[Bug] comsenseqa infer/eval bug about opencompass HOT 3 CLOSED

but, i use python tools/prompt_viewer.py configs/eval_test_comsenseqa.py -n -a

from mmengine.config import read_base
from opencompass.partitioners import SizePartitioner, NaivePartitioner
from opencompass.runners import LocalRunner
from opencompass.tasks import OpenICLInferTask, OpenICLEvalTask
with read_base():
    from .datasets.commonsenseqa.commonsenseqa_ppl_5545e2 import commonsenseqa_datasets

len of self.index_ds in commonsenseqa : 9741
len of self.test_ds in commonsenseqa : 1221
but the len of ice_idx_list is 1221. It is equal to len of self.test_ds.

from opencompass.

There is a partitioner machenism during launching tasks, that would explain the differences between 611 and 1221. But I'm not sure why there is a out of bound error...

from opencompass.

It‘s my fault. I saved the intermediate results of the index as binary files, but did not consider partitioner machenism. Thanks.

from opencompass.