jmmmm

Excuse me, could you please release the model weights which are not fine-tuned on COCO-Stuff training set?

Excuse me, could you please release the model weights which are not fine-tuned on COCO-Stuff training set?

Thank you for your work. If the input size of the network is 256 x 256, should I increase one encoder and decoder layer to keep the bottleneck layer's feature...

请问您的hand_feature融合入bert模型前后的效果是否对比了呀？我将您的hand_feature方法融入另一个深度模型中，训练过程中出现train_acc正常上升趋势，相比融入之前有所提升，但valid_acc随着epoch的增加不断下降，数据集是相同的。