search

ymy-k / DPText-DETR

[AAAI'23 Oral] DPText-DETR: Towards Better Scene Text Detection with Dynamic Points in Transformer
Other
170 stars 22 forks source link

training time #24

Open yangxcccscsa opened 1 year ago

yangxcccscsa commented 1 year ago

Hello, thanks for your work. Can you tell me how long pre-training and fine-tuning need to be trained on what configuration of the GPU?

babble2 commented 11 months ago

about 15h with a 3090 during the pre-training, and fine-tuning about 3 hours