cross attention — Postimages
Recently, the above model has been tried based on TensorFlow keras, and the problem of data classification accuracy has dropped significantly.
Want to perform fusion training of cross attention between two feature data，But the training classification accuracy is only 49%，however，The accuracy before fusion is 94%.
I am very confused, please give some adjustment suggestions from professional friends.
Hi @urnotcoward
Welcome to the TensorFlow Forum!
The model accuracy also depends on the dataset type, dataset preprocessing and which optimizer, loss function has been used for model compilation. The given information is not enough to understand the issue. Please provide minimal reproducible code to replicate the error and to fix this issue. Thank you
1 Like