kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-01-14 06:27:45 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

[None][test] rename llm_perf_full to llm_perf_core and add missing cases (#7899)

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 2025-09-23 14:04:34 +08:00 committed by GitHub
parent a4b4ed4535
commit 05bec3bf0f
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
tests/integration/test_lists/qa/llm_perf_full.yml → tests/integration/test_lists/qa/llm_perf_core.yml
View File

@ -1,5 +1,5 @@
version: 0.0.1
llm_perf_full:
llm_perf_core:
# one gpu test
- condition:
ranges:
@ -74,6 +74,8 @@ llm_perf_full:
- 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-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)
@ -382,6 +384,7 @@ llm_perf_full:
#llama_v3.1_70b
- perf/test_perf.py::test_perf[llama_v3.1_70b-bench-pytorch-bfloat16-input_output_len:128,128-tp:2-gpus:2]
- perf/test_perf.py::test_perf[llama_v3.1_70b_instruct_fp8-bench-pytorch-float8-input_output_len:128,128]
- perf/test_perf.py::test_perf[llama_v3.1_70b-bench-pytorch-bfloat16-maxbs:1-input_output_len:128,128-reqs:10-gpus:2]
- perf/test_perf.py::test_perf[llama_v3.1_70b_instruct_fp8-bench-pytorch-float8-input_output_len:512,32-kv_cache_dtype:fp8]
- perf/test_perf.py::test_perf[llama_v3.3_70b_instruct_fp4-bench-pytorch-float4-input_output_len:128,128-tp:2-gpus:2]
- perf/test_perf.py::test_perf[llama_v3.3_70b_instruct_fp4-bench-pytorch-float4-input_output_len:512,32-kv_cache_dtype:fp8-tp:2-gpus:2]

1
tests/integration/test_lists/qa/llm_perf_nim.yml
View File

@ -308,6 +308,7 @@ llm_perf_nim:
- perf/test_perf.py::test_perf[llama_v3.1_70b_instruct-bench-bfloat16-input_output_len:200,2000-reqs:64-con:200-gpus:8]
- perf/test_perf.py::test_perf[llama_v3.1_70b_instruct-bench-bfloat16-input_output_len:200,2000-reqs:8-con:1-gpus:8] # timeout for h20, move to l2 test
- perf/test_perf.py::test_perf[llama_v3.1_70b_instruct-bench-bfloat16-input_output_len:2000,200-reqs:64-gpus:8]
- perf/test_perf.py::test_perf[llama_v3.1_70b-bench-bfloat16-maxbs:1-maxnt:544-input_output_len:512,32-quant:fp8-gpus:8]
- perf/test_perf.py::test_perf[llama_v3.3_70b_instruct-bench-bfloat16-input_output_len:128,128-gpus:8]
- perf/test_perf.py::test_perf[llama_v3.3_70b_instruct-bench-bfloat16-maxbs:16-maxnt:5000-input_output_len:5000,500-reqs:64-con:250-gpus:8]

2
tests/integration/test_lists/qa/llm_release_perf_multinode_test.txt
View File

@ -1,2 +0,0 @@
perf/test_perf.py::test_perf[llama_v3.1_70b-cppmanager-exe-plugin_ifb-float16-input_output_len:1024,1024-quant:fp8-tp:8-pp:2]
perf/test_perf.py::test_perf[mixtral_8x7b_v0.1-cppmanager-exe-plugin_ifb-float16-input_output_len:512,512-quant:fp8-tp:8-pp:2]