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

[https://nvbugs/5433545][fix] TestPhi4MiniInstruct::test_auto_dtype - Use max_seq_len=4096 to fallback to the short RoPE factor (#6895)

Browse Source 
Signed-off-by: Michal Guzek <mguzek@nvidia.com>
This commit is contained in:
Michal Guzek 2025-08-22 10:28:09 -07:00 committed by GitHub
parent 4b6cca0662
commit 7ea53ff516
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tests/integration/defs/accuracy/test_llm_api_pytorch.py
View File

@ -2334,7 +2334,7 @@ class TestPhi4MiniInstruct(LlmapiAccuracyTestHarness):
MODEL_PATH = f"{llm_models_root()}/Phi-4-mini-instruct"
def test_auto_dtype(self):
with LLM(self.MODEL_PATH) as llm:
with LLM(self.MODEL_PATH, max_seq_len=4096) as llm:
task = CnnDailymail(self.MODEL_NAME)
task.evaluate(llm)
task = MMLU(self.MODEL_NAME)