yitu-opensource / T2T-ViT

ICCV2021, Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet
Other
1.14k stars 177 forks source link

colab #75

Open SMaryamHosseini opened 1 year ago

SMaryamHosseini commented 1 year ago

Hi I want to train your model on my dataset that made from 15000 sample train image and 120000 train image.also I want to use google colab,is it possible to train this model in colab?How do i run your github model in colab?is it better to train model in sample train image or train image?