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
resume training with My own dataset acc problem #105
Comments
Update: based on close issues, I edited this lines:
but it give an error how can i correct that |
Update:
|
after training finished, the accuracy of
What is your opinion? |
Thanks for your attention. |
Ok thanks. |
Please modify the strategy to evaluate and save your model: Lines 241 to 245 in 2cd40c4
|
very thanks Dear. |
My problem solved! and get beter results. thanks for your supports: |
thanks. my goal is to resume training with my own dataset. My dataset has only one special character +
0-9
anda-z
(38 char's). Mytraining
images are316,000
andvalidation
images are144,000
images. duringtraining
, this line ofmain.py
print("correct / total: %d / %d, " % (n_correct, n_total))
printsn_total
:64,000
. why 64000? None of mytraining
andtesting
images are64,000
. other note is thatacc
during training reach99%
but when i load trained model and test some images withdemo.py
,acc
is0
and all of predictions arewrong
. Iresume
training with yourdemo.pth
but actually, results ofdemo.pth
is very very better than new trained models.Can you help me what was wrong? thanks
The text was updated successfully, but these errors were encountered: