How to calculate f1 score and blue score

Viewed 38

I want to calculate f1_score and blue_score of my caption model. when I feed an image to my model it predicts text captions according to the data it is being trained on.

image = "/content/ttt.png"
matches = find_matches(t_embeddings, 
                   [image], 
                   normalize=True)[1]

for i in range(9):
   print((matches[i]))

Prediction

No acute pulmonary findings.

and the actual sentence is

No acute pulmonary found.

Now I want to calculate f1_score and blue score between the predicted and actual caption. But I have no idea.

0 Answers
Related