[AMD][CI] Fix VLM MMMU nightly max_tokens for CoT prompt (#41979)
This commit is contained in:
@@ -229,7 +229,10 @@ class TestNightlyVLMMmmuEvalAMD(unittest.TestCase):
|
||||
eval_name="mmmu",
|
||||
num_examples=100,
|
||||
num_threads=64,
|
||||
max_tokens=30,
|
||||
# The MMMU prompt asks for step-by-step reasoning ending in
|
||||
# "Answer: X" (#21841); 30 tokens truncated it before the
|
||||
# answer, so the parser fell back to "A" for every model.
|
||||
max_tokens=1024,
|
||||
)
|
||||
|
||||
# Run evaluation with timing
|
||||
|
||||
Reference in New Issue
Block a user