Closed SilvesterSun closed 3 years ago
我是单独test,然后再eval的时候可以,但是在Ocean.yaml中设置TEST: ISTRUE: True 以及TUNE: ISTRUE: True时无法正常的评估和微调,报的错也是max() arg is an empty sequence
好的,谢谢。我也试了,test和eval分开就可以了。
------------------ Original ------------------ From: xiaofeinot <notifications@github.com> Date: Sat,Oct 31,2020 9:16 AM To: researchmm/TracKit <TracKit@noreply.github.com> Cc: SilvesterSun <983082671@qq.com>, Author <author@noreply.github.com> Subject: Re: [researchmm/TracKit] eval_vot (#43)
我是单独test,然后再eval的时候可以,但是在Ocean.yaml中设置TEST: ISTRUE: True 以及TUNE: ISTRUE: True时无法正常的评估和微调,报的错也是max() arg is an empty sequence
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.
好的,谢谢。我也试了,test和eval分开就可以了。 … ------------------ Original ------------------ From: xiaofeinot <notifications@github.com> Date: Sat,Oct 31,2020 9:16 AM To: researchmm/TracKit <TracKit@noreply.github.com> Cc: SilvesterSun <983082671@qq.com>, Author <author@noreply.github.com> Subject: Re: [researchmm/TracKit] eval_vot (#43) 我是单独test,然后再eval的时候可以,但是在Ocean.yaml中设置TEST: ISTRUE: True 以及TUNE: ISTRUE: True时无法正常的评估和微调,报的错也是max() arg is an empty sequence — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.
那你可以在某个数据集上微调吗?原作者只给了单独test和eval的命令,没有单独tune的命令
似乎你需要吧result放到result文件夹里面才能评估
好的,谢谢。我也试了,test和eval分开就可以了。 … ------------------ Original ------------------ From: xiaofeinot <notifications@github.com> Date: Sat,Oct 31,2020 9:16 AM To: researchmm/TracKit <TracKit@noreply.github.com> Cc: SilvesterSun <983082671@qq.com>, Author <author@noreply.github.com> Subject: Re: [researchmm/TracKit] eval_vot (#43) 我是单独test,然后再eval的时候可以,但是在Ocean.yaml中设置TEST: ISTRUE: True 以及TUNE: ISTRUE: True时无法正常的评估和微调,报的错也是max() arg is an empty sequence — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.
你好,请问可以详细说一下你是怎么解决的吗?
作者您好,当我在评估vot测验结果的时候,报错如下: File "/home/ubuntu/TracKit/lib/eval_toolkit/bin/../../eval_toolkit/pysot/evaluation/eao_benchmark.py", line 110, in _calculate_eao max_len = max([len(x) for x in all_overlaps]) ValueError: max() arg is an empty sequence 请问可以解答一下什么原因吗?谢谢