crp94 / personalised-aesthetic-assessment-using-residual-adapters Goto Github PK

I am a newer and have no idea to set the k in this script .
would you mind showing me the best performed script ？thanks !

Hello, thank you for your work, when I run the file I have encountered the following problems：
Adapters.py
Problem1：
try:
running_loss += loss.data[0]
except:
print('unexpected error, could not calculate loss or do a sum.')

During training, it outputs 'unexpected error, could not calculate loss or do a sum.'
So I changed running_loss + = loss.data [0] to running_loss + = loss.item ()

Problem2：
ratings_i=[(list(np.float_([j for j in i]))) for i in ratings]
predictions_i=[(list(np.float_([j for j in i]))) for i in predictions]
ratings_i=np.vstack(ratings)
predictions_i=np.vstack(predictions)

ratings_i has been assigned twice, I do n’t know what should be correct here，and the following problems occur during runtime
TypeError: can't convert CUDA tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.
I tried modifying ratings_i = np.vstack (ratings) to ratings_i = np.vstack (ratings.cpu ()) but it didn't seem to solve the problem.
Can you help, thank you very much.