search

zzzDavid / ICDAR-2019-SROIE

ICDAR 2019 Robust Reading Challenge on Scanned Receipts OCR and Information Extraction
MIT License
384 stars 132 forks source link

Code to predict and evaluate the model for task 1? #16

Closed cuongngm closed 3 years ago

cuongngm commented 3 years ago

I am training CTPN model in pytorch based your task 1 code, it is clear and help me more understand the original paper . But i see 2 problem. -First, when i trained with SROIE 2019 dataset by your code, i met with "cuda out of memory" after first epoch, and my solution is add "with torch.no_grad()" in line 64 of file "train.py". -Second, i trained your code with 10 epochs and loss values don't converge. After, i changed the optimizer from "adagrad" to "SGD" with lr=1e-3. And this work. Hope this help anyone.

I had the weights of model and i have a question for author. Can you share the code to predict and evaluate the model? Tks

patrick22414 commented 3 years ago

Hi! Thank you for the interest and help. I'm glad that the code has helped you.

I'm really sorry to say that I cannot give you the predict/eval code because I haven't written them. I must admit I have half-abandoned the revamp since my work got busy.

If you would like to add your own predict/eval code, you could submit a pull request and I'll be sure to check it out. I believe I left it around NMS and visualisation.

Again, I appreciate your effort and I'm sorry that I cannot give you a satisfactory answer.

cuongngm commented 3 years ago

Tks for your reply. I will attempt. Wish you all the best in your work.

Vào Th 5, 3 thg 12, 2020 vào lúc 22:22 Patrick Yang < notifications@github.com> đã viết:

Hi! Thank you for the interest and help. I'm glad that the code has helped you.

I'm really sorry to say that I cannot give you the predict/eval code because I haven't written them. I must admit I have half-abandoned the revamp since my work got busy.

If you would like to add your own predict/eval code, you could submit a pull request and I'll be sure to check it out. I believe I left it around NMS and visualisation.

Again, I appreciate your effort and I'm sorry that I cannot give you a satisfactory answer.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/zzzDavid/ICDAR-2019-SROIE/issues/16#issuecomment-738075085, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMXE2E4GGWFTGDNEMINSZYTSS6ULVANCNFSM4T7FED3A .