Skip to content

[None][test] Test trtllm-bench AD vs, PT BEs on H100 single gpu #6487

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Aug 11, 2025
Merged

[None][test] Test trtllm-bench AD vs, PT BEs on H100 single gpu #6487

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
695d012
added AD perf test
MrGeva Jul 23, 2025
97732d9
set free_mem_ration to 0.5
MrGeva Jul 30, 2025
35e87cc
modified the test params
MrGeva Jul 30, 2025
a896d76
allow to run the test only on h100 and not on a30
MrGeva Jul 30, 2025
bbe4513
mark the test no_xdist, filter it out on b200
MrGeva Jul 31, 2025
806394c
add timeout to subprocess calls for CI safety
galagam Aug 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 1 addition & 2 deletions tests/integration/test_lists/test-db/l0_a30.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,8 +18,7 @@ l0_a30:
- unittest/_torch/modeling -k "modeling_phi3"
- unittest/_torch/modeling -k "modeling_qwen"
- unittest/_torch/modeling -k "modeling_qwen_moe"
- unittest/_torch/modeling -k "modeling_exaone4"
- unittest/_torch/auto_deploy/unit/singlegpu
- unittest/_torch/auto_deploy/unit/singlegpu -k "not test_trtllm_bench_backend_comparison"
- unittest/_torch/test_beam_search.py
- condition:
ranges:
Expand Down
2 changes: 1 addition & 1 deletion tests/integration/test_lists/test-db/l0_b200.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -67,7 +67,7 @@ l0_b200:
- unittest/_torch/modeling -k "modeling_mixtral"
- unittest/_torch/modeling -k "modeling_deepseek"
- unittest/_torch/modeling -k "modeling_gpt_oss"
- unittest/_torch/auto_deploy/unit/singlegpu
- unittest/_torch/auto_deploy/unit/singlegpu -k "not test_trtllm_bench_backend_comparison"
- unittest/_torch/speculative/test_eagle3.py
- unittest/_torch/speculative/test_kv_cache_reuse.py
- unittest/_torch/speculative/test_dynamic_spec_decode.py
Expand Down
1 change: 1 addition & 0 deletions tests/integration/test_lists/test-db/l0_h100.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,7 @@ l0_h100:
tests:
# ------------- PyTorch tests ---------------
# Only key models in H100: llama/mixtral/nemotron/deepseek
- unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py::test_trtllm_bench_backend_comparison
- unittest/_torch -k "not (modeling or multi_gpu or auto_deploy)" TIMEOUT (90)
- unittest/_torch -k "modeling_llama"
- unittest/_torch/modeling -k "modeling_mixtral"
Expand Down
Loading