-
Notifications
You must be signed in to change notification settings - Fork 111
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
details about the base model and the fine-tuned #18
Comments
@futureisatyourhand Hi~How did you evaluate the base model?Did you execute |
Hello, I have solved the problem, but the AP is much higher than the author's, the randomness is very strong. |
i've met similar problem, could you please tell us how you solve it? @futureisatyourhand |
|
Thanks for your reply! I still have some questions: |
Hi~How did you perform the code when evaluating the model? |
The weight file is your best weights saved, and the "results/path/to/comp4_det_test_" is your results containing boxes and label. when the first instruction executed successfully, and the the "results/path/to/comp4_det_test_" is generated. You should scan source code. |
Is the first instruction written like this? python valid_ensemble.py cfg/metayolo.data cfg/darknet_dynamic.cfg cfg/reweighting_net.cfg backup/metayolo/000010.weights
I want to know the best weights is darknet19_448.cfg or backup/metayolo/****.weights?
Because when I use darknet19_448.cfg the code can't run
…------------------ 原始邮件 ------------------
发件人: "bingykang/Fewshot_Detection" ***@***.***>;
发送时间: 2021年9月23日(星期四) 上午9:19
***@***.***>;
***@***.******@***.***>;
主题: Re: [bingykang/Fewshot_Detection] details about the base model and the fine-tuned (#18)
Hi~How did you perform the code when evaluating the model?
“python valid_ensemble.py cfg/metayolo.data cfg/darknet_dynamic.cfg cfg/reweighting_net.cfg path/toweightfile
python scripts/voc_eval.py results/path/to/comp4_det_test_”
I don’t know the “path/toweightfile” and “results/path/to/comp4_det_test_”
The weight file is your best weights saved, and the "results/path/to/comp4_det_test_" is your results containing boxes and label. when the first instruction executed successfully, and the the "results/path/to/comp4_det_test_" is generated. You should scan source code.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications on the go with GitHub Mobile for iOS or Android.
|
the
the instruction is to test, so it is no relation to darknet19_448.cfg, the darknet19_448.weights is pretrained weights to get backup/metayolo/000010.weights!!! |
I want to know why the ***.txt is empty?It has no information.
…------------------ 原始邮件 ------------------
发件人: "bingykang/Fewshot_Detection" ***@***.***>;
发送时间: 2021年9月23日(星期四) 中午11:40
***@***.***>;
***@***.******@***.***>;
主题: Re: [bingykang/Fewshot_Detection] details about the base model and the fine-tuned (#18)
Is the first instruction written like this? python valid_ensemble.py cfg/metayolo.data cfg/darknet_dynamic.cfg cfg/reweighting_net.cfg backup/metayolo/000010.weights I want to know the best weights is darknet19_448.cfg or backup/metayolo/.weights? Because when I use darknet19_448.cfg the code can't run
…
------------------ 原始邮件 ------------------ 发件人: "bingykang/Fewshot_Detection" @.>; 发送时间: 2021年9月23日(星期四) 上午9:19 @.>; @.@.**>; 主题: Re: [bingykang/Fewshot_Detection] details about the base model and the fine-tuned (#18) Hi~How did you perform the code when evaluating the model? “python valid_ensemble.py cfg/metayolo.data cfg/darknet_dynamic.cfg cfg/reweighting_net.cfg path/toweightfile python scripts/voc_eval.py results/path/to/comp4_det_test_” I don’t know the “path/toweightfile” and “results/path/to/comp4_det_test_” The weight file is your best weights saved, and the "results/path/to/comp4_det_test_" is your results containing boxes and label. when the first instruction executed successfully, and the the "results/path/to/comp4_det_test_" is generated. You should scan source code. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.
the
Is the first instruction written like this? python valid_ensemble.py cfg/metayolo.data cfg/darknet_dynamic.cfg cfg/reweighting_net.cfg backup/metayolo/000010.weights I want to know the best weights is darknet19_448.cfg or backup/metayolo/.weights? Because when I use darknet19_448.cfg the code can't run
…
------------------ 原始邮件 ------------------ 发件人: "bingykang/Fewshot_Detection" @.>; 发送时间: 2021年9月23日(星期四) 上午9:19 @.>; @.@.**>; 主题: Re: [bingykang/Fewshot_Detection] details about the base model and the fine-tuned (#18) Hi~How did you perform the code when evaluating the model? “python valid_ensemble.py cfg/metayolo.data cfg/darknet_dynamic.cfg cfg/reweighting_net.cfg path/toweightfile python scripts/voc_eval.py results/path/to/comp4_det_test_” I don’t know the “path/toweightfile” and “results/path/to/comp4_det_test_” The weight file is your best weights saved, and the "results/path/to/comp4_det_test_" is your results containing boxes and label. when the first instruction executed successfully, and the the "results/path/to/comp4_det_test_" is generated. You should scan source code. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.
the instruction is to test, so it is no relation to darknet19_448.cfg, the darknet19_448.weights is pretrained weights to get backup/metayolo/000010.weights!!!
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub, or unsubscribe.
Triage notifications on the go with GitHub Mobile for iOS or Android.
|
your result is none! You must look code. |
I use two gpus, the other configurations is the same as the author, why is my performance poor?
The following figure is the eval of 500 epochs for base model.
![500](https://user-images.githubusercontent.com/40520052/69598113-441fa800-1043-11ea-9645-b28b7b751a8f.png)
The following figure is the eval of 5 epochs for the fine-tune.
![5](https://user-images.githubusercontent.com/40520052/69598120-4aae1f80-1043-11ea-98c4-5df4267ab1f2.png)
The following figure is the eval of 10 epochs for the fine-tune.
![10](https://user-images.githubusercontent.com/40520052/69598130-50a40080-1043-11ea-8662-7449b4e415b1.png)
@bingykang
The text was updated successfully, but these errors were encountered: