about eval.py

[imotas30]

hello ,I can't get the same result using the model you give. For example, when I use my result of UAV123 , I get the lower 4% evaluation than that in readme.markdown you give. Even if I use the result of UAV123 you offer ,I get the lower 2% than that in readme.markdown.

[dhy1222]

hello ,I can't get the same result using the model you give. For example, when I use my result of UAV123 , I get the lower 4% evaluation than that in readme.markdown you give. Even if I use the result of UAV123 you offer ,I get the lower 2% than that in readme.markdown.

Have you figured it out? I meet the same question.