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

Merge commit 'ed27a72bcf71f7ab0e7137f7999988c9de82386f' into feat/b300_cu13

Browse Source 
Signed-off-by: Xiwen Yu <13230610+VALLIS-NERIA@users.noreply.github.com>
This commit is contained in:
Xiwen Yu 2025-09-08 21:58:43 +08:00
commit 4cf9fed1e7
2 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
jenkins/L0_Test.groovy
View File

@ -380,7 +380,7 @@ def runLLMTestlistOnSlurm(pipeline, platform, testList, config=VANILLA_CONFIG, p
// Wait 10 minutes to check status of the node again
sleep(time: 10, unit: 'MINUTES')
// Avoid the node being stuck in the held state.
Utils.exec(pipeline, Utils.sshUserCmd(remote, "\"scontrol release ${slurmJobID} || true\""))
Utils.exec(pipeline, script: Utils.sshUserCmd(remote, "\"scontrol release ${slurmJobID} || true\""), numRetries: 3)
counter++
}
}

4
tests/integration/defs/accuracy/references/mmlu.yaml
View File

@ -208,10 +208,10 @@ Qwen3/Qwen3-30B-A3B:
accuracy: 79.53
- quant_algo: FP8
kv_cache_quant_algo: FP8
accuracy: 80.65
accuracy: 79.53
- quant_algo: NVFP4
kv_cache_quant_algo: FP8
accuracy: 80.65
accuracy: 79.53
- quant_algo: W4A8_MXFP4_FP8
accuracy: 79.78
- quant_algo: W4A8_MXFP4_MXFP8