-
Notifications
You must be signed in to change notification settings - Fork 7.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
微调det_v4,hmean很低 #12101
Comments
2000张训练数据,400张验证,batch_size是8,lr是0.0001,到21个epoch后hmean一直无法提升 |
可以先infer一下看看模型输出,查找原因 |
是因为模型太大很难调出效果吗,数据集是根据原模型检测不出来的情况准备的,专门微调检测这一种情况。现在是22个epoch,hmean还是0.04 |
数据量大吗 |
2000张训练数据,400张验证,现在是到27个epoch,hmean还是0.04 |
请问你设置cal_metric_during_train: false了么,然后这个还会评估精度么?我的都是0 |
设置了, cal_metric_during_train: false,有eval数据,每到一定轮数评估一次。我的loss在持续降低,但是precision,recall,hmean没怎么提高,都很低 |
那你模型导出时,导出了student、student2、teacher 3个文件夹的推理模型文件么? |
我用的配置是student.yml,模型是best_accuracy,微调生成best_accuracy模型。加载的时候有两个告警:The pretrained params backbone.conv1.hardswish.scale not in model,The pretrained params backbone.conv1.hardswish.bias not in model,参数不能完全加载 |
好吧,我用的ch_PP-OCRv4_det_cml.yml,这个丢失的更多。而且搜到说尽量别用这个yml |
一般不会掉这么点的,可能还是需要检查下哪里出问题了 |
@UserWangZz |
差很多是指什么情况呢? |
@UserWangZz 这有个loss一直为0 |
cbnloss在这个yml中是不起作用的 |
微调det_v4,loss从降低4.9降到1.2,但是precision,recall,hmean都很低,基本是0.04,这种情况原因是啥?
The text was updated successfully, but these errors were encountered: