kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-02-03 09:41:30 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

[https://nvbugs/5564465][test] ensure deepseek_v3_lite isl + osl < max_seq_len (#8565)

Browse Source 
Signed-off-by: Ruodi Lu <ruodil@users.noreply.github.com>
Co-authored-by: Ruodi Lu <ruodil@users.noreply.github.com>
This commit is contained in:
ruodil 2025-10-28 15:25:52 +08:00 committed by GitHub
parent bf72eb045e
commit 6b9b73ee27
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tests/integration/test_lists/qa/llm_perf_cluster.yml
View File

@ -17,7 +17,7 @@ llm_perf_cluster:
- perf/test_perf.py::test_perf[mistral_7b_v0.1-bench-float16-input_output_len:500,2000-quant:fp8]
- perf/test_perf.py::test_perf[deepseek_v3_lite_nvfp4-bench-pytorch-streaming-float4-maxbs:2048-maxnt:8192-input_output_len:256,256-reqs:200]
# 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:5000,500-reqs:200]
- 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]
# Phi-4-multimodal-instruct
- perf/test_perf.py::test_perf[phi_4_multimodal_instruct-bench-pytorch-bfloat16-input_output_len:500,2000-con:250]
- perf/test_perf.py::test_perf[phi_4_multimodal_instruct-bench-pytorch-bfloat16-input_output_len:1000,1000-con:250]

2
tests/integration/test_lists/qa/llm_perf_cluster_nim.yml
View File

@ -17,7 +17,7 @@ llm_perf_cluster_nim:
- perf/test_perf.py::test_perf[mistral_7b_v0.1-bench-float16-input_output_len:500,2000-quant:fp8]
- perf/test_perf.py::test_perf[deepseek_v3_lite_nvfp4-bench-pytorch-streaming-float4-maxbs:2048-maxnt:8192-input_output_len:256,256-reqs:200]
# 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:5000,500-reqs:200]
- 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]
# Phi-4-multimodal-instruct
- perf/test_perf.py::test_perf[phi_4_multimodal_instruct-bench-pytorch-bfloat16-input_output_len:500,2000-con:250]
- perf/test_perf.py::test_perf[phi_4_multimodal_instruct-bench-pytorch-bfloat16-input_output_len:1000,1000-con:250]

2
tests/integration/test_lists/qa/llm_perf_core.yml
View File

@ -80,7 +80,7 @@ llm_perf_core:
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-float8-input_output_len:2000,500]
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-streaming-float8-input_output_len:2000,500]
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-float8-input_output_len:500,2000]
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-float8-maxbs:512-maxnt:2048-kv_frac:0.85-input_output_len:5000,500-reqs:200]
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-float8-maxbs:512-maxnt:2048-kv_frac:0.85-input_output_len:3000,500-reqs:200]
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-streaming-float8-input_output_len:128,128]
- perf/test_perf.py::test_perf[mistral_small_v3.1_24b-bench-pytorch-bfloat16-maxbs:1-maxnt:20000-input_output_len:20000,2000-reqs:8-con:1] TIMEOUT(120)
- perf/test_perf.py::test_perf[mistral_small_v3.1_24b-bench-pytorch-bfloat16-maxbs:4096-maxnt:20000-input_output_len:20000,2000-reqs:500-con:200] TIMEOUT(120)

2
tests/integration/test_lists/qa/llm_perf_sanity.yml
View File

@ -166,7 +166,7 @@ llm_perf_sanity:
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-streaming-float8-input_output_len:128,128]
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-streaming-float8-input_output_len:2000,500]
# for chunked prefill cases
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-float8-maxbs:512-maxnt:2048-kv_frac:0.85-input_output_len:5000,500-reqs:200]
- perf/test_perf.py::test_perf[deepseek_v3_lite_fp8-bench-pytorch-float8-maxbs:512-maxnt:2048-kv_frac:0.85-input_output_len:3000,500-reqs:200]
- perf/test_perf.py::test_perf[llama_v3.1_405b_instruct_fp8-bench-pytorch-float8-input_output_len:128,128-tp:8-gpus:8]
- perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-input_output_len:512,32-ep:8-tp:8-gpus:8]
- perf/test_perf.py::test_perf[llama_v4_maverick_17b_128e_instruct_fp8-bench-pytorch-float8-maxbs:1024-maxnt:20000-kv_frac:0.6-input_output_len:20000,2000-reqs:1000-ep:8-tp:8-gpus:8]