/qa/llm_perf_core.yml | 49 ++++++++++++++----- 1 file changed, 37 insertions(+), 12 deletions(-) diff --git a/tests/integration/test_lists/qa/llm_perf_core.yml b/tests/integration/test_lists/qa/llm_perf_core.yml index 76d67941a6..a89c181f70 100644 --- a/tests/integration/test_lists/qa/llm_perf_core.yml +++ b/tests/integration/test_lists/qa/llm_perf_core.yml @@ -251,11 +251,11 @@ llm_perf_core: #Mistral-Small-3.1-24B-Instruct-2503 - perf/test_perf.py::test_perf[mistral_small_v3.1_24b-bench-pytorch-bfloat16-maxbs:4096-maxnt:20000-input_output_len:20000,2000-reqs:300-con:200-gpus:2] TIMEOUT(120) - perf/test_perf.py::test_perf[starcoder_15b-bench-float16-input_output_len:512,200-gpus:4] - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:512-input_output_len:128,128-ep:4-tp:4-gpus:4] - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-streaming-float4-maxbs:512-input_output_len:128,128-ep:4-tp:4-gpus:4] - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:1000,1000-reqs:2000-ep:4-tp:4-gpus:4] TIMEOUT(120) - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:1000,2000-reqs:3000-ep:4-tp:4-gpus:4] TIMEOUT(120) - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:1000-maxnt:5000-kv_frac:0.85-input_output_len:5000,500-reqs:20000-ep:4-tp:4-gpus:4] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:512-input_output_len:128,128-ep:4-tp:4-gpus:4] + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-streaming-float4-maxbs:512-input_output_len:128,128-ep:4-tp:4-gpus:4] + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:1000,1000-reqs:2000-ep:4-tp:4-gpus:4] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:1000,2000-reqs:3000-ep:4-tp:4-gpus:4] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:1000-maxnt:5000-kv_frac:0.85-input_output_len:5000,500-reqs:20000-ep:4-tp:4-gpus:4] TIMEOUT(120) #llama_v3.1_405b_instruct_fp4 - perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:128,128-gpus:4] - perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:1000,1000-gpus:4] @@ -285,10 +285,10 @@ llm_perf_core: tests: # for chunked prefill cases - perf/test_perf.py::test_perf[deepseek_v3_lite_nvfp4-bench-pytorch-float4-maxbs:512-maxnt:2048-kv_frac:0.85-input_output_len:3000,500-reqs:200] - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:512-maxnt:2048-kv_frac:0.85-input_output_len:5000,500-reqs:200-ep:4-tp:4-gpus:4] TIMEOUT(120) - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:256-maxnt:1024-kv_frac:0.85-input_output_len:2000,2000-reqs:200-ep:4-tp:4-gpus:4] TIMEOUT(120) - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:1000-maxnt:5000-kv_frac:0.85-input_output_len:5000,500-reqs:2000-ep:4-tp:4-gpus:4] TIMEOUT(120) - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:32-maxnt:32768-input_output_len:8192,1024-reqs:20-con:1-ep:1-tp:4-gpus:4] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:512-maxnt:2048-kv_frac:0.85-input_output_len:5000,500-reqs:200-ep:4-tp:4-gpus:4] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:256-maxnt:1024-kv_frac:0.85-input_output_len:2000,2000-reqs:200-ep:4-tp:4-gpus:4] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:1000-maxnt:5000-kv_frac:0.85-input_output_len:5000,500-reqs:2000-ep:4-tp:4-gpus:4] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:32-maxnt:32768-input_output_len:8192,1024-reqs:20-con:1-ep:1-tp:4-gpus:4] TIMEOUT(120) # 12: B200, B300 test cases - condition: @@ -304,6 +304,31 @@ llm_perf_core: - perf/test_perf.py::test_perf[kimi_k2_nvfp4-bench-pytorch-float4-maxbs:512-maxnt:2048-kv_frac:0.6-input_output_len:1000,1000-ep:8-tp:8-gpus:8] - perf/test_perf.py::test_perf[kimi_k2_nvfp4-bench-pytorch-float4-maxbs:512-maxnt:2048-kv_frac:0.6-input_output_len:5000,500-reqs:2000-ep:8-tp:8-gpus:8] TIMEOUT(120) + # llama_v3.1_405b_instruct_fp4 + - perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:128,128-tp:8-gpus:8] + - perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:512,32-tp:8-gpus:8] + - perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp4-bench-pytorch-float4-maxbs:1024-maxnt:4096-kv_frac:0.85-input_output_len:500,2000-tp:8-gpus:8] + - perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp4-bench-pytorch-float4-maxbs:1024-maxnt:4096-kv_frac:0.85-input_output_len:1000,1000-tp:8-gpus:8] + - perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp4-bench-pytorch-float4-maxbs:4096-maxnt:20000-kv_frac:0.85-input_output_len:20000,2000-tp:8-gpus:8] TIMEOUT(120) + + # deepseek_r1_fp8 + - perf/test_perf.py::test_perf[deepseek_r1_fp8-bench-pytorch-float8-maxbs:512-input_output_len:128,128-ep:8-tp:8-gpus:8] + - perf/test_perf.py::test_perf[deepseek_r1_fp8-bench-pytorch-float8-maxbs:1-input_output_len:1000,2000-reqs:10-ep:4-tp:8-gpus:8] + - perf/test_perf.py::test_perf[deepseek_r1_fp8-bench-pytorch-float8-maxbs:384-maxnt:1536-input_output_len:1000,2000-reqs:49152-con:3072-ep:8-tp:8-gpus:8] + + # deepseek_r1_0528_fp4 + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:1000,1000-reqs:20000-ep:8-tp:8-gpus:8] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-kv_frac:0.85-input_output_len:1000,2000-reqs:3000-ep:8-tp:8-gpus:8] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:1000-maxnt:5000-kv_frac:0.85-input_output_len:5000,500-reqs:20000-ep:4-tp:4-gpus:4] TIMEOUT(120) + + # gpt_oss_120b_fp4 + - perf/test_perf.py::test_perf[gpt_oss_120b_fp4-bench-pytorch-float4-maxbs:720-maxnt:16384-input_output_len:1024,1024-reqs:1280-con:256-ep:8-tp:8-gpus:8] + - perf/test_perf.py::test_perf[gpt_oss_120b_fp4-bench-pytorch-float4-maxbs:720-maxnt:16384-input_output_len:1024,1024-reqs:2560-con:512-ep:8-tp:8-gpus:8] + - perf/test_perf.py::test_perf[gpt_oss_120b_fp4-bench-pytorch-float4-maxbs:720-maxnt:16384-input_output_len:1024,1024-reqs:5120-con:1024-ep:8-tp:8-gpus:8] TIMEOUT(120) + - perf/test_perf.py::test_perf[gpt_oss_120b_fp4-bench-pytorch-float4-maxbs:720-maxnt:16384-input_output_len:1024,1024-reqs:20480-con:4096-ep:8-tp:8-gpus:8] TIMEOUT(180) + - perf/test_perf.py::test_perf[gpt_oss_120b_fp4-bench-pytorch-float4-maxbs:720-maxnt:16384-input_output_len:1024,1024-reqs:8-con:1-ep:8-tp:8-gpus:8] + - perf/test_perf.py::test_perf[gpt_oss_120b_fp4-bench-pytorch-float4-maxbs:720-maxnt:16384-input_output_len:1024,1024-reqs:100-con:32-ep:8-tp:8-gpus:8] + # 13: H100, H20, H200, B200, B300 test cases - condition: ranges: @@ -443,8 +468,8 @@ llm_perf_core: - perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-streaming-float8-input_output_len:2000,500-ep:8-tp:8-gpus:8] - perf/test_perf.py::test_perf[qwen3_235b_a22b_fp4-bench-pytorch-float4-input_output_len:1000,2000-con:8-ep:8-tp:8-gpus:8] - perf/test_perf.py::test_perf[qwen3_235b_a22b_fp4-bench-pytorch-float4-input_output_len:1000,2000-con:512-ep:8-tp:8-gpus:8] - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:1-input_output_len:1000,2000-reqs:10-ep:4-tp:8-gpus:8] TIMEOUT(120) - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-float4-maxbs:384-maxnt:1536-input_output_len:1000,2000-reqs:10000-con:3072-ep:8-tp:8-gpus:8] TIMEOUT(120) #max throughput test - - perf/test_perf.py::test_perf[deepseek_r1_nvfp4-bench-pytorch-streaming-float4-maxbs:512-maxnt:5220-input_output_len:4000,2000-reqs:512-ep:8-tp:8-gpus:8] + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:1-input_output_len:1000,2000-reqs:10-ep:4-tp:8-gpus:8] TIMEOUT(120) + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-float4-maxbs:384-maxnt:1536-input_output_len:1000,2000-reqs:10000-con:3072-ep:8-tp:8-gpus:8] TIMEOUT(120) #max throughput test + - perf/test_perf.py::test_perf[deepseek_r1_0528_fp4-bench-pytorch-streaming-float4-maxbs:512-maxnt:5220-input_output_len:4000,2000-reqs:512-ep:8-tp:8-gpus:8] - perf/test_perf.py::test_perf[deepseek_v3_lite_nvfp4-bench-pytorch-float4-maxbs:1-input_output_len:1000,2000-reqs:10-ep:4-tp:8-gpus:8] - perf/test_perf.py::test_perf[deepseek_v3_lite_nvfp4-bench-pytorch-float4-maxbs:384-maxnt:1536-input_output_len:1000,2000-reqs:10000-con:3072-ep:8-tp:8-gpus:8] TIMEOUT(120)