kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-01-13 22:18:36 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Merge d7b3cd4ec5 into 38296a472b

Browse Source
This commit is contained in:
Chuang Zhu 2026-01-13 19:43:46 +08:00 committed by GitHub
commit 33c1106030
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 9 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
tests/integration/defs/disaggregated/test_disaggregated.py
View File

@ -365,8 +365,13 @@ def run_disaggregated_test(example_dir,
"""Run disaggregated test with given configuration."""
cleanup_output_files()
run_env = env.copy()
run_env["UCX_TLS"] = "^ib"
# on some CI nodes , we set UCX_TLS to "^ib" to avoid the issue that IB equipped but not available.
# we set UCX_MM_ERROR_HANDLING to "y" to avoid the issue that NIXL cannot use IB or TCP for notify on some CI nodes,
# setting it to "y" will enable NIXL to use system memory for notify.
run_env["UCX_TLS"] = "^ib"
run_env["UCX_MM_ERROR_HANDLING"] = "y"
num_ranks, config_file = get_test_config(test_desc, example_dir,
os.path.dirname(__file__))
@ -1190,6 +1195,7 @@ def test_disaggregated_deepseek_v3_lite_fp8_nixl(disaggregated_test_root,
env = llm_venv._new_env.copy()
env["TRTLLM_USE_NIXL_KVCACHE"] = "1"
env["UCX_TLS"] = "^ib"
env["UCX_MM_ERROR_HANDLING"] = "y"
run_disaggregated_test(disaggregated_example_root,
"deepseek_v3_lite_fp8_nixl",
env=env,
@ -1497,6 +1503,7 @@ def run_disaggregated_benchmark(example_dir,
"""Run disaggregated test with given configuration."""
run_env = env.copy()
run_env["UCX_TLS"] = "^ib"
run_env["UCX_MM_ERROR_HANDLING"] = "y"
workers_cmd = [
'mpirun', '--allow-run-as-root', '--oversubscribe', '-n',
str(num_ranks), 'trtllm-serve', 'disaggregated_mpi_worker', '-c',
@ -1677,6 +1684,7 @@ def run_disaggregated_aiperf(config_file,
cleanup_output_files()
run_env = env.copy()
run_env["UCX_TLS"] = "^ib"
run_env["UCX_MM_ERROR_HANDLING"] = "y"
workers_cmd = [
'mpirun', '--allow-run-as-root', '--oversubscribe', '-n',

3
tests/integration/test_lists/waives.txt
View File

@ -343,9 +343,6 @@ triton_server/test_triton.py::test_gpt_ib_lad[gpt-ib-lad] SKIP (https://nvbugs/5
unittest/_torch/modules/test_fused_moe.py::test_fused_moe_fp8_blockwise_cute_dsl_multi_gpu[MoEWeightLoadingMode.FUSED_GATE_UP_PROJ-DefaultMoeRoutingMethod-1] SKIP (https://nvbugs/5775256)
unittest/_torch/auto_deploy/unit/multigpu/transformations/library/test_ep_sharding.py::test_ep_shard[3-2] SKIP (https://nvbugs/5777041)
unittest/_torch/auto_deploy/unit/multigpu/transformations/library/test_ep_sharding.py::test_ep_shard[8-2] SKIP (https://nvbugs/5777041)
disaggregated/test_disaggregated.py::test_disaggregated_benchmark_on_diff_backends[llama-3.1-8b-instruct-hf-fp8] SKIP (https://nvbugs/5769890)
disaggregated/test_disaggregated.py::test_disaggregated_benchmark_on_diff_backends[DeepSeek-V3-Lite-bf16] SKIP (https://nvbugs/5769890)
disaggregated/test_disaggregated.py::test_disaggregated_benchmark_on_diff_backends[llama-v3-8b-hf] SKIP (https://nvbugs/5769890,https://nvbugs/5748683)
accuracy/test_llm_api_pytorch.py::TestDeepSeekR1::test_nvfp4_multi_gpus[throughput_pp4_mtp] SKIP (https://nvbugs/5779536)
unittest/_torch/attention/test_flashinfer_star_attn.py::TestStarAttention::test_flashinfer_star_attention[num_layers:2-num_heads:32-num_kv_heads:8-head_dim:64-anchor_size:64-block_size:64-dtype:torch.float16] SKIP (https://nvbugs/5781389)
unittest/_torch/ray_orchestrator/multi_gpu/test_ops.py::test_reducescatter_pg_op[var_len:True-seqlen:16-hidden:128] SKIP (https://nvbugs/5781383)