Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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
2 changes: 1 addition & 1 deletion cpp/include/tensorrt_llm/deep_gemm/scheduler.cuh
Original file line number Diff line number Diff line change
Expand Up @@ -379,7 +379,7 @@ struct GroupedMaskedScheduler
}
};

// Need to keep the same as the one in tests/unittest/_torch/thop/deep_gemm_tests.py
// Need to keep the same as the one in tests/unittest/_torch/thop/parallel/deep_gemm_tests.py
template <typename T_offset, typename T_index>
__host__ __device__ __forceinline__ T_offset compute_padded_offset(T_offset offset, T_index problem_idx)
{
Expand Down
3 changes: 2 additions & 1 deletion tests/integration/defs/.test_durations
Original file line number Diff line number Diff line change
Expand Up @@ -146,7 +146,8 @@
"test_unittests.py::test_unittests_v2[unittest/_torch/multimodal]": 23.54,
"test_unittests.py::test_unittests_v2[unittest/_torch/sampler]": 107.66,
"test_unittests.py::test_unittests_v2[unittest/_torch/speculative]": 1850.16,
"test_unittests.py::test_unittests_v2[unittest/_torch/thop]": 852.56,
"test_unittests.py::test_unittests_v2[unittest/_torch/thop/parallel]": 311.58,
"test_unittests.py::test_unittests_v2[unittest/_torch/thop/serial]": 18.96,
"test_unittests.py::test_unittests_v2[unittest/_torch/modeling -k \"modeling_mixtral\"]": 208.1838396479725,
"test_unittests.py::test_unittests_v2[unittest/_torch/multi_gpu_modeling -k \"deepseek\"]": 393.0210295501165,
"test_e2e.py::test_ptp_quickstart_bert[TRTLLM-BertForSequenceClassification-bert/bert-base-uncased-yelp-polarity]": 21.019993914989755,
Expand Down
12 changes: 10 additions & 2 deletions tests/integration/defs/agg_unit_mem_df.csv
Original file line number Diff line number Diff line change
Expand Up @@ -101,9 +101,17 @@ unittest/trt/model/test_mamba.py,NVIDIA H100,10,
"unittest/trt/attention/test_gpt_attention.py -k ""partition2""",NVIDIA L40,6,
"unittest/trt/attention/test_gpt_attention.py -k ""partition3""",NVIDIA L40,6,
"unittest/trt/attention/test_gpt_attention.py -k ""xqa_generic""",NVIDIA L40,3,
unittest/_torch/attention,NVIDIA Graphics Device,4,B200 Bring Up Board
unittest/_torch/misc,NVIDIA Graphics Device,4,B200 Bring Up Board
unittest/_torch/speculative,NVIDIA Graphics Device,4,B200 Bring Up Board
unittest/_torch/thop,NVIDIA Graphics Device,32,B200 Bring Up Board
unittest/_torch/thop/parallel,NVIDIA Graphics Device,4,B200 Bring Up Board
"unittest/_torch/auto_deploy/unit/singlegpu -k ""not test_trtllm_bench_backend_comparison""",NVIDIA Graphics Device,4,B200 Bring Up Board
unittest/_torch/attention,NVIDIA B200,4,
unittest/_torch/misc,NVIDIA B200,4,
unittest/_torch/speculative,NVIDIA B200,4,
unittest/_torch/thop,NVIDIA B200,32,
unittest/_torch/thop/parallel,NVIDIA B200,4,
"unittest/_torch/auto_deploy/unit/singlegpu -k ""not test_trtllm_bench_backend_comparison""",NVIDIA B200,4,
unittest/_torch/attention,NVIDIA H100,4,
unittest/_torch/misc,NVIDIA H100,4,
unittest/_torch/speculative,NVIDIA H100,2,
unittest/_torch/thop/parallel,NVIDIA H100,4,
3 changes: 2 additions & 1 deletion tests/integration/test_lists/test-db/l0_b200.yml
Original file line number Diff line number Diff line change
Expand Up @@ -51,7 +51,8 @@ l0_b200:
- unittest/_torch/multimodal
- unittest/_torch/sampler
- unittest/_torch/speculative
- unittest/_torch/thop
- unittest/_torch/thop/parallel
- unittest/_torch/thop/serial
- unittest/_torch/modeling -k "modeling_llama"
- unittest/_torch/modeling -k "modeling_mixtral"
- unittest/_torch/modeling -k "modeling_deepseek"
Expand Down
3 changes: 2 additions & 1 deletion tests/integration/test_lists/test-db/l0_h100.yml
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,8 @@ l0_h100:
- unittest/_torch/multimodal
- unittest/_torch/sampler
- unittest/_torch/speculative
- unittest/_torch/thop
- unittest/_torch/thop/parallel
- unittest/_torch/thop/serial
# Only key models in H100: llama/mixtral/nemotron/deepseek
- unittest/_torch/modeling -k "modeling_llama"
- unittest/_torch/modeling -k "modeling_mixtral"
Expand Down